diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..cbc16a281d44fa3cd26691e97819efe9aa5d73e4 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..abe358e6c598d7bbdb15fdb56d82543ce8d39c6a --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e1fb2e150457cd55e42aa984e76445c7c1975148 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ebcd1eb5cc1f7f5acdec496da85c435704fab875 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..f0ce490e5a86265ca06bbf433ba85151c44fc52e --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..419636ed9ebc47a0aac457adc4119b4a4696fa16 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..dcc482e172d6d2483eb55cf9611fc891502a93a9 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..0e25e1c9dc675f9bf268ee38368b8f2d11e1fe5b --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer3/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..6ac9d80f989deb17d1699dca3da7f5ae516f2132 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..5e1365115ff2e1e7c04df668c974946a9b6d8b9a --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..83279c896f104506f39be4ca33e2617c68e7e716 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b89fd001c4710b2db5fba3aa867355223bc23be5 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..8d45f9a6576f8b54e47015cdec9f971844520277 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e31ead982ed274eb5c1fa253b3c4cf7cf7b6812d --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1a6edaef08454bac25e4283e9f1ed2e61edcd498 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..b65520475a8eb2a0d2af9046d6714f4021b5bcbe --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..652c2b5223b952919e4467624f20aa53fdf645b9 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e82e6313dc2dcabc7b60db4ec4a5534e4d2cfce4 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..6f01773597daad147ce761eb0867030a543a4fce --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..ba9fd34b1d8b4c44e65deeca3c3f14df46f719bc --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9b4ee06e73d4add0e85cc3e59e81898ae653fe0c --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..f6f55d9b3c2e95e4318a470c6178ad511eaec087 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..1ac4b17e6c9124a1778204a042a397405a5cd419 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer4/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b5011e858455e9d32a3f5f5a0210601c7f29b574 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..a9d0c95cc209065c8fe79bfb775e5a644b3f5a3c --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..808dde63059ac0dfc60cba59b01af6936237a505 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f50022d9cf02a1f941c9a27d14759a64436cb39f --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..064ddebec9c74134b045bccd6e82913afd1427fb --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..c1f713c163bd3888ef3dd59d1edb6e5ad99479bb --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..2c1e103df6824671029a86991d52c3bf971f3476 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..383e528236481cd5a33c55a6475c40fed5b8d526 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..b143c53103eb807dd2215899ead8e9bf39cce5d1 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..ab91207f04607000e15d1e1c9a67294d77c571f0 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d111ab27e41776f5eed2e58df6cd5c14a54094d0 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e83c8a2bfa81968d7e065c4a82db203348fab963 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..99f31ececb8e9c9e01a4467a583eb79cc529a780 --- /dev/null +++ b/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 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d133dbd9b7f4f73680ded45a79035d611ef805e5 --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..91782a1e1310545666aea6dc7290c60873193adc --- /dev/null +++ b/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer5/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..657108f3455e08ff7c2b3de3481106495a53c9f6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3831237180110736 +INFO:root: +[Valid][0][UTT][ MSE = 0.3132 | LCC = 0.8272 | SRCC = 0.8344 ] +INFO:root:Average loss=0.25501078768418384 +INFO:root: +[Valid][1][UTT][ MSE = 0.2953 | LCC = 0.8379 | SRCC = 0.8495 ] +INFO:root:Average loss=0.22087167298946625 +INFO:root: +[Valid][2][UTT][ MSE = 0.2506 | LCC = 0.8597 | SRCC = 0.8603 ] +INFO:root:Average loss=0.19115176987953675 +INFO:root: +[Valid][3][UTT][ MSE = 0.2340 | LCC = 0.8610 | SRCC = 0.8620 ] +INFO:root:Average loss=0.17888337889542946 +INFO:root: +[Valid][4][UTT][ MSE = 0.2335 | LCC = 0.8636 | SRCC = 0.8659 ] +INFO:root:Average loss=0.1679643659064403 +INFO:root: +[Valid][5][UTT][ MSE = 0.2357 | LCC = 0.8631 | SRCC = 0.8640 ] +INFO:root:Average loss=0.15120859071612358 +INFO:root: +[Valid][6][UTT][ MSE = 0.2303 | LCC = 0.8630 | SRCC = 0.8634 ] +INFO:root:Average loss=0.13979135616085467 +INFO:root: +[Valid][7][UTT][ MSE = 0.2256 | LCC = 0.8621 | SRCC = 0.8630 ] +INFO:root:Average loss=0.13150826574136049 +INFO:root: +[Valid][8][UTT][ MSE = 0.2333 | LCC = 0.8590 | SRCC = 0.8606 ] +INFO:root:Average loss=0.13127750349350464 +INFO:root: +[Valid][9][UTT][ MSE = 0.2284 | LCC = 0.8560 | SRCC = 0.8570 ] +INFO:root:Average loss=0.11460986418219712 +INFO:root: +[Valid][10][UTT][ MSE = 0.2505 | LCC = 0.8545 | SRCC = 0.8547 ] +INFO:root:Average loss=0.11074295085974228 +INFO:root: +[Valid][11][UTT][ MSE = 0.2451 | LCC = 0.8527 | SRCC = 0.8556 ] +INFO:root:Average loss=0.10010256885718076 +INFO:root: +[Valid][12][UTT][ MSE = 0.2425 | LCC = 0.8500 | SRCC = 0.8495 ] +INFO:root:Average loss=0.09126249395119838 +INFO:root: +[Valid][13][UTT][ MSE = 0.2333 | LCC = 0.8513 | SRCC = 0.8537 ] +INFO:root:Average loss=0.08062022413389805 +INFO:root: +[Valid][14][UTT][ MSE = 0.2375 | LCC = 0.8504 | SRCC = 0.8518 ] +INFO:root:Average loss=0.08052675965695809 +INFO:root: +[Valid][15][UTT][ MSE = 0.2548 | LCC = 0.8471 | SRCC = 0.8478 ] +INFO:root:Average loss=0.08025678548102196 +INFO:root: +[Valid][16][UTT][ MSE = 0.2412 | LCC = 0.8471 | SRCC = 0.8496 ] +INFO:root:Average loss=0.07302775148015755 +INFO:root: +[Valid][17][UTT][ MSE = 0.2364 | LCC = 0.8489 | SRCC = 0.8508 ] +INFO:root:Average loss=0.07180747456657581 +INFO:root: +[Valid][18][UTT][ MSE = 0.2453 | LCC = 0.8449 | SRCC = 0.8471 ] +INFO:root:Average loss=0.06627320455243954 +INFO:root: +[Valid][19][UTT][ MSE = 0.2378 | LCC = 0.8497 | SRCC = 0.8502 ] +INFO:root:Average loss=0.061709498126919456 +INFO:root: +[Valid][20][UTT][ MSE = 0.2489 | LCC = 0.8452 | SRCC = 0.8480 ] +INFO:root:Average loss=0.05846950578957032 +INFO:root: +[Valid][21][UTT][ MSE = 0.2363 | LCC = 0.8466 | SRCC = 0.8488 ] +INFO:root:Average loss=0.06001436056043857 +INFO:root: +[Valid][22][UTT][ MSE = 0.2530 | LCC = 0.8421 | SRCC = 0.8476 ] +INFO:root:Average loss=0.05532627376035238 +INFO:root: +[Valid][23][UTT][ MSE = 0.2381 | LCC = 0.8465 | SRCC = 0.8483 ] +INFO:root:Average loss=0.053260249205124684 +INFO:root: +[Valid][24][UTT][ MSE = 0.2415 | LCC = 0.8476 | SRCC = 0.8491 ] +INFO:root:Average loss=0.05074558703181071 +INFO:root: +[Valid][25][UTT][ MSE = 0.2471 | LCC = 0.8517 | SRCC = 0.8538 ] +INFO:root:Average loss=0.049996789783621445 +INFO:root: +[Valid][26][UTT][ MSE = 0.2363 | LCC = 0.8488 | SRCC = 0.8520 ] +INFO:root:Average loss=0.045570363983129844 +INFO:root: +[Valid][27][UTT][ MSE = 0.2420 | LCC = 0.8457 | SRCC = 0.8492 ] +INFO:root:Average loss=0.043971052226156764 +INFO:root: +[Valid][28][UTT][ MSE = 0.2352 | LCC = 0.8491 | SRCC = 0.8509 ] +INFO:root:Average loss=0.04267909248860983 +INFO:root: +[Valid][29][UTT][ MSE = 0.2394 | LCC = 0.8478 | SRCC = 0.8513 ] +INFO:root:Average loss=0.04252502756814162 +INFO:root: +[Valid][30][UTT][ MSE = 0.2413 | LCC = 0.8484 | SRCC = 0.8494 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2352 | LCC = 0.8611 | SRCC = 0.8612 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..b1a34d3d12955770138560614f58dbb486080392 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29700f0a9f88d305e2cf97e3006062a4d40599faf580a499caf17f29e97358d5 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..67114c2eac03d6d42b15434f543495fcbe6d0246 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4bbe91c625271b1e774579b85794ff1dcdf81ba84bebe0288fcd3263cc71f815 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..635a05df60933e510bbeda3c5b08e28490343f88 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.378720261156559 +INFO:root: +[Valid][0][UTT][ MSE = 0.3266 | LCC = 0.8261 | SRCC = 0.8270 ] +INFO:root:Average loss=0.24996941422040647 +INFO:root: +[Valid][1][UTT][ MSE = 0.3019 | LCC = 0.8477 | SRCC = 0.8465 ] +INFO:root:Average loss=0.228211294955168 +INFO:root: +[Valid][2][UTT][ MSE = 0.2705 | LCC = 0.8504 | SRCC = 0.8529 ] +INFO:root:Average loss=0.20478647698958716 +INFO:root: +[Valid][3][UTT][ MSE = 0.2728 | LCC = 0.8607 | SRCC = 0.8578 ] +INFO:root:Average loss=0.19582041247915 +INFO:root: +[Valid][4][UTT][ MSE = 0.2797 | LCC = 0.8596 | SRCC = 0.8576 ] +INFO:root:Average loss=0.17698439525870177 +INFO:root: +[Valid][5][UTT][ MSE = 0.2671 | LCC = 0.8621 | SRCC = 0.8602 ] +INFO:root:Average loss=0.16398125160963106 +INFO:root: +[Valid][6][UTT][ MSE = 0.2397 | LCC = 0.8613 | SRCC = 0.8602 ] +INFO:root:Average loss=0.16039399506571966 +INFO:root: +[Valid][7][UTT][ MSE = 0.2539 | LCC = 0.8593 | SRCC = 0.8574 ] +INFO:root:Average loss=0.1519507414255387 +INFO:root: +[Valid][8][UTT][ MSE = 0.2412 | LCC = 0.8604 | SRCC = 0.8603 ] +INFO:root:Average loss=0.1313275776994534 +INFO:root: +[Valid][9][UTT][ MSE = 0.2426 | LCC = 0.8573 | SRCC = 0.8569 ] +INFO:root:Average loss=0.12803760572121695 +INFO:root: +[Valid][10][UTT][ MSE = 0.2497 | LCC = 0.8540 | SRCC = 0.8561 ] +INFO:root:Average loss=0.11442299617024568 +INFO:root: +[Valid][11][UTT][ MSE = 0.2407 | LCC = 0.8557 | SRCC = 0.8554 ] +INFO:root:Average loss=0.10827693543755092 +INFO:root: +[Valid][12][UTT][ MSE = 0.2520 | LCC = 0.8540 | SRCC = 0.8545 ] +INFO:root:Average loss=0.10286672957814656 +INFO:root: +[Valid][13][UTT][ MSE = 0.2589 | LCC = 0.8516 | SRCC = 0.8533 ] +INFO:root:Average loss=0.09772942339380582 +INFO:root: +[Valid][14][UTT][ MSE = 0.2497 | LCC = 0.8504 | SRCC = 0.8511 ] +INFO:root:Average loss=0.09115855400569928 +INFO:root: +[Valid][15][UTT][ MSE = 0.2744 | LCC = 0.8509 | SRCC = 0.8517 ] +INFO:root:Average loss=0.08543974103835913 +INFO:root: +[Valid][16][UTT][ MSE = 0.2580 | LCC = 0.8491 | SRCC = 0.8506 ] +INFO:root:Average loss=0.07857877670381314 +INFO:root: +[Valid][17][UTT][ MSE = 0.2457 | LCC = 0.8542 | SRCC = 0.8549 ] +INFO:root:Average loss=0.07571838180033061 +INFO:root: +[Valid][18][UTT][ MSE = 0.2383 | LCC = 0.8510 | SRCC = 0.8523 ] +INFO:root:Average loss=0.07937544393233764 +INFO:root: +[Valid][19][UTT][ MSE = 0.2396 | LCC = 0.8524 | SRCC = 0.8533 ] +INFO:root:Average loss=0.06652220176198544 +INFO:root: +[Valid][20][UTT][ MSE = 0.2461 | LCC = 0.8530 | SRCC = 0.8541 ] +INFO:root:Average loss=0.06603705142744076 +INFO:root: +[Valid][21][UTT][ MSE = 0.2502 | LCC = 0.8516 | SRCC = 0.8525 ] +INFO:root:Average loss=0.0618230023731788 +INFO:root: +[Valid][22][UTT][ MSE = 0.2301 | LCC = 0.8563 | SRCC = 0.8571 ] +INFO:root:Average loss=0.05861382205516864 +INFO:root: +[Valid][23][UTT][ MSE = 0.2512 | LCC = 0.8508 | SRCC = 0.8517 ] +INFO:root:Average loss=0.05585944363608574 +INFO:root: +[Valid][24][UTT][ MSE = 0.2440 | LCC = 0.8522 | SRCC = 0.8546 ] +INFO:root:Average loss=0.054580213502049446 +INFO:root: +[Valid][25][UTT][ MSE = 0.2395 | LCC = 0.8539 | SRCC = 0.8558 ] +INFO:root:Average loss=0.054663908309661426 +INFO:root: +[Valid][26][UTT][ MSE = 0.2405 | LCC = 0.8525 | SRCC = 0.8550 ] +INFO:root:Average loss=0.05221030612786611 +INFO:root: +[Valid][27][UTT][ MSE = 0.2373 | LCC = 0.8555 | SRCC = 0.8572 ] +INFO:root:Average loss=0.04847100584839399 +INFO:root: +[Valid][28][UTT][ MSE = 0.2475 | LCC = 0.8497 | SRCC = 0.8512 ] +INFO:root:Average loss=0.04701485974379839 +INFO:root: +[Valid][29][UTT][ MSE = 0.2404 | LCC = 0.8498 | SRCC = 0.8506 ] +INFO:root:Average loss=0.04767208202527119 +INFO:root: +[Valid][30][UTT][ MSE = 0.2595 | LCC = 0.8487 | SRCC = 0.8492 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2646 | LCC = 0.8587 | SRCC = 0.8621 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..6a512b4bda5f08380d563eb8ec06ac5dd7bd3b0b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3be3ea81d39fb6e465451f2d40f5d316f2b408898dccbecaf7574f55104f53a5 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e710c333cdbd172639678388b79787dc54d69ea7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2265d43e546973c0f7c1f39adc6ee599e909c945e5fb0be2e76329eddba45a73 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..fcb38bee00e426df603d67a50469a14b68fa8a66 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3507161612311999 +INFO:root: +[Valid][0][UTT][ MSE = 0.2940 | LCC = 0.8288 | SRCC = 0.8328 ] +INFO:root:Average loss=0.25065553685029346 +INFO:root: +[Valid][1][UTT][ MSE = 0.2846 | LCC = 0.8525 | SRCC = 0.8508 ] +INFO:root:Average loss=0.22239575668787345 +INFO:root: +[Valid][2][UTT][ MSE = 0.2542 | LCC = 0.8580 | SRCC = 0.8583 ] +INFO:root:Average loss=0.20350113320044982 +INFO:root: +[Valid][3][UTT][ MSE = 0.3009 | LCC = 0.8596 | SRCC = 0.8601 ] +INFO:root:Average loss=0.19417511098659956 +INFO:root: +[Valid][4][UTT][ MSE = 0.2999 | LCC = 0.8598 | SRCC = 0.8610 ] +INFO:root:Average loss=0.1777204226415891 +INFO:root: +[Valid][5][UTT][ MSE = 0.2915 | LCC = 0.8629 | SRCC = 0.8641 ] +INFO:root:Average loss=0.1583129677634973 +INFO:root: +[Valid][6][UTT][ MSE = 0.2356 | LCC = 0.8623 | SRCC = 0.8638 ] +INFO:root:Average loss=0.14729099243115157 +INFO:root: +[Valid][7][UTT][ MSE = 0.2751 | LCC = 0.8574 | SRCC = 0.8581 ] +INFO:root:Average loss=0.14345991210295603 +INFO:root: +[Valid][8][UTT][ MSE = 0.2335 | LCC = 0.8594 | SRCC = 0.8608 ] +INFO:root:Average loss=0.12716778558798325 +INFO:root: +[Valid][9][UTT][ MSE = 0.2462 | LCC = 0.8625 | SRCC = 0.8626 ] +INFO:root:Average loss=0.11774037315104252 +INFO:root: +[Valid][10][UTT][ MSE = 0.2633 | LCC = 0.8607 | SRCC = 0.8605 ] +INFO:root:Average loss=0.11114028277687538 +INFO:root: +[Valid][11][UTT][ MSE = 0.2634 | LCC = 0.8543 | SRCC = 0.8560 ] +INFO:root:Average loss=0.10451319651343884 +INFO:root: +[Valid][12][UTT][ MSE = 0.2599 | LCC = 0.8537 | SRCC = 0.8558 ] +INFO:root:Average loss=0.09663593133863731 +INFO:root: +[Valid][13][UTT][ MSE = 0.2412 | LCC = 0.8553 | SRCC = 0.8572 ] +INFO:root:Average loss=0.09527754391997288 +INFO:root: +[Valid][14][UTT][ MSE = 0.2491 | LCC = 0.8534 | SRCC = 0.8547 ] +INFO:root:Average loss=0.0914269206711115 +INFO:root: +[Valid][15][UTT][ MSE = 0.2364 | LCC = 0.8535 | SRCC = 0.8541 ] +INFO:root:Average loss=0.0927127484136667 +INFO:root: +[Valid][16][UTT][ MSE = 0.2578 | LCC = 0.8555 | SRCC = 0.8576 ] +INFO:root:Average loss=0.08650923190781704 +INFO:root: +[Valid][17][UTT][ MSE = 0.2384 | LCC = 0.8536 | SRCC = 0.8549 ] +INFO:root:Average loss=0.07556040723545429 +INFO:root: +[Valid][18][UTT][ MSE = 0.2720 | LCC = 0.8515 | SRCC = 0.8535 ] +INFO:root:Average loss=0.07145752891516075 +INFO:root: +[Valid][19][UTT][ MSE = 0.2386 | LCC = 0.8531 | SRCC = 0.8535 ] +INFO:root:Average loss=0.06601839488706528 +INFO:root: +[Valid][20][UTT][ MSE = 0.2517 | LCC = 0.8504 | SRCC = 0.8520 ] +INFO:root:Average loss=0.06794279001844235 +INFO:root: +[Valid][21][UTT][ MSE = 0.2457 | LCC = 0.8480 | SRCC = 0.8480 ] +INFO:root:Average loss=0.06335268690226933 +INFO:root: +[Valid][22][UTT][ MSE = 0.2710 | LCC = 0.8481 | SRCC = 0.8497 ] +INFO:root:Average loss=0.06154606596399576 +INFO:root: +[Valid][23][UTT][ MSE = 0.2592 | LCC = 0.8465 | SRCC = 0.8477 ] +INFO:root:Average loss=0.05958865329814263 +INFO:root: +[Valid][24][UTT][ MSE = 0.2654 | LCC = 0.8530 | SRCC = 0.8536 ] +INFO:root:Average loss=0.05639162100851536 +INFO:root: +[Valid][25][UTT][ MSE = 0.2498 | LCC = 0.8490 | SRCC = 0.8491 ] +INFO:root:Average loss=0.050788035950599573 +INFO:root: +[Valid][26][UTT][ MSE = 0.2432 | LCC = 0.8523 | SRCC = 0.8520 ] +INFO:root:Average loss=0.04797415167857439 +INFO:root: +[Valid][27][UTT][ MSE = 0.2446 | LCC = 0.8513 | SRCC = 0.8509 ] +INFO:root:Average loss=0.05251868940794315 +INFO:root: +[Valid][28][UTT][ MSE = 0.2571 | LCC = 0.8482 | SRCC = 0.8486 ] +INFO:root:Average loss=0.0496730314902006 +INFO:root: +[Valid][29][UTT][ MSE = 0.2740 | LCC = 0.8482 | SRCC = 0.8482 ] +INFO:root:Average loss=0.04700042073352215 +INFO:root: +[Valid][30][UTT][ MSE = 0.2546 | LCC = 0.8483 | SRCC = 0.8488 ] +INFO:root: +[Test][31][UTT][ MSE = 0.3394 | LCC = 0.8532 | SRCC = 0.8616 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..355b8651d5a6e9db1665e33a4c26c05f996b936c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d55ea55c3b3f14435cbf9201b1d629b23f87202380d999daae6dd3a80b5323c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..cf99aaf42d003e7dd27eba111d518cf8a7d35b09 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fcddd9b8ff72881a2282b706920d8815f0f5ad50096d42abec6d1056c18fb14c +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..93d00f3969032b092391dfc65994108f75332ecb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.372464782343461 +INFO:root: +[Valid][0][UTT][ MSE = 0.3123 | LCC = 0.8355 | SRCC = 0.8345 ] +INFO:root:Average loss=0.24211036203763422 +INFO:root: +[Valid][1][UTT][ MSE = 0.3396 | LCC = 0.8483 | SRCC = 0.8494 ] +INFO:root:Average loss=0.21982192993164062 +INFO:root: +[Valid][2][UTT][ MSE = 0.2861 | LCC = 0.8591 | SRCC = 0.8568 ] +INFO:root:Average loss=0.20098297240642402 +INFO:root: +[Valid][3][UTT][ MSE = 0.2411 | LCC = 0.8639 | SRCC = 0.8636 ] +INFO:root:Average loss=0.1842764725860877 +INFO:root: +[Valid][4][UTT][ MSE = 0.2814 | LCC = 0.8630 | SRCC = 0.8618 ] +INFO:root:Average loss=0.17466745773951212 +INFO:root: +[Valid][5][UTT][ MSE = 0.2472 | LCC = 0.8691 | SRCC = 0.8687 ] +INFO:root:Average loss=0.1601611819022741 +INFO:root: +[Valid][6][UTT][ MSE = 0.2544 | LCC = 0.8690 | SRCC = 0.8688 ] +INFO:root:Average loss=0.15247779396864083 +INFO:root: +[Valid][7][UTT][ MSE = 0.2381 | LCC = 0.8671 | SRCC = 0.8680 ] +INFO:root:Average loss=0.1369944052436413 +INFO:root: +[Valid][8][UTT][ MSE = 0.2489 | LCC = 0.8602 | SRCC = 0.8641 ] +INFO:root:Average loss=0.13775745970316422 +INFO:root: +[Valid][9][UTT][ MSE = 0.2506 | LCC = 0.8636 | SRCC = 0.8639 ] +INFO:root:Average loss=0.12498104199767113 +INFO:root: +[Valid][10][UTT][ MSE = 0.2445 | LCC = 0.8574 | SRCC = 0.8592 ] +INFO:root:Average loss=0.11680494544979854 +INFO:root: +[Valid][11][UTT][ MSE = 0.2682 | LCC = 0.8581 | SRCC = 0.8620 ] +INFO:root:Average loss=0.10374500487859432 +INFO:root: +[Valid][12][UTT][ MSE = 0.2546 | LCC = 0.8573 | SRCC = 0.8601 ] +INFO:root:Average loss=0.09824156589232959 +INFO:root: +[Valid][13][UTT][ MSE = 0.2501 | LCC = 0.8570 | SRCC = 0.8594 ] +INFO:root:Average loss=0.09103322210602272 +INFO:root: +[Valid][14][UTT][ MSE = 0.2626 | LCC = 0.8516 | SRCC = 0.8548 ] +INFO:root:Average loss=0.08912647453447183 +INFO:root: +[Valid][15][UTT][ MSE = 0.2425 | LCC = 0.8569 | SRCC = 0.8585 ] +INFO:root:Average loss=0.08378784979383151 +INFO:root: +[Valid][16][UTT][ MSE = 0.2658 | LCC = 0.8538 | SRCC = 0.8557 ] +INFO:root:Average loss=0.07779893369819874 +INFO:root: +[Valid][17][UTT][ MSE = 0.2447 | LCC = 0.8524 | SRCC = 0.8557 ] +INFO:root:Average loss=0.07227410533680366 +INFO:root: +[Valid][18][UTT][ MSE = 0.2526 | LCC = 0.8529 | SRCC = 0.8548 ] +INFO:root:Average loss=0.0712522429246933 +INFO:root: +[Valid][19][UTT][ MSE = 0.2573 | LCC = 0.8521 | SRCC = 0.8554 ] +INFO:root:Average loss=0.06612720779883556 +INFO:root: +[Valid][20][UTT][ MSE = 0.2680 | LCC = 0.8532 | SRCC = 0.8546 ] +INFO:root:Average loss=0.06205460209494982 +INFO:root: +[Valid][21][UTT][ MSE = 0.2540 | LCC = 0.8476 | SRCC = 0.8497 ] +INFO:root:Average loss=0.0609915199665687 +INFO:root: +[Valid][22][UTT][ MSE = 0.2567 | LCC = 0.8472 | SRCC = 0.8520 ] +INFO:root:Average loss=0.05713487000992665 +INFO:root: +[Valid][23][UTT][ MSE = 0.2503 | LCC = 0.8457 | SRCC = 0.8464 ] +INFO:root:Average loss=0.06806574948132038 +INFO:root: +[Valid][24][UTT][ MSE = 0.2376 | LCC = 0.8530 | SRCC = 0.8537 ] +INFO:root:Average loss=0.05705983377993107 +INFO:root: +[Valid][25][UTT][ MSE = 0.2389 | LCC = 0.8528 | SRCC = 0.8536 ] +INFO:root:Average loss=0.05496639151794788 +INFO:root: +[Valid][26][UTT][ MSE = 0.2435 | LCC = 0.8518 | SRCC = 0.8541 ] +INFO:root:Average loss=0.05581863802403976 +INFO:root: +[Valid][27][UTT][ MSE = 0.2650 | LCC = 0.8445 | SRCC = 0.8470 ] +INFO:root:Average loss=0.0503495645064574 +INFO:root: +[Valid][28][UTT][ MSE = 0.2519 | LCC = 0.8482 | SRCC = 0.8495 ] +INFO:root:Average loss=0.04655972139862103 +INFO:root: +[Valid][29][UTT][ MSE = 0.2476 | LCC = 0.8482 | SRCC = 0.8498 ] +INFO:root:Average loss=0.04412623215466738 +INFO:root: +[Valid][30][UTT][ MSE = 0.2511 | LCC = 0.8460 | SRCC = 0.8465 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2668 | LCC = 0.8552 | SRCC = 0.8561 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..03251d74f6f6cb4a9a69bded1a450579e8997193 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb456879e6030c48df1523936e1141e5fc90f50538886990443bf458e3b1bc14 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..dfec936e7913227c8801421b96e5ee35b390bd9e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb0f79f36a7dec6fbca2b6bedd08839dac451171848c3acd7817a63696d076b0 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e23de0372f526a0bbd2cef4e8330dd8d7c48ed16 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3974172503520281 +INFO:root: +[Valid][0][UTT][ MSE = 0.3185 | LCC = 0.8255 | SRCC = 0.8334 ] +INFO:root:Average loss=0.25446235292997116 +INFO:root: +[Valid][1][UTT][ MSE = 0.3046 | LCC = 0.8491 | SRCC = 0.8497 ] +INFO:root:Average loss=0.22255847392938075 +INFO:root: +[Valid][2][UTT][ MSE = 0.2548 | LCC = 0.8611 | SRCC = 0.8611 ] +INFO:root:Average loss=0.20596562211330122 +INFO:root: +[Valid][3][UTT][ MSE = 0.2418 | LCC = 0.8635 | SRCC = 0.8631 ] +INFO:root:Average loss=0.1899465500162198 +INFO:root: +[Valid][4][UTT][ MSE = 0.2475 | LCC = 0.8615 | SRCC = 0.8608 ] +INFO:root:Average loss=0.1711982975785549 +INFO:root: +[Valid][5][UTT][ MSE = 0.2564 | LCC = 0.8622 | SRCC = 0.8618 ] +INFO:root:Average loss=0.15843762992284235 +INFO:root: +[Valid][6][UTT][ MSE = 0.2485 | LCC = 0.8608 | SRCC = 0.8612 ] +INFO:root:Average loss=0.14913654050383812 +INFO:root: +[Valid][7][UTT][ MSE = 0.2463 | LCC = 0.8605 | SRCC = 0.8618 ] +INFO:root:Average loss=0.1425474402614129 +INFO:root: +[Valid][8][UTT][ MSE = 0.2411 | LCC = 0.8579 | SRCC = 0.8581 ] +INFO:root:Average loss=0.12734107415263468 +INFO:root: +[Valid][9][UTT][ MSE = 0.2435 | LCC = 0.8561 | SRCC = 0.8565 ] +INFO:root:Average loss=0.11863793776585506 +INFO:root: +[Valid][10][UTT][ MSE = 0.2489 | LCC = 0.8568 | SRCC = 0.8585 ] +INFO:root:Average loss=0.11144940001078141 +INFO:root: +[Valid][11][UTT][ MSE = 0.2413 | LCC = 0.8595 | SRCC = 0.8588 ] +INFO:root:Average loss=0.10869134542269585 +INFO:root: +[Valid][12][UTT][ MSE = 0.2561 | LCC = 0.8571 | SRCC = 0.8577 ] +INFO:root:Average loss=0.10559073238609692 +INFO:root: +[Valid][13][UTT][ MSE = 0.2476 | LCC = 0.8526 | SRCC = 0.8532 ] +INFO:root:Average loss=0.09692289107120954 +INFO:root: +[Valid][14][UTT][ MSE = 0.2648 | LCC = 0.8494 | SRCC = 0.8500 ] +INFO:root:Average loss=0.09044719511308731 +INFO:root: +[Valid][15][UTT][ MSE = 0.2551 | LCC = 0.8500 | SRCC = 0.8512 ] +INFO:root:Average loss=0.08468268477381805 +INFO:root: +[Valid][16][UTT][ MSE = 0.2438 | LCC = 0.8480 | SRCC = 0.8497 ] +INFO:root:Average loss=0.0783796789458929 +INFO:root: +[Valid][17][UTT][ MSE = 0.2502 | LCC = 0.8469 | SRCC = 0.8479 ] +INFO:root:Average loss=0.07381653026319467 +INFO:root: +[Valid][18][UTT][ MSE = 0.2518 | LCC = 0.8478 | SRCC = 0.8482 ] +INFO:root:Average loss=0.07185546213235611 +INFO:root: +[Valid][19][UTT][ MSE = 0.2792 | LCC = 0.8436 | SRCC = 0.8441 ] +INFO:root:Average loss=0.06938657580086818 +INFO:root: +[Valid][20][UTT][ MSE = 0.2574 | LCC = 0.8479 | SRCC = 0.8478 ] +INFO:root:Average loss=0.06655950166093998 +INFO:root: +[Valid][21][UTT][ MSE = 0.2599 | LCC = 0.8467 | SRCC = 0.8481 ] +INFO:root:Average loss=0.05936948802226629 +INFO:root: +[Valid][22][UTT][ MSE = 0.2507 | LCC = 0.8472 | SRCC = 0.8469 ] +INFO:root:Average loss=0.05908569110891758 +INFO:root: +[Valid][23][UTT][ MSE = 0.2507 | LCC = 0.8448 | SRCC = 0.8452 ] +INFO:root:Average loss=0.058501418823233016 +INFO:root: +[Valid][24][UTT][ MSE = 0.2492 | LCC = 0.8466 | SRCC = 0.8469 ] +INFO:root:Average loss=0.05367949538123913 +INFO:root: +[Valid][25][UTT][ MSE = 0.2521 | LCC = 0.8449 | SRCC = 0.8459 ] +INFO:root:Average loss=0.053429921802419886 +INFO:root: +[Valid][26][UTT][ MSE = 0.2519 | LCC = 0.8429 | SRCC = 0.8432 ] +INFO:root:Average loss=0.05497577762565552 +INFO:root: +[Valid][27][UTT][ MSE = 0.2579 | LCC = 0.8455 | SRCC = 0.8462 ] +INFO:root:Average loss=0.050961461109228626 +INFO:root: +[Valid][28][UTT][ MSE = 0.2517 | LCC = 0.8438 | SRCC = 0.8445 ] +INFO:root:Average loss=0.05155435469574653 +INFO:root: +[Valid][29][UTT][ MSE = 0.2649 | LCC = 0.8389 | SRCC = 0.8397 ] +INFO:root:Average loss=0.04527646420189203 +INFO:root: +[Valid][30][UTT][ MSE = 0.2400 | LCC = 0.8457 | SRCC = 0.8471 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2553 | LCC = 0.8604 | SRCC = 0.8676 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..a45c63943083c258b1c47bbdcfc8f8939c26c16e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6b15c261435101ddaed80cd2fdcefceae3e9b8e0727be508821844bc1da6b09 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..945dd36432df30914c8afe2738d9c1fe6c2ec555 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2c2f1bbe8663eb9e1762d84b774f7ab9229020c845a1f83579fcc474d723c45 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer6/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4b0985cbbeeb4b6aa8d228255471de77a2e2c73a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3902181801505578 +INFO:root: +[Valid][0][UTT][ MSE = 0.3553 | LCC = 0.8273 | SRCC = 0.8326 ] +INFO:root:Average loss=0.2525960867985701 +INFO:root: +[Valid][1][UTT][ MSE = 0.3024 | LCC = 0.8518 | SRCC = 0.8502 ] +INFO:root:Average loss=0.2217088051331349 +INFO:root: +[Valid][2][UTT][ MSE = 0.3368 | LCC = 0.8594 | SRCC = 0.8588 ] +INFO:root:Average loss=0.20215964775819045 +INFO:root: +[Valid][3][UTT][ MSE = 0.2468 | LCC = 0.8633 | SRCC = 0.8620 ] +INFO:root:Average loss=0.19138672814155236 +INFO:root: +[Valid][4][UTT][ MSE = 0.3051 | LCC = 0.8634 | SRCC = 0.8632 ] +INFO:root:Average loss=0.17653619450254318 +INFO:root: +[Valid][5][UTT][ MSE = 0.2991 | LCC = 0.8632 | SRCC = 0.8628 ] +INFO:root:Average loss=0.16108658499060532 +INFO:root: +[Valid][6][UTT][ MSE = 0.2958 | LCC = 0.8589 | SRCC = 0.8617 ] +INFO:root:Average loss=0.1559470874758867 +INFO:root: +[Valid][7][UTT][ MSE = 0.2622 | LCC = 0.8571 | SRCC = 0.8579 ] +INFO:root:Average loss=0.13905299684176078 +INFO:root: +[Valid][8][UTT][ MSE = 0.2493 | LCC = 0.8612 | SRCC = 0.8614 ] +INFO:root:Average loss=0.13135979544275847 +INFO:root: +[Valid][9][UTT][ MSE = 0.3616 | LCC = 0.8544 | SRCC = 0.8566 ] +INFO:root:Average loss=0.12678713504320535 +INFO:root: +[Valid][10][UTT][ MSE = 0.3081 | LCC = 0.8526 | SRCC = 0.8544 ] +INFO:root:Average loss=0.11175016313791275 +INFO:root: +[Valid][11][UTT][ MSE = 0.2583 | LCC = 0.8556 | SRCC = 0.8577 ] +INFO:root:Average loss=0.10902668478397223 +INFO:root: +[Valid][12][UTT][ MSE = 0.2689 | LCC = 0.8548 | SRCC = 0.8565 ] +INFO:root:Average loss=0.09689251109002492 +INFO:root: +[Valid][13][UTT][ MSE = 0.2977 | LCC = 0.8490 | SRCC = 0.8507 ] +INFO:root:Average loss=0.09768260724078386 +INFO:root: +[Valid][14][UTT][ MSE = 0.2676 | LCC = 0.8563 | SRCC = 0.8590 ] +INFO:root:Average loss=0.09266159143776466 +INFO:root: +[Valid][15][UTT][ MSE = 0.2423 | LCC = 0.8574 | SRCC = 0.8591 ] +INFO:root:Average loss=0.08121539695331684 +INFO:root: +[Valid][16][UTT][ MSE = 0.2504 | LCC = 0.8511 | SRCC = 0.8535 ] +INFO:root:Average loss=0.07761010088217564 +INFO:root: +[Valid][17][UTT][ MSE = 0.2413 | LCC = 0.8497 | SRCC = 0.8521 ] +INFO:root:Average loss=0.07700301544406475 +INFO:root: +[Valid][18][UTT][ MSE = 0.2762 | LCC = 0.8544 | SRCC = 0.8557 ] +INFO:root:Average loss=0.07277327169401523 +INFO:root: +[Valid][19][UTT][ MSE = 0.2546 | LCC = 0.8499 | SRCC = 0.8511 ] +INFO:root:Average loss=0.06655246062347522 +INFO:root: +[Valid][20][UTT][ MSE = 0.2902 | LCC = 0.8461 | SRCC = 0.8482 ] +INFO:root:Average loss=0.0627551161421415 +INFO:root: +[Valid][21][UTT][ MSE = 0.2829 | LCC = 0.8483 | SRCC = 0.8485 ] +INFO:root:Average loss=0.061603728586282484 +INFO:root: +[Valid][22][UTT][ MSE = 0.2746 | LCC = 0.8487 | SRCC = 0.8504 ] +INFO:root:Average loss=0.05611191790264387 +INFO:root: +[Valid][23][UTT][ MSE = 0.2753 | LCC = 0.8502 | SRCC = 0.8520 ] +INFO:root:Average loss=0.05439722308745751 +INFO:root: +[Valid][24][UTT][ MSE = 0.2632 | LCC = 0.8423 | SRCC = 0.8457 ] +INFO:root:Average loss=0.054214923905256465 +INFO:root: +[Valid][25][UTT][ MSE = 0.2680 | LCC = 0.8421 | SRCC = 0.8436 ] +INFO:root:Average loss=0.050335786424768277 +INFO:root: +[Valid][26][UTT][ MSE = 0.2508 | LCC = 0.8484 | SRCC = 0.8494 ] +INFO:root:Average loss=0.0477841412409758 +INFO:root: +[Valid][27][UTT][ MSE = 0.2885 | LCC = 0.8437 | SRCC = 0.8464 ] +INFO:root:Average loss=0.04718658461784705 +INFO:root: +[Valid][28][UTT][ MSE = 0.2895 | LCC = 0.8475 | SRCC = 0.8482 ] +INFO:root:Average loss=0.047267679745952286 +INFO:root: +[Valid][29][UTT][ MSE = 0.2674 | LCC = 0.8434 | SRCC = 0.8446 ] +INFO:root:Average loss=0.043926538565219976 +INFO:root: +[Valid][30][UTT][ MSE = 0.2427 | LCC = 0.8498 | SRCC = 0.8515 ] +INFO:root: +[Test][31][UTT][ MSE = 0.3134 | LCC = 0.8606 | SRCC = 0.8626 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..3c92add812fb0eafd8ac835c51261b6f9432b758 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01674facd34a8f091d2c7d2840f1a442a4bd7a5e90b63717804e885a175f7fa7 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..f5232306a887b3b6da198ac6c6c67226d940f219 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a3649a0fd7219a0ff77ecf4d77abd494536ca7f947a5d1fd9abecb6bcf6291d +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..ae6aab271468cb63c0bd183903ece437d73f5f26 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.39960386871527404 +INFO:root: +[Valid][0][UTT][ MSE = 0.3577 | LCC = 0.8278 | SRCC = 0.8284 ] +INFO:root:Average loss=0.2477858943434862 +INFO:root: +[Valid][1][UTT][ MSE = 0.3316 | LCC = 0.8549 | SRCC = 0.8519 ] +INFO:root:Average loss=0.23362882186969122 +INFO:root: +[Valid][2][UTT][ MSE = 0.2940 | LCC = 0.8591 | SRCC = 0.8574 ] +INFO:root:Average loss=0.2109610847173593 +INFO:root: +[Valid][3][UTT][ MSE = 0.2716 | LCC = 0.8613 | SRCC = 0.8601 ] +INFO:root:Average loss=0.19807169681940323 +INFO:root: +[Valid][4][UTT][ MSE = 0.2549 | LCC = 0.8625 | SRCC = 0.8606 ] +INFO:root:Average loss=0.18113941317185378 +INFO:root: +[Valid][5][UTT][ MSE = 0.2847 | LCC = 0.8617 | SRCC = 0.8606 ] +INFO:root:Average loss=0.17357614846565786 +INFO:root: +[Valid][6][UTT][ MSE = 0.2766 | LCC = 0.8650 | SRCC = 0.8628 ] +INFO:root:Average loss=0.15981253503988951 +INFO:root: +[Valid][7][UTT][ MSE = 0.2677 | LCC = 0.8634 | SRCC = 0.8624 ] +INFO:root:Average loss=0.14575828124697393 +INFO:root: +[Valid][8][UTT][ MSE = 0.2861 | LCC = 0.8645 | SRCC = 0.8622 ] +INFO:root:Average loss=0.13708572318920723 +INFO:root: +[Valid][9][UTT][ MSE = 0.2990 | LCC = 0.8604 | SRCC = 0.8587 ] +INFO:root:Average loss=0.1293827310586587 +INFO:root: +[Valid][10][UTT][ MSE = 0.3014 | LCC = 0.8560 | SRCC = 0.8544 ] +INFO:root:Average loss=0.12601815479306075 +INFO:root: +[Valid][11][UTT][ MSE = 0.2882 | LCC = 0.8510 | SRCC = 0.8511 ] +INFO:root:Average loss=0.11856841783110912 +INFO:root: +[Valid][12][UTT][ MSE = 0.2777 | LCC = 0.8565 | SRCC = 0.8555 ] +INFO:root:Average loss=0.10499270203021857 +INFO:root: +[Valid][13][UTT][ MSE = 0.2558 | LCC = 0.8523 | SRCC = 0.8520 ] +INFO:root:Average loss=0.09926985428692439 +INFO:root: +[Valid][14][UTT][ MSE = 0.2898 | LCC = 0.8510 | SRCC = 0.8507 ] +INFO:root:Average loss=0.09514975098845287 +INFO:root: +[Valid][15][UTT][ MSE = 0.2809 | LCC = 0.8516 | SRCC = 0.8499 ] +INFO:root:Average loss=0.09005638675238842 +INFO:root: +[Valid][16][UTT][ MSE = 0.2724 | LCC = 0.8501 | SRCC = 0.8500 ] +INFO:root:Average loss=0.08379584135344395 +INFO:root: +[Valid][17][UTT][ MSE = 0.2559 | LCC = 0.8500 | SRCC = 0.8500 ] +INFO:root:Average loss=0.08558906275683488 +INFO:root: +[Valid][18][UTT][ MSE = 0.2743 | LCC = 0.8492 | SRCC = 0.8486 ] +INFO:root:Average loss=0.07543446438816878 +INFO:root: +[Valid][19][UTT][ MSE = 0.2716 | LCC = 0.8483 | SRCC = 0.8475 ] +INFO:root:Average loss=0.07403995655477047 +INFO:root: +[Valid][20][UTT][ MSE = 0.2877 | LCC = 0.8495 | SRCC = 0.8497 ] +INFO:root:Average loss=0.07137104496359825 +INFO:root: +[Valid][21][UTT][ MSE = 0.3180 | LCC = 0.8504 | SRCC = 0.8509 ] +INFO:root:Average loss=0.07023470213588996 +INFO:root: +[Valid][22][UTT][ MSE = 0.2843 | LCC = 0.8471 | SRCC = 0.8463 ] +INFO:root:Average loss=0.06386024041626698 +INFO:root: +[Valid][23][UTT][ MSE = 0.2729 | LCC = 0.8491 | SRCC = 0.8489 ] +INFO:root:Average loss=0.06062525697052479 +INFO:root: +[Valid][24][UTT][ MSE = 0.2822 | LCC = 0.8475 | SRCC = 0.8472 ] +INFO:root:Average loss=0.059425110140672095 +INFO:root: +[Valid][25][UTT][ MSE = 0.2581 | LCC = 0.8483 | SRCC = 0.8484 ] +INFO:root:Average loss=0.059447238699365884 +INFO:root: +[Valid][26][UTT][ MSE = 0.2580 | LCC = 0.8446 | SRCC = 0.8449 ] +INFO:root:Average loss=0.05396419223875572 +INFO:root: +[Valid][27][UTT][ MSE = 0.3106 | LCC = 0.8457 | SRCC = 0.8453 ] +INFO:root:Average loss=0.05302743876400667 +INFO:root: +[Valid][28][UTT][ MSE = 0.2978 | LCC = 0.8483 | SRCC = 0.8478 ] +INFO:root:Average loss=0.05138668895531923 +INFO:root: +[Valid][29][UTT][ MSE = 0.2647 | LCC = 0.8451 | SRCC = 0.8453 ] +INFO:root:Average loss=0.04799689216395983 +INFO:root: +[Valid][30][UTT][ MSE = 0.2707 | LCC = 0.8524 | SRCC = 0.8527 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2597 | LCC = 0.8622 | SRCC = 0.8624 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e8241bb1b723c96ed4eac7b1eeabb5bc59549f39 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53ad0920894516aba95404e11bca159f58a586a963d903b3a1bed5293dd31d21 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..8af238e38c048c4345213cbb2ebe7c745dd969cd --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:737721e26deb124b925a8aa02979f088d7e11cfb4f734d8bce20add444dbec82 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..75a2dd10d0f1b160528a952ed8987a51c531eaa9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3752569678502205 +INFO:root: +[Valid][0][UTT][ MSE = 0.3299 | LCC = 0.8334 | SRCC = 0.8342 ] +INFO:root:Average loss=0.24599897517607763 +INFO:root: +[Valid][1][UTT][ MSE = 0.2915 | LCC = 0.8577 | SRCC = 0.8558 ] +INFO:root:Average loss=0.22173113901263627 +INFO:root: +[Valid][2][UTT][ MSE = 0.2689 | LCC = 0.8639 | SRCC = 0.8626 ] +INFO:root:Average loss=0.20271934979619124 +INFO:root: +[Valid][3][UTT][ MSE = 0.2639 | LCC = 0.8688 | SRCC = 0.8664 ] +INFO:root:Average loss=0.18722012505317345 +INFO:root: +[Valid][4][UTT][ MSE = 0.2462 | LCC = 0.8653 | SRCC = 0.8640 ] +INFO:root:Average loss=0.17474895553329053 +INFO:root: +[Valid][5][UTT][ MSE = 0.2681 | LCC = 0.8668 | SRCC = 0.8654 ] +INFO:root:Average loss=0.16309035798678032 +INFO:root: +[Valid][6][UTT][ MSE = 0.2806 | LCC = 0.8670 | SRCC = 0.8673 ] +INFO:root:Average loss=0.15091286628292158 +INFO:root: +[Valid][7][UTT][ MSE = 0.2452 | LCC = 0.8643 | SRCC = 0.8649 ] +INFO:root:Average loss=0.14275324220458666 +INFO:root: +[Valid][8][UTT][ MSE = 0.2485 | LCC = 0.8623 | SRCC = 0.8606 ] +INFO:root:Average loss=0.1447782085683101 +INFO:root: +[Valid][9][UTT][ MSE = 0.2312 | LCC = 0.8620 | SRCC = 0.8627 ] +INFO:root:Average loss=0.12330724852971542 +INFO:root: +[Valid][10][UTT][ MSE = 0.2731 | LCC = 0.8573 | SRCC = 0.8564 ] +INFO:root:Average loss=0.12501454410644677 +INFO:root: +[Valid][11][UTT][ MSE = 0.2654 | LCC = 0.8536 | SRCC = 0.8577 ] +INFO:root:Average loss=0.11001602073128407 +INFO:root: +[Valid][12][UTT][ MSE = 0.2504 | LCC = 0.8576 | SRCC = 0.8576 ] +INFO:root:Average loss=0.1009626400967439 +INFO:root: +[Valid][13][UTT][ MSE = 0.2650 | LCC = 0.8510 | SRCC = 0.8512 ] +INFO:root:Average loss=0.09427091254828832 +INFO:root: +[Valid][14][UTT][ MSE = 0.2622 | LCC = 0.8470 | SRCC = 0.8485 ] +INFO:root:Average loss=0.09060916376228516 +INFO:root: +[Valid][15][UTT][ MSE = 0.2467 | LCC = 0.8588 | SRCC = 0.8598 ] +INFO:root:Average loss=0.08621610390643279 +INFO:root: +[Valid][16][UTT][ MSE = 0.2559 | LCC = 0.8521 | SRCC = 0.8521 ] +INFO:root:Average loss=0.07929890449994649 +INFO:root: +[Valid][17][UTT][ MSE = 0.2748 | LCC = 0.8491 | SRCC = 0.8500 ] +INFO:root:Average loss=0.0763796758479797 +INFO:root: +[Valid][18][UTT][ MSE = 0.2600 | LCC = 0.8535 | SRCC = 0.8531 ] +INFO:root:Average loss=0.07292073692839879 +INFO:root: +[Valid][19][UTT][ MSE = 0.2448 | LCC = 0.8573 | SRCC = 0.8577 ] +INFO:root:Average loss=0.07124764744478923 +INFO:root: +[Valid][20][UTT][ MSE = 0.2604 | LCC = 0.8545 | SRCC = 0.8546 ] +INFO:root:Average loss=0.07029503311675328 +INFO:root: +[Valid][21][UTT][ MSE = 0.2535 | LCC = 0.8502 | SRCC = 0.8490 ] +INFO:root:Average loss=0.06901883496305881 +INFO:root: +[Valid][22][UTT][ MSE = 0.2387 | LCC = 0.8567 | SRCC = 0.8557 ] +INFO:root:Average loss=0.06170855815976094 +INFO:root: +[Valid][23][UTT][ MSE = 0.2463 | LCC = 0.8559 | SRCC = 0.8542 ] +INFO:root:Average loss=0.06080059711940777 +INFO:root: +[Valid][24][UTT][ MSE = 0.2658 | LCC = 0.8514 | SRCC = 0.8505 ] +INFO:root:Average loss=0.057137058761257395 +INFO:root: +[Valid][25][UTT][ MSE = 0.2472 | LCC = 0.8497 | SRCC = 0.8501 ] +INFO:root:Average loss=0.05293954875415716 +INFO:root: +[Valid][26][UTT][ MSE = 0.2444 | LCC = 0.8504 | SRCC = 0.8498 ] +INFO:root:Average loss=0.05541191087701382 +INFO:root: +[Valid][27][UTT][ MSE = 0.2604 | LCC = 0.8531 | SRCC = 0.8526 ] +INFO:root:Average loss=0.05185968404970108 +INFO:root: +[Valid][28][UTT][ MSE = 0.2401 | LCC = 0.8518 | SRCC = 0.8518 ] +INFO:root:Average loss=0.049416145094885275 +INFO:root: +[Valid][29][UTT][ MSE = 0.2626 | LCC = 0.8466 | SRCC = 0.8465 ] +INFO:root:Average loss=0.04761177136634405 +INFO:root: +[Valid][30][UTT][ MSE = 0.2561 | LCC = 0.8486 | SRCC = 0.8490 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2663 | LCC = 0.8561 | SRCC = 0.8578 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..3c77f7bc25032c0703aca7a5d61dcb4a3073d4e9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:526f29c7d77459ff95b381f5d9fd5d40b395b5386e49994cce05d3d13837e20a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c4b495e3f52239db94cda2716347d335cc61b4a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e141747636210f39c44917e9ed8066d9d43e31ca40f22ec8eae0f65cb47fbd1 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..98bea03214e66ee4db7d0ccf8e4c8b15ea7587ac --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.49001851696998644 +INFO:root: +[Valid][0][UTT][ MSE = 0.3500 | LCC = 0.8134 | SRCC = 0.8206 ] +INFO:root:Average loss=0.26011820137500763 +INFO:root: +[Valid][1][UTT][ MSE = 0.3094 | LCC = 0.8479 | SRCC = 0.8476 ] +INFO:root:Average loss=0.22876064880536154 +INFO:root: +[Valid][2][UTT][ MSE = 0.2607 | LCC = 0.8568 | SRCC = 0.8579 ] +INFO:root:Average loss=0.2025522985137426 +INFO:root: +[Valid][3][UTT][ MSE = 0.2493 | LCC = 0.8616 | SRCC = 0.8613 ] +INFO:root:Average loss=0.19637662592606667 +INFO:root: +[Valid][4][UTT][ MSE = 0.2838 | LCC = 0.8606 | SRCC = 0.8625 ] +INFO:root:Average loss=0.18181149212595743 +INFO:root: +[Valid][5][UTT][ MSE = 0.2438 | LCC = 0.8634 | SRCC = 0.8639 ] +INFO:root:Average loss=0.1676370851122416 +INFO:root: +[Valid][6][UTT][ MSE = 0.2437 | LCC = 0.8640 | SRCC = 0.8641 ] +INFO:root:Average loss=0.15469872368833956 +INFO:root: +[Valid][7][UTT][ MSE = 0.2312 | LCC = 0.8659 | SRCC = 0.8664 ] +INFO:root:Average loss=0.15373220848731506 +INFO:root: +[Valid][8][UTT][ MSE = 0.2323 | LCC = 0.8672 | SRCC = 0.8677 ] +INFO:root:Average loss=0.1360960265573783 +INFO:root: +[Valid][9][UTT][ MSE = 0.2354 | LCC = 0.8628 | SRCC = 0.8628 ] +INFO:root:Average loss=0.13262648660785112 +INFO:root: +[Valid][10][UTT][ MSE = 0.2519 | LCC = 0.8634 | SRCC = 0.8627 ] +INFO:root:Average loss=0.12090425040477361 +INFO:root: +[Valid][11][UTT][ MSE = 0.2337 | LCC = 0.8621 | SRCC = 0.8615 ] +INFO:root:Average loss=0.10914526917995551 +INFO:root: +[Valid][12][UTT][ MSE = 0.2542 | LCC = 0.8576 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1034282960761816 +INFO:root: +[Valid][13][UTT][ MSE = 0.2384 | LCC = 0.8575 | SRCC = 0.8565 ] +INFO:root:Average loss=0.09801353265841801 +INFO:root: +[Valid][14][UTT][ MSE = 0.2512 | LCC = 0.8611 | SRCC = 0.8597 ] +INFO:root:Average loss=0.09147002586187461 +INFO:root: +[Valid][15][UTT][ MSE = 0.2210 | LCC = 0.8597 | SRCC = 0.8605 ] +INFO:root:Average loss=0.08814752569947487 +INFO:root: +[Valid][16][UTT][ MSE = 0.2433 | LCC = 0.8541 | SRCC = 0.8544 ] +INFO:root:Average loss=0.08457906620624737 +INFO:root: +[Valid][17][UTT][ MSE = 0.2455 | LCC = 0.8501 | SRCC = 0.8502 ] +INFO:root:Average loss=0.07892483191039318 +INFO:root: +[Valid][18][UTT][ MSE = 0.2592 | LCC = 0.8504 | SRCC = 0.8503 ] +INFO:root:Average loss=0.07531784828274678 +INFO:root: +[Valid][19][UTT][ MSE = 0.2456 | LCC = 0.8498 | SRCC = 0.8484 ] +INFO:root:Average loss=0.07597047119186474 +INFO:root: +[Valid][20][UTT][ MSE = 0.2443 | LCC = 0.8492 | SRCC = 0.8497 ] +INFO:root:Average loss=0.06510175315615459 +INFO:root: +[Valid][21][UTT][ MSE = 0.2440 | LCC = 0.8510 | SRCC = 0.8502 ] +INFO:root:Average loss=0.06398671903671363 +INFO:root: +[Valid][22][UTT][ MSE = 0.2490 | LCC = 0.8467 | SRCC = 0.8474 ] +INFO:root:Average loss=0.06594825464372452 +INFO:root: +[Valid][23][UTT][ MSE = 0.2463 | LCC = 0.8508 | SRCC = 0.8518 ] +INFO:root:Average loss=0.05493569345428394 +INFO:root: +[Valid][24][UTT][ MSE = 0.2515 | LCC = 0.8467 | SRCC = 0.8472 ] +INFO:root:Average loss=0.055665518515385114 +INFO:root: +[Valid][25][UTT][ MSE = 0.2575 | LCC = 0.8463 | SRCC = 0.8474 ] +INFO:root:Average loss=0.054137762516545944 +INFO:root: +[Valid][26][UTT][ MSE = 0.2488 | LCC = 0.8453 | SRCC = 0.8467 ] +INFO:root:Average loss=0.05135851792800121 +INFO:root: +[Valid][27][UTT][ MSE = 0.2582 | LCC = 0.8442 | SRCC = 0.8445 ] +INFO:root:Average loss=0.04896121314511849 +INFO:root: +[Valid][28][UTT][ MSE = 0.2492 | LCC = 0.8467 | SRCC = 0.8467 ] +INFO:root:Average loss=0.04554560360236046 +INFO:root: +[Valid][29][UTT][ MSE = 0.2447 | LCC = 0.8490 | SRCC = 0.8507 ] +INFO:root:Average loss=0.048725919654736154 +INFO:root: +[Valid][30][UTT][ MSE = 0.2445 | LCC = 0.8454 | SRCC = 0.8475 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2392 | LCC = 0.8599 | SRCC = 0.8620 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e9868115101b402af7f41663d9fcecb84ea4b0d4 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:236fc2e8677d34061786531ea44e605e951833a996c57ea4db2bfe774d0a313c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..26860452cb5d5d22d6851d93b6fb374d9eb93f9b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6826403cad9cf6f6336766ebe1e6b09662115cb4741c4edcd51c4e28848601ad +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..8399a99e851a3f351ffbbee45c260309294d9db9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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=0.3646835579704016 +INFO:root: +[Valid][0][UTT][ MSE = 0.3078 | LCC = 0.8324 | SRCC = 0.8364 ] +INFO:root:Average loss=0.2480829982803418 +INFO:root: +[Valid][1][UTT][ MSE = 0.2810 | LCC = 0.8515 | SRCC = 0.8535 ] +INFO:root:Average loss=0.22146330487269622 +INFO:root: +[Valid][2][UTT][ MSE = 0.2783 | LCC = 0.8575 | SRCC = 0.8610 ] +INFO:root:Average loss=0.20117930189157143 +INFO:root: +[Valid][3][UTT][ MSE = 0.2559 | LCC = 0.8643 | SRCC = 0.8628 ] +INFO:root:Average loss=0.18754297533096412 +INFO:root: +[Valid][4][UTT][ MSE = 0.2540 | LCC = 0.8680 | SRCC = 0.8661 ] +INFO:root:Average loss=0.17753720971254203 +INFO:root: +[Valid][5][UTT][ MSE = 0.2389 | LCC = 0.8652 | SRCC = 0.8644 ] +INFO:root:Average loss=0.16447620313519087 +INFO:root: +[Valid][6][UTT][ MSE = 0.2414 | LCC = 0.8656 | SRCC = 0.8643 ] +INFO:root:Average loss=0.15088252274271768 +INFO:root: +[Valid][7][UTT][ MSE = 0.2477 | LCC = 0.8653 | SRCC = 0.8653 ] +INFO:root:Average loss=0.14132687468559313 +INFO:root: +[Valid][8][UTT][ MSE = 0.2520 | LCC = 0.8649 | SRCC = 0.8633 ] +INFO:root:Average loss=0.14039636756747198 +INFO:root: +[Valid][9][UTT][ MSE = 0.2265 | LCC = 0.8646 | SRCC = 0.8657 ] +INFO:root:Average loss=0.12504318308753845 +INFO:root: +[Valid][10][UTT][ MSE = 0.2476 | LCC = 0.8565 | SRCC = 0.8565 ] +INFO:root:Average loss=0.11885184317063062 +INFO:root: +[Valid][11][UTT][ MSE = 0.2240 | LCC = 0.8641 | SRCC = 0.8649 ] +INFO:root:Average loss=0.110701967890446 +INFO:root: +[Valid][12][UTT][ MSE = 0.2298 | LCC = 0.8614 | SRCC = 0.8624 ] +INFO:root:Average loss=0.10282754597182457 +INFO:root: +[Valid][13][UTT][ MSE = 0.2678 | LCC = 0.8599 | SRCC = 0.8610 ] +INFO:root:Average loss=0.09790028717655402 +INFO:root: +[Valid][14][UTT][ MSE = 0.2462 | LCC = 0.8589 | SRCC = 0.8586 ] +INFO:root:Average loss=0.09150363433246429 +INFO:root: +[Valid][15][UTT][ MSE = 0.2268 | LCC = 0.8596 | SRCC = 0.8602 ] +INFO:root:Average loss=0.08854124155373146 +INFO:root: +[Valid][16][UTT][ MSE = 0.2426 | LCC = 0.8578 | SRCC = 0.8575 ] +INFO:root:Average loss=0.08293748952639408 +INFO:root: +[Valid][17][UTT][ MSE = 0.2365 | LCC = 0.8558 | SRCC = 0.8570 ] +INFO:root:Average loss=0.07781148467881557 +INFO:root: +[Valid][18][UTT][ MSE = 0.2449 | LCC = 0.8550 | SRCC = 0.8574 ] +INFO:root:Average loss=0.07511231136054565 +INFO:root: +[Valid][19][UTT][ MSE = 0.2369 | LCC = 0.8550 | SRCC = 0.8555 ] +INFO:root:Average loss=0.07114294123573181 +INFO:root: +[Valid][20][UTT][ MSE = 0.2423 | LCC = 0.8536 | SRCC = 0.8536 ] +INFO:root:Average loss=0.06904426580056167 +INFO:root: +[Valid][21][UTT][ MSE = 0.2431 | LCC = 0.8527 | SRCC = 0.8531 ] +INFO:root:Average loss=0.06501126375335914 +INFO:root: +[Valid][22][UTT][ MSE = 0.2510 | LCC = 0.8493 | SRCC = 0.8503 ] +INFO:root:Average loss=0.06387794103759986 +INFO:root: +[Valid][23][UTT][ MSE = 0.2566 | LCC = 0.8458 | SRCC = 0.8459 ] +INFO:root:Average loss=0.059514359260598816 +INFO:root: +[Valid][24][UTT][ MSE = 0.2357 | LCC = 0.8547 | SRCC = 0.8557 ] +INFO:root:Average loss=0.05899956664786889 +INFO:root: +[Valid][25][UTT][ MSE = 0.2334 | LCC = 0.8569 | SRCC = 0.8576 ] +INFO:root:Average loss=0.05503422494691152 +INFO:root: +[Valid][26][UTT][ MSE = 0.2373 | LCC = 0.8542 | SRCC = 0.8537 ] +INFO:root:Average loss=0.04957995458673208 +INFO:root: +[Valid][27][UTT][ MSE = 0.2352 | LCC = 0.8523 | SRCC = 0.8520 ] +INFO:root:Average loss=0.050884801106384166 +INFO:root: +[Valid][28][UTT][ MSE = 0.2356 | LCC = 0.8519 | SRCC = 0.8528 ] +INFO:root:Average loss=0.050861418581543826 +INFO:root: +[Valid][29][UTT][ MSE = 0.2332 | LCC = 0.8537 | SRCC = 0.8522 ] +INFO:root:Average loss=0.047371361715098224 +INFO:root: +[Valid][30][UTT][ MSE = 0.2413 | LCC = 0.8538 | SRCC = 0.8532 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2607 | LCC = 0.8639 | SRCC = 0.8649 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..aee96e034f77dee925bf318f8dafd6b40149a87c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0fc1da87d84d02a20f44066e4e9564234b05e0e4fc1d52baebc0eca65f51db6a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..ade5e83d8ecf5ac01f2bdd7abc30805960a72641 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f66f311c96bf2724c0b349f9261f812a09c3b3403986425cae6ed372f23e723 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9229fb15397ef1aa155fc735cb2c4b94a9939bc8 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer7/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=0.3432107560145549 +INFO:root: +[Valid][0][UTT][ MSE = 0.3394 | LCC = 0.8291 | SRCC = 0.8364 ] +INFO:root:Average loss=0.24676508666613164 +INFO:root: +[Valid][1][UTT][ MSE = 0.2804 | LCC = 0.8541 | SRCC = 0.8527 ] +INFO:root:Average loss=0.21936350067456564 +INFO:root: +[Valid][2][UTT][ MSE = 0.2599 | LCC = 0.8613 | SRCC = 0.8593 ] +INFO:root:Average loss=0.20506425889638755 +INFO:root: +[Valid][3][UTT][ MSE = 0.2318 | LCC = 0.8616 | SRCC = 0.8635 ] +INFO:root:Average loss=0.19808018312622339 +INFO:root: +[Valid][4][UTT][ MSE = 0.2590 | LCC = 0.8673 | SRCC = 0.8659 ] +INFO:root:Average loss=0.17864308343865934 +INFO:root: +[Valid][5][UTT][ MSE = 0.2393 | LCC = 0.8678 | SRCC = 0.8668 ] +INFO:root:Average loss=0.16424568847585946 +INFO:root: +[Valid][6][UTT][ MSE = 0.2351 | LCC = 0.8684 | SRCC = 0.8677 ] +INFO:root:Average loss=0.15404899236865532 +INFO:root: +[Valid][7][UTT][ MSE = 0.2570 | LCC = 0.8682 | SRCC = 0.8667 ] +INFO:root:Average loss=0.1423460671152824 +INFO:root: +[Valid][8][UTT][ MSE = 0.2364 | LCC = 0.8635 | SRCC = 0.8649 ] +INFO:root:Average loss=0.13478828995273665 +INFO:root: +[Valid][9][UTT][ MSE = 0.2307 | LCC = 0.8657 | SRCC = 0.8643 ] +INFO:root:Average loss=0.12877172671067408 +INFO:root: +[Valid][10][UTT][ MSE = 0.2357 | LCC = 0.8643 | SRCC = 0.8637 ] +INFO:root:Average loss=0.12300507991741864 +INFO:root: +[Valid][11][UTT][ MSE = 0.2232 | LCC = 0.8638 | SRCC = 0.8651 ] +INFO:root:Average loss=0.11085751881966224 +INFO:root: +[Valid][12][UTT][ MSE = 0.2522 | LCC = 0.8571 | SRCC = 0.8600 ] +INFO:root:Average loss=0.10285454229093514 +INFO:root: +[Valid][13][UTT][ MSE = 0.2665 | LCC = 0.8564 | SRCC = 0.8567 ] +INFO:root:Average loss=0.09646791907457206 +INFO:root: +[Valid][14][UTT][ MSE = 0.2338 | LCC = 0.8529 | SRCC = 0.8534 ] +INFO:root:Average loss=0.09337378083131252 +INFO:root: +[Valid][15][UTT][ MSE = 0.2476 | LCC = 0.8539 | SRCC = 0.8532 ] +INFO:root:Average loss=0.08426550909494743 +INFO:root: +[Valid][16][UTT][ MSE = 0.2410 | LCC = 0.8553 | SRCC = 0.8565 ] +INFO:root:Average loss=0.08547945454334602 +INFO:root: +[Valid][17][UTT][ MSE = 0.2381 | LCC = 0.8550 | SRCC = 0.8561 ] +INFO:root:Average loss=0.07687298466379826 +INFO:root: +[Valid][18][UTT][ MSE = 0.2451 | LCC = 0.8539 | SRCC = 0.8540 ] +INFO:root:Average loss=0.07559859700118884 +INFO:root: +[Valid][19][UTT][ MSE = 0.2461 | LCC = 0.8512 | SRCC = 0.8504 ] +INFO:root:Average loss=0.0683107564273553 +INFO:root: +[Valid][20][UTT][ MSE = 0.2540 | LCC = 0.8506 | SRCC = 0.8508 ] +INFO:root:Average loss=0.07247042130583371 +INFO:root: +[Valid][21][UTT][ MSE = 0.2445 | LCC = 0.8514 | SRCC = 0.8513 ] +INFO:root:Average loss=0.06710269096761177 +INFO:root: +[Valid][22][UTT][ MSE = 0.2364 | LCC = 0.8519 | SRCC = 0.8526 ] +INFO:root:Average loss=0.06065760739147663 +INFO:root: +[Valid][23][UTT][ MSE = 0.2501 | LCC = 0.8495 | SRCC = 0.8494 ] +INFO:root:Average loss=0.05840931832790375 +INFO:root: +[Valid][24][UTT][ MSE = 0.2511 | LCC = 0.8483 | SRCC = 0.8484 ] +INFO:root:Average loss=0.05865004529746679 +INFO:root: +[Valid][25][UTT][ MSE = 0.2448 | LCC = 0.8479 | SRCC = 0.8483 ] +INFO:root:Average loss=0.05649067466266644 +INFO:root: +[Valid][26][UTT][ MSE = 0.2471 | LCC = 0.8464 | SRCC = 0.8474 ] +INFO:root:Average loss=0.051083472270805105 +INFO:root: +[Valid][27][UTT][ MSE = 0.2508 | LCC = 0.8475 | SRCC = 0.8474 ] +INFO:root:Average loss=0.05222678392265852 +INFO:root: +[Valid][28][UTT][ MSE = 0.2480 | LCC = 0.8458 | SRCC = 0.8466 ] +INFO:root:Average loss=0.0502773553180771 +INFO:root: +[Valid][29][UTT][ MSE = 0.2457 | LCC = 0.8444 | SRCC = 0.8454 ] +INFO:root:Average loss=0.045577824067993045 +INFO:root: +[Valid][30][UTT][ MSE = 0.2625 | LCC = 0.8394 | SRCC = 0.8389 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2288 | LCC = 0.8656 | SRCC = 0.8655 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..758e04cac66c9a1aa3f5ea514636d714e234f036 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:460dbc6fd54b1525329bc521de2d7ef5186c1c8194a7166111cacdd311da2cb4 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4e125c49c9a899e40fa6527f27deac0b2fc68a16 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1159d1b3bb53cc41873ddc03ed2d19fb1bec53e6fc18d18fed2feb6e8e4514b0 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..c116b4cb08910d7d82af1996a0b074fdacad590a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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=0.3614535010777987 +INFO:root: +[Valid][0][UTT][ MSE = 0.3047 | LCC = 0.8323 | SRCC = 0.8356 ] +INFO:root:Average loss=0.24450552673676076 +INFO:root: +[Valid][1][UTT][ MSE = 0.2658 | LCC = 0.8543 | SRCC = 0.8539 ] +INFO:root:Average loss=0.23114972542493772 +INFO:root: +[Valid][2][UTT][ MSE = 0.2629 | LCC = 0.8589 | SRCC = 0.8583 ] +INFO:root:Average loss=0.2120393186998673 +INFO:root: +[Valid][3][UTT][ MSE = 0.2809 | LCC = 0.8650 | SRCC = 0.8627 ] +INFO:root:Average loss=0.1979809939288176 +INFO:root: +[Valid][4][UTT][ MSE = 0.2591 | LCC = 0.8665 | SRCC = 0.8657 ] +INFO:root:Average loss=0.19468684303454864 +INFO:root: +[Valid][5][UTT][ MSE = 0.2679 | LCC = 0.8613 | SRCC = 0.8665 ] +INFO:root:Average loss=0.1743511761037203 +INFO:root: +[Valid][6][UTT][ MSE = 0.2367 | LCC = 0.8677 | SRCC = 0.8695 ] +INFO:root:Average loss=0.16736831000218025 +INFO:root: +[Valid][7][UTT][ MSE = 0.2282 | LCC = 0.8706 | SRCC = 0.8707 ] +INFO:root:Average loss=0.16272649102104017 +INFO:root: +[Valid][8][UTT][ MSE = 0.2398 | LCC = 0.8666 | SRCC = 0.8669 ] +INFO:root:Average loss=0.1522993799776603 +INFO:root: +[Valid][9][UTT][ MSE = 0.2616 | LCC = 0.8687 | SRCC = 0.8693 ] +INFO:root:Average loss=0.14300388145523193 +INFO:root: +[Valid][10][UTT][ MSE = 0.2328 | LCC = 0.8679 | SRCC = 0.8680 ] +INFO:root:Average loss=0.13358230764667192 +INFO:root: +[Valid][11][UTT][ MSE = 0.2479 | LCC = 0.8647 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1269704527579821 +INFO:root: +[Valid][12][UTT][ MSE = 0.2262 | LCC = 0.8656 | SRCC = 0.8652 ] +INFO:root:Average loss=0.12070045687067203 +INFO:root: +[Valid][13][UTT][ MSE = 0.2562 | LCC = 0.8648 | SRCC = 0.8644 ] +INFO:root:Average loss=0.11248713340132664 +INFO:root: +[Valid][14][UTT][ MSE = 0.2373 | LCC = 0.8630 | SRCC = 0.8663 ] +INFO:root:Average loss=0.10893294893396206 +INFO:root: +[Valid][15][UTT][ MSE = 0.2443 | LCC = 0.8651 | SRCC = 0.8636 ] +INFO:root:Average loss=0.101377946157486 +INFO:root: +[Valid][16][UTT][ MSE = 0.2439 | LCC = 0.8582 | SRCC = 0.8593 ] +INFO:root:Average loss=0.103970400033853 +INFO:root: +[Valid][17][UTT][ MSE = 0.2425 | LCC = 0.8603 | SRCC = 0.8597 ] +INFO:root:Average loss=0.09671227423808514 +INFO:root: +[Valid][18][UTT][ MSE = 0.2486 | LCC = 0.8595 | SRCC = 0.8592 ] +INFO:root:Average loss=0.09099070961849812 +INFO:root: +[Valid][19][UTT][ MSE = 0.2629 | LCC = 0.8529 | SRCC = 0.8540 ] +INFO:root:Average loss=0.08292181412569988 +INFO:root: +[Valid][20][UTT][ MSE = 0.2491 | LCC = 0.8549 | SRCC = 0.8553 ] +INFO:root:Average loss=0.08267299291224052 +INFO:root: +[Valid][21][UTT][ MSE = 0.2382 | LCC = 0.8557 | SRCC = 0.8575 ] +INFO:root:Average loss=0.0801040454266163 +INFO:root: +[Valid][22][UTT][ MSE = 0.2637 | LCC = 0.8483 | SRCC = 0.8501 ] +INFO:root:Average loss=0.07588939669613655 +INFO:root: +[Valid][23][UTT][ MSE = 0.2774 | LCC = 0.8483 | SRCC = 0.8545 ] +INFO:root:Average loss=0.07100357697942318 +INFO:root: +[Valid][24][UTT][ MSE = 0.2577 | LCC = 0.8485 | SRCC = 0.8477 ] +INFO:root:Average loss=0.0680277115928057 +INFO:root: +[Valid][25][UTT][ MSE = 0.2629 | LCC = 0.8524 | SRCC = 0.8544 ] +INFO:root:Average loss=0.06421923159788816 +INFO:root: +[Valid][26][UTT][ MSE = 0.2426 | LCC = 0.8540 | SRCC = 0.8565 ] +INFO:root:Average loss=0.06240973549966629 +INFO:root: +[Valid][27][UTT][ MSE = 0.2388 | LCC = 0.8554 | SRCC = 0.8560 ] +INFO:root:Average loss=0.05970427814202431 +INFO:root: +[Valid][28][UTT][ MSE = 0.2533 | LCC = 0.8505 | SRCC = 0.8536 ] +INFO:root:Average loss=0.058480461629537434 +INFO:root: +[Valid][29][UTT][ MSE = 0.2540 | LCC = 0.8510 | SRCC = 0.8518 ] +INFO:root:Average loss=0.05766998145442743 +INFO:root: +[Valid][30][UTT][ MSE = 0.2545 | LCC = 0.8512 | SRCC = 0.8538 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2498 | LCC = 0.8612 | SRCC = 0.8656 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..bb0aeef3bac1e109b0d8152b9e2466c10c80d776 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a61f39b2746dc7025813d54fab1db89b34f6cf9d5e4bd2c1baed41b633ecbb5 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..3a513977032ab800677ecfc97af861e40715fecf --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4fb6f177cbf73628178a733744d67685865627d5b948ce9a8df32e39e5d81abb +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..aff66d50d77433910bd3f9544da16d521df2fb12 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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=0.3516380987488307 +INFO:root: +[Valid][0][UTT][ MSE = 0.3453 | LCC = 0.8341 | SRCC = 0.8357 ] +INFO:root:Average loss=0.2529864685657697 +INFO:root: +[Valid][1][UTT][ MSE = 0.2905 | LCC = 0.8524 | SRCC = 0.8508 ] +INFO:root:Average loss=0.22398477047681808 +INFO:root: +[Valid][2][UTT][ MSE = 0.2661 | LCC = 0.8553 | SRCC = 0.8543 ] +INFO:root:Average loss=0.2076014396853936 +INFO:root: +[Valid][3][UTT][ MSE = 0.2978 | LCC = 0.8639 | SRCC = 0.8635 ] +INFO:root:Average loss=0.19490004006104591 +INFO:root: +[Valid][4][UTT][ MSE = 0.2476 | LCC = 0.8619 | SRCC = 0.8636 ] +INFO:root:Average loss=0.17966025236707467 +INFO:root: +[Valid][5][UTT][ MSE = 0.2608 | LCC = 0.8644 | SRCC = 0.8654 ] +INFO:root:Average loss=0.17245814271080187 +INFO:root: +[Valid][6][UTT][ MSE = 0.2556 | LCC = 0.8648 | SRCC = 0.8671 ] +INFO:root:Average loss=0.1652795757429722 +INFO:root: +[Valid][7][UTT][ MSE = 0.2452 | LCC = 0.8686 | SRCC = 0.8695 ] +INFO:root:Average loss=0.15274065971756592 +INFO:root: +[Valid][8][UTT][ MSE = 0.2469 | LCC = 0.8670 | SRCC = 0.8683 ] +INFO:root:Average loss=0.14398109483031127 +INFO:root: +[Valid][9][UTT][ MSE = 0.2554 | LCC = 0.8665 | SRCC = 0.8679 ] +INFO:root:Average loss=0.12957979117830595 +INFO:root: +[Valid][10][UTT][ MSE = 0.2263 | LCC = 0.8672 | SRCC = 0.8687 ] +INFO:root:Average loss=0.12925797605361694 +INFO:root: +[Valid][11][UTT][ MSE = 0.2502 | LCC = 0.8633 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1196739280071014 +INFO:root: +[Valid][12][UTT][ MSE = 0.2350 | LCC = 0.8654 | SRCC = 0.8668 ] +INFO:root:Average loss=0.11689529028267433 +INFO:root: +[Valid][13][UTT][ MSE = 0.2504 | LCC = 0.8622 | SRCC = 0.8635 ] +INFO:root:Average loss=0.10949666435137773 +INFO:root: +[Valid][14][UTT][ MSE = 0.2589 | LCC = 0.8609 | SRCC = 0.8623 ] +INFO:root:Average loss=0.1051233158661769 +INFO:root: +[Valid][15][UTT][ MSE = 0.2326 | LCC = 0.8606 | SRCC = 0.8623 ] +INFO:root:Average loss=0.10063403777969189 +INFO:root: +[Valid][16][UTT][ MSE = 0.2354 | LCC = 0.8598 | SRCC = 0.8617 ] +INFO:root:Average loss=0.09258067421615124 +INFO:root: +[Valid][17][UTT][ MSE = 0.2574 | LCC = 0.8560 | SRCC = 0.8577 ] +INFO:root:Average loss=0.09082944481036602 +INFO:root: +[Valid][18][UTT][ MSE = 0.2419 | LCC = 0.8567 | SRCC = 0.8594 ] +INFO:root:Average loss=0.0820172998863153 +INFO:root: +[Valid][19][UTT][ MSE = 0.2493 | LCC = 0.8576 | SRCC = 0.8590 ] +INFO:root:Average loss=0.08595115285462294 +INFO:root: +[Valid][20][UTT][ MSE = 0.2459 | LCC = 0.8587 | SRCC = 0.8621 ] +INFO:root:Average loss=0.07643845271414672 +INFO:root: +[Valid][21][UTT][ MSE = 0.2317 | LCC = 0.8585 | SRCC = 0.8621 ] +INFO:root:Average loss=0.0732080611662987 +INFO:root: +[Valid][22][UTT][ MSE = 0.2424 | LCC = 0.8576 | SRCC = 0.8598 ] +INFO:root:Average loss=0.07361755553537454 +INFO:root: +[Valid][23][UTT][ MSE = 0.2422 | LCC = 0.8546 | SRCC = 0.8566 ] +INFO:root:Average loss=0.06607579630918992 +INFO:root: +[Valid][24][UTT][ MSE = 0.2310 | LCC = 0.8559 | SRCC = 0.8596 ] +INFO:root:Average loss=0.06237181328619138 +INFO:root: +[Valid][25][UTT][ MSE = 0.2417 | LCC = 0.8532 | SRCC = 0.8573 ] +INFO:root:Average loss=0.06111018650997908 +INFO:root: +[Valid][26][UTT][ MSE = 0.2399 | LCC = 0.8555 | SRCC = 0.8593 ] +INFO:root:Average loss=0.05993199305465588 +INFO:root: +[Valid][27][UTT][ MSE = 0.2577 | LCC = 0.8520 | SRCC = 0.8539 ] +INFO:root:Average loss=0.057586636633062974 +INFO:root: +[Valid][28][UTT][ MSE = 0.2385 | LCC = 0.8543 | SRCC = 0.8577 ] +INFO:root:Average loss=0.055165342174661465 +INFO:root: +[Valid][29][UTT][ MSE = 0.2383 | LCC = 0.8526 | SRCC = 0.8554 ] +INFO:root:Average loss=0.05349643538013483 +INFO:root: +[Valid][30][UTT][ MSE = 0.2509 | LCC = 0.8508 | SRCC = 0.8539 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2444 | LCC = 0.8655 | SRCC = 0.8643 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..bb8f5cf420e35f2487060c94b8ab98ee9dbc4f0e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bef75c6078c4d230738ae435848be6750fcc8a4611b3ebc7871e3d8838ba664b +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..c3e1208634b7e45f52b51d8f5c1b21c5d71ffc78 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:acc83709534e109081a94a01af645f7c67c40bc1b2df3aac1d9605f8cff6d87f +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..5a97b449f3a4d2aca04c08a53ab53190bee9dc76 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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=0.3543617761670015 +INFO:root: +[Valid][0][UTT][ MSE = 0.3016 | LCC = 0.8227 | SRCC = 0.8353 ] +INFO:root:Average loss=0.24898730027369964 +INFO:root: +[Valid][1][UTT][ MSE = 0.2779 | LCC = 0.8482 | SRCC = 0.8550 ] +INFO:root:Average loss=0.23017283414418882 +INFO:root: +[Valid][2][UTT][ MSE = 0.2434 | LCC = 0.8610 | SRCC = 0.8635 ] +INFO:root:Average loss=0.21876525066984007 +INFO:root: +[Valid][3][UTT][ MSE = 0.2530 | LCC = 0.8649 | SRCC = 0.8655 ] +INFO:root:Average loss=0.20051592416488206 +INFO:root: +[Valid][4][UTT][ MSE = 0.2482 | LCC = 0.8666 | SRCC = 0.8681 ] +INFO:root:Average loss=0.1911004660412287 +INFO:root: +[Valid][5][UTT][ MSE = 0.2452 | LCC = 0.8701 | SRCC = 0.8745 ] +INFO:root:Average loss=0.18059722534739053 +INFO:root: +[Valid][6][UTT][ MSE = 0.2298 | LCC = 0.8701 | SRCC = 0.8729 ] +INFO:root:Average loss=0.16304381153522393 +INFO:root: +[Valid][7][UTT][ MSE = 0.2240 | LCC = 0.8639 | SRCC = 0.8712 ] +INFO:root:Average loss=0.15888740485295272 +INFO:root: +[Valid][8][UTT][ MSE = 0.2716 | LCC = 0.8629 | SRCC = 0.8694 ] +INFO:root:Average loss=0.1459781476893486 +INFO:root: +[Valid][9][UTT][ MSE = 0.2347 | LCC = 0.8686 | SRCC = 0.8719 ] +INFO:root:Average loss=0.1380547611759259 +INFO:root: +[Valid][10][UTT][ MSE = 0.2288 | LCC = 0.8684 | SRCC = 0.8711 ] +INFO:root:Average loss=0.13053262119109815 +INFO:root: +[Valid][11][UTT][ MSE = 0.2817 | LCC = 0.8625 | SRCC = 0.8698 ] +INFO:root:Average loss=0.12499953089998318 +INFO:root: +[Valid][12][UTT][ MSE = 0.2254 | LCC = 0.8649 | SRCC = 0.8674 ] +INFO:root:Average loss=0.11754965973206055 +INFO:root: +[Valid][13][UTT][ MSE = 0.2376 | LCC = 0.8620 | SRCC = 0.8680 ] +INFO:root:Average loss=0.11114036435118088 +INFO:root: +[Valid][14][UTT][ MSE = 0.2329 | LCC = 0.8623 | SRCC = 0.8664 ] +INFO:root:Average loss=0.10420576234658559 +INFO:root: +[Valid][15][UTT][ MSE = 0.2274 | LCC = 0.8577 | SRCC = 0.8603 ] +INFO:root:Average loss=0.09982132338560544 +INFO:root: +[Valid][16][UTT][ MSE = 0.2361 | LCC = 0.8589 | SRCC = 0.8623 ] +INFO:root:Average loss=0.09831118383086644 +INFO:root: +[Valid][17][UTT][ MSE = 0.2302 | LCC = 0.8627 | SRCC = 0.8642 ] +INFO:root:Average loss=0.08882360255871063 +INFO:root: +[Valid][18][UTT][ MSE = 0.2591 | LCC = 0.8577 | SRCC = 0.8591 ] +INFO:root:Average loss=0.08872593060517922 +INFO:root: +[Valid][19][UTT][ MSE = 0.2591 | LCC = 0.8543 | SRCC = 0.8604 ] +INFO:root:Average loss=0.08280329664166157 +INFO:root: +[Valid][20][UTT][ MSE = 0.2283 | LCC = 0.8609 | SRCC = 0.8636 ] +INFO:root:Average loss=0.07818585643783593 +INFO:root: +[Valid][21][UTT][ MSE = 0.2359 | LCC = 0.8540 | SRCC = 0.8585 ] +INFO:root:Average loss=0.07252786470911442 +INFO:root: +[Valid][22][UTT][ MSE = 0.2366 | LCC = 0.8568 | SRCC = 0.8587 ] +INFO:root:Average loss=0.06832333467900753 +INFO:root: +[Valid][23][UTT][ MSE = 0.2310 | LCC = 0.8562 | SRCC = 0.8590 ] +INFO:root:Average loss=0.06754801088036635 +INFO:root: +[Valid][24][UTT][ MSE = 0.2337 | LCC = 0.8552 | SRCC = 0.8585 ] +INFO:root:Average loss=0.06510524475612701 +INFO:root: +[Valid][25][UTT][ MSE = 0.2384 | LCC = 0.8548 | SRCC = 0.8591 ] +INFO:root:Average loss=0.06340143810480069 +INFO:root: +[Valid][26][UTT][ MSE = 0.2311 | LCC = 0.8559 | SRCC = 0.8606 ] +INFO:root:Average loss=0.06325821382686114 +INFO:root: +[Valid][27][UTT][ MSE = 0.2398 | LCC = 0.8536 | SRCC = 0.8566 ] +INFO:root:Average loss=0.0599535236087365 +INFO:root: +[Valid][28][UTT][ MSE = 0.2307 | LCC = 0.8585 | SRCC = 0.8615 ] +INFO:root:Average loss=0.05446200631558895 +INFO:root: +[Valid][29][UTT][ MSE = 0.2401 | LCC = 0.8541 | SRCC = 0.8571 ] +INFO:root:Average loss=0.05421274518355345 +INFO:root: +[Valid][30][UTT][ MSE = 0.2477 | LCC = 0.8536 | SRCC = 0.8565 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2474 | LCC = 0.8673 | SRCC = 0.8667 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..1fd3ba50db03b622009dcb495b1b452f42beb033 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb5a7a943e796720ef2a3fda3fb0c209d2ca1763253fa123c5b40ec6adb18346 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..9957a42c93138bfc73bf0df2ed6636d5a0ff57c0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:690c999b6f3d99fc3165550c9412b58dae2d9d5235b877fb127ad90a234513bd +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..41f70b21266fab552875dfc2a1936a0312fceaa0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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=0.39195935638287127 +INFO:root: +[Valid][0][UTT][ MSE = 0.3359 | LCC = 0.8273 | SRCC = 0.8286 ] +INFO:root:Average loss=0.25427304475735396 +INFO:root: +[Valid][1][UTT][ MSE = 0.3183 | LCC = 0.8493 | SRCC = 0.8476 ] +INFO:root:Average loss=0.2300855435239963 +INFO:root: +[Valid][2][UTT][ MSE = 0.2700 | LCC = 0.8601 | SRCC = 0.8586 ] +INFO:root:Average loss=0.21263381514029625 +INFO:root: +[Valid][3][UTT][ MSE = 0.2682 | LCC = 0.8600 | SRCC = 0.8598 ] +INFO:root:Average loss=0.1963432357670405 +INFO:root: +[Valid][4][UTT][ MSE = 0.2591 | LCC = 0.8638 | SRCC = 0.8640 ] +INFO:root:Average loss=0.18962331536488655 +INFO:root: +[Valid][5][UTT][ MSE = 0.2840 | LCC = 0.8659 | SRCC = 0.8667 ] +INFO:root:Average loss=0.1779774279357531 +INFO:root: +[Valid][6][UTT][ MSE = 0.2329 | LCC = 0.8609 | SRCC = 0.8649 ] +INFO:root:Average loss=0.16882672848609778 +INFO:root: +[Valid][7][UTT][ MSE = 0.2642 | LCC = 0.8637 | SRCC = 0.8644 ] +INFO:root:Average loss=0.16010035927861166 +INFO:root: +[Valid][8][UTT][ MSE = 0.2285 | LCC = 0.8602 | SRCC = 0.8646 ] +INFO:root:Average loss=0.14956344052767143 +INFO:root: +[Valid][9][UTT][ MSE = 0.2330 | LCC = 0.8592 | SRCC = 0.8626 ] +INFO:root:Average loss=0.1378292754674569 +INFO:root: +[Valid][10][UTT][ MSE = 0.2462 | LCC = 0.8573 | SRCC = 0.8600 ] +INFO:root:Average loss=0.1336530318053869 +INFO:root: +[Valid][11][UTT][ MSE = 0.2434 | LCC = 0.8580 | SRCC = 0.8616 ] +INFO:root:Average loss=0.1254052891372106 +INFO:root: +[Valid][12][UTT][ MSE = 0.2282 | LCC = 0.8611 | SRCC = 0.8637 ] +INFO:root:Average loss=0.11554864029853772 +INFO:root: +[Valid][13][UTT][ MSE = 0.2236 | LCC = 0.8645 | SRCC = 0.8655 ] +INFO:root:Average loss=0.11332445056774677 +INFO:root: +[Valid][14][UTT][ MSE = 0.2380 | LCC = 0.8576 | SRCC = 0.8607 ] +INFO:root:Average loss=0.10492495036660096 +INFO:root: +[Valid][15][UTT][ MSE = 0.2409 | LCC = 0.8567 | SRCC = 0.8594 ] +INFO:root:Average loss=0.09740550219057462 +INFO:root: +[Valid][16][UTT][ MSE = 0.2447 | LCC = 0.8549 | SRCC = 0.8556 ] +INFO:root:Average loss=0.0936104351033767 +INFO:root: +[Valid][17][UTT][ MSE = 0.2348 | LCC = 0.8552 | SRCC = 0.8568 ] +INFO:root:Average loss=0.08477469571889976 +INFO:root: +[Valid][18][UTT][ MSE = 0.2399 | LCC = 0.8540 | SRCC = 0.8558 ] +INFO:root:Average loss=0.09058046006621459 +INFO:root: +[Valid][19][UTT][ MSE = 0.2312 | LCC = 0.8562 | SRCC = 0.8584 ] +INFO:root:Average loss=0.0814922176874601 +INFO:root: +[Valid][20][UTT][ MSE = 0.2350 | LCC = 0.8548 | SRCC = 0.8562 ] +INFO:root:Average loss=0.07906855270266533 +INFO:root: +[Valid][21][UTT][ MSE = 0.2484 | LCC = 0.8551 | SRCC = 0.8536 ] +INFO:root:Average loss=0.07465051563504414 +INFO:root: +[Valid][22][UTT][ MSE = 0.2393 | LCC = 0.8532 | SRCC = 0.8547 ] +INFO:root:Average loss=0.07002473966433452 +INFO:root: +[Valid][23][UTT][ MSE = 0.2447 | LCC = 0.8506 | SRCC = 0.8506 ] +INFO:root:Average loss=0.06908781038453946 +INFO:root: +[Valid][24][UTT][ MSE = 0.2416 | LCC = 0.8505 | SRCC = 0.8518 ] +INFO:root:Average loss=0.062375281006097794 +INFO:root: +[Valid][25][UTT][ MSE = 0.2419 | LCC = 0.8490 | SRCC = 0.8488 ] +INFO:root:Average loss=0.06363879120311676 +INFO:root: +[Valid][26][UTT][ MSE = 0.2425 | LCC = 0.8533 | SRCC = 0.8540 ] +INFO:root:Average loss=0.059381319830814995 +INFO:root: +[Valid][27][UTT][ MSE = 0.2336 | LCC = 0.8539 | SRCC = 0.8552 ] +INFO:root:Average loss=0.059819579124450684 +INFO:root: +[Valid][28][UTT][ MSE = 0.2407 | LCC = 0.8536 | SRCC = 0.8527 ] +INFO:root:Average loss=0.054399034151664145 +INFO:root: +[Valid][29][UTT][ MSE = 0.2571 | LCC = 0.8496 | SRCC = 0.8500 ] +INFO:root:Average loss=0.051598511301936247 +INFO:root: +[Valid][30][UTT][ MSE = 0.2339 | LCC = 0.8520 | SRCC = 0.8530 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2564 | LCC = 0.8671 | SRCC = 0.8671 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..af1e96279dc24191a715009efa335d2e34055f1d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b530d338a8a983d4d0dc1945fe5b616241a117b2d2fa17ddb65cf014310f8e8 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..3d16de0b42b222b0c3607785bf8cd99cae9c525d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86826e79580ea16e7fec14f21b8c4e8400b4c66270f1469d9f20ccc7bad29924 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f11177c194b65b86b0d4f96747cf310c5863a388 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer8/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=0.38400148275570994 +INFO:root: +[Valid][0][UTT][ MSE = 0.3345 | LCC = 0.8232 | SRCC = 0.8295 ] +INFO:root:Average loss=0.2554006077922307 +INFO:root: +[Valid][1][UTT][ MSE = 0.2915 | LCC = 0.8489 | SRCC = 0.8478 ] +INFO:root:Average loss=0.22843630401751933 +INFO:root: +[Valid][2][UTT][ MSE = 0.2467 | LCC = 0.8574 | SRCC = 0.8586 ] +INFO:root:Average loss=0.20907616290526512 +INFO:root: +[Valid][3][UTT][ MSE = 0.2765 | LCC = 0.8647 | SRCC = 0.8634 ] +INFO:root:Average loss=0.1928497364696784 +INFO:root: +[Valid][4][UTT][ MSE = 0.2561 | LCC = 0.8641 | SRCC = 0.8627 ] +INFO:root:Average loss=0.18419032400617233 +INFO:root: +[Valid][5][UTT][ MSE = 0.2415 | LCC = 0.8604 | SRCC = 0.8615 ] +INFO:root:Average loss=0.17062823856488252 +INFO:root: +[Valid][6][UTT][ MSE = 0.2405 | LCC = 0.8663 | SRCC = 0.8654 ] +INFO:root:Average loss=0.17525942528095 +INFO:root: +[Valid][7][UTT][ MSE = 0.2658 | LCC = 0.8629 | SRCC = 0.8623 ] +INFO:root:Average loss=0.1511209069345242 +INFO:root: +[Valid][8][UTT][ MSE = 0.2555 | LCC = 0.8672 | SRCC = 0.8658 ] +INFO:root:Average loss=0.14197916719011772 +INFO:root: +[Valid][9][UTT][ MSE = 0.2281 | LCC = 0.8653 | SRCC = 0.8671 ] +INFO:root:Average loss=0.13528820862754798 +INFO:root: +[Valid][10][UTT][ MSE = 0.2492 | LCC = 0.8618 | SRCC = 0.8613 ] +INFO:root:Average loss=0.13205580804974604 +INFO:root: +[Valid][11][UTT][ MSE = 0.2375 | LCC = 0.8628 | SRCC = 0.8619 ] +INFO:root:Average loss=0.1233714161775051 +INFO:root: +[Valid][12][UTT][ MSE = 0.2314 | LCC = 0.8631 | SRCC = 0.8643 ] +INFO:root:Average loss=0.11587841216570292 +INFO:root: +[Valid][13][UTT][ MSE = 0.2509 | LCC = 0.8601 | SRCC = 0.8618 ] +INFO:root:Average loss=0.10912908842930427 +INFO:root: +[Valid][14][UTT][ MSE = 0.2268 | LCC = 0.8616 | SRCC = 0.8634 ] +INFO:root:Average loss=0.10805998666164203 +INFO:root: +[Valid][15][UTT][ MSE = 0.2452 | LCC = 0.8570 | SRCC = 0.8590 ] +INFO:root:Average loss=0.1015613344617379 +INFO:root: +[Valid][16][UTT][ MSE = 0.2529 | LCC = 0.8550 | SRCC = 0.8566 ] +INFO:root:Average loss=0.09125200148003224 +INFO:root: +[Valid][17][UTT][ MSE = 0.2421 | LCC = 0.8571 | SRCC = 0.8590 ] +INFO:root:Average loss=0.08972409931130898 +INFO:root: +[Valid][18][UTT][ MSE = 0.2321 | LCC = 0.8583 | SRCC = 0.8610 ] +INFO:root:Average loss=0.08706526658855952 +INFO:root: +[Valid][19][UTT][ MSE = 0.2360 | LCC = 0.8578 | SRCC = 0.8586 ] +INFO:root:Average loss=0.0860665598167823 +INFO:root: +[Valid][20][UTT][ MSE = 0.2703 | LCC = 0.8601 | SRCC = 0.8612 ] +INFO:root:Average loss=0.08008361306901161 +INFO:root: +[Valid][21][UTT][ MSE = 0.2354 | LCC = 0.8552 | SRCC = 0.8574 ] +INFO:root:Average loss=0.08046908566775994 +INFO:root: +[Valid][22][UTT][ MSE = 0.2431 | LCC = 0.8517 | SRCC = 0.8530 ] +INFO:root:Average loss=0.07337328586249779 +INFO:root: +[Valid][23][UTT][ MSE = 0.2480 | LCC = 0.8517 | SRCC = 0.8531 ] +INFO:root:Average loss=0.07073683258241568 +INFO:root: +[Valid][24][UTT][ MSE = 0.2474 | LCC = 0.8514 | SRCC = 0.8519 ] +INFO:root:Average loss=0.06866148572701675 +INFO:root: +[Valid][25][UTT][ MSE = 0.2489 | LCC = 0.8548 | SRCC = 0.8564 ] +INFO:root:Average loss=0.07080787076399876 +INFO:root: +[Valid][26][UTT][ MSE = 0.2481 | LCC = 0.8555 | SRCC = 0.8555 ] +INFO:root:Average loss=0.06434181776757424 +INFO:root: +[Valid][27][UTT][ MSE = 0.2493 | LCC = 0.8498 | SRCC = 0.8510 ] +INFO:root:Average loss=0.06219903938472271 +INFO:root: +[Valid][28][UTT][ MSE = 0.2400 | LCC = 0.8549 | SRCC = 0.8552 ] +INFO:root:Average loss=0.05991233312166654 +INFO:root: +[Valid][29][UTT][ MSE = 0.2294 | LCC = 0.8569 | SRCC = 0.8588 ] +INFO:root:Average loss=0.05633872489516552 +INFO:root: +[Valid][30][UTT][ MSE = 0.2404 | LCC = 0.8538 | SRCC = 0.8549 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2729 | LCC = 0.8553 | SRCC = 0.8588 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..b019b0c6c48f9d864cfc7a88276f5803d13c8302 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfeee2d526fb3104f8cb17a560d55390eb81cc7969abac0c8235b630f034404f +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..8ba90eb2ee8af0eb1a8e1c320ac43693fb4f58bd --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32c74d95e578c0f03f76e64a745823153228599df9fb8b0ab5ce3b8e09ed3c07 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e9c9c08888e6f8b9d6faa82ed4425f5be60521e6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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=0.32327583546821886 +INFO:root: +[Valid][0][UTT][ MSE = 0.3365 | LCC = 0.8311 | SRCC = 0.8285 ] +INFO:root:Average loss=0.2633518773393753 +INFO:root: +[Valid][1][UTT][ MSE = 0.3142 | LCC = 0.8467 | SRCC = 0.8443 ] +INFO:root:Average loss=0.24265167117118835 +INFO:root: +[Valid][2][UTT][ MSE = 0.2966 | LCC = 0.8494 | SRCC = 0.8513 ] +INFO:root:Average loss=0.24045826819462654 +INFO:root: +[Valid][3][UTT][ MSE = 0.3105 | LCC = 0.8522 | SRCC = 0.8514 ] +INFO:root:Average loss=0.2183155344846921 +INFO:root: +[Valid][4][UTT][ MSE = 0.2588 | LCC = 0.8586 | SRCC = 0.8595 ] +INFO:root:Average loss=0.21141697963078818 +INFO:root: +[Valid][5][UTT][ MSE = 0.2509 | LCC = 0.8624 | SRCC = 0.8629 ] +INFO:root:Average loss=0.2088816945369427 +INFO:root: +[Valid][6][UTT][ MSE = 0.2517 | LCC = 0.8593 | SRCC = 0.8607 ] +INFO:root:Average loss=0.1994120068848133 +INFO:root: +[Valid][7][UTT][ MSE = 0.2496 | LCC = 0.8643 | SRCC = 0.8652 ] +INFO:root:Average loss=0.1917733806066024 +INFO:root: +[Valid][8][UTT][ MSE = 0.2388 | LCC = 0.8657 | SRCC = 0.8665 ] +INFO:root:Average loss=0.1952084711728952 +INFO:root: +[Valid][9][UTT][ MSE = 0.2779 | LCC = 0.8668 | SRCC = 0.8668 ] +INFO:root:Average loss=0.18093604546708938 +INFO:root: +[Valid][10][UTT][ MSE = 0.2436 | LCC = 0.8647 | SRCC = 0.8659 ] +INFO:root:Average loss=0.17939605668951303 +INFO:root: +[Valid][11][UTT][ MSE = 0.2514 | LCC = 0.8612 | SRCC = 0.8631 ] +INFO:root:Average loss=0.17084043826430273 +INFO:root: +[Valid][12][UTT][ MSE = 0.2641 | LCC = 0.8659 | SRCC = 0.8663 ] +INFO:root:Average loss=0.16727976147563028 +INFO:root: +[Valid][13][UTT][ MSE = 0.2594 | LCC = 0.8623 | SRCC = 0.8631 ] +INFO:root:Average loss=0.15963676906166932 +INFO:root: +[Valid][14][UTT][ MSE = 0.2518 | LCC = 0.8649 | SRCC = 0.8668 ] +INFO:root:Average loss=0.1549871197113624 +INFO:root: +[Valid][15][UTT][ MSE = 0.2778 | LCC = 0.8640 | SRCC = 0.8630 ] +INFO:root:Average loss=0.15601059975914466 +INFO:root: +[Valid][16][UTT][ MSE = 0.2578 | LCC = 0.8627 | SRCC = 0.8636 ] +INFO:root:Average loss=0.14357589958952025 +INFO:root: +[Valid][17][UTT][ MSE = 0.2291 | LCC = 0.8610 | SRCC = 0.8612 ] +INFO:root:Average loss=0.14174985541747168 +INFO:root: +[Valid][18][UTT][ MSE = 0.2555 | LCC = 0.8656 | SRCC = 0.8657 ] +INFO:root:Average loss=0.1380036978576428 +INFO:root: +[Valid][19][UTT][ MSE = 0.2591 | LCC = 0.8597 | SRCC = 0.8608 ] +INFO:root:Average loss=0.13158248680142257 +INFO:root: +[Valid][20][UTT][ MSE = 0.2386 | LCC = 0.8620 | SRCC = 0.8605 ] +INFO:root:Average loss=0.12959174095438078 +INFO:root: +[Valid][21][UTT][ MSE = 0.2449 | LCC = 0.8596 | SRCC = 0.8599 ] +INFO:root:Average loss=0.12718118679447052 +INFO:root: +[Valid][22][UTT][ MSE = 0.2372 | LCC = 0.8604 | SRCC = 0.8597 ] +INFO:root:Average loss=0.11856315724360637 +INFO:root: +[Valid][23][UTT][ MSE = 0.2444 | LCC = 0.8619 | SRCC = 0.8626 ] +INFO:root:Average loss=0.11721538556500888 +INFO:root: +[Valid][24][UTT][ MSE = 0.2312 | LCC = 0.8634 | SRCC = 0.8634 ] +INFO:root:Average loss=0.11831410458454719 +INFO:root: +[Valid][25][UTT][ MSE = 0.2489 | LCC = 0.8631 | SRCC = 0.8621 ] +INFO:root:Average loss=0.1109315992738956 +INFO:root: +[Valid][26][UTT][ MSE = 0.2466 | LCC = 0.8575 | SRCC = 0.8574 ] +INFO:root:Average loss=0.11470362243170922 +INFO:root: +[Valid][27][UTT][ MSE = 0.2638 | LCC = 0.8580 | SRCC = 0.8593 ] +INFO:root:Average loss=0.10675549793701905 +INFO:root: +[Valid][28][UTT][ MSE = 0.2507 | LCC = 0.8543 | SRCC = 0.8551 ] +INFO:root:Average loss=0.10173687529869568 +INFO:root: +[Valid][29][UTT][ MSE = 0.2489 | LCC = 0.8565 | SRCC = 0.8565 ] +INFO:root:Average loss=0.0998828078691776 +INFO:root: +[Valid][30][UTT][ MSE = 0.2805 | LCC = 0.8425 | SRCC = 0.8451 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2717 | LCC = 0.8647 | SRCC = 0.8655 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..09e8a092576efe6ccfa8060ea8bdb410a0cc3b79 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61baa71a2eaf646224e3548a077c925a847eb797a3641dafb0b22e4ae510d71e +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7a2aa778ee106f5ba687b9392b9d999b1822f50d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ccfb98a1473fdbb06af669630a67603cf011163eb6b903afcc9bb6391bff7afb +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..3e40b6ba274597d2f46186a3955601abd3a0b83c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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=0.38379344707116103 +INFO:root: +[Valid][0][UTT][ MSE = 0.3624 | LCC = 0.8270 | SRCC = 0.8238 ] +INFO:root:Average loss=0.26301115598434055 +INFO:root: +[Valid][1][UTT][ MSE = 0.2842 | LCC = 0.8331 | SRCC = 0.8387 ] +INFO:root:Average loss=0.253959039464975 +INFO:root: +[Valid][2][UTT][ MSE = 0.2932 | LCC = 0.8460 | SRCC = 0.8459 ] +INFO:root:Average loss=0.23923041633306405 +INFO:root: +[Valid][3][UTT][ MSE = 0.2897 | LCC = 0.8510 | SRCC = 0.8505 ] +INFO:root:Average loss=0.2220643829458799 +INFO:root: +[Valid][4][UTT][ MSE = 0.3002 | LCC = 0.8576 | SRCC = 0.8552 ] +INFO:root:Average loss=0.2158811360788651 +INFO:root: +[Valid][5][UTT][ MSE = 0.2683 | LCC = 0.8588 | SRCC = 0.8585 ] +INFO:root:Average loss=0.20960695353838113 +INFO:root: +[Valid][6][UTT][ MSE = 0.2594 | LCC = 0.8583 | SRCC = 0.8631 ] +INFO:root:Average loss=0.20220466340199494 +INFO:root: +[Valid][7][UTT][ MSE = 0.2338 | LCC = 0.8612 | SRCC = 0.8645 ] +INFO:root:Average loss=0.19431573477311012 +INFO:root: +[Valid][8][UTT][ MSE = 0.2653 | LCC = 0.8647 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1859378789861997 +INFO:root: +[Valid][9][UTT][ MSE = 0.2598 | LCC = 0.8630 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1830394718891535 +INFO:root: +[Valid][10][UTT][ MSE = 0.2435 | LCC = 0.8620 | SRCC = 0.8637 ] +INFO:root:Average loss=0.17636510414572862 +INFO:root: +[Valid][11][UTT][ MSE = 0.2711 | LCC = 0.8663 | SRCC = 0.8694 ] +INFO:root:Average loss=0.17615581838748393 +INFO:root: +[Valid][12][UTT][ MSE = 0.2314 | LCC = 0.8662 | SRCC = 0.8684 ] +INFO:root:Average loss=0.1669312251301912 +INFO:root: +[Valid][13][UTT][ MSE = 0.2438 | LCC = 0.8630 | SRCC = 0.8655 ] +INFO:root:Average loss=0.1611374416985573 +INFO:root: +[Valid][14][UTT][ MSE = 0.2638 | LCC = 0.8649 | SRCC = 0.8652 ] +INFO:root:Average loss=0.15599869220302656 +INFO:root: +[Valid][15][UTT][ MSE = 0.2769 | LCC = 0.8654 | SRCC = 0.8655 ] +INFO:root:Average loss=0.14848595494643235 +INFO:root: +[Valid][16][UTT][ MSE = 0.3013 | LCC = 0.8504 | SRCC = 0.8580 ] +INFO:root:Average loss=0.1486570072861818 +INFO:root: +[Valid][17][UTT][ MSE = 0.2448 | LCC = 0.8623 | SRCC = 0.8628 ] +INFO:root:Average loss=0.14005419592826795 +INFO:root: +[Valid][18][UTT][ MSE = 0.2821 | LCC = 0.8614 | SRCC = 0.8593 ] +INFO:root:Average loss=0.13866762874218133 +INFO:root: +[Valid][19][UTT][ MSE = 0.2264 | LCC = 0.8631 | SRCC = 0.8662 ] +INFO:root:Average loss=0.13147579448727462 +INFO:root: +[Valid][20][UTT][ MSE = 0.2610 | LCC = 0.8618 | SRCC = 0.8627 ] +INFO:root:Average loss=0.13023037539842802 +INFO:root: +[Valid][21][UTT][ MSE = 0.2445 | LCC = 0.8631 | SRCC = 0.8632 ] +INFO:root:Average loss=0.12417347786518243 +INFO:root: +[Valid][22][UTT][ MSE = 0.2811 | LCC = 0.8606 | SRCC = 0.8608 ] +INFO:root:Average loss=0.12369941050807635 +INFO:root: +[Valid][23][UTT][ MSE = 0.2351 | LCC = 0.8609 | SRCC = 0.8623 ] +INFO:root:Average loss=0.11348461168698776 +INFO:root: +[Valid][24][UTT][ MSE = 0.2480 | LCC = 0.8585 | SRCC = 0.8583 ] +INFO:root:Average loss=0.11737068169392072 +INFO:root: +[Valid][25][UTT][ MSE = 0.2408 | LCC = 0.8590 | SRCC = 0.8592 ] +INFO:root:Average loss=0.10852837008543503 +INFO:root: +[Valid][26][UTT][ MSE = 0.2590 | LCC = 0.8549 | SRCC = 0.8554 ] +INFO:root:Average loss=0.10546889800864917 +INFO:root: +[Valid][27][UTT][ MSE = 0.2488 | LCC = 0.8543 | SRCC = 0.8552 ] +INFO:root:Average loss=0.10275620699692996 +INFO:root: +[Valid][28][UTT][ MSE = 0.2280 | LCC = 0.8605 | SRCC = 0.8622 ] +INFO:root:Average loss=0.09508142094963636 +INFO:root: +[Valid][29][UTT][ MSE = 0.2517 | LCC = 0.8514 | SRCC = 0.8504 ] +INFO:root:Average loss=0.09873722675137031 +INFO:root: +[Valid][30][UTT][ MSE = 0.2526 | LCC = 0.8532 | SRCC = 0.8539 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2575 | LCC = 0.8665 | SRCC = 0.8640 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ec3ac5baeaae9def5255736b00ebd603883ef838 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d720ac43347a1dde81e35d21a7b030a0ddd03ec36b832b9364078eb3c2eee63 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..32f3d4e6c90aeaefff8bfd3a2b26acd3566f8219 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a62d070db3ca06f8f49bb2a2460b5747d627759d5f71c062ee1c2e3d59cab343 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..174d57f8865e96b0c8a83b40adfea6a602339bb1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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=0.34807988160695785 +INFO:root: +[Valid][0][UTT][ MSE = 0.3389 | LCC = 0.8285 | SRCC = 0.8289 ] +INFO:root:Average loss=0.26272860570595813 +INFO:root: +[Valid][1][UTT][ MSE = 0.3350 | LCC = 0.8443 | SRCC = 0.8423 ] +INFO:root:Average loss=0.2531656257999249 +INFO:root: +[Valid][2][UTT][ MSE = 0.3211 | LCC = 0.8443 | SRCC = 0.8449 ] +INFO:root:Average loss=0.2368129421121035 +INFO:root: +[Valid][3][UTT][ MSE = 0.3350 | LCC = 0.8435 | SRCC = 0.8504 ] +INFO:root:Average loss=0.2261238165008716 +INFO:root: +[Valid][4][UTT][ MSE = 0.2929 | LCC = 0.8588 | SRCC = 0.8564 ] +INFO:root:Average loss=0.21668150868171301 +INFO:root: +[Valid][5][UTT][ MSE = 0.2585 | LCC = 0.8624 | SRCC = 0.8614 ] +INFO:root:Average loss=0.21101694936171556 +INFO:root: +[Valid][6][UTT][ MSE = 0.2549 | LCC = 0.8610 | SRCC = 0.8604 ] +INFO:root:Average loss=0.20423394995622146 +INFO:root: +[Valid][7][UTT][ MSE = 0.2676 | LCC = 0.8636 | SRCC = 0.8635 ] +INFO:root:Average loss=0.19333516366970846 +INFO:root: +[Valid][8][UTT][ MSE = 0.2515 | LCC = 0.8672 | SRCC = 0.8672 ] +INFO:root:Average loss=0.18898884025521767 +INFO:root: +[Valid][9][UTT][ MSE = 0.2744 | LCC = 0.8677 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1800963893914834 +INFO:root: +[Valid][10][UTT][ MSE = 0.2991 | LCC = 0.8636 | SRCC = 0.8640 ] +INFO:root:Average loss=0.18335473814453834 +INFO:root: +[Valid][11][UTT][ MSE = 0.2461 | LCC = 0.8673 | SRCC = 0.8664 ] +INFO:root:Average loss=0.16944046127490509 +INFO:root: +[Valid][12][UTT][ MSE = 0.2876 | LCC = 0.8657 | SRCC = 0.8636 ] +INFO:root:Average loss=0.17110358226375702 +INFO:root: +[Valid][13][UTT][ MSE = 0.3070 | LCC = 0.8607 | SRCC = 0.8638 ] +INFO:root:Average loss=0.1685297559851255 +INFO:root: +[Valid][14][UTT][ MSE = 0.2881 | LCC = 0.8587 | SRCC = 0.8606 ] +INFO:root:Average loss=0.15736717854936919 +INFO:root: +[Valid][15][UTT][ MSE = 0.3224 | LCC = 0.8616 | SRCC = 0.8621 ] +INFO:root:Average loss=0.16322396924862495 +INFO:root: +[Valid][16][UTT][ MSE = 0.2994 | LCC = 0.8604 | SRCC = 0.8568 ] +INFO:root:Average loss=0.1463312451273967 +INFO:root: +[Valid][17][UTT][ MSE = 0.2801 | LCC = 0.8646 | SRCC = 0.8627 ] +INFO:root:Average loss=0.144158381013534 +INFO:root: +[Valid][18][UTT][ MSE = 0.2810 | LCC = 0.8623 | SRCC = 0.8611 ] +INFO:root:Average loss=0.14296907282028443 +INFO:root: +[Valid][19][UTT][ MSE = 0.2519 | LCC = 0.8656 | SRCC = 0.8646 ] +INFO:root:Average loss=0.13462834556897482 +INFO:root: +[Valid][20][UTT][ MSE = 0.2569 | LCC = 0.8630 | SRCC = 0.8649 ] +INFO:root:Average loss=0.13140001988563782 +INFO:root: +[Valid][21][UTT][ MSE = 0.2585 | LCC = 0.8634 | SRCC = 0.8630 ] +INFO:root:Average loss=0.12587775366428572 +INFO:root: +[Valid][22][UTT][ MSE = 0.2583 | LCC = 0.8563 | SRCC = 0.8576 ] +INFO:root:Average loss=0.1226643426104998 +INFO:root: +[Valid][23][UTT][ MSE = 0.3122 | LCC = 0.8603 | SRCC = 0.8598 ] +INFO:root:Average loss=0.12059139813750218 +INFO:root: +[Valid][24][UTT][ MSE = 0.2690 | LCC = 0.8584 | SRCC = 0.8574 ] +INFO:root:Average loss=0.10991388330092797 +INFO:root: +[Valid][25][UTT][ MSE = 0.2820 | LCC = 0.8594 | SRCC = 0.8592 ] +INFO:root:Average loss=0.11490287775030503 +INFO:root: +[Valid][26][UTT][ MSE = 0.2931 | LCC = 0.8566 | SRCC = 0.8544 ] +INFO:root:Average loss=0.1079370453953743 +INFO:root: +[Valid][27][UTT][ MSE = 0.2853 | LCC = 0.8516 | SRCC = 0.8498 ] +INFO:root:Average loss=0.10328078695023671 +INFO:root: +[Valid][28][UTT][ MSE = 0.2853 | LCC = 0.8541 | SRCC = 0.8534 ] +INFO:root:Average loss=0.10226308654707211 +INFO:root: +[Valid][29][UTT][ MSE = 0.2596 | LCC = 0.8473 | SRCC = 0.8467 ] +INFO:root:Average loss=0.09903597559493321 +INFO:root: +[Valid][30][UTT][ MSE = 0.2597 | LCC = 0.8555 | SRCC = 0.8551 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2603 | LCC = 0.8612 | SRCC = 0.8610 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d4b2bc203d472bfccb72e5a9a809b5e9ca3e1d5f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0d4fca7fd06a772050d87a128af1c8153eda597af3973ceeda5227f24cd1d0c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..167293df183e15c9b14bb0212ba2d798e03a7796 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b532d2c4b749dafe8ee6ad3ebbef52b9ee73b0e237888e8646e5688cdb450f84 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..7626f22ee9d75104d3f8b748b4519c5952039025 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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=0.33572408174857116 +INFO:root: +[Valid][0][UTT][ MSE = 0.3271 | LCC = 0.8196 | SRCC = 0.8277 ] +INFO:root:Average loss=0.2607962905596464 +INFO:root: +[Valid][1][UTT][ MSE = 0.3114 | LCC = 0.8373 | SRCC = 0.8415 ] +INFO:root:Average loss=0.25571536559325 +INFO:root: +[Valid][2][UTT][ MSE = 0.3314 | LCC = 0.8472 | SRCC = 0.8478 ] +INFO:root:Average loss=0.23711200268604818 +INFO:root: +[Valid][3][UTT][ MSE = 0.3221 | LCC = 0.8567 | SRCC = 0.8539 ] +INFO:root:Average loss=0.2282086614614878 +INFO:root: +[Valid][4][UTT][ MSE = 0.2828 | LCC = 0.8563 | SRCC = 0.8558 ] +INFO:root:Average loss=0.217584572159327 +INFO:root: +[Valid][5][UTT][ MSE = 0.2683 | LCC = 0.8597 | SRCC = 0.8606 ] +INFO:root:Average loss=0.21095306540911013 +INFO:root: +[Valid][6][UTT][ MSE = 0.3031 | LCC = 0.8605 | SRCC = 0.8637 ] +INFO:root:Average loss=0.20741724070066062 +INFO:root: +[Valid][7][UTT][ MSE = 0.2547 | LCC = 0.8614 | SRCC = 0.8646 ] +INFO:root:Average loss=0.21496365754268107 +INFO:root: +[Valid][8][UTT][ MSE = 0.2695 | LCC = 0.8674 | SRCC = 0.8658 ] +INFO:root:Average loss=0.19657498464370385 +INFO:root: +[Valid][9][UTT][ MSE = 0.2880 | LCC = 0.8645 | SRCC = 0.8657 ] +INFO:root:Average loss=0.18608106749180037 +INFO:root: +[Valid][10][UTT][ MSE = 0.2629 | LCC = 0.8645 | SRCC = 0.8668 ] +INFO:root:Average loss=0.18716502361572707 +INFO:root: +[Valid][11][UTT][ MSE = 0.2326 | LCC = 0.8653 | SRCC = 0.8698 ] +INFO:root:Average loss=0.17580555856992036 +INFO:root: +[Valid][12][UTT][ MSE = 0.2757 | LCC = 0.8676 | SRCC = 0.8689 ] +INFO:root:Average loss=0.17966002655717042 +INFO:root: +[Valid][13][UTT][ MSE = 0.2475 | LCC = 0.8634 | SRCC = 0.8664 ] +INFO:root:Average loss=0.16809996504050034 +INFO:root: +[Valid][14][UTT][ MSE = 0.2364 | LCC = 0.8677 | SRCC = 0.8682 ] +INFO:root:Average loss=0.16727023533521554 +INFO:root: +[Valid][15][UTT][ MSE = 0.2723 | LCC = 0.8670 | SRCC = 0.8659 ] +INFO:root:Average loss=0.15903615139615843 +INFO:root: +[Valid][16][UTT][ MSE = 0.2519 | LCC = 0.8613 | SRCC = 0.8634 ] +INFO:root:Average loss=0.15642962279992226 +INFO:root: +[Valid][17][UTT][ MSE = 0.2486 | LCC = 0.8601 | SRCC = 0.8613 ] +INFO:root:Average loss=0.15542899215450653 +INFO:root: +[Valid][18][UTT][ MSE = 0.3215 | LCC = 0.8610 | SRCC = 0.8661 ] +INFO:root:Average loss=0.15081003126807702 +INFO:root: +[Valid][19][UTT][ MSE = 0.2356 | LCC = 0.8615 | SRCC = 0.8674 ] +INFO:root:Average loss=0.14102343899699357 +INFO:root: +[Valid][20][UTT][ MSE = 0.2555 | LCC = 0.8615 | SRCC = 0.8623 ] +INFO:root:Average loss=0.13383311711442777 +INFO:root: +[Valid][21][UTT][ MSE = 0.2488 | LCC = 0.8630 | SRCC = 0.8644 ] +INFO:root:Average loss=0.13279335563763595 +INFO:root: +[Valid][22][UTT][ MSE = 0.2633 | LCC = 0.8622 | SRCC = 0.8613 ] +INFO:root:Average loss=0.1268676981711999 +INFO:root: +[Valid][23][UTT][ MSE = 0.2578 | LCC = 0.8578 | SRCC = 0.8586 ] +INFO:root:Average loss=0.12510163633105081 +INFO:root: +[Valid][24][UTT][ MSE = 0.2895 | LCC = 0.8613 | SRCC = 0.8626 ] +INFO:root:Average loss=0.12210713355587079 +INFO:root: +[Valid][25][UTT][ MSE = 0.2625 | LCC = 0.8522 | SRCC = 0.8545 ] +INFO:root:Average loss=0.1183473269144694 +INFO:root: +[Valid][26][UTT][ MSE = 0.2667 | LCC = 0.8582 | SRCC = 0.8609 ] +INFO:root:Average loss=0.11706660095697795 +INFO:root: +[Valid][27][UTT][ MSE = 0.2379 | LCC = 0.8566 | SRCC = 0.8594 ] +INFO:root:Average loss=0.11852137648906463 +INFO:root: +[Valid][28][UTT][ MSE = 0.2708 | LCC = 0.8521 | SRCC = 0.8551 ] +INFO:root:Average loss=0.1101952177973894 +INFO:root: +[Valid][29][UTT][ MSE = 0.2598 | LCC = 0.8539 | SRCC = 0.8544 ] +INFO:root:Average loss=0.10542302172726546 +INFO:root: +[Valid][30][UTT][ MSE = 0.2628 | LCC = 0.8488 | SRCC = 0.8513 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2439 | LCC = 0.8600 | SRCC = 0.8630 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..a174007f2f60188cabaf8a7c7b508a5bd03c2c33 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d4d46c5fdd88e5cb986d35526cf45d5ba47d2f5df640dafd5cb09f7536add051 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4c795e9608103eeb12bf8bdd4c4d6e8c9b395f69 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e4870587c0328fc0db7f6488a98c86b6a0b40c712c36d1a840d7249c72cbe1a +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..dc505ae19a1bbd932e0241216b85eb7f80447991 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_4epoch/layer9/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=0.3594962752018219 +INFO:root: +[Valid][0][UTT][ MSE = 0.3253 | LCC = 0.8220 | SRCC = 0.8237 ] +INFO:root:Average loss=0.26300340795364135 +INFO:root: +[Valid][1][UTT][ MSE = 0.3216 | LCC = 0.8418 | SRCC = 0.8424 ] +INFO:root:Average loss=0.24747894742549995 +INFO:root: +[Valid][2][UTT][ MSE = 0.2814 | LCC = 0.8502 | SRCC = 0.8514 ] +INFO:root:Average loss=0.24279388059408236 +INFO:root: +[Valid][3][UTT][ MSE = 0.3115 | LCC = 0.8534 | SRCC = 0.8511 ] +INFO:root:Average loss=0.22355610590714675 +INFO:root: +[Valid][4][UTT][ MSE = 0.3310 | LCC = 0.8590 | SRCC = 0.8571 ] +INFO:root:Average loss=0.2150093368612803 +INFO:root: +[Valid][5][UTT][ MSE = 0.2676 | LCC = 0.8598 | SRCC = 0.8608 ] +INFO:root:Average loss=0.21969481099110383 +INFO:root: +[Valid][6][UTT][ MSE = 0.2569 | LCC = 0.8611 | SRCC = 0.8616 ] +INFO:root:Average loss=0.20725039087044886 +INFO:root: +[Valid][7][UTT][ MSE = 0.2465 | LCC = 0.8560 | SRCC = 0.8599 ] +INFO:root:Average loss=0.1974759592841833 +INFO:root: +[Valid][8][UTT][ MSE = 0.2572 | LCC = 0.8668 | SRCC = 0.8670 ] +INFO:root:Average loss=0.18985353037714958 +INFO:root: +[Valid][9][UTT][ MSE = 0.2875 | LCC = 0.8542 | SRCC = 0.8629 ] +INFO:root:Average loss=0.18659593919530892 +INFO:root: +[Valid][10][UTT][ MSE = 0.2491 | LCC = 0.8650 | SRCC = 0.8643 ] +INFO:root:Average loss=0.18174462927839694 +INFO:root: +[Valid][11][UTT][ MSE = 0.2571 | LCC = 0.8661 | SRCC = 0.8655 ] +INFO:root:Average loss=0.17756588308092874 +INFO:root: +[Valid][12][UTT][ MSE = 0.2398 | LCC = 0.8648 | SRCC = 0.8666 ] +INFO:root:Average loss=0.1689560043697174 +INFO:root: +[Valid][13][UTT][ MSE = 0.2500 | LCC = 0.8647 | SRCC = 0.8640 ] +INFO:root:Average loss=0.16853030771017075 +INFO:root: +[Valid][14][UTT][ MSE = 0.2475 | LCC = 0.8675 | SRCC = 0.8659 ] +INFO:root:Average loss=0.160929625806136 +INFO:root: +[Valid][15][UTT][ MSE = 0.2980 | LCC = 0.8643 | SRCC = 0.8628 ] +INFO:root:Average loss=0.1624273732304573 +INFO:root: +[Valid][16][UTT][ MSE = 0.2833 | LCC = 0.8612 | SRCC = 0.8646 ] +INFO:root:Average loss=0.15295614588719147 +INFO:root: +[Valid][17][UTT][ MSE = 0.2433 | LCC = 0.8642 | SRCC = 0.8650 ] +INFO:root:Average loss=0.1500207085448962 +INFO:root: +[Valid][18][UTT][ MSE = 0.2453 | LCC = 0.8625 | SRCC = 0.8632 ] +INFO:root:Average loss=0.1436832005587908 +INFO:root: +[Valid][19][UTT][ MSE = 0.2479 | LCC = 0.8618 | SRCC = 0.8636 ] +INFO:root:Average loss=0.13956734700462756 +INFO:root: +[Valid][20][UTT][ MSE = 0.2430 | LCC = 0.8582 | SRCC = 0.8612 ] +INFO:root:Average loss=0.1323820356375132 +INFO:root: +[Valid][21][UTT][ MSE = 0.2407 | LCC = 0.8628 | SRCC = 0.8635 ] +INFO:root:Average loss=0.1327209185140255 +INFO:root: +[Valid][22][UTT][ MSE = 0.2624 | LCC = 0.8606 | SRCC = 0.8603 ] +INFO:root:Average loss=0.12915716568628946 +INFO:root: +[Valid][23][UTT][ MSE = 0.2494 | LCC = 0.8617 | SRCC = 0.8611 ] +INFO:root:Average loss=0.12379108340694354 +INFO:root: +[Valid][24][UTT][ MSE = 0.2303 | LCC = 0.8590 | SRCC = 0.8620 ] +INFO:root:Average loss=0.12001354190019461 +INFO:root: +[Valid][25][UTT][ MSE = 0.2475 | LCC = 0.8618 | SRCC = 0.8610 ] +INFO:root:Average loss=0.11578379237116912 +INFO:root: +[Valid][26][UTT][ MSE = 0.2385 | LCC = 0.8569 | SRCC = 0.8569 ] +INFO:root:Average loss=0.11233727624401069 +INFO:root: +[Valid][27][UTT][ MSE = 0.2536 | LCC = 0.8600 | SRCC = 0.8604 ] +INFO:root:Average loss=0.10915664086739223 +INFO:root: +[Valid][28][UTT][ MSE = 0.2302 | LCC = 0.8611 | SRCC = 0.8619 ] +INFO:root:Average loss=0.10720498028856057 +INFO:root: +[Valid][29][UTT][ MSE = 0.2552 | LCC = 0.8536 | SRCC = 0.8535 ] +INFO:root:Average loss=0.10871991113974498 +INFO:root: +[Valid][30][UTT][ MSE = 0.2455 | LCC = 0.8528 | SRCC = 0.8550 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2472 | LCC = 0.8645 | SRCC = 0.8626 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..403ebc3140e5cf06039da5e9e8d90980d00d11e0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3eae6e971036dabb904b086ce185e0cae2858fb2e3cea6df52af8e4309a30b7c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..1c658316717de7ca40626be0553033cb589c1a6c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:887c137b9ee29b6d5fd17c6fea17db0da44f3afc4296aedd302e6247c13825c9 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4ad525936bb63e8cae80881d75fae4491f64e677 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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=0.5960703408106779 +INFO:root: +[Valid][0][UTT][ MSE = 0.4061 | LCC = 0.7353 | SRCC = 0.7371 ] +INFO:root:Average loss=0.3127645028707309 +INFO:root: +[Valid][1][UTT][ MSE = 0.3191 | LCC = 0.7957 | SRCC = 0.7983 ] +INFO:root:Average loss=0.2558143083483745 +INFO:root: +[Valid][2][UTT][ MSE = 0.3014 | LCC = 0.8076 | SRCC = 0.8118 ] +INFO:root:Average loss=0.24831331750521293 +INFO:root: +[Valid][3][UTT][ MSE = 0.2859 | LCC = 0.8211 | SRCC = 0.8188 ] +INFO:root:Average loss=0.22444840558828452 +INFO:root: +[Valid][4][UTT][ MSE = 0.2823 | LCC = 0.8221 | SRCC = 0.8242 ] +INFO:root:Average loss=0.21001127476875597 +INFO:root: +[Valid][5][UTT][ MSE = 0.3057 | LCC = 0.8292 | SRCC = 0.8288 ] +INFO:root:Average loss=0.20050436831437624 +INFO:root: +[Valid][6][UTT][ MSE = 0.2970 | LCC = 0.8345 | SRCC = 0.8324 ] +INFO:root:Average loss=0.1892753182313381 +INFO:root: +[Valid][7][UTT][ MSE = 0.2671 | LCC = 0.8347 | SRCC = 0.8324 ] +INFO:root:Average loss=0.18359179412707305 +INFO:root: +[Valid][8][UTT][ MSE = 0.2730 | LCC = 0.8337 | SRCC = 0.8300 ] +INFO:root:Average loss=0.16770407442863172 +INFO:root: +[Valid][9][UTT][ MSE = 0.2723 | LCC = 0.8308 | SRCC = 0.8306 ] +INFO:root:Average loss=0.1781671289832164 +INFO:root: +[Valid][10][UTT][ MSE = 0.2977 | LCC = 0.8412 | SRCC = 0.8393 ] +INFO:root:Average loss=0.16042332924329317 +INFO:root: +[Valid][11][UTT][ MSE = 0.2644 | LCC = 0.8361 | SRCC = 0.8343 ] +INFO:root:Average loss=0.15069590251033121 +INFO:root: +[Valid][12][UTT][ MSE = 0.2666 | LCC = 0.8364 | SRCC = 0.8334 ] +INFO:root:Average loss=0.14913045443021333 +INFO:root: +[Valid][13][UTT][ MSE = 0.2513 | LCC = 0.8397 | SRCC = 0.8391 ] +INFO:root:Average loss=0.1390658604602019 +INFO:root: +[Valid][14][UTT][ MSE = 0.2681 | LCC = 0.8353 | SRCC = 0.8397 ] +INFO:root:Average loss=0.13445298115794474 +INFO:root: +[Valid][15][UTT][ MSE = 0.2634 | LCC = 0.8482 | SRCC = 0.8458 ] +INFO:root:Average loss=0.12418157912981816 +INFO:root: +[Valid][16][UTT][ MSE = 0.2475 | LCC = 0.8407 | SRCC = 0.8409 ] +INFO:root:Average loss=0.12290711586291973 +INFO:root: +[Valid][17][UTT][ MSE = 0.2647 | LCC = 0.8409 | SRCC = 0.8404 ] +INFO:root:Average loss=0.11860500018184002 +INFO:root: +[Valid][18][UTT][ MSE = 0.2503 | LCC = 0.8415 | SRCC = 0.8404 ] +INFO:root:Average loss=0.11121973232963146 +INFO:root: +[Valid][19][UTT][ MSE = 0.2515 | LCC = 0.8459 | SRCC = 0.8454 ] +INFO:root:Average loss=0.10697556124665798 +INFO:root: +[Valid][20][UTT][ MSE = 0.2419 | LCC = 0.8468 | SRCC = 0.8456 ] +INFO:root:Average loss=0.10454285264206238 +INFO:root: +[Valid][21][UTT][ MSE = 0.2459 | LCC = 0.8405 | SRCC = 0.8419 ] +INFO:root:Average loss=0.10595151581443273 +INFO:root: +[Valid][22][UTT][ MSE = 0.2416 | LCC = 0.8467 | SRCC = 0.8461 ] +INFO:root:Average loss=0.10001899206485504 +INFO:root: +[Valid][23][UTT][ MSE = 0.2473 | LCC = 0.8400 | SRCC = 0.8422 ] +INFO:root:Average loss=0.0977102655153244 +INFO:root: +[Valid][24][UTT][ MSE = 0.2512 | LCC = 0.8394 | SRCC = 0.8403 ] +INFO:root:Average loss=0.08969869010914595 +INFO:root: +[Valid][25][UTT][ MSE = 0.2483 | LCC = 0.8433 | SRCC = 0.8423 ] +INFO:root:Average loss=0.09513216761824413 +INFO:root: +[Valid][26][UTT][ MSE = 0.2563 | LCC = 0.8465 | SRCC = 0.8456 ] +INFO:root:Average loss=0.08799929510897551 +INFO:root: +[Valid][27][UTT][ MSE = 0.2372 | LCC = 0.8471 | SRCC = 0.8481 ] +INFO:root:Average loss=0.0859949563940366 +INFO:root: +[Valid][28][UTT][ MSE = 0.2435 | LCC = 0.8470 | SRCC = 0.8456 ] +INFO:root:Average loss=0.08100173603265713 +INFO:root: +[Valid][29][UTT][ MSE = 0.2367 | LCC = 0.8474 | SRCC = 0.8470 ] +INFO:root:Average loss=0.0784848085962809 +INFO:root: +[Valid][30][UTT][ MSE = 0.2515 | LCC = 0.8458 | SRCC = 0.8459 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2747 | LCC = 0.8556 | SRCC = 0.8556 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ba869328afbf96710555b9ecec75d5665f69af0b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2bd905b156113b0f4cc34bfc5dd3397ffc485f0f7937822efd6290dd5f3f288e +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..9608370b865523b8b6e4d2d547096cb75a4407f0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2ded82ff2c6f2a7b128d697beca88192ad16134202cd6595b1f5923c71c2142 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4154b60b3258a7e0ce42d5665e98133e613ba0a7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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=0.5119304431554599 +INFO:root: +[Valid][0][UTT][ MSE = 0.3659 | LCC = 0.7643 | SRCC = 0.7663 ] +INFO:root:Average loss=0.2997244963279137 +INFO:root: +[Valid][1][UTT][ MSE = 0.3330 | LCC = 0.7996 | SRCC = 0.8016 ] +INFO:root:Average loss=0.27291951271203846 +INFO:root: +[Valid][2][UTT][ MSE = 0.3020 | LCC = 0.8126 | SRCC = 0.8129 ] +INFO:root:Average loss=0.2518412071542862 +INFO:root: +[Valid][3][UTT][ MSE = 0.3356 | LCC = 0.8146 | SRCC = 0.8177 ] +INFO:root:Average loss=0.22543862204139048 +INFO:root: +[Valid][4][UTT][ MSE = 0.2906 | LCC = 0.8253 | SRCC = 0.8249 ] +INFO:root:Average loss=0.22898551286795202 +INFO:root: +[Valid][5][UTT][ MSE = 0.2751 | LCC = 0.8287 | SRCC = 0.8275 ] +INFO:root:Average loss=0.20736417728356826 +INFO:root: +[Valid][6][UTT][ MSE = 0.2620 | LCC = 0.8393 | SRCC = 0.8383 ] +INFO:root:Average loss=0.19834035854691115 +INFO:root: +[Valid][7][UTT][ MSE = 0.2692 | LCC = 0.8355 | SRCC = 0.8355 ] +INFO:root:Average loss=0.18353565132770783 +INFO:root: +[Valid][8][UTT][ MSE = 0.2519 | LCC = 0.8416 | SRCC = 0.8392 ] +INFO:root:Average loss=0.1782539505033921 +INFO:root: +[Valid][9][UTT][ MSE = 0.2650 | LCC = 0.8438 | SRCC = 0.8407 ] +INFO:root:Average loss=0.16147862718655512 +INFO:root: +[Valid][10][UTT][ MSE = 0.2699 | LCC = 0.8437 | SRCC = 0.8401 ] +INFO:root:Average loss=0.16192402155735555 +INFO:root: +[Valid][11][UTT][ MSE = 0.2448 | LCC = 0.8419 | SRCC = 0.8417 ] +INFO:root:Average loss=0.15371757249037424 +INFO:root: +[Valid][12][UTT][ MSE = 0.2617 | LCC = 0.8409 | SRCC = 0.8348 ] +INFO:root:Average loss=0.14456071437169343 +INFO:root: +[Valid][13][UTT][ MSE = 0.2546 | LCC = 0.8432 | SRCC = 0.8381 ] +INFO:root:Average loss=0.14000531180928916 +INFO:root: +[Valid][14][UTT][ MSE = 0.2723 | LCC = 0.8461 | SRCC = 0.8415 ] +INFO:root:Average loss=0.1446456998013533 +INFO:root: +[Valid][15][UTT][ MSE = 0.2626 | LCC = 0.8484 | SRCC = 0.8453 ] +INFO:root:Average loss=0.13735919216504464 +INFO:root: +[Valid][16][UTT][ MSE = 0.2446 | LCC = 0.8414 | SRCC = 0.8420 ] +INFO:root:Average loss=0.12453650377499752 +INFO:root: +[Valid][17][UTT][ MSE = 0.2453 | LCC = 0.8483 | SRCC = 0.8455 ] +INFO:root:Average loss=0.12255396120823346 +INFO:root: +[Valid][18][UTT][ MSE = 0.2441 | LCC = 0.8479 | SRCC = 0.8465 ] +INFO:root:Average loss=0.12004041719513062 +INFO:root: +[Valid][19][UTT][ MSE = 0.2812 | LCC = 0.8431 | SRCC = 0.8422 ] +INFO:root:Average loss=0.11091009556100918 +INFO:root: +[Valid][20][UTT][ MSE = 0.2504 | LCC = 0.8475 | SRCC = 0.8445 ] +INFO:root:Average loss=0.11170638276216312 +INFO:root: +[Valid][21][UTT][ MSE = 0.2595 | LCC = 0.8460 | SRCC = 0.8436 ] +INFO:root:Average loss=0.1006958466500808 +INFO:root: +[Valid][22][UTT][ MSE = 0.2424 | LCC = 0.8443 | SRCC = 0.8415 ] +INFO:root:Average loss=0.10102860696422748 +INFO:root: +[Valid][23][UTT][ MSE = 0.2444 | LCC = 0.8505 | SRCC = 0.8475 ] +INFO:root:Average loss=0.10012606994654888 +INFO:root: +[Valid][24][UTT][ MSE = 0.2768 | LCC = 0.8427 | SRCC = 0.8438 ] +INFO:root:Average loss=0.09473842611679664 +INFO:root: +[Valid][25][UTT][ MSE = 0.2425 | LCC = 0.8471 | SRCC = 0.8446 ] +INFO:root:Average loss=0.09198780133365056 +INFO:root: +[Valid][26][UTT][ MSE = 0.2516 | LCC = 0.8481 | SRCC = 0.8444 ] +INFO:root:Average loss=0.08251088026624459 +INFO:root: +[Valid][27][UTT][ MSE = 0.2403 | LCC = 0.8494 | SRCC = 0.8457 ] +INFO:root:Average loss=0.0839204575197819 +INFO:root: +[Valid][28][UTT][ MSE = 0.2642 | LCC = 0.8423 | SRCC = 0.8420 ] +INFO:root:Average loss=0.08656372000964788 +INFO:root: +[Valid][29][UTT][ MSE = 0.2382 | LCC = 0.8506 | SRCC = 0.8495 ] +INFO:root:Average loss=0.07958602379911985 +INFO:root: +[Valid][30][UTT][ MSE = 0.2439 | LCC = 0.8485 | SRCC = 0.8450 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2413 | LCC = 0.8628 | SRCC = 0.8638 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..47a03239ce85270a3efa4d690a0ffa11e7dac3d2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57717547ff4a361965fa57af818ac4baeac546db369973d906b961d6c81fb9af +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..8c464c8180ff8cd825a5ecb481e8c8ac53c46e80 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:027b668bb34757bce840a75d0530cb085563febdba86e5503144968baddd57ec +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d4e454706a0a78cd0b240968bf28077ea98f01af --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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=0.6418334520780123 +INFO:root: +[Valid][0][UTT][ MSE = 0.4368 | LCC = 0.7096 | SRCC = 0.7195 ] +INFO:root:Average loss=0.31574674390065366 +INFO:root: +[Valid][1][UTT][ MSE = 0.3206 | LCC = 0.7915 | SRCC = 0.7952 ] +INFO:root:Average loss=0.27041512231032055 +INFO:root: +[Valid][2][UTT][ MSE = 0.3104 | LCC = 0.8031 | SRCC = 0.8087 ] +INFO:root:Average loss=0.24709904193878174 +INFO:root: +[Valid][3][UTT][ MSE = 0.2880 | LCC = 0.8172 | SRCC = 0.8199 ] +INFO:root:Average loss=0.21710066000620523 +INFO:root: +[Valid][4][UTT][ MSE = 0.2766 | LCC = 0.8283 | SRCC = 0.8260 ] +INFO:root:Average loss=0.2095627207786609 +INFO:root: +[Valid][5][UTT][ MSE = 0.2795 | LCC = 0.8218 | SRCC = 0.8291 ] +INFO:root:Average loss=0.19018907864124346 +INFO:root: +[Valid][6][UTT][ MSE = 0.2674 | LCC = 0.8328 | SRCC = 0.8330 ] +INFO:root:Average loss=0.18640683877926606 +INFO:root: +[Valid][7][UTT][ MSE = 0.2778 | LCC = 0.8306 | SRCC = 0.8319 ] +INFO:root:Average loss=0.17895266728905532 +INFO:root: +[Valid][8][UTT][ MSE = 0.2622 | LCC = 0.8358 | SRCC = 0.8373 ] +INFO:root:Average loss=0.16967898378005394 +INFO:root: +[Valid][9][UTT][ MSE = 0.2779 | LCC = 0.8384 | SRCC = 0.8382 ] +INFO:root:Average loss=0.15764866712001654 +INFO:root: +[Valid][10][UTT][ MSE = 0.2752 | LCC = 0.8409 | SRCC = 0.8409 ] +INFO:root:Average loss=0.15517678293280113 +INFO:root: +[Valid][11][UTT][ MSE = 0.2658 | LCC = 0.8394 | SRCC = 0.8388 ] +INFO:root:Average loss=0.14758101153450134 +INFO:root: +[Valid][12][UTT][ MSE = 0.2560 | LCC = 0.8394 | SRCC = 0.8394 ] +INFO:root:Average loss=0.1364254416563572 +INFO:root: +[Valid][13][UTT][ MSE = 0.2686 | LCC = 0.8401 | SRCC = 0.8395 ] +INFO:root:Average loss=0.13244806908262083 +INFO:root: +[Valid][14][UTT][ MSE = 0.2466 | LCC = 0.8428 | SRCC = 0.8423 ] +INFO:root:Average loss=0.12509061778203034 +INFO:root: +[Valid][15][UTT][ MSE = 0.2461 | LCC = 0.8437 | SRCC = 0.8438 ] +INFO:root:Average loss=0.11863764251271884 +INFO:root: +[Valid][16][UTT][ MSE = 0.2476 | LCC = 0.8414 | SRCC = 0.8402 ] +INFO:root:Average loss=0.11598269402598724 +INFO:root: +[Valid][17][UTT][ MSE = 0.2620 | LCC = 0.8361 | SRCC = 0.8368 ] +INFO:root:Average loss=0.12367690937259258 +INFO:root: +[Valid][18][UTT][ MSE = 0.2473 | LCC = 0.8435 | SRCC = 0.8424 ] +INFO:root:Average loss=0.10613737675624016 +INFO:root: +[Valid][19][UTT][ MSE = 0.2506 | LCC = 0.8440 | SRCC = 0.8420 ] +INFO:root:Average loss=0.1025849927503329 +INFO:root: +[Valid][20][UTT][ MSE = 0.2427 | LCC = 0.8454 | SRCC = 0.8455 ] +INFO:root:Average loss=0.09619503859908153 +INFO:root: +[Valid][21][UTT][ MSE = 0.2421 | LCC = 0.8446 | SRCC = 0.8426 ] +INFO:root:Average loss=0.09757718969231997 +INFO:root: +[Valid][22][UTT][ MSE = 0.2645 | LCC = 0.8445 | SRCC = 0.8445 ] +INFO:root:Average loss=0.09456389789015819 +INFO:root: +[Valid][23][UTT][ MSE = 0.2428 | LCC = 0.8448 | SRCC = 0.8443 ] +INFO:root:Average loss=0.09095317440537307 +INFO:root: +[Valid][24][UTT][ MSE = 0.2423 | LCC = 0.8427 | SRCC = 0.8408 ] +INFO:root:Average loss=0.09447970594733189 +INFO:root: +[Valid][25][UTT][ MSE = 0.2448 | LCC = 0.8435 | SRCC = 0.8439 ] +INFO:root:Average loss=0.08549970899445888 +INFO:root: +[Valid][26][UTT][ MSE = 0.2459 | LCC = 0.8456 | SRCC = 0.8453 ] +INFO:root:Average loss=0.09048202409385106 +INFO:root: +[Valid][27][UTT][ MSE = 0.2446 | LCC = 0.8416 | SRCC = 0.8409 ] +INFO:root:Average loss=0.07691479918475334 +INFO:root: +[Valid][28][UTT][ MSE = 0.2450 | LCC = 0.8419 | SRCC = 0.8396 ] +INFO:root:Average loss=0.07752726117196755 +INFO:root: +[Valid][29][UTT][ MSE = 0.2602 | LCC = 0.8384 | SRCC = 0.8385 ] +INFO:root:Average loss=0.07609479057674225 +INFO:root: +[Valid][30][UTT][ MSE = 0.2427 | LCC = 0.8413 | SRCC = 0.8407 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2593 | LCC = 0.8527 | SRCC = 0.8550 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ae8255600c6e851a1ddbcca2fc5f25ae4da01b02 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:385ee8e63ff345452fbc1c66b5d292c7f8f80ed820ae714db55f72d93ca2df89 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..9924530ab75078fd015da0c51c4df3576915133e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86fcd72fba4775947a8da8645ee0a4307c17d20a7d1e29c5a29eb2b35cf34f2a +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..26ce1e1c4a6b877db5b1f3ee9570150187dcb565 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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=0.4932205428679784 +INFO:root: +[Valid][0][UTT][ MSE = 0.4044 | LCC = 0.7488 | SRCC = 0.7574 ] +INFO:root:Average loss=0.31500307757120866 +INFO:root: +[Valid][1][UTT][ MSE = 0.3340 | LCC = 0.7910 | SRCC = 0.7955 ] +INFO:root:Average loss=0.27896627745567226 +INFO:root: +[Valid][2][UTT][ MSE = 0.3010 | LCC = 0.8127 | SRCC = 0.8141 ] +INFO:root:Average loss=0.250183981580612 +INFO:root: +[Valid][3][UTT][ MSE = 0.2990 | LCC = 0.8176 | SRCC = 0.8200 ] +INFO:root:Average loss=0.22998157296425256 +INFO:root: +[Valid][4][UTT][ MSE = 0.2744 | LCC = 0.8267 | SRCC = 0.8281 ] +INFO:root:Average loss=0.22712785311234304 +INFO:root: +[Valid][5][UTT][ MSE = 0.2782 | LCC = 0.8324 | SRCC = 0.8321 ] +INFO:root:Average loss=0.20589212481027994 +INFO:root: +[Valid][6][UTT][ MSE = 0.2940 | LCC = 0.8356 | SRCC = 0.8342 ] +INFO:root:Average loss=0.19384761756429306 +INFO:root: +[Valid][7][UTT][ MSE = 0.2661 | LCC = 0.8341 | SRCC = 0.8312 ] +INFO:root:Average loss=0.19826353140748465 +INFO:root: +[Valid][8][UTT][ MSE = 0.2948 | LCC = 0.8267 | SRCC = 0.8305 ] +INFO:root:Average loss=0.17846870412811255 +INFO:root: +[Valid][9][UTT][ MSE = 0.2635 | LCC = 0.8336 | SRCC = 0.8356 ] +INFO:root:Average loss=0.16995830671527448 +INFO:root: +[Valid][10][UTT][ MSE = 0.2621 | LCC = 0.8361 | SRCC = 0.8362 ] +INFO:root:Average loss=0.16248943475194466 +INFO:root: +[Valid][11][UTT][ MSE = 0.2694 | LCC = 0.8396 | SRCC = 0.8381 ] +INFO:root:Average loss=0.1571786153392914 +INFO:root: +[Valid][12][UTT][ MSE = 0.2565 | LCC = 0.8405 | SRCC = 0.8397 ] +INFO:root:Average loss=0.14827534346244273 +INFO:root: +[Valid][13][UTT][ MSE = 0.2636 | LCC = 0.8353 | SRCC = 0.8329 ] +INFO:root:Average loss=0.1495579670255001 +INFO:root: +[Valid][14][UTT][ MSE = 0.2611 | LCC = 0.8374 | SRCC = 0.8370 ] +INFO:root:Average loss=0.13987739498798663 +INFO:root: +[Valid][15][UTT][ MSE = 0.2527 | LCC = 0.8395 | SRCC = 0.8397 ] +INFO:root:Average loss=0.14543951245454642 +INFO:root: +[Valid][16][UTT][ MSE = 0.2621 | LCC = 0.8382 | SRCC = 0.8382 ] +INFO:root:Average loss=0.12457984131880295 +INFO:root: +[Valid][17][UTT][ MSE = 0.2582 | LCC = 0.8388 | SRCC = 0.8399 ] +INFO:root:Average loss=0.12391242489982875 +INFO:root: +[Valid][18][UTT][ MSE = 0.2610 | LCC = 0.8386 | SRCC = 0.8378 ] +INFO:root:Average loss=0.11367042897603451 +INFO:root: +[Valid][19][UTT][ MSE = 0.2804 | LCC = 0.8362 | SRCC = 0.8357 ] +INFO:root:Average loss=0.11454035790684895 +INFO:root: +[Valid][20][UTT][ MSE = 0.2603 | LCC = 0.8396 | SRCC = 0.8395 ] +INFO:root:Average loss=0.12032755254170834 +INFO:root: +[Valid][21][UTT][ MSE = 0.2500 | LCC = 0.8408 | SRCC = 0.8402 ] +INFO:root:Average loss=0.10656944910685222 +INFO:root: +[Valid][22][UTT][ MSE = 0.2627 | LCC = 0.8430 | SRCC = 0.8423 ] +INFO:root:Average loss=0.10325921651644585 +INFO:root: +[Valid][23][UTT][ MSE = 0.2500 | LCC = 0.8413 | SRCC = 0.8403 ] +INFO:root:Average loss=0.09834023154316804 +INFO:root: +[Valid][24][UTT][ MSE = 0.2789 | LCC = 0.8415 | SRCC = 0.8408 ] +INFO:root:Average loss=0.0977439297697483 +INFO:root: +[Valid][25][UTT][ MSE = 0.2553 | LCC = 0.8390 | SRCC = 0.8354 ] +INFO:root:Average loss=0.09807379476917095 +INFO:root: +[Valid][26][UTT][ MSE = 0.2576 | LCC = 0.8422 | SRCC = 0.8422 ] +INFO:root:Average loss=0.0893989021006303 +INFO:root: +[Valid][27][UTT][ MSE = 0.2729 | LCC = 0.8417 | SRCC = 0.8412 ] +INFO:root:Average loss=0.0924553659099799 +INFO:root: +[Valid][28][UTT][ MSE = 0.2541 | LCC = 0.8440 | SRCC = 0.8431 ] +INFO:root:Average loss=0.08636994946461457 +INFO:root: +[Valid][29][UTT][ MSE = 0.2450 | LCC = 0.8435 | SRCC = 0.8441 ] +INFO:root:Average loss=0.0773644588696651 +INFO:root: +[Valid][30][UTT][ MSE = 0.2466 | LCC = 0.8446 | SRCC = 0.8436 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2575 | LCC = 0.8546 | SRCC = 0.8555 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..c519fd8dee569e83cbef6b02a541e07096334a58 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d69ad429122c58255cfe729edca9ac950429e37a83ca8add2d61c97a85882e08 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7430f449fe7e7349c310a6a5cb2b940ad84382b0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9532fdc56854c06641c2bb8b51abff277c85b4a02574ccfde72db7a1378f108 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..5e1be512d6f2559fd0e6eb7dc9f638f31e9d1cb4 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer0/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=0.5558533851916974 +INFO:root: +[Valid][0][UTT][ MSE = 0.4102 | LCC = 0.7454 | SRCC = 0.7460 ] +INFO:root:Average loss=0.314694028825332 +INFO:root: +[Valid][1][UTT][ MSE = 0.3260 | LCC = 0.7946 | SRCC = 0.7945 ] +INFO:root:Average loss=0.26737395788614565 +INFO:root: +[Valid][2][UTT][ MSE = 0.3133 | LCC = 0.8127 | SRCC = 0.8141 ] +INFO:root:Average loss=0.24034474904720599 +INFO:root: +[Valid][3][UTT][ MSE = 0.2903 | LCC = 0.8191 | SRCC = 0.8218 ] +INFO:root:Average loss=0.22300590727573785 +INFO:root: +[Valid][4][UTT][ MSE = 0.2866 | LCC = 0.8276 | SRCC = 0.8268 ] +INFO:root:Average loss=0.21366133502660653 +INFO:root: +[Valid][5][UTT][ MSE = 0.2773 | LCC = 0.8271 | SRCC = 0.8275 ] +INFO:root:Average loss=0.19769211189869124 +INFO:root: +[Valid][6][UTT][ MSE = 0.2896 | LCC = 0.8281 | SRCC = 0.8263 ] +INFO:root:Average loss=0.20410021127034456 +INFO:root: +[Valid][7][UTT][ MSE = 0.2799 | LCC = 0.8287 | SRCC = 0.8292 ] +INFO:root:Average loss=0.18709636871249247 +INFO:root: +[Valid][8][UTT][ MSE = 0.2716 | LCC = 0.8319 | SRCC = 0.8305 ] +INFO:root:Average loss=0.1720253894917476 +INFO:root: +[Valid][9][UTT][ MSE = 0.2793 | LCC = 0.8320 | SRCC = 0.8295 ] +INFO:root:Average loss=0.16957806346890253 +INFO:root: +[Valid][10][UTT][ MSE = 0.2819 | LCC = 0.8286 | SRCC = 0.8303 ] +INFO:root:Average loss=0.15952460076182318 +INFO:root: +[Valid][11][UTT][ MSE = 0.2718 | LCC = 0.8305 | SRCC = 0.8326 ] +INFO:root:Average loss=0.15475672961045533 +INFO:root: +[Valid][12][UTT][ MSE = 0.2771 | LCC = 0.8305 | SRCC = 0.8320 ] +INFO:root:Average loss=0.14190535915967745 +INFO:root: +[Valid][13][UTT][ MSE = 0.2670 | LCC = 0.8378 | SRCC = 0.8346 ] +INFO:root:Average loss=0.14454697607419428 +INFO:root: +[Valid][14][UTT][ MSE = 0.2610 | LCC = 0.8367 | SRCC = 0.8349 ] +INFO:root:Average loss=0.13601268207033476 +INFO:root: +[Valid][15][UTT][ MSE = 0.2592 | LCC = 0.8366 | SRCC = 0.8337 ] +INFO:root:Average loss=0.1324301183414765 +INFO:root: +[Valid][16][UTT][ MSE = 0.2697 | LCC = 0.8365 | SRCC = 0.8357 ] +INFO:root:Average loss=0.1263054269246566 +INFO:root: +[Valid][17][UTT][ MSE = 0.2689 | LCC = 0.8347 | SRCC = 0.8320 ] +INFO:root:Average loss=0.12044030781357716 +INFO:root: +[Valid][18][UTT][ MSE = 0.2685 | LCC = 0.8339 | SRCC = 0.8334 ] +INFO:root:Average loss=0.11341449637443592 +INFO:root: +[Valid][19][UTT][ MSE = 0.2599 | LCC = 0.8299 | SRCC = 0.8293 ] +INFO:root:Average loss=0.11230582008377099 +INFO:root: +[Valid][20][UTT][ MSE = 0.2601 | LCC = 0.8348 | SRCC = 0.8342 ] +INFO:root:Average loss=0.11140689440071583 +INFO:root: +[Valid][21][UTT][ MSE = 0.2672 | LCC = 0.8340 | SRCC = 0.8304 ] +INFO:root:Average loss=0.10007085737127525 +INFO:root: +[Valid][22][UTT][ MSE = 0.2646 | LCC = 0.8361 | SRCC = 0.8318 ] +INFO:root:Average loss=0.09780419359986599 +INFO:root: +[Valid][23][UTT][ MSE = 0.2714 | LCC = 0.8338 | SRCC = 0.8327 ] +INFO:root:Average loss=0.09931402032574017 +INFO:root: +[Valid][24][UTT][ MSE = 0.2510 | LCC = 0.8394 | SRCC = 0.8378 ] +INFO:root:Average loss=0.09226875642362313 +INFO:root: +[Valid][25][UTT][ MSE = 0.2564 | LCC = 0.8349 | SRCC = 0.8330 ] +INFO:root:Average loss=0.09218659170736106 +INFO:root: +[Valid][26][UTT][ MSE = 0.2839 | LCC = 0.8306 | SRCC = 0.8285 ] +INFO:root:Average loss=0.08850150120755036 +INFO:root: +[Valid][27][UTT][ MSE = 0.2644 | LCC = 0.8346 | SRCC = 0.8333 ] +INFO:root:Average loss=0.0818972688836929 +INFO:root: +[Valid][28][UTT][ MSE = 0.2450 | LCC = 0.8398 | SRCC = 0.8383 ] +INFO:root:Average loss=0.07843304248765492 +INFO:root: +[Valid][29][UTT][ MSE = 0.2512 | LCC = 0.8375 | SRCC = 0.8356 ] +INFO:root:Average loss=0.08115213216306308 +INFO:root: +[Valid][30][UTT][ MSE = 0.2465 | LCC = 0.8400 | SRCC = 0.8383 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2364 | LCC = 0.8572 | SRCC = 0.8552 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..03c25c6868c4e78cd3f9d058bb7d79cd9801debf --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b3ff41150441c14f5cf8b528ff5f5f90c8b3162bd47885b24e38a5650a51896f +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..71210aaa6b5b5250952c7cc1f863875a592b6b22 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8556f8712ae5440f2d587431fca5dd25e43180d66cf312bdfa81758ff378cb7 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..61b73a09c5519a47adac9493cd6a3b4f7f598ba5 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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=0.47028993929807955 +INFO:root: +[Valid][0][UTT][ MSE = 0.3445 | LCC = 0.7748 | SRCC = 0.7815 ] +INFO:root:Average loss=0.2739105142461948 +INFO:root: +[Valid][1][UTT][ MSE = 0.2877 | LCC = 0.8180 | SRCC = 0.8221 ] +INFO:root:Average loss=0.22821743289629617 +INFO:root: +[Valid][2][UTT][ MSE = 0.2740 | LCC = 0.8368 | SRCC = 0.8382 ] +INFO:root:Average loss=0.20952318417720306 +INFO:root: +[Valid][3][UTT][ MSE = 0.2628 | LCC = 0.8412 | SRCC = 0.8454 ] +INFO:root:Average loss=0.19267830959497353 +INFO:root: +[Valid][4][UTT][ MSE = 0.2504 | LCC = 0.8489 | SRCC = 0.8496 ] +INFO:root:Average loss=0.1756745854822489 +INFO:root: +[Valid][5][UTT][ MSE = 0.2369 | LCC = 0.8512 | SRCC = 0.8503 ] +INFO:root:Average loss=0.16747307720092627 +INFO:root: +[Valid][6][UTT][ MSE = 0.2337 | LCC = 0.8542 | SRCC = 0.8523 ] +INFO:root:Average loss=0.1567717747619519 +INFO:root: +[Valid][7][UTT][ MSE = 0.2452 | LCC = 0.8537 | SRCC = 0.8518 ] +INFO:root:Average loss=0.14242968479028115 +INFO:root: +[Valid][8][UTT][ MSE = 0.2309 | LCC = 0.8551 | SRCC = 0.8547 ] +INFO:root:Average loss=0.13862422041786024 +INFO:root: +[Valid][9][UTT][ MSE = 0.2269 | LCC = 0.8586 | SRCC = 0.8561 ] +INFO:root:Average loss=0.13333122871625117 +INFO:root: +[Valid][10][UTT][ MSE = 0.2257 | LCC = 0.8561 | SRCC = 0.8561 ] +INFO:root:Average loss=0.12466936119091816 +INFO:root: +[Valid][11][UTT][ MSE = 0.2593 | LCC = 0.8512 | SRCC = 0.8520 ] +INFO:root:Average loss=0.12887407218416533 +INFO:root: +[Valid][12][UTT][ MSE = 0.2306 | LCC = 0.8561 | SRCC = 0.8549 ] +INFO:root:Average loss=0.11457514495421679 +INFO:root: +[Valid][13][UTT][ MSE = 0.2549 | LCC = 0.8503 | SRCC = 0.8499 ] +INFO:root:Average loss=0.11050517656482182 +INFO:root: +[Valid][14][UTT][ MSE = 0.2467 | LCC = 0.8578 | SRCC = 0.8574 ] +INFO:root:Average loss=0.10088201301793258 +INFO:root: +[Valid][15][UTT][ MSE = 0.2266 | LCC = 0.8566 | SRCC = 0.8554 ] +INFO:root:Average loss=0.09751168657571842 +INFO:root: +[Valid][16][UTT][ MSE = 0.2328 | LCC = 0.8579 | SRCC = 0.8569 ] +INFO:root:Average loss=0.09755696217792156 +INFO:root: +[Valid][17][UTT][ MSE = 0.2585 | LCC = 0.8578 | SRCC = 0.8564 ] +INFO:root:Average loss=0.0927538542697827 +INFO:root: +[Valid][18][UTT][ MSE = 0.2304 | LCC = 0.8580 | SRCC = 0.8559 ] +INFO:root:Average loss=0.09212391761442025 +INFO:root: +[Valid][19][UTT][ MSE = 0.2353 | LCC = 0.8591 | SRCC = 0.8578 ] +INFO:root:Average loss=0.08220753178764613 +INFO:root: +[Valid][20][UTT][ MSE = 0.2264 | LCC = 0.8584 | SRCC = 0.8560 ] +INFO:root:Average loss=0.07576050079212739 +INFO:root: +[Valid][21][UTT][ MSE = 0.2404 | LCC = 0.8527 | SRCC = 0.8515 ] +INFO:root:Average loss=0.0761243208574179 +INFO:root: +[Valid][22][UTT][ MSE = 0.2224 | LCC = 0.8559 | SRCC = 0.8544 ] +INFO:root:Average loss=0.07392243424860331 +INFO:root: +[Valid][23][UTT][ MSE = 0.2193 | LCC = 0.8603 | SRCC = 0.8578 ] +INFO:root:Average loss=0.07203307652320617 +INFO:root: +[Valid][24][UTT][ MSE = 0.2243 | LCC = 0.8592 | SRCC = 0.8572 ] +INFO:root:Average loss=0.06811546352811348 +INFO:root: +[Valid][25][UTT][ MSE = 0.2224 | LCC = 0.8572 | SRCC = 0.8544 ] +INFO:root:Average loss=0.06275242337813744 +INFO:root: +[Valid][26][UTT][ MSE = 0.2222 | LCC = 0.8588 | SRCC = 0.8565 ] +INFO:root:Average loss=0.061504712710395835 +INFO:root: +[Valid][27][UTT][ MSE = 0.2274 | LCC = 0.8578 | SRCC = 0.8560 ] +INFO:root:Average loss=0.06154036913544704 +INFO:root: +[Valid][28][UTT][ MSE = 0.2242 | LCC = 0.8563 | SRCC = 0.8559 ] +INFO:root:Average loss=0.057513228545968346 +INFO:root: +[Valid][29][UTT][ MSE = 0.2190 | LCC = 0.8607 | SRCC = 0.8584 ] +INFO:root:Average loss=0.058831147610759124 +INFO:root: +[Valid][30][UTT][ MSE = 0.2225 | LCC = 0.8603 | SRCC = 0.8587 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2346 | LCC = 0.8634 | SRCC = 0.8642 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..da396255db22aaf05893904dab645e845802a9aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb358e00b84b4c1c03829f36f0533850e89f69f4e64c5e17d1247d4e14500e38 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..ec786beb0fea52eec6e9c5a745cc6abc1059b446 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89c78b8f9618fa2862a7b892030cd2b1d3fdc8729062a581b7d729ed4cbb268a +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..60e789db0d44bb1d63bbd78b58e1b4ac85bd9d70 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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=0.5035896314642369 +INFO:root: +[Valid][0][UTT][ MSE = 0.3375 | LCC = 0.7812 | SRCC = 0.7848 ] +INFO:root:Average loss=0.26980930203810716 +INFO:root: +[Valid][1][UTT][ MSE = 0.2951 | LCC = 0.8277 | SRCC = 0.8276 ] +INFO:root:Average loss=0.23443900316189498 +INFO:root: +[Valid][2][UTT][ MSE = 0.2670 | LCC = 0.8430 | SRCC = 0.8435 ] +INFO:root:Average loss=0.20995070899908358 +INFO:root: +[Valid][3][UTT][ MSE = 0.2595 | LCC = 0.8485 | SRCC = 0.8490 ] +INFO:root:Average loss=0.2012450750439595 +INFO:root: +[Valid][4][UTT][ MSE = 0.2850 | LCC = 0.8509 | SRCC = 0.8500 ] +INFO:root:Average loss=0.19862034964637879 +INFO:root: +[Valid][5][UTT][ MSE = 0.2570 | LCC = 0.8550 | SRCC = 0.8545 ] +INFO:root:Average loss=0.1742247729920424 +INFO:root: +[Valid][6][UTT][ MSE = 0.2601 | LCC = 0.8581 | SRCC = 0.8584 ] +INFO:root:Average loss=0.1544427909912207 +INFO:root: +[Valid][7][UTT][ MSE = 0.2469 | LCC = 0.8591 | SRCC = 0.8582 ] +INFO:root:Average loss=0.14937100960658148 +INFO:root: +[Valid][8][UTT][ MSE = 0.2330 | LCC = 0.8585 | SRCC = 0.8562 ] +INFO:root:Average loss=0.14613696034902182 +INFO:root: +[Valid][9][UTT][ MSE = 0.2617 | LCC = 0.8623 | SRCC = 0.8605 ] +INFO:root:Average loss=0.14194764349705133 +INFO:root: +[Valid][10][UTT][ MSE = 0.2373 | LCC = 0.8626 | SRCC = 0.8625 ] +INFO:root:Average loss=0.12733262395247436 +INFO:root: +[Valid][11][UTT][ MSE = 0.2633 | LCC = 0.8588 | SRCC = 0.8593 ] +INFO:root:Average loss=0.12238265870091243 +INFO:root: +[Valid][12][UTT][ MSE = 0.2877 | LCC = 0.8571 | SRCC = 0.8580 ] +INFO:root:Average loss=0.1164811735925002 +INFO:root: +[Valid][13][UTT][ MSE = 0.2529 | LCC = 0.8609 | SRCC = 0.8605 ] +INFO:root:Average loss=0.10788544792777453 +INFO:root: +[Valid][14][UTT][ MSE = 0.2462 | LCC = 0.8609 | SRCC = 0.8602 ] +INFO:root:Average loss=0.10067938678921798 +INFO:root: +[Valid][15][UTT][ MSE = 0.2451 | LCC = 0.8599 | SRCC = 0.8600 ] +INFO:root:Average loss=0.09618646178681117 +INFO:root: +[Valid][16][UTT][ MSE = 0.2544 | LCC = 0.8629 | SRCC = 0.8632 ] +INFO:root:Average loss=0.09215998558852917 +INFO:root: +[Valid][17][UTT][ MSE = 0.2313 | LCC = 0.8635 | SRCC = 0.8633 ] +INFO:root:Average loss=0.09373146338531604 +INFO:root: +[Valid][18][UTT][ MSE = 0.2685 | LCC = 0.8604 | SRCC = 0.8614 ] +INFO:root:Average loss=0.09518104199415599 +INFO:root: +[Valid][19][UTT][ MSE = 0.2764 | LCC = 0.8597 | SRCC = 0.8604 ] +INFO:root:Average loss=0.08395814781005566 +INFO:root: +[Valid][20][UTT][ MSE = 0.2221 | LCC = 0.8646 | SRCC = 0.8649 ] +INFO:root:Average loss=0.08192523306187911 +INFO:root: +[Valid][21][UTT][ MSE = 0.2506 | LCC = 0.8615 | SRCC = 0.8609 ] +INFO:root:Average loss=0.07688132243660781 +INFO:root: +[Valid][22][UTT][ MSE = 0.2287 | LCC = 0.8604 | SRCC = 0.8601 ] +INFO:root:Average loss=0.07188534411864403 +INFO:root: +[Valid][23][UTT][ MSE = 0.2489 | LCC = 0.8610 | SRCC = 0.8595 ] +INFO:root:Average loss=0.0704778078943491 +INFO:root: +[Valid][24][UTT][ MSE = 0.2252 | LCC = 0.8592 | SRCC = 0.8588 ] +INFO:root:Average loss=0.06671010253903194 +INFO:root: +[Valid][25][UTT][ MSE = 0.2485 | LCC = 0.8604 | SRCC = 0.8603 ] +INFO:root:Average loss=0.07177903923468712 +INFO:root: +[Valid][26][UTT][ MSE = 0.2361 | LCC = 0.8616 | SRCC = 0.8606 ] +INFO:root:Average loss=0.06538005550511372 +INFO:root: +[Valid][27][UTT][ MSE = 0.2204 | LCC = 0.8644 | SRCC = 0.8636 ] +INFO:root:Average loss=0.06196800905924577 +INFO:root: +[Valid][28][UTT][ MSE = 0.2220 | LCC = 0.8640 | SRCC = 0.8628 ] +INFO:root:Average loss=0.058270050738102354 +INFO:root: +[Valid][29][UTT][ MSE = 0.2195 | LCC = 0.8648 | SRCC = 0.8649 ] +INFO:root:Average loss=0.05635286332705082 +INFO:root: +[Valid][30][UTT][ MSE = 0.2216 | LCC = 0.8621 | SRCC = 0.8618 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2274 | LCC = 0.8608 | SRCC = 0.8627 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..6886aeae4b81db840b5350b7bb1fbeb72e8fa0db --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec7aa303498b9989be8b7fa9fc0f92908cfc78f9299a4c5439a52c72a5b16393 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..b3814a37633dae10b212fad8efc5f8dd283df95f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e56d9e0c4c17966d394fd1cef66e52cc742b8ea40f6fb0ed5ba1faa473b4677d +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..32767a291d66d90ae7313b24604e44b263b812d2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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=0.4860981844174556 +INFO:root: +[Valid][0][UTT][ MSE = 0.3424 | LCC = 0.7779 | SRCC = 0.7820 ] +INFO:root:Average loss=0.27407641384081965 +INFO:root: +[Valid][1][UTT][ MSE = 0.2768 | LCC = 0.8259 | SRCC = 0.8263 ] +INFO:root:Average loss=0.23034982325939032 +INFO:root: +[Valid][2][UTT][ MSE = 0.2678 | LCC = 0.8374 | SRCC = 0.8361 ] +INFO:root:Average loss=0.2105509883318192 +INFO:root: +[Valid][3][UTT][ MSE = 0.2549 | LCC = 0.8440 | SRCC = 0.8425 ] +INFO:root:Average loss=0.19492521519080186 +INFO:root: +[Valid][4][UTT][ MSE = 0.2489 | LCC = 0.8457 | SRCC = 0.8436 ] +INFO:root:Average loss=0.18145151503193072 +INFO:root: +[Valid][5][UTT][ MSE = 0.2447 | LCC = 0.8506 | SRCC = 0.8484 ] +INFO:root:Average loss=0.16871883595983186 +INFO:root: +[Valid][6][UTT][ MSE = 0.2423 | LCC = 0.8541 | SRCC = 0.8522 ] +INFO:root:Average loss=0.16162793300090691 +INFO:root: +[Valid][7][UTT][ MSE = 0.2345 | LCC = 0.8573 | SRCC = 0.8553 ] +INFO:root:Average loss=0.1493010958417868 +INFO:root: +[Valid][8][UTT][ MSE = 0.2418 | LCC = 0.8544 | SRCC = 0.8561 ] +INFO:root:Average loss=0.13896675341022321 +INFO:root: +[Valid][9][UTT][ MSE = 0.2238 | LCC = 0.8582 | SRCC = 0.8588 ] +INFO:root:Average loss=0.13175110194163445 +INFO:root: +[Valid][10][UTT][ MSE = 0.2525 | LCC = 0.8537 | SRCC = 0.8534 ] +INFO:root:Average loss=0.13188382333669907 +INFO:root: +[Valid][11][UTT][ MSE = 0.2228 | LCC = 0.8596 | SRCC = 0.8588 ] +INFO:root:Average loss=0.11658947656934078 +INFO:root: +[Valid][12][UTT][ MSE = 0.2230 | LCC = 0.8610 | SRCC = 0.8596 ] +INFO:root:Average loss=0.11116113131627059 +INFO:root: +[Valid][13][UTT][ MSE = 0.2238 | LCC = 0.8614 | SRCC = 0.8573 ] +INFO:root:Average loss=0.11423144618479106 +INFO:root: +[Valid][14][UTT][ MSE = 0.2438 | LCC = 0.8603 | SRCC = 0.8586 ] +INFO:root:Average loss=0.10079593903934345 +INFO:root: +[Valid][15][UTT][ MSE = 0.2157 | LCC = 0.8608 | SRCC = 0.8595 ] +INFO:root:Average loss=0.09578091278672218 +INFO:root: +[Valid][16][UTT][ MSE = 0.2553 | LCC = 0.8592 | SRCC = 0.8594 ] +INFO:root:Average loss=0.09197040900396995 +INFO:root: +[Valid][17][UTT][ MSE = 0.2276 | LCC = 0.8584 | SRCC = 0.8572 ] +INFO:root:Average loss=0.08586762243738541 +INFO:root: +[Valid][18][UTT][ MSE = 0.2302 | LCC = 0.8599 | SRCC = 0.8588 ] +INFO:root:Average loss=0.08448653042507477 +INFO:root: +[Valid][19][UTT][ MSE = 0.2175 | LCC = 0.8602 | SRCC = 0.8587 ] +INFO:root:Average loss=0.08634200682624793 +INFO:root: +[Valid][20][UTT][ MSE = 0.2342 | LCC = 0.8576 | SRCC = 0.8548 ] +INFO:root:Average loss=0.07574437539546917 +INFO:root: +[Valid][21][UTT][ MSE = 0.2283 | LCC = 0.8565 | SRCC = 0.8542 ] +INFO:root:Average loss=0.08243599572242835 +INFO:root: +[Valid][22][UTT][ MSE = 0.2324 | LCC = 0.8593 | SRCC = 0.8573 ] +INFO:root:Average loss=0.07046673103020741 +INFO:root: +[Valid][23][UTT][ MSE = 0.2307 | LCC = 0.8572 | SRCC = 0.8540 ] +INFO:root:Average loss=0.06950717142377144 +INFO:root: +[Valid][24][UTT][ MSE = 0.2766 | LCC = 0.8524 | SRCC = 0.8519 ] +INFO:root:Average loss=0.06781585342608966 +INFO:root: +[Valid][25][UTT][ MSE = 0.2466 | LCC = 0.8552 | SRCC = 0.8538 ] +INFO:root:Average loss=0.061189356403282054 +INFO:root: +[Valid][26][UTT][ MSE = 0.2238 | LCC = 0.8587 | SRCC = 0.8568 ] +INFO:root:Average loss=0.06340145472532664 +INFO:root: +[Valid][27][UTT][ MSE = 0.2272 | LCC = 0.8610 | SRCC = 0.8603 ] +INFO:root:Average loss=0.06128537024442966 +INFO:root: +[Valid][28][UTT][ MSE = 0.2206 | LCC = 0.8608 | SRCC = 0.8593 ] +INFO:root:Average loss=0.053619290821445294 +INFO:root: +[Valid][29][UTT][ MSE = 0.2372 | LCC = 0.8568 | SRCC = 0.8564 ] +INFO:root:Average loss=0.05898569466976019 +INFO:root: +[Valid][30][UTT][ MSE = 0.2274 | LCC = 0.8592 | SRCC = 0.8573 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2269 | LCC = 0.8633 | SRCC = 0.8624 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e72cc920707f6d17f149a60c0ac4f2d14fec858e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df6d6364ad261a5f94504ebb1b5908401a76b2833b6ad1722ff13d57f3579476 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..ffdbffafc96ccdb7b46c1bc7bd014ae4c1033d00 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b77de37861d88f5ebd4aba7c011cc39f539f314ad7d3304a6b5b1fc50de80416 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4944715f8d3d85f7f0b7ef7577c673a999cd0938 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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=0.48392338000046903 +INFO:root: +[Valid][0][UTT][ MSE = 0.3435 | LCC = 0.7747 | SRCC = 0.7766 ] +INFO:root:Average loss=0.28698584819451356 +INFO:root: +[Valid][1][UTT][ MSE = 0.3020 | LCC = 0.8180 | SRCC = 0.8221 ] +INFO:root:Average loss=0.23830052064015314 +INFO:root: +[Valid][2][UTT][ MSE = 0.2791 | LCC = 0.8334 | SRCC = 0.8351 ] +INFO:root:Average loss=0.21695177180644792 +INFO:root: +[Valid][3][UTT][ MSE = 0.2782 | LCC = 0.8450 | SRCC = 0.8446 ] +INFO:root:Average loss=0.20218652104720092 +INFO:root: +[Valid][4][UTT][ MSE = 0.2493 | LCC = 0.8485 | SRCC = 0.8489 ] +INFO:root:Average loss=0.18359380989120558 +INFO:root: +[Valid][5][UTT][ MSE = 0.2503 | LCC = 0.8548 | SRCC = 0.8530 ] +INFO:root:Average loss=0.17048615093032518 +INFO:root: +[Valid][6][UTT][ MSE = 0.2352 | LCC = 0.8583 | SRCC = 0.8573 ] +INFO:root:Average loss=0.15753455450519538 +INFO:root: +[Valid][7][UTT][ MSE = 0.2265 | LCC = 0.8618 | SRCC = 0.8594 ] +INFO:root:Average loss=0.1463976938946125 +INFO:root: +[Valid][8][UTT][ MSE = 0.2464 | LCC = 0.8597 | SRCC = 0.8599 ] +INFO:root:Average loss=0.14024258634218803 +INFO:root: +[Valid][9][UTT][ MSE = 0.2337 | LCC = 0.8608 | SRCC = 0.8593 ] +INFO:root:Average loss=0.13191830748930955 +INFO:root: +[Valid][10][UTT][ MSE = 0.2179 | LCC = 0.8621 | SRCC = 0.8615 ] +INFO:root:Average loss=0.12857772858861166 +INFO:root: +[Valid][11][UTT][ MSE = 0.2405 | LCC = 0.8618 | SRCC = 0.8605 ] +INFO:root:Average loss=0.1344869332626844 +INFO:root: +[Valid][12][UTT][ MSE = 0.2264 | LCC = 0.8646 | SRCC = 0.8626 ] +INFO:root:Average loss=0.1165979092892928 +INFO:root: +[Valid][13][UTT][ MSE = 0.2564 | LCC = 0.8623 | SRCC = 0.8622 ] +INFO:root:Average loss=0.11422119203668374 +INFO:root: +[Valid][14][UTT][ MSE = 0.2271 | LCC = 0.8642 | SRCC = 0.8642 ] +INFO:root:Average loss=0.10832411299149196 +INFO:root: +[Valid][15][UTT][ MSE = 0.2469 | LCC = 0.8632 | SRCC = 0.8628 ] +INFO:root:Average loss=0.10270269348835334 +INFO:root: +[Valid][16][UTT][ MSE = 0.2239 | LCC = 0.8632 | SRCC = 0.8628 ] +INFO:root:Average loss=0.09854948587524585 +INFO:root: +[Valid][17][UTT][ MSE = 0.2260 | LCC = 0.8656 | SRCC = 0.8638 ] +INFO:root:Average loss=0.09307961949171165 +INFO:root: +[Valid][18][UTT][ MSE = 0.2266 | LCC = 0.8652 | SRCC = 0.8634 ] +INFO:root:Average loss=0.08290323113592771 +INFO:root: +[Valid][19][UTT][ MSE = 0.2203 | LCC = 0.8639 | SRCC = 0.8625 ] +INFO:root:Average loss=0.0862569255897632 +INFO:root: +[Valid][20][UTT][ MSE = 0.2196 | LCC = 0.8590 | SRCC = 0.8586 ] +INFO:root:Average loss=0.0812414853523175 +INFO:root: +[Valid][21][UTT][ MSE = 0.2397 | LCC = 0.8607 | SRCC = 0.8612 ] +INFO:root:Average loss=0.07965629757978977 +INFO:root: +[Valid][22][UTT][ MSE = 0.2165 | LCC = 0.8637 | SRCC = 0.8625 ] +INFO:root:Average loss=0.07561662033773385 +INFO:root: +[Valid][23][UTT][ MSE = 0.2239 | LCC = 0.8627 | SRCC = 0.8621 ] +INFO:root:Average loss=0.07121086645966922 +INFO:root: +[Valid][24][UTT][ MSE = 0.2242 | LCC = 0.8630 | SRCC = 0.8620 ] +INFO:root:Average loss=0.06768073509327877 +INFO:root: +[Valid][25][UTT][ MSE = 0.2251 | LCC = 0.8639 | SRCC = 0.8624 ] +INFO:root:Average loss=0.07047050704176609 +INFO:root: +[Valid][26][UTT][ MSE = 0.2296 | LCC = 0.8639 | SRCC = 0.8637 ] +INFO:root:Average loss=0.06793932623874682 +INFO:root: +[Valid][27][UTT][ MSE = 0.2268 | LCC = 0.8629 | SRCC = 0.8608 ] +INFO:root:Average loss=0.06646584829267783 +INFO:root: +[Valid][28][UTT][ MSE = 0.2167 | LCC = 0.8666 | SRCC = 0.8656 ] +INFO:root:Average loss=0.06053964320856791 +INFO:root: +[Valid][29][UTT][ MSE = 0.2316 | LCC = 0.8636 | SRCC = 0.8622 ] +INFO:root:Average loss=0.06211291626095772 +INFO:root: +[Valid][30][UTT][ MSE = 0.2238 | LCC = 0.8621 | SRCC = 0.8603 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2339 | LCC = 0.8644 | SRCC = 0.8656 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..11234ee9c75ba5520e80f016b8dc646926bd488b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29e2365d2ca611a45d0edb05d7e5dd85739d0e97e3aac0448fb5fddcb55fffd2 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..6685df8f28cdbceea491656f7e1baacb4950ce14 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1e23d80ec2190a56bde9e030840fe8b6042bc668ac05d742786752d7169bd326 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..dedf96e975502658220591f5616134a86c0b9ac0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer1/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=0.5350919033472354 +INFO:root: +[Valid][0][UTT][ MSE = 0.3645 | LCC = 0.7675 | SRCC = 0.7737 ] +INFO:root:Average loss=0.2695529799048717 +INFO:root: +[Valid][1][UTT][ MSE = 0.2829 | LCC = 0.8216 | SRCC = 0.8295 ] +INFO:root:Average loss=0.23013242286367294 +INFO:root: +[Valid][2][UTT][ MSE = 0.2824 | LCC = 0.8416 | SRCC = 0.8445 ] +INFO:root:Average loss=0.2039399538666774 +INFO:root: +[Valid][3][UTT][ MSE = 0.2421 | LCC = 0.8532 | SRCC = 0.8531 ] +INFO:root:Average loss=0.18874159627235854 +INFO:root: +[Valid][4][UTT][ MSE = 0.2453 | LCC = 0.8586 | SRCC = 0.8571 ] +INFO:root:Average loss=0.1719995882266607 +INFO:root: +[Valid][5][UTT][ MSE = 0.2356 | LCC = 0.8616 | SRCC = 0.8620 ] +INFO:root:Average loss=0.16328568355395243 +INFO:root: +[Valid][6][UTT][ MSE = 0.2510 | LCC = 0.8636 | SRCC = 0.8616 ] +INFO:root:Average loss=0.14879061004672295 +INFO:root: +[Valid][7][UTT][ MSE = 0.2357 | LCC = 0.8646 | SRCC = 0.8616 ] +INFO:root:Average loss=0.1410846314751185 +INFO:root: +[Valid][8][UTT][ MSE = 0.2402 | LCC = 0.8593 | SRCC = 0.8548 ] +INFO:root:Average loss=0.1272498023433563 +INFO:root: +[Valid][9][UTT][ MSE = 0.2363 | LCC = 0.8638 | SRCC = 0.8621 ] +INFO:root:Average loss=0.15256370881047004 +INFO:root: +[Valid][10][UTT][ MSE = 0.2253 | LCC = 0.8672 | SRCC = 0.8646 ] +INFO:root:Average loss=0.12042084460457166 +INFO:root: +[Valid][11][UTT][ MSE = 0.2250 | LCC = 0.8665 | SRCC = 0.8656 ] +INFO:root:Average loss=0.11425618664958538 +INFO:root: +[Valid][12][UTT][ MSE = 0.2449 | LCC = 0.8616 | SRCC = 0.8608 ] +INFO:root:Average loss=0.11842619302945259 +INFO:root: +[Valid][13][UTT][ MSE = 0.2474 | LCC = 0.8669 | SRCC = 0.8656 ] +INFO:root:Average loss=0.09689268250113879 +INFO:root: +[Valid][14][UTT][ MSE = 0.2179 | LCC = 0.8684 | SRCC = 0.8649 ] +INFO:root:Average loss=0.09891317880306488 +INFO:root: +[Valid][15][UTT][ MSE = 0.2113 | LCC = 0.8688 | SRCC = 0.8669 ] +INFO:root:Average loss=0.09220971983785813 +INFO:root: +[Valid][16][UTT][ MSE = 0.2218 | LCC = 0.8711 | SRCC = 0.8680 ] +INFO:root:Average loss=0.0907900024874088 +INFO:root: +[Valid][17][UTT][ MSE = 0.2541 | LCC = 0.8645 | SRCC = 0.8615 ] +INFO:root:Average loss=0.08367992130418618 +INFO:root: +[Valid][18][UTT][ MSE = 0.2183 | LCC = 0.8687 | SRCC = 0.8655 ] +INFO:root:Average loss=0.07653574258662187 +INFO:root: +[Valid][19][UTT][ MSE = 0.2099 | LCC = 0.8711 | SRCC = 0.8673 ] +INFO:root:Average loss=0.07563196356670979 +INFO:root: +[Valid][20][UTT][ MSE = 0.2188 | LCC = 0.8673 | SRCC = 0.8638 ] +INFO:root:Average loss=0.07298640123544595 +INFO:root: +[Valid][21][UTT][ MSE = 0.2138 | LCC = 0.8695 | SRCC = 0.8662 ] +INFO:root:Average loss=0.0700682014800035 +INFO:root: +[Valid][22][UTT][ MSE = 0.2253 | LCC = 0.8677 | SRCC = 0.8652 ] +INFO:root:Average loss=0.06922046792430755 +INFO:root: +[Valid][23][UTT][ MSE = 0.2228 | LCC = 0.8670 | SRCC = 0.8639 ] +INFO:root:Average loss=0.06898166688206868 +INFO:root: +[Valid][24][UTT][ MSE = 0.2161 | LCC = 0.8682 | SRCC = 0.8647 ] +INFO:root:Average loss=0.07659558760814178 +INFO:root: +[Valid][25][UTT][ MSE = 0.2116 | LCC = 0.8705 | SRCC = 0.8674 ] +INFO:root:Average loss=0.061977494890109085 +INFO:root: +[Valid][26][UTT][ MSE = 0.2104 | LCC = 0.8693 | SRCC = 0.8667 ] +INFO:root:Average loss=0.0627120113812196 +INFO:root: +[Valid][27][UTT][ MSE = 0.2401 | LCC = 0.8669 | SRCC = 0.8641 ] +INFO:root:Average loss=0.06006219037450277 +INFO:root: +[Valid][28][UTT][ MSE = 0.2220 | LCC = 0.8691 | SRCC = 0.8675 ] +INFO:root:Average loss=0.055434658836859926 +INFO:root: +[Valid][29][UTT][ MSE = 0.2200 | LCC = 0.8661 | SRCC = 0.8622 ] +INFO:root:Average loss=0.055666244039550804 +INFO:root: +[Valid][30][UTT][ MSE = 0.2219 | LCC = 0.8703 | SRCC = 0.8662 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2424 | LCC = 0.8648 | SRCC = 0.8651 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..86f558f05a0edd1f8ef0ac0d35a26a4aa6883e00 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:958c2b336ed7496d347edbce773ed924008f7cd98d68118525a237c85e644ee3 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..b5d486ff5b3521a84717afdc9dd425670eb07bd6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7c5314013ab2f3d64377e4aaff5fe1ce087aaa92e5a449b80db7980e021b905 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f2447c93c1a082cc2a66a0130b4c83e064673639 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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=0.3693487953681212 +INFO:root: +[Valid][0][UTT][ MSE = 0.4088 | LCC = 0.8126 | SRCC = 0.8232 ] +INFO:root:Average loss=0.27153219798436534 +INFO:root: +[Valid][1][UTT][ MSE = 0.3199 | LCC = 0.8309 | SRCC = 0.8304 ] +INFO:root:Average loss=0.25249966444113314 +INFO:root: +[Valid][2][UTT][ MSE = 0.3083 | LCC = 0.8419 | SRCC = 0.8417 ] +INFO:root:Average loss=0.24451472170842 +INFO:root: +[Valid][3][UTT][ MSE = 0.2804 | LCC = 0.8466 | SRCC = 0.8459 ] +INFO:root:Average loss=0.23783861559170943 +INFO:root: +[Valid][4][UTT][ MSE = 0.2992 | LCC = 0.8511 | SRCC = 0.8504 ] +INFO:root:Average loss=0.22769803229050758 +INFO:root: +[Valid][5][UTT][ MSE = 0.3318 | LCC = 0.8515 | SRCC = 0.8502 ] +INFO:root:Average loss=0.23634772231945625 +INFO:root: +[Valid][6][UTT][ MSE = 0.2744 | LCC = 0.8516 | SRCC = 0.8519 ] +INFO:root:Average loss=0.22434344792213196 +INFO:root: +[Valid][7][UTT][ MSE = 0.2942 | LCC = 0.8519 | SRCC = 0.8509 ] +INFO:root:Average loss=0.2150281289449105 +INFO:root: +[Valid][8][UTT][ MSE = 0.2639 | LCC = 0.8523 | SRCC = 0.8533 ] +INFO:root:Average loss=0.21677453166399246 +INFO:root: +[Valid][9][UTT][ MSE = 0.2871 | LCC = 0.8575 | SRCC = 0.8563 ] +INFO:root:Average loss=0.21301312362536406 +INFO:root: +[Valid][10][UTT][ MSE = 0.2644 | LCC = 0.8579 | SRCC = 0.8586 ] +INFO:root:Average loss=0.21081833016031828 +INFO:root: +[Valid][11][UTT][ MSE = 0.2913 | LCC = 0.8560 | SRCC = 0.8559 ] +INFO:root:Average loss=0.2127571136523516 +INFO:root: +[Valid][12][UTT][ MSE = 0.3259 | LCC = 0.8628 | SRCC = 0.8605 ] +INFO:root:Average loss=0.2034555373665614 +INFO:root: +[Valid][13][UTT][ MSE = 0.2705 | LCC = 0.8578 | SRCC = 0.8563 ] +INFO:root:Average loss=0.19689843440667176 +INFO:root: +[Valid][14][UTT][ MSE = 0.2638 | LCC = 0.8559 | SRCC = 0.8565 ] +INFO:root:Average loss=0.1989323019217222 +INFO:root: +[Valid][15][UTT][ MSE = 0.2391 | LCC = 0.8530 | SRCC = 0.8612 ] +INFO:root:Average loss=0.19617604445188475 +INFO:root: +[Valid][16][UTT][ MSE = 0.2515 | LCC = 0.8581 | SRCC = 0.8612 ] +INFO:root:Average loss=0.1952202320098877 +INFO:root: +[Valid][17][UTT][ MSE = 0.2881 | LCC = 0.8608 | SRCC = 0.8590 ] +INFO:root:Average loss=0.18920334953910264 +INFO:root: +[Valid][18][UTT][ MSE = 0.2473 | LCC = 0.8629 | SRCC = 0.8615 ] +INFO:root:Average loss=0.1829192007963474 +INFO:root: +[Valid][19][UTT][ MSE = 0.2542 | LCC = 0.8600 | SRCC = 0.8600 ] +INFO:root:Average loss=0.17802939735926115 +INFO:root: +[Valid][20][UTT][ MSE = 0.2656 | LCC = 0.8545 | SRCC = 0.8577 ] +INFO:root:Average loss=0.17652130346649733 +INFO:root: +[Valid][21][UTT][ MSE = 0.2756 | LCC = 0.8594 | SRCC = 0.8601 ] +INFO:root:Average loss=0.17873991519594803 +INFO:root: +[Valid][22][UTT][ MSE = 0.2558 | LCC = 0.8576 | SRCC = 0.8604 ] +INFO:root:Average loss=0.1772308634259762 +INFO:root: +[Valid][23][UTT][ MSE = 0.2676 | LCC = 0.8548 | SRCC = 0.8563 ] +INFO:root:Average loss=0.17494274379733282 +INFO:root: +[Valid][24][UTT][ MSE = 0.2516 | LCC = 0.8567 | SRCC = 0.8601 ] +INFO:root:Average loss=0.1716010430111335 +INFO:root: +[Valid][25][UTT][ MSE = 0.2706 | LCC = 0.8580 | SRCC = 0.8577 ] +INFO:root:Average loss=0.1686271850306254 +INFO:root: +[Valid][26][UTT][ MSE = 0.2616 | LCC = 0.8581 | SRCC = 0.8563 ] +INFO:root:Average loss=0.16588553442404821 +INFO:root: +[Valid][27][UTT][ MSE = 0.2817 | LCC = 0.8587 | SRCC = 0.8572 ] +INFO:root:Average loss=0.16589842001215005 +INFO:root: +[Valid][28][UTT][ MSE = 0.2371 | LCC = 0.8555 | SRCC = 0.8577 ] +INFO:root:Average loss=0.1565935118840291 +INFO:root: +[Valid][29][UTT][ MSE = 0.2661 | LCC = 0.8542 | SRCC = 0.8523 ] +INFO:root:Average loss=0.1554831586396083 +INFO:root: +[Valid][30][UTT][ MSE = 0.2383 | LCC = 0.8572 | SRCC = 0.8584 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2485 | LCC = 0.8624 | SRCC = 0.8632 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..f28f919aa3b5604abe061bc833e8f4936b47cb25 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7d6cf1a69587ae9cf3d66b34b915169a0b90e31a81d73514fbad5a564ba7631 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..38b7457788846f319d7f7aec23ffe621b833db51 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c48333f8d2076760a581670d40191ac729c11da0c1b43f1693c9a61eb6019ec3 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..8429c995a43b628ce589266b8682eb14a18bdba2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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=0.32443578426654524 +INFO:root: +[Valid][0][UTT][ MSE = 0.3442 | LCC = 0.8233 | SRCC = 0.8224 ] +INFO:root:Average loss=0.2671893991720982 +INFO:root: +[Valid][1][UTT][ MSE = 0.3033 | LCC = 0.8350 | SRCC = 0.8359 ] +INFO:root:Average loss=0.24777412701111573 +INFO:root: +[Valid][2][UTT][ MSE = 0.3367 | LCC = 0.8451 | SRCC = 0.8438 ] +INFO:root:Average loss=0.24350597843145713 +INFO:root: +[Valid][3][UTT][ MSE = 0.3054 | LCC = 0.8508 | SRCC = 0.8480 ] +INFO:root:Average loss=0.23018980217285645 +INFO:root: +[Valid][4][UTT][ MSE = 0.2802 | LCC = 0.8515 | SRCC = 0.8506 ] +INFO:root:Average loss=0.22753922870525947 +INFO:root: +[Valid][5][UTT][ MSE = 0.2919 | LCC = 0.8516 | SRCC = 0.8518 ] +INFO:root:Average loss=0.24086362066177222 +INFO:root: +[Valid][6][UTT][ MSE = 0.2747 | LCC = 0.8515 | SRCC = 0.8531 ] +INFO:root:Average loss=0.22716541511890215 +INFO:root: +[Valid][7][UTT][ MSE = 0.3951 | LCC = 0.8549 | SRCC = 0.8548 ] +INFO:root:Average loss=0.22524452228576708 +INFO:root: +[Valid][8][UTT][ MSE = 0.3002 | LCC = 0.8466 | SRCC = 0.8500 ] +INFO:root:Average loss=0.2230625739082312 +INFO:root: +[Valid][9][UTT][ MSE = 0.2554 | LCC = 0.8508 | SRCC = 0.8562 ] +INFO:root:Average loss=0.21566604708249754 +INFO:root: +[Valid][10][UTT][ MSE = 0.3133 | LCC = 0.8472 | SRCC = 0.8518 ] +INFO:root:Average loss=0.21112006501509592 +INFO:root: +[Valid][11][UTT][ MSE = 0.2912 | LCC = 0.8598 | SRCC = 0.8588 ] +INFO:root:Average loss=0.20989795525868735 +INFO:root: +[Valid][12][UTT][ MSE = 0.2690 | LCC = 0.8615 | SRCC = 0.8602 ] +INFO:root:Average loss=0.20094985992480546 +INFO:root: +[Valid][13][UTT][ MSE = 0.2693 | LCC = 0.8584 | SRCC = 0.8579 ] +INFO:root:Average loss=0.20742017613389555 +INFO:root: +[Valid][14][UTT][ MSE = 0.2876 | LCC = 0.8619 | SRCC = 0.8613 ] +INFO:root:Average loss=0.19295683675087416 +INFO:root: +[Valid][15][UTT][ MSE = 0.2872 | LCC = 0.8629 | SRCC = 0.8618 ] +INFO:root:Average loss=0.19477870993507215 +INFO:root: +[Valid][16][UTT][ MSE = 0.2600 | LCC = 0.8576 | SRCC = 0.8570 ] +INFO:root:Average loss=0.19428497304519018 +INFO:root: +[Valid][17][UTT][ MSE = 0.3141 | LCC = 0.8596 | SRCC = 0.8579 ] +INFO:root:Average loss=0.19322007359602514 +INFO:root: +[Valid][18][UTT][ MSE = 0.2854 | LCC = 0.8618 | SRCC = 0.8628 ] +INFO:root:Average loss=0.1996678494108029 +INFO:root: +[Valid][19][UTT][ MSE = 0.3198 | LCC = 0.8592 | SRCC = 0.8580 ] +INFO:root:Average loss=0.18551855725355637 +INFO:root: +[Valid][20][UTT][ MSE = 0.3309 | LCC = 0.8567 | SRCC = 0.8549 ] +INFO:root:Average loss=0.18664560849085832 +INFO:root: +[Valid][21][UTT][ MSE = 0.2435 | LCC = 0.8571 | SRCC = 0.8569 ] +INFO:root:Average loss=0.1794741304639058 +INFO:root: +[Valid][22][UTT][ MSE = 0.2559 | LCC = 0.8575 | SRCC = 0.8572 ] +INFO:root:Average loss=0.17813968085325682 +INFO:root: +[Valid][23][UTT][ MSE = 0.2556 | LCC = 0.8591 | SRCC = 0.8585 ] +INFO:root:Average loss=0.17648586267844224 +INFO:root: +[Valid][24][UTT][ MSE = 0.2848 | LCC = 0.8509 | SRCC = 0.8505 ] +INFO:root:Average loss=0.17688758336962798 +INFO:root: +[Valid][25][UTT][ MSE = 0.2564 | LCC = 0.8588 | SRCC = 0.8584 ] +INFO:root:Average loss=0.16786899751959702 +INFO:root: +[Valid][26][UTT][ MSE = 0.2931 | LCC = 0.8527 | SRCC = 0.8511 ] +INFO:root:Average loss=0.1707303686401783 +INFO:root: +[Valid][27][UTT][ MSE = 0.2642 | LCC = 0.8543 | SRCC = 0.8564 ] +INFO:root:Average loss=0.16256082860323098 +INFO:root: +[Valid][28][UTT][ MSE = 0.2723 | LCC = 0.8581 | SRCC = 0.8574 ] +INFO:root:Average loss=0.16085506946994707 +INFO:root: +[Valid][29][UTT][ MSE = 0.2408 | LCC = 0.8564 | SRCC = 0.8559 ] +INFO:root:Average loss=0.15798046965247545 +INFO:root: +[Valid][30][UTT][ MSE = 0.2901 | LCC = 0.8536 | SRCC = 0.8550 ] +INFO:root: +[Test][31][UTT][ MSE = 0.3008 | LCC = 0.8589 | SRCC = 0.8625 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..804c2f6129bcca32efe04013d44143dec05ab7cc --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6aea0f14e9fa940e0dd45ec996ad9fbd748da8271dc327b0ce8f2745fdbcb0ad +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e9ba2198893ff8067bc698d31cdd752bc24f3d92 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c37701ab579f5c2eb2a1a977bf0fc4c4da2f694c5bf69c846c700612f6d52b51 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4308e000cfca9c4d2059c4d3d35e13a9a4c8c32e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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=0.31853577437309116 +INFO:root: +[Valid][0][UTT][ MSE = 0.3601 | LCC = 0.8200 | SRCC = 0.8181 ] +INFO:root:Average loss=0.2685626670718193 +INFO:root: +[Valid][1][UTT][ MSE = 0.3412 | LCC = 0.8346 | SRCC = 0.8324 ] +INFO:root:Average loss=0.2581852674484253 +INFO:root: +[Valid][2][UTT][ MSE = 0.3169 | LCC = 0.8412 | SRCC = 0.8392 ] +INFO:root:Average loss=0.24311401599492782 +INFO:root: +[Valid][3][UTT][ MSE = 0.2943 | LCC = 0.8375 | SRCC = 0.8446 ] +INFO:root:Average loss=0.23545899700659972 +INFO:root: +[Valid][4][UTT][ MSE = 0.3208 | LCC = 0.8466 | SRCC = 0.8483 ] +INFO:root:Average loss=0.22986211608617735 +INFO:root: +[Valid][5][UTT][ MSE = 0.3013 | LCC = 0.8473 | SRCC = 0.8505 ] +INFO:root:Average loss=0.22514174859493208 +INFO:root: +[Valid][6][UTT][ MSE = 0.2877 | LCC = 0.8534 | SRCC = 0.8536 ] +INFO:root:Average loss=0.22336194759760147 +INFO:root: +[Valid][7][UTT][ MSE = 0.3073 | LCC = 0.8504 | SRCC = 0.8516 ] +INFO:root:Average loss=0.22127393709543425 +INFO:root: +[Valid][8][UTT][ MSE = 0.2903 | LCC = 0.8538 | SRCC = 0.8543 ] +INFO:root:Average loss=0.21120928017756876 +INFO:root: +[Valid][9][UTT][ MSE = 0.2881 | LCC = 0.8546 | SRCC = 0.8555 ] +INFO:root:Average loss=0.213949866210803 +INFO:root: +[Valid][10][UTT][ MSE = 0.2834 | LCC = 0.8539 | SRCC = 0.8562 ] +INFO:root:Average loss=0.21417527798658761 +INFO:root: +[Valid][11][UTT][ MSE = 0.2712 | LCC = 0.8541 | SRCC = 0.8559 ] +INFO:root:Average loss=0.20637179662783942 +INFO:root: +[Valid][12][UTT][ MSE = 0.2926 | LCC = 0.8582 | SRCC = 0.8579 ] +INFO:root:Average loss=0.20763606081406275 +INFO:root: +[Valid][13][UTT][ MSE = 0.2962 | LCC = 0.8583 | SRCC = 0.8565 ] +INFO:root:Average loss=0.19961075484752655 +INFO:root: +[Valid][14][UTT][ MSE = 0.2890 | LCC = 0.8559 | SRCC = 0.8607 ] +INFO:root:Average loss=0.1951364095394428 +INFO:root: +[Valid][15][UTT][ MSE = 0.2923 | LCC = 0.8580 | SRCC = 0.8577 ] +INFO:root:Average loss=0.1920493713174111 +INFO:root: +[Valid][16][UTT][ MSE = 0.2589 | LCC = 0.8612 | SRCC = 0.8611 ] +INFO:root:Average loss=0.19209646834776953 +INFO:root: +[Valid][17][UTT][ MSE = 0.2754 | LCC = 0.8564 | SRCC = 0.8590 ] +INFO:root:Average loss=0.19033783798416457 +INFO:root: +[Valid][18][UTT][ MSE = 0.2588 | LCC = 0.8619 | SRCC = 0.8634 ] +INFO:root:Average loss=0.1850184440994874 +INFO:root: +[Valid][19][UTT][ MSE = 0.2767 | LCC = 0.8618 | SRCC = 0.8622 ] +INFO:root:Average loss=0.18485063667862844 +INFO:root: +[Valid][20][UTT][ MSE = 0.2782 | LCC = 0.8576 | SRCC = 0.8613 ] +INFO:root:Average loss=0.18061125364441138 +INFO:root: +[Valid][21][UTT][ MSE = 0.2741 | LCC = 0.8593 | SRCC = 0.8610 ] +INFO:root:Average loss=0.17924500180360597 +INFO:root: +[Valid][22][UTT][ MSE = 0.3119 | LCC = 0.8596 | SRCC = 0.8597 ] +INFO:root:Average loss=0.19132453652146536 +INFO:root: +[Valid][23][UTT][ MSE = 0.2632 | LCC = 0.8642 | SRCC = 0.8665 ] +INFO:root:Average loss=0.17506988728657746 +INFO:root: +[Valid][24][UTT][ MSE = 0.2749 | LCC = 0.8601 | SRCC = 0.8589 ] +INFO:root:Average loss=0.17565786590178809 +INFO:root: +[Valid][25][UTT][ MSE = 0.2711 | LCC = 0.8581 | SRCC = 0.8625 ] +INFO:root:Average loss=0.17032265787323317 +INFO:root: +[Valid][26][UTT][ MSE = 0.2648 | LCC = 0.8620 | SRCC = 0.8616 ] +INFO:root:Average loss=0.16672236797137138 +INFO:root: +[Valid][27][UTT][ MSE = 0.2490 | LCC = 0.8642 | SRCC = 0.8639 ] +INFO:root:Average loss=0.16418729884884295 +INFO:root: +[Valid][28][UTT][ MSE = 0.2694 | LCC = 0.8576 | SRCC = 0.8600 ] +INFO:root:Average loss=0.16355540555639145 +INFO:root: +[Valid][29][UTT][ MSE = 0.2931 | LCC = 0.8500 | SRCC = 0.8528 ] +INFO:root:Average loss=0.16241018464549994 +INFO:root: +[Valid][30][UTT][ MSE = 0.2929 | LCC = 0.8599 | SRCC = 0.8580 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2510 | LCC = 0.8577 | SRCC = 0.8581 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..fa99fe584f6dfbdee46951d908bdb992d67bcf7e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19dc298b91307c52fae3cbe77c08d3d0d9f9bc84efe06771d89c575cb7db1969 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..328fa57c3cc620009ae7cd0ef4ec1f8c56af7fc6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0a2730c42afc1a5a65ca650d4a9c8405e1a21e097cfaac920281746cbb77f9e4 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..76b240af272d6e420f8d91bb77990b0567690196 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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=0.3809534498514273 +INFO:root: +[Valid][0][UTT][ MSE = 0.3753 | LCC = 0.8057 | SRCC = 0.8165 ] +INFO:root:Average loss=0.26749358689173675 +INFO:root: +[Valid][1][UTT][ MSE = 0.3236 | LCC = 0.8223 | SRCC = 0.8278 ] +INFO:root:Average loss=0.2539805135665796 +INFO:root: +[Valid][2][UTT][ MSE = 0.3263 | LCC = 0.8429 | SRCC = 0.8437 ] +INFO:root:Average loss=0.2445374513283754 +INFO:root: +[Valid][3][UTT][ MSE = 0.3407 | LCC = 0.8503 | SRCC = 0.8471 ] +INFO:root:Average loss=0.2370534680592708 +INFO:root: +[Valid][4][UTT][ MSE = 0.2851 | LCC = 0.8477 | SRCC = 0.8497 ] +INFO:root:Average loss=0.23087215862977198 +INFO:root: +[Valid][5][UTT][ MSE = 0.2884 | LCC = 0.8444 | SRCC = 0.8461 ] +INFO:root:Average loss=0.2239012953180533 +INFO:root: +[Valid][6][UTT][ MSE = 0.3023 | LCC = 0.8552 | SRCC = 0.8530 ] +INFO:root:Average loss=0.2276908938701336 +INFO:root: +[Valid][7][UTT][ MSE = 0.2941 | LCC = 0.8550 | SRCC = 0.8549 ] +INFO:root:Average loss=0.21903494038642982 +INFO:root: +[Valid][8][UTT][ MSE = 0.2996 | LCC = 0.8512 | SRCC = 0.8502 ] +INFO:root:Average loss=0.2172511040400236 +INFO:root: +[Valid][9][UTT][ MSE = 0.2924 | LCC = 0.8493 | SRCC = 0.8519 ] +INFO:root:Average loss=0.22441953019453928 +INFO:root: +[Valid][10][UTT][ MSE = 0.3044 | LCC = 0.8520 | SRCC = 0.8530 ] +INFO:root:Average loss=0.20526853748239005 +INFO:root: +[Valid][11][UTT][ MSE = 0.3001 | LCC = 0.8538 | SRCC = 0.8536 ] +INFO:root:Average loss=0.21113530966715935 +INFO:root: +[Valid][12][UTT][ MSE = 0.3153 | LCC = 0.8522 | SRCC = 0.8518 ] +INFO:root:Average loss=0.20612128098041582 +INFO:root: +[Valid][13][UTT][ MSE = 0.2689 | LCC = 0.8587 | SRCC = 0.8587 ] +INFO:root:Average loss=0.20187107989421257 +INFO:root: +[Valid][14][UTT][ MSE = 0.2712 | LCC = 0.8574 | SRCC = 0.8577 ] +INFO:root:Average loss=0.20011824254806226 +INFO:root: +[Valid][15][UTT][ MSE = 0.2526 | LCC = 0.8534 | SRCC = 0.8577 ] +INFO:root:Average loss=0.1990479160195742 +INFO:root: +[Valid][16][UTT][ MSE = 0.2768 | LCC = 0.8586 | SRCC = 0.8588 ] +INFO:root:Average loss=0.1925294209176149 +INFO:root: +[Valid][17][UTT][ MSE = 0.2677 | LCC = 0.8564 | SRCC = 0.8587 ] +INFO:root:Average loss=0.1873142333366932 +INFO:root: +[Valid][18][UTT][ MSE = 0.3224 | LCC = 0.8561 | SRCC = 0.8552 ] +INFO:root:Average loss=0.19009197178559425 +INFO:root: +[Valid][19][UTT][ MSE = 0.2892 | LCC = 0.8563 | SRCC = 0.8561 ] +INFO:root:Average loss=0.18385422841096535 +INFO:root: +[Valid][20][UTT][ MSE = 0.2698 | LCC = 0.8550 | SRCC = 0.8585 ] +INFO:root:Average loss=0.1880989494996193 +INFO:root: +[Valid][21][UTT][ MSE = 0.2627 | LCC = 0.8573 | SRCC = 0.8561 ] +INFO:root:Average loss=0.17769901034159538 +INFO:root: +[Valid][22][UTT][ MSE = 0.2672 | LCC = 0.8578 | SRCC = 0.8581 ] +INFO:root:Average loss=0.17926795341265508 +INFO:root: +[Valid][23][UTT][ MSE = 0.3044 | LCC = 0.8569 | SRCC = 0.8531 ] +INFO:root:Average loss=0.17730675408473381 +INFO:root: +[Valid][24][UTT][ MSE = 0.2631 | LCC = 0.8541 | SRCC = 0.8557 ] +INFO:root:Average loss=0.16830008782637426 +INFO:root: +[Valid][25][UTT][ MSE = 0.2592 | LCC = 0.8573 | SRCC = 0.8586 ] +INFO:root:Average loss=0.17045478494121477 +INFO:root: +[Valid][26][UTT][ MSE = 0.2809 | LCC = 0.8503 | SRCC = 0.8546 ] +INFO:root:Average loss=0.16727118614392403 +INFO:root: +[Valid][27][UTT][ MSE = 0.2663 | LCC = 0.8493 | SRCC = 0.8539 ] +INFO:root:Average loss=0.16927625535008234 +INFO:root: +[Valid][28][UTT][ MSE = 0.2453 | LCC = 0.8531 | SRCC = 0.8573 ] +INFO:root:Average loss=0.16228054244166765 +INFO:root: +[Valid][29][UTT][ MSE = 0.2673 | LCC = 0.8552 | SRCC = 0.8575 ] +INFO:root:Average loss=0.15797862181296715 +INFO:root: +[Valid][30][UTT][ MSE = 0.2990 | LCC = 0.8530 | SRCC = 0.8485 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2546 | LCC = 0.8618 | SRCC = 0.8630 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..87160ccc7d5047b246385ed610be7403b97adbfa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:15b5c217dcafb21c86cbce5120108fb47464d6f3f5487c7ddade61fc9422070c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..44d9ad2bb7755a4d9b98046945b4e550044eb7d7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05e7341678809b28963fab81fef87a52c9bf09cc566f9198e11be83d867a7db8 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9d21ef374de0c0cabd5a4efb743ee94f45bc0996 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer10/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=0.32787801974858993 +INFO:root: +[Valid][0][UTT][ MSE = 0.3504 | LCC = 0.8246 | SRCC = 0.8223 ] +INFO:root:Average loss=0.26671737604416335 +INFO:root: +[Valid][1][UTT][ MSE = 0.3457 | LCC = 0.8326 | SRCC = 0.8324 ] +INFO:root:Average loss=0.25409154746776974 +INFO:root: +[Valid][2][UTT][ MSE = 0.3295 | LCC = 0.8428 | SRCC = 0.8429 ] +INFO:root:Average loss=0.23883862831653693 +INFO:root: +[Valid][3][UTT][ MSE = 0.3198 | LCC = 0.8497 | SRCC = 0.8483 ] +INFO:root:Average loss=0.24023308070042196 +INFO:root: +[Valid][4][UTT][ MSE = 0.2980 | LCC = 0.8466 | SRCC = 0.8478 ] +INFO:root:Average loss=0.22850284859155998 +INFO:root: +[Valid][5][UTT][ MSE = 0.2989 | LCC = 0.8549 | SRCC = 0.8524 ] +INFO:root:Average loss=0.22343768924474716 +INFO:root: +[Valid][6][UTT][ MSE = 0.3265 | LCC = 0.8544 | SRCC = 0.8540 ] +INFO:root:Average loss=0.21743756131483957 +INFO:root: +[Valid][7][UTT][ MSE = 0.2782 | LCC = 0.8526 | SRCC = 0.8543 ] +INFO:root:Average loss=0.22428951355127189 +INFO:root: +[Valid][8][UTT][ MSE = 0.3055 | LCC = 0.8432 | SRCC = 0.8504 ] +INFO:root:Average loss=0.21679180344709983 +INFO:root: +[Valid][9][UTT][ MSE = 0.2981 | LCC = 0.8532 | SRCC = 0.8538 ] +INFO:root:Average loss=0.20907769753382757 +INFO:root: +[Valid][10][UTT][ MSE = 0.2737 | LCC = 0.8547 | SRCC = 0.8583 ] +INFO:root:Average loss=0.21023442424260652 +INFO:root: +[Valid][11][UTT][ MSE = 0.2962 | LCC = 0.8540 | SRCC = 0.8558 ] +INFO:root:Average loss=0.2064553161080067 +INFO:root: +[Valid][12][UTT][ MSE = 0.2687 | LCC = 0.8527 | SRCC = 0.8569 ] +INFO:root:Average loss=0.20571286976337433 +INFO:root: +[Valid][13][UTT][ MSE = 0.2765 | LCC = 0.8476 | SRCC = 0.8539 ] +INFO:root:Average loss=0.20231740042949334 +INFO:root: +[Valid][14][UTT][ MSE = 0.2916 | LCC = 0.8585 | SRCC = 0.8590 ] +INFO:root:Average loss=0.20534234111889815 +INFO:root: +[Valid][15][UTT][ MSE = 0.2849 | LCC = 0.8537 | SRCC = 0.8566 ] +INFO:root:Average loss=0.1976072265742681 +INFO:root: +[Valid][16][UTT][ MSE = 0.2966 | LCC = 0.8567 | SRCC = 0.8606 ] +INFO:root:Average loss=0.19791036032331297 +INFO:root: +[Valid][17][UTT][ MSE = 0.2987 | LCC = 0.8521 | SRCC = 0.8561 ] +INFO:root:Average loss=0.19032588467383996 +INFO:root: +[Valid][18][UTT][ MSE = 0.2621 | LCC = 0.8614 | SRCC = 0.8612 ] +INFO:root:Average loss=0.18743273749565467 +INFO:root: +[Valid][19][UTT][ MSE = 0.2429 | LCC = 0.8578 | SRCC = 0.8639 ] +INFO:root:Average loss=0.18986364473135042 +INFO:root: +[Valid][20][UTT][ MSE = 0.2703 | LCC = 0.8580 | SRCC = 0.8601 ] +INFO:root:Average loss=0.18349238676138413 +INFO:root: +[Valid][21][UTT][ MSE = 0.2533 | LCC = 0.8577 | SRCC = 0.8619 ] +INFO:root:Average loss=0.17685365619567725 +INFO:root: +[Valid][22][UTT][ MSE = 0.2708 | LCC = 0.8616 | SRCC = 0.8600 ] +INFO:root:Average loss=0.18126156381689584 +INFO:root: +[Valid][23][UTT][ MSE = 0.2897 | LCC = 0.8605 | SRCC = 0.8595 ] +INFO:root:Average loss=0.17828758318836874 +INFO:root: +[Valid][24][UTT][ MSE = 0.2791 | LCC = 0.8588 | SRCC = 0.8611 ] +INFO:root:Average loss=0.17609841720415995 +INFO:root: +[Valid][25][UTT][ MSE = 0.2718 | LCC = 0.8556 | SRCC = 0.8593 ] +INFO:root:Average loss=0.17238655514442003 +INFO:root: +[Valid][26][UTT][ MSE = 0.2773 | LCC = 0.8603 | SRCC = 0.8604 ] +INFO:root:Average loss=0.1687513780899537 +INFO:root: +[Valid][27][UTT][ MSE = 0.2561 | LCC = 0.8598 | SRCC = 0.8621 ] +INFO:root:Average loss=0.16584994682134727 +INFO:root: +[Valid][28][UTT][ MSE = 0.2562 | LCC = 0.8555 | SRCC = 0.8546 ] +INFO:root:Average loss=0.16584944533996093 +INFO:root: +[Valid][29][UTT][ MSE = 0.2639 | LCC = 0.8580 | SRCC = 0.8582 ] +INFO:root:Average loss=0.1582831204510652 +INFO:root: +[Valid][30][UTT][ MSE = 0.2917 | LCC = 0.8546 | SRCC = 0.8564 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2718 | LCC = 0.8564 | SRCC = 0.8554 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..a31750421e1aafff690124ee85ce5a69a1cb51df --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f2093fbbd4637d2b396f3ab156ad218919be755e86783bf3d0c1ec203159d75 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..6b52f30ce1456618c7f3970871b0abdb3f3f7d01 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60623e7e8ad9f32e032a915583bae57bd4963f62a9e02da14da3c389e5430850 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b015e170812c2b6581813874494cd38d73d6dc6d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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=0.3498677204434688 +INFO:root: +[Valid][0][UTT][ MSE = 0.3920 | LCC = 0.8071 | SRCC = 0.8067 ] +INFO:root:Average loss=0.2871877719194461 +INFO:root: +[Valid][1][UTT][ MSE = 0.3342 | LCC = 0.8158 | SRCC = 0.8207 ] +INFO:root:Average loss=0.27770928064217937 +INFO:root: +[Valid][2][UTT][ MSE = 0.3863 | LCC = 0.8229 | SRCC = 0.8265 ] +INFO:root:Average loss=0.26529207596412074 +INFO:root: +[Valid][3][UTT][ MSE = 0.3275 | LCC = 0.8340 | SRCC = 0.8337 ] +INFO:root:Average loss=0.2605136464803647 +INFO:root: +[Valid][4][UTT][ MSE = 0.3201 | LCC = 0.8323 | SRCC = 0.8355 ] +INFO:root:Average loss=0.25534905712956035 +INFO:root: +[Valid][5][UTT][ MSE = 0.3003 | LCC = 0.8361 | SRCC = 0.8401 ] +INFO:root:Average loss=0.25071667707883394 +INFO:root: +[Valid][6][UTT][ MSE = 0.3413 | LCC = 0.8460 | SRCC = 0.8428 ] +INFO:root:Average loss=0.252388525276612 +INFO:root: +[Valid][7][UTT][ MSE = 0.2879 | LCC = 0.8410 | SRCC = 0.8441 ] +INFO:root:Average loss=0.24631624745252806 +INFO:root: +[Valid][8][UTT][ MSE = 0.3066 | LCC = 0.8410 | SRCC = 0.8408 ] +INFO:root:Average loss=0.2475110607651564 +INFO:root: +[Valid][9][UTT][ MSE = 0.3433 | LCC = 0.8479 | SRCC = 0.8459 ] +INFO:root:Average loss=0.23883708470906967 +INFO:root: +[Valid][10][UTT][ MSE = 0.3164 | LCC = 0.8460 | SRCC = 0.8466 ] +INFO:root:Average loss=0.24245563034827894 +INFO:root: +[Valid][11][UTT][ MSE = 0.3044 | LCC = 0.8507 | SRCC = 0.8481 ] +INFO:root:Average loss=0.23160485885082147 +INFO:root: +[Valid][12][UTT][ MSE = 0.3032 | LCC = 0.8493 | SRCC = 0.8473 ] +INFO:root:Average loss=0.23330133924117455 +INFO:root: +[Valid][13][UTT][ MSE = 0.3052 | LCC = 0.8535 | SRCC = 0.8512 ] +INFO:root:Average loss=0.22728052658912462 +INFO:root: +[Valid][14][UTT][ MSE = 0.2747 | LCC = 0.8498 | SRCC = 0.8504 ] +INFO:root:Average loss=0.22113222304062966 +INFO:root: +[Valid][15][UTT][ MSE = 0.3031 | LCC = 0.8529 | SRCC = 0.8539 ] +INFO:root:Average loss=0.22970679708016226 +INFO:root: +[Valid][16][UTT][ MSE = 0.2964 | LCC = 0.8478 | SRCC = 0.8499 ] +INFO:root:Average loss=0.22293068525882867 +INFO:root: +[Valid][17][UTT][ MSE = 0.2972 | LCC = 0.8545 | SRCC = 0.8531 ] +INFO:root:Average loss=0.2224022428958844 +INFO:root: +[Valid][18][UTT][ MSE = 0.2977 | LCC = 0.8540 | SRCC = 0.8542 ] +INFO:root:Average loss=0.22086821897671774 +INFO:root: +[Valid][19][UTT][ MSE = 0.2782 | LCC = 0.8529 | SRCC = 0.8522 ] +INFO:root:Average loss=0.21911440025537443 +INFO:root: +[Valid][20][UTT][ MSE = 0.2745 | LCC = 0.8538 | SRCC = 0.8576 ] +INFO:root:Average loss=0.22470463124605325 +INFO:root: +[Valid][21][UTT][ MSE = 0.3035 | LCC = 0.8574 | SRCC = 0.8553 ] +INFO:root:Average loss=0.21254396457702684 +INFO:root: +[Valid][22][UTT][ MSE = 0.2712 | LCC = 0.8487 | SRCC = 0.8521 ] +INFO:root:Average loss=0.21471589918319994 +INFO:root: +[Valid][23][UTT][ MSE = 0.2861 | LCC = 0.8568 | SRCC = 0.8556 ] +INFO:root:Average loss=0.22117155561080346 +INFO:root: +[Valid][24][UTT][ MSE = 0.2996 | LCC = 0.8532 | SRCC = 0.8521 ] +INFO:root:Average loss=0.21247489463824493 +INFO:root: +[Valid][25][UTT][ MSE = 0.2923 | LCC = 0.8535 | SRCC = 0.8554 ] +INFO:root:Average loss=0.20804383223637557 +INFO:root: +[Valid][26][UTT][ MSE = 0.2702 | LCC = 0.8544 | SRCC = 0.8559 ] +INFO:root:Average loss=0.20646470087843063 +INFO:root: +[Valid][27][UTT][ MSE = 0.2745 | LCC = 0.8567 | SRCC = 0.8557 ] +INFO:root:Average loss=0.2089455966383983 +INFO:root: +[Valid][28][UTT][ MSE = 0.2707 | LCC = 0.8515 | SRCC = 0.8568 ] +INFO:root:Average loss=0.2081168660750756 +INFO:root: +[Valid][29][UTT][ MSE = 0.2531 | LCC = 0.8589 | SRCC = 0.8599 ] +INFO:root:Average loss=0.20748582749794692 +INFO:root: +[Valid][30][UTT][ MSE = 0.2679 | LCC = 0.8556 | SRCC = 0.8580 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2610 | LCC = 0.8568 | SRCC = 0.8619 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..9fbc88670912440e243cbbb743416c728bfddeec --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2bbff1ca43c26356ed0673398d4d65fcdb76ce281820a100d5a02bb128b2518 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..494b115f5b35072ece0d46f22400d3e2d650f6eb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:21eddc9107d1e283d512ac7287af8585bae2d981feed764cafe0121166557228 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..9f3c95a39fbaee7d230de93efd4d0612b2ae25ec --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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=0.3357546810920422 +INFO:root: +[Valid][0][UTT][ MSE = 0.3461 | LCC = 0.8103 | SRCC = 0.8122 ] +INFO:root:Average loss=0.29355879548268443 +INFO:root: +[Valid][1][UTT][ MSE = 0.3462 | LCC = 0.8256 | SRCC = 0.8236 ] +INFO:root:Average loss=0.2803091893975551 +INFO:root: +[Valid][2][UTT][ MSE = 0.3496 | LCC = 0.8293 | SRCC = 0.8266 ] +INFO:root:Average loss=0.26766665795674693 +INFO:root: +[Valid][3][UTT][ MSE = 0.3316 | LCC = 0.8330 | SRCC = 0.8337 ] +INFO:root:Average loss=0.26763656869148594 +INFO:root: +[Valid][4][UTT][ MSE = 0.3056 | LCC = 0.8348 | SRCC = 0.8361 ] +INFO:root:Average loss=0.25941084497250044 +INFO:root: +[Valid][5][UTT][ MSE = 0.3537 | LCC = 0.8404 | SRCC = 0.8397 ] +INFO:root:Average loss=0.2576942334954555 +INFO:root: +[Valid][6][UTT][ MSE = 0.2986 | LCC = 0.8384 | SRCC = 0.8421 ] +INFO:root:Average loss=0.24174000781316024 +INFO:root: +[Valid][7][UTT][ MSE = 0.3204 | LCC = 0.8451 | SRCC = 0.8445 ] +INFO:root:Average loss=0.24560502897470426 +INFO:root: +[Valid][8][UTT][ MSE = 0.2815 | LCC = 0.8394 | SRCC = 0.8444 ] +INFO:root:Average loss=0.23926564917350426 +INFO:root: +[Valid][9][UTT][ MSE = 0.3238 | LCC = 0.8471 | SRCC = 0.8485 ] +INFO:root:Average loss=0.23461820681889853 +INFO:root: +[Valid][10][UTT][ MSE = 0.2831 | LCC = 0.8348 | SRCC = 0.8443 ] +INFO:root:Average loss=0.23401594887941313 +INFO:root: +[Valid][11][UTT][ MSE = 0.3034 | LCC = 0.8524 | SRCC = 0.8497 ] +INFO:root:Average loss=0.24092318977300936 +INFO:root: +[Valid][12][UTT][ MSE = 0.3044 | LCC = 0.8515 | SRCC = 0.8516 ] +INFO:root:Average loss=0.2333331394654054 +INFO:root: +[Valid][13][UTT][ MSE = 0.2919 | LCC = 0.8483 | SRCC = 0.8508 ] +INFO:root:Average loss=0.22356079748043647 +INFO:root: +[Valid][14][UTT][ MSE = 0.3080 | LCC = 0.8485 | SRCC = 0.8512 ] +INFO:root:Average loss=0.23436596827247205 +INFO:root: +[Valid][15][UTT][ MSE = 0.2740 | LCC = 0.8439 | SRCC = 0.8507 ] +INFO:root:Average loss=0.2339254373159164 +INFO:root: +[Valid][16][UTT][ MSE = 0.3018 | LCC = 0.8516 | SRCC = 0.8503 ] +INFO:root:Average loss=0.2312856272627146 +INFO:root: +[Valid][17][UTT][ MSE = 0.2773 | LCC = 0.8446 | SRCC = 0.8509 ] +INFO:root:Average loss=0.22060217956701914 +INFO:root: +[Valid][18][UTT][ MSE = 0.2694 | LCC = 0.8475 | SRCC = 0.8514 ] +INFO:root:Average loss=0.22078480093907088 +INFO:root: +[Valid][19][UTT][ MSE = 0.2717 | LCC = 0.8473 | SRCC = 0.8491 ] +INFO:root:Average loss=0.2145096588975344 +INFO:root: +[Valid][20][UTT][ MSE = 0.2772 | LCC = 0.8523 | SRCC = 0.8547 ] +INFO:root:Average loss=0.21466333189835915 +INFO:root: +[Valid][21][UTT][ MSE = 0.2670 | LCC = 0.8578 | SRCC = 0.8574 ] +INFO:root:Average loss=0.22296918145357034 +INFO:root: +[Valid][22][UTT][ MSE = 0.2677 | LCC = 0.8536 | SRCC = 0.8559 ] +INFO:root:Average loss=0.21877762751701552 +INFO:root: +[Valid][23][UTT][ MSE = 0.2627 | LCC = 0.8480 | SRCC = 0.8544 ] +INFO:root:Average loss=0.20865256453935915 +INFO:root: +[Valid][24][UTT][ MSE = 0.2578 | LCC = 0.8545 | SRCC = 0.8555 ] +INFO:root:Average loss=0.21400413853235734 +INFO:root: +[Valid][25][UTT][ MSE = 0.2607 | LCC = 0.8512 | SRCC = 0.8548 ] +INFO:root:Average loss=0.21586118772243842 +INFO:root: +[Valid][26][UTT][ MSE = 0.2763 | LCC = 0.8544 | SRCC = 0.8556 ] +INFO:root:Average loss=0.21367784322072297 +INFO:root: +[Valid][27][UTT][ MSE = 0.2733 | LCC = 0.8561 | SRCC = 0.8583 ] +INFO:root:Average loss=0.2176807014605938 +INFO:root: +[Valid][28][UTT][ MSE = 0.2595 | LCC = 0.8507 | SRCC = 0.8577 ] +INFO:root:Average loss=0.21189736670408493 +INFO:root: +[Valid][29][UTT][ MSE = 0.2679 | LCC = 0.8571 | SRCC = 0.8566 ] +INFO:root:Average loss=0.20505159806746703 +INFO:root: +[Valid][30][UTT][ MSE = 0.2652 | LCC = 0.8552 | SRCC = 0.8582 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2707 | LCC = 0.8566 | SRCC = 0.8610 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e2d928f4ba8acbedffcafdeed1042b5c02d471e9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d72f8b6bf5a86405dc6ffc984f8542a04b4c6a16ad2385e07aa58473842041c5 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..d77736a35fa756667616bf7bc2b364c073a1dc09 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f6c966d127747469e7274f6c00e816d3aa00c8fa2fda1c418a9dcda9d2487cc +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..a67d7b7b5990c97d4a6b9530a659ef7421bb6718 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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=0.3573500414689382 +INFO:root: +[Valid][0][UTT][ MSE = 0.3837 | LCC = 0.8041 | SRCC = 0.8038 ] +INFO:root:Average loss=0.3004629203142264 +INFO:root: +[Valid][1][UTT][ MSE = 0.4070 | LCC = 0.8202 | SRCC = 0.8176 ] +INFO:root:Average loss=0.2777913336952527 +INFO:root: +[Valid][2][UTT][ MSE = 0.3454 | LCC = 0.8224 | SRCC = 0.8243 ] +INFO:root:Average loss=0.265379555332355 +INFO:root: +[Valid][3][UTT][ MSE = 0.3175 | LCC = 0.8308 | SRCC = 0.8325 ] +INFO:root:Average loss=0.2613722641880696 +INFO:root: +[Valid][4][UTT][ MSE = 0.3448 | LCC = 0.8390 | SRCC = 0.8372 ] +INFO:root:Average loss=0.24908943684437337 +INFO:root: +[Valid][5][UTT][ MSE = 0.3290 | LCC = 0.8350 | SRCC = 0.8384 ] +INFO:root:Average loss=0.24711873917243418 +INFO:root: +[Valid][6][UTT][ MSE = 0.3418 | LCC = 0.8343 | SRCC = 0.8423 ] +INFO:root:Average loss=0.25122319543972993 +INFO:root: +[Valid][7][UTT][ MSE = 0.3064 | LCC = 0.8464 | SRCC = 0.8475 ] +INFO:root:Average loss=0.24329156199326882 +INFO:root: +[Valid][8][UTT][ MSE = 0.3179 | LCC = 0.8361 | SRCC = 0.8437 ] +INFO:root:Average loss=0.23901490905346015 +INFO:root: +[Valid][9][UTT][ MSE = 0.2932 | LCC = 0.8391 | SRCC = 0.8445 ] +INFO:root:Average loss=0.23581804908238924 +INFO:root: +[Valid][10][UTT][ MSE = 0.3106 | LCC = 0.8510 | SRCC = 0.8485 ] +INFO:root:Average loss=0.23571150539777216 +INFO:root: +[Valid][11][UTT][ MSE = 0.3093 | LCC = 0.8486 | SRCC = 0.8475 ] +INFO:root:Average loss=0.2292786811788877 +INFO:root: +[Valid][12][UTT][ MSE = 0.2824 | LCC = 0.8474 | SRCC = 0.8484 ] +INFO:root:Average loss=0.22692329111771706 +INFO:root: +[Valid][13][UTT][ MSE = 0.2866 | LCC = 0.8507 | SRCC = 0.8506 ] +INFO:root:Average loss=0.2214563813729164 +INFO:root: +[Valid][14][UTT][ MSE = 0.2963 | LCC = 0.8511 | SRCC = 0.8543 ] +INFO:root:Average loss=0.22143000211471167 +INFO:root: +[Valid][15][UTT][ MSE = 0.2807 | LCC = 0.8488 | SRCC = 0.8514 ] +INFO:root:Average loss=0.22660414779033416 +INFO:root: +[Valid][16][UTT][ MSE = 0.2818 | LCC = 0.8488 | SRCC = 0.8520 ] +INFO:root:Average loss=0.22344534624463472 +INFO:root: +[Valid][17][UTT][ MSE = 0.2810 | LCC = 0.8511 | SRCC = 0.8522 ] +INFO:root:Average loss=0.22922055442363787 +INFO:root: +[Valid][18][UTT][ MSE = 0.3006 | LCC = 0.8504 | SRCC = 0.8510 ] +INFO:root:Average loss=0.2185929379402063 +INFO:root: +[Valid][19][UTT][ MSE = 0.2887 | LCC = 0.8569 | SRCC = 0.8564 ] +INFO:root:Average loss=0.2263107425891436 +INFO:root: +[Valid][20][UTT][ MSE = 0.2984 | LCC = 0.8531 | SRCC = 0.8550 ] +INFO:root:Average loss=0.2113683059429511 +INFO:root: +[Valid][21][UTT][ MSE = 0.2932 | LCC = 0.8526 | SRCC = 0.8534 ] +INFO:root:Average loss=0.2164573453557797 +INFO:root: +[Valid][22][UTT][ MSE = 0.2999 | LCC = 0.8543 | SRCC = 0.8572 ] +INFO:root:Average loss=0.22483454931240815 +INFO:root: +[Valid][23][UTT][ MSE = 0.2895 | LCC = 0.8554 | SRCC = 0.8554 ] +INFO:root:Average loss=0.21055445667260733 +INFO:root: +[Valid][24][UTT][ MSE = 0.2793 | LCC = 0.8527 | SRCC = 0.8551 ] +INFO:root:Average loss=0.21028724723519424 +INFO:root: +[Valid][25][UTT][ MSE = 0.3183 | LCC = 0.8564 | SRCC = 0.8543 ] +INFO:root:Average loss=0.2083548529025836 +INFO:root: +[Valid][26][UTT][ MSE = 0.2671 | LCC = 0.8552 | SRCC = 0.8569 ] +INFO:root:Average loss=0.21019324564780945 +INFO:root: +[Valid][27][UTT][ MSE = 0.2760 | LCC = 0.8598 | SRCC = 0.8578 ] +INFO:root:Average loss=0.2047891962604645 +INFO:root: +[Valid][28][UTT][ MSE = 0.2769 | LCC = 0.8548 | SRCC = 0.8574 ] +INFO:root:Average loss=0.21132392971179423 +INFO:root: +[Valid][29][UTT][ MSE = 0.2820 | LCC = 0.8542 | SRCC = 0.8566 ] +INFO:root:Average loss=0.20584440613404298 +INFO:root: +[Valid][30][UTT][ MSE = 0.3120 | LCC = 0.8548 | SRCC = 0.8556 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2657 | LCC = 0.8605 | SRCC = 0.8607 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..3759bd10c8e58ed65c4788cbd6324d0c5275a8e4 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aec50957f51ecfb904c0dc07c58e5c337503e4e620829fba425135099d086e1a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..52eb82ba677381a6455b3f1435c344ee2b97672c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:392a944651f81536762f047fe96ef7eae59498f210cabb903afc33b1fa924656 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..223c8a3de273a29c78c348c2c4a33c0e4f4b5c3d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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=0.35247679379505986 +INFO:root: +[Valid][0][UTT][ MSE = 0.3895 | LCC = 0.7991 | SRCC = 0.8061 ] +INFO:root:Average loss=0.2844558243568127 +INFO:root: +[Valid][1][UTT][ MSE = 0.3762 | LCC = 0.8264 | SRCC = 0.8230 ] +INFO:root:Average loss=0.27488976334914184 +INFO:root: +[Valid][2][UTT][ MSE = 0.3649 | LCC = 0.8289 | SRCC = 0.8280 ] +INFO:root:Average loss=0.25700792364585096 +INFO:root: +[Valid][3][UTT][ MSE = 0.3220 | LCC = 0.8317 | SRCC = 0.8383 ] +INFO:root:Average loss=0.2641121131869463 +INFO:root: +[Valid][4][UTT][ MSE = 0.3698 | LCC = 0.8306 | SRCC = 0.8358 ] +INFO:root:Average loss=0.2475983364841877 +INFO:root: +[Valid][5][UTT][ MSE = 0.3243 | LCC = 0.8423 | SRCC = 0.8436 ] +INFO:root:Average loss=0.25166317935173327 +INFO:root: +[Valid][6][UTT][ MSE = 0.3242 | LCC = 0.8406 | SRCC = 0.8430 ] +INFO:root:Average loss=0.24490525000370467 +INFO:root: +[Valid][7][UTT][ MSE = 0.3059 | LCC = 0.8390 | SRCC = 0.8445 ] +INFO:root:Average loss=0.24681974871036333 +INFO:root: +[Valid][8][UTT][ MSE = 0.3101 | LCC = 0.8466 | SRCC = 0.8464 ] +INFO:root:Average loss=0.2381217842682814 +INFO:root: +[Valid][9][UTT][ MSE = 0.3077 | LCC = 0.8409 | SRCC = 0.8487 ] +INFO:root:Average loss=0.23177969321990624 +INFO:root: +[Valid][10][UTT][ MSE = 0.3327 | LCC = 0.8462 | SRCC = 0.8472 ] +INFO:root:Average loss=0.22922194367035842 +INFO:root: +[Valid][11][UTT][ MSE = 0.3129 | LCC = 0.8519 | SRCC = 0.8508 ] +INFO:root:Average loss=0.23035486787557602 +INFO:root: +[Valid][12][UTT][ MSE = 0.3163 | LCC = 0.8447 | SRCC = 0.8458 ] +INFO:root:Average loss=0.22499829951005104 +INFO:root: +[Valid][13][UTT][ MSE = 0.3427 | LCC = 0.8506 | SRCC = 0.8502 ] +INFO:root:Average loss=0.22630186436268 +INFO:root: +[Valid][14][UTT][ MSE = 0.3291 | LCC = 0.8459 | SRCC = 0.8500 ] +INFO:root:Average loss=0.22547572927597243 +INFO:root: +[Valid][15][UTT][ MSE = 0.3063 | LCC = 0.8551 | SRCC = 0.8553 ] +INFO:root:Average loss=0.22424740420702177 +INFO:root: +[Valid][16][UTT][ MSE = 0.3216 | LCC = 0.8465 | SRCC = 0.8483 ] +INFO:root:Average loss=0.2238525507541803 +INFO:root: +[Valid][17][UTT][ MSE = 0.2842 | LCC = 0.8523 | SRCC = 0.8533 ] +INFO:root:Average loss=0.2196105617361191 +INFO:root: +[Valid][18][UTT][ MSE = 0.3164 | LCC = 0.8488 | SRCC = 0.8522 ] +INFO:root:Average loss=0.22745927098469856 +INFO:root: +[Valid][19][UTT][ MSE = 0.2829 | LCC = 0.8469 | SRCC = 0.8512 ] +INFO:root:Average loss=0.21336027005544075 +INFO:root: +[Valid][20][UTT][ MSE = 0.3133 | LCC = 0.8555 | SRCC = 0.8570 ] +INFO:root:Average loss=0.23036031081126288 +INFO:root: +[Valid][21][UTT][ MSE = 0.3118 | LCC = 0.8573 | SRCC = 0.8558 ] +INFO:root:Average loss=0.20995879669984183 +INFO:root: +[Valid][22][UTT][ MSE = 0.3036 | LCC = 0.8542 | SRCC = 0.8531 ] +INFO:root:Average loss=0.2126657920770156 +INFO:root: +[Valid][23][UTT][ MSE = 0.3057 | LCC = 0.8556 | SRCC = 0.8545 ] +INFO:root:Average loss=0.20798588076081032 +INFO:root: +[Valid][24][UTT][ MSE = 0.2976 | LCC = 0.8559 | SRCC = 0.8573 ] +INFO:root:Average loss=0.20950717287950027 +INFO:root: +[Valid][25][UTT][ MSE = 0.2999 | LCC = 0.8526 | SRCC = 0.8543 ] +INFO:root:Average loss=0.20802895973126093 +INFO:root: +[Valid][26][UTT][ MSE = 0.2844 | LCC = 0.8552 | SRCC = 0.8553 ] +INFO:root:Average loss=0.20003918491494962 +INFO:root: +[Valid][27][UTT][ MSE = 0.2888 | LCC = 0.8547 | SRCC = 0.8543 ] +INFO:root:Average loss=0.20190977868743432 +INFO:root: +[Valid][28][UTT][ MSE = 0.2699 | LCC = 0.8526 | SRCC = 0.8552 ] +INFO:root:Average loss=0.20106139769538856 +INFO:root: +[Valid][29][UTT][ MSE = 0.3181 | LCC = 0.8525 | SRCC = 0.8555 ] +INFO:root:Average loss=0.2022155626462056 +INFO:root: +[Valid][30][UTT][ MSE = 0.2524 | LCC = 0.8556 | SRCC = 0.8597 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2912 | LCC = 0.8649 | SRCC = 0.8639 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..a91edbbb487e922ec16c12141fcf5d5d1e317d06 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66c3986351bc32877bb01572a5e4a27a371286e7cc222cfe357ddd54c50bdaab +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..a94ed525804136412f6f5012875f9c3999cca070 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73f3f766b0e72101d325c93bc386b1e656e5ce88ff8e5c640dcf8a0f003d21a8 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..847e905db1156d13713285e9bc3290b40456bd73 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer11/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=0.35611927929597026 +INFO:root: +[Valid][0][UTT][ MSE = 0.3487 | LCC = 0.8053 | SRCC = 0.8079 ] +INFO:root:Average loss=0.2878641025760235 +INFO:root: +[Valid][1][UTT][ MSE = 0.3443 | LCC = 0.8240 | SRCC = 0.8221 ] +INFO:root:Average loss=0.2715938889827484 +INFO:root: +[Valid][2][UTT][ MSE = 0.3296 | LCC = 0.8303 | SRCC = 0.8307 ] +INFO:root:Average loss=0.26022405807788557 +INFO:root: +[Valid][3][UTT][ MSE = 0.3227 | LCC = 0.8380 | SRCC = 0.8370 ] +INFO:root:Average loss=0.26535990318426717 +INFO:root: +[Valid][4][UTT][ MSE = 0.3538 | LCC = 0.8430 | SRCC = 0.8415 ] +INFO:root:Average loss=0.261014176102785 +INFO:root: +[Valid][5][UTT][ MSE = 0.2911 | LCC = 0.8432 | SRCC = 0.8419 ] +INFO:root:Average loss=0.2521813122125772 +INFO:root: +[Valid][6][UTT][ MSE = 0.3162 | LCC = 0.8427 | SRCC = 0.8435 ] +INFO:root:Average loss=0.2507454982170692 +INFO:root: +[Valid][7][UTT][ MSE = 0.3464 | LCC = 0.8457 | SRCC = 0.8459 ] +INFO:root:Average loss=0.2514904108949197 +INFO:root: +[Valid][8][UTT][ MSE = 0.3399 | LCC = 0.8409 | SRCC = 0.8418 ] +INFO:root:Average loss=0.24518005511699578 +INFO:root: +[Valid][9][UTT][ MSE = 0.3262 | LCC = 0.8446 | SRCC = 0.8476 ] +INFO:root:Average loss=0.23434360736073592 +INFO:root: +[Valid][10][UTT][ MSE = 0.2972 | LCC = 0.8509 | SRCC = 0.8487 ] +INFO:root:Average loss=0.22748110462457705 +INFO:root: +[Valid][11][UTT][ MSE = 0.3310 | LCC = 0.8486 | SRCC = 0.8483 ] +INFO:root:Average loss=0.23707101532281974 +INFO:root: +[Valid][12][UTT][ MSE = 0.2953 | LCC = 0.8459 | SRCC = 0.8462 ] +INFO:root:Average loss=0.2314557060599327 +INFO:root: +[Valid][13][UTT][ MSE = 0.3331 | LCC = 0.8439 | SRCC = 0.8439 ] +INFO:root:Average loss=0.22814995566239724 +INFO:root: +[Valid][14][UTT][ MSE = 0.2863 | LCC = 0.8477 | SRCC = 0.8494 ] +INFO:root:Average loss=0.22926439765172127 +INFO:root: +[Valid][15][UTT][ MSE = 0.3108 | LCC = 0.8506 | SRCC = 0.8499 ] +INFO:root:Average loss=0.2300266982653202 +INFO:root: +[Valid][16][UTT][ MSE = 0.2953 | LCC = 0.8440 | SRCC = 0.8520 ] +INFO:root:Average loss=0.22788987117700088 +INFO:root: +[Valid][17][UTT][ MSE = 0.2721 | LCC = 0.8502 | SRCC = 0.8513 ] +INFO:root:Average loss=0.21819452731273112 +INFO:root: +[Valid][18][UTT][ MSE = 0.3006 | LCC = 0.8431 | SRCC = 0.8506 ] +INFO:root:Average loss=0.22365357726812363 +INFO:root: +[Valid][19][UTT][ MSE = 0.2851 | LCC = 0.8523 | SRCC = 0.8547 ] +INFO:root:Average loss=0.2165249391244008 +INFO:root: +[Valid][20][UTT][ MSE = 0.2865 | LCC = 0.8516 | SRCC = 0.8521 ] +INFO:root:Average loss=0.21395557259137815 +INFO:root: +[Valid][21][UTT][ MSE = 0.2826 | LCC = 0.8515 | SRCC = 0.8556 ] +INFO:root:Average loss=0.2137471343844365 +INFO:root: +[Valid][22][UTT][ MSE = 0.2765 | LCC = 0.8513 | SRCC = 0.8531 ] +INFO:root:Average loss=0.2160283883030598 +INFO:root: +[Valid][23][UTT][ MSE = 0.2934 | LCC = 0.8574 | SRCC = 0.8553 ] +INFO:root:Average loss=0.21312226125827202 +INFO:root: +[Valid][24][UTT][ MSE = 0.2757 | LCC = 0.8528 | SRCC = 0.8544 ] +INFO:root:Average loss=0.21504421933339193 +INFO:root: +[Valid][25][UTT][ MSE = 0.2636 | LCC = 0.8546 | SRCC = 0.8546 ] +INFO:root:Average loss=0.20571426321298647 +INFO:root: +[Valid][26][UTT][ MSE = 0.3045 | LCC = 0.8596 | SRCC = 0.8573 ] +INFO:root:Average loss=0.20549300733285072 +INFO:root: +[Valid][27][UTT][ MSE = 0.2663 | LCC = 0.8592 | SRCC = 0.8595 ] +INFO:root:Average loss=0.20659546181559563 +INFO:root: +[Valid][28][UTT][ MSE = 0.2902 | LCC = 0.8560 | SRCC = 0.8540 ] +INFO:root:Average loss=0.205801317993647 +INFO:root: +[Valid][29][UTT][ MSE = 0.2751 | LCC = 0.8590 | SRCC = 0.8594 ] +INFO:root:Average loss=0.20145443588113174 +INFO:root: +[Valid][30][UTT][ MSE = 0.2719 | LCC = 0.8596 | SRCC = 0.8608 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2824 | LCC = 0.8563 | SRCC = 0.8596 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..8a5ee763d114f77eb96536c38e8ff7c935690c47 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1db17d41811e463d661985b257205c5f55653efe2f24325b91c34eb790b5827d +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..da5372009992e6ea4ee07843c1ef8a069130de0e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:17ee4e3ceee9d479f0714e7cdc6aee101e88380585736fda25960fab80290e6a +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..01075adb2490e7121cff797225980af94b3729ad --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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=0.5755800174979063 +INFO:root: +[Valid][0][UTT][ MSE = 0.3311 | LCC = 0.8036 | SRCC = 0.8103 ] +INFO:root:Average loss=0.24553625285625458 +INFO:root: +[Valid][1][UTT][ MSE = 0.2892 | LCC = 0.8446 | SRCC = 0.8433 ] +INFO:root:Average loss=0.21585702857910058 +INFO:root: +[Valid][2][UTT][ MSE = 0.2512 | LCC = 0.8503 | SRCC = 0.8552 ] +INFO:root:Average loss=0.19887409875026116 +INFO:root: +[Valid][3][UTT][ MSE = 0.2591 | LCC = 0.8530 | SRCC = 0.8557 ] +INFO:root:Average loss=0.1835118588537742 +INFO:root: +[Valid][4][UTT][ MSE = 0.2353 | LCC = 0.8587 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1719862780509851 +INFO:root: +[Valid][5][UTT][ MSE = 0.2366 | LCC = 0.8609 | SRCC = 0.8612 ] +INFO:root:Average loss=0.16326061989634466 +INFO:root: +[Valid][6][UTT][ MSE = 0.2280 | LCC = 0.8656 | SRCC = 0.8655 ] +INFO:root:Average loss=0.15126964086905503 +INFO:root: +[Valid][7][UTT][ MSE = 0.2184 | LCC = 0.8665 | SRCC = 0.8670 ] +INFO:root:Average loss=0.1427211791085891 +INFO:root: +[Valid][8][UTT][ MSE = 0.2324 | LCC = 0.8627 | SRCC = 0.8627 ] +INFO:root:Average loss=0.137933998917922 +INFO:root: +[Valid][9][UTT][ MSE = 0.2762 | LCC = 0.8660 | SRCC = 0.8675 ] +INFO:root:Average loss=0.13058648745600993 +INFO:root: +[Valid][10][UTT][ MSE = 0.2218 | LCC = 0.8648 | SRCC = 0.8656 ] +INFO:root:Average loss=0.1156923645772995 +INFO:root: +[Valid][11][UTT][ MSE = 0.2199 | LCC = 0.8648 | SRCC = 0.8649 ] +INFO:root:Average loss=0.11455836701087463 +INFO:root: +[Valid][12][UTT][ MSE = 0.2198 | LCC = 0.8630 | SRCC = 0.8636 ] +INFO:root:Average loss=0.11014084441539569 +INFO:root: +[Valid][13][UTT][ MSE = 0.2300 | LCC = 0.8678 | SRCC = 0.8680 ] +INFO:root:Average loss=0.09617698866014297 +INFO:root: +[Valid][14][UTT][ MSE = 0.2280 | LCC = 0.8653 | SRCC = 0.8644 ] +INFO:root:Average loss=0.0930055456283765 +INFO:root: +[Valid][15][UTT][ MSE = 0.2239 | LCC = 0.8637 | SRCC = 0.8655 ] +INFO:root:Average loss=0.09004347895582517 +INFO:root: +[Valid][16][UTT][ MSE = 0.2201 | LCC = 0.8689 | SRCC = 0.8693 ] +INFO:root:Average loss=0.08631087853931464 +INFO:root: +[Valid][17][UTT][ MSE = 0.2213 | LCC = 0.8684 | SRCC = 0.8687 ] +INFO:root:Average loss=0.07969817533515967 +INFO:root: +[Valid][18][UTT][ MSE = 0.2286 | LCC = 0.8623 | SRCC = 0.8625 ] +INFO:root:Average loss=0.0778472525282548 +INFO:root: +[Valid][19][UTT][ MSE = 0.2239 | LCC = 0.8652 | SRCC = 0.8650 ] +INFO:root:Average loss=0.07282580726612838 +INFO:root: +[Valid][20][UTT][ MSE = 0.2152 | LCC = 0.8648 | SRCC = 0.8661 ] +INFO:root:Average loss=0.07048238699252789 +INFO:root: +[Valid][21][UTT][ MSE = 0.2172 | LCC = 0.8664 | SRCC = 0.8669 ] +INFO:root:Average loss=0.07466754216987354 +INFO:root: +[Valid][22][UTT][ MSE = 0.2237 | LCC = 0.8657 | SRCC = 0.8654 ] +INFO:root:Average loss=0.06779841307359628 +INFO:root: +[Valid][23][UTT][ MSE = 0.2241 | LCC = 0.8626 | SRCC = 0.8630 ] +INFO:root:Average loss=0.06371982037447967 +INFO:root: +[Valid][24][UTT][ MSE = 0.2277 | LCC = 0.8621 | SRCC = 0.8620 ] +INFO:root:Average loss=0.05989357203436203 +INFO:root: +[Valid][25][UTT][ MSE = 0.2335 | LCC = 0.8633 | SRCC = 0.8630 ] +INFO:root:Average loss=0.05993656451121355 +INFO:root: +[Valid][26][UTT][ MSE = 0.2148 | LCC = 0.8644 | SRCC = 0.8644 ] +INFO:root:Average loss=0.059263804354346715 +INFO:root: +[Valid][27][UTT][ MSE = 0.2248 | LCC = 0.8646 | SRCC = 0.8648 ] +INFO:root:Average loss=0.05905124339728783 +INFO:root: +[Valid][28][UTT][ MSE = 0.2422 | LCC = 0.8635 | SRCC = 0.8636 ] +INFO:root:Average loss=0.056386042529573806 +INFO:root: +[Valid][29][UTT][ MSE = 0.2185 | LCC = 0.8655 | SRCC = 0.8655 ] +INFO:root:Average loss=0.05472555441352037 +INFO:root: +[Valid][30][UTT][ MSE = 0.2231 | LCC = 0.8663 | SRCC = 0.8666 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2418 | LCC = 0.8702 | SRCC = 0.8712 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..db58b6c0f1fcde2caffeb1761ebee1d0826f24ea --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:413775086fb94784c3f043114e7e55048ef2ed91d77854394f61c1064c7b67b3 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..577c5596d251e0e8e15d98d4ef05561d0da31ef9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0bc72c0013ffdfc54558cc735eccf3db4ac3157047b2fedda9144998d0a808fb +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..58891dc14c401fab5d23099cf14713acd745bada --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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=0.42613556828254306 +INFO:root: +[Valid][0][UTT][ MSE = 0.3339 | LCC = 0.8046 | SRCC = 0.8173 ] +INFO:root:Average loss=0.2442050366065441 +INFO:root: +[Valid][1][UTT][ MSE = 0.3288 | LCC = 0.8367 | SRCC = 0.8413 ] +INFO:root:Average loss=0.22009821341205865 +INFO:root: +[Valid][2][UTT][ MSE = 0.2523 | LCC = 0.8528 | SRCC = 0.8524 ] +INFO:root:Average loss=0.19535815161772263 +INFO:root: +[Valid][3][UTT][ MSE = 0.2473 | LCC = 0.8585 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1778540286498192 +INFO:root: +[Valid][4][UTT][ MSE = 0.2387 | LCC = 0.8607 | SRCC = 0.8595 ] +INFO:root:Average loss=0.17400876614145744 +INFO:root: +[Valid][5][UTT][ MSE = 0.2444 | LCC = 0.8612 | SRCC = 0.8604 ] +INFO:root:Average loss=0.15864905991997474 +INFO:root: +[Valid][6][UTT][ MSE = 0.2453 | LCC = 0.8598 | SRCC = 0.8607 ] +INFO:root:Average loss=0.1514845483769209 +INFO:root: +[Valid][7][UTT][ MSE = 0.2349 | LCC = 0.8675 | SRCC = 0.8663 ] +INFO:root:Average loss=0.14585120670306376 +INFO:root: +[Valid][8][UTT][ MSE = 0.2355 | LCC = 0.8636 | SRCC = 0.8654 ] +INFO:root:Average loss=0.13831403020482796 +INFO:root: +[Valid][9][UTT][ MSE = 0.2323 | LCC = 0.8680 | SRCC = 0.8698 ] +INFO:root:Average loss=0.12585434250724623 +INFO:root: +[Valid][10][UTT][ MSE = 0.2218 | LCC = 0.8683 | SRCC = 0.8685 ] +INFO:root:Average loss=0.11994030660925767 +INFO:root: +[Valid][11][UTT][ MSE = 0.2256 | LCC = 0.8679 | SRCC = 0.8674 ] +INFO:root:Average loss=0.1115102010468642 +INFO:root: +[Valid][12][UTT][ MSE = 0.2202 | LCC = 0.8678 | SRCC = 0.8667 ] +INFO:root:Average loss=0.11346313987786953 +INFO:root: +[Valid][13][UTT][ MSE = 0.2983 | LCC = 0.8616 | SRCC = 0.8652 ] +INFO:root:Average loss=0.10211200639605522 +INFO:root: +[Valid][14][UTT][ MSE = 0.2143 | LCC = 0.8697 | SRCC = 0.8700 ] +INFO:root:Average loss=0.09530460204069431 +INFO:root: +[Valid][15][UTT][ MSE = 0.2306 | LCC = 0.8671 | SRCC = 0.8692 ] +INFO:root:Average loss=0.0959089566499759 +INFO:root: +[Valid][16][UTT][ MSE = 0.2690 | LCC = 0.8701 | SRCC = 0.8709 ] +INFO:root:Average loss=0.09350262376933526 +INFO:root: +[Valid][17][UTT][ MSE = 0.2272 | LCC = 0.8689 | SRCC = 0.8690 ] +INFO:root:Average loss=0.08216841189334026 +INFO:root: +[Valid][18][UTT][ MSE = 0.2193 | LCC = 0.8705 | SRCC = 0.8710 ] +INFO:root:Average loss=0.0824619995859953 +INFO:root: +[Valid][19][UTT][ MSE = 0.2304 | LCC = 0.8675 | SRCC = 0.8674 ] +INFO:root:Average loss=0.07179319003644662 +INFO:root: +[Valid][20][UTT][ MSE = 0.2316 | LCC = 0.8676 | SRCC = 0.8674 ] +INFO:root:Average loss=0.07369422234403782 +INFO:root: +[Valid][21][UTT][ MSE = 0.2492 | LCC = 0.8667 | SRCC = 0.8668 ] +INFO:root:Average loss=0.0650312714278698 +INFO:root: +[Valid][22][UTT][ MSE = 0.2303 | LCC = 0.8644 | SRCC = 0.8636 ] +INFO:root:Average loss=0.06843935965727536 +INFO:root: +[Valid][23][UTT][ MSE = 0.2186 | LCC = 0.8679 | SRCC = 0.8674 ] +INFO:root:Average loss=0.06255641288291185 +INFO:root: +[Valid][24][UTT][ MSE = 0.2165 | LCC = 0.8671 | SRCC = 0.8667 ] +INFO:root:Average loss=0.06139328483587656 +INFO:root: +[Valid][25][UTT][ MSE = 0.2203 | LCC = 0.8665 | SRCC = 0.8655 ] +INFO:root:Average loss=0.06953091733157635 +INFO:root: +[Valid][26][UTT][ MSE = 0.2137 | LCC = 0.8689 | SRCC = 0.8682 ] +INFO:root:Average loss=0.06081899446554673 +INFO:root: +[Valid][27][UTT][ MSE = 0.2371 | LCC = 0.8662 | SRCC = 0.8662 ] +INFO:root:Average loss=0.054146875985539876 +INFO:root: +[Valid][28][UTT][ MSE = 0.2409 | LCC = 0.8646 | SRCC = 0.8641 ] +INFO:root:Average loss=0.055704907060433656 +INFO:root: +[Valid][29][UTT][ MSE = 0.2157 | LCC = 0.8683 | SRCC = 0.8676 ] +INFO:root:Average loss=0.057436032411761775 +INFO:root: +[Valid][30][UTT][ MSE = 0.2682 | LCC = 0.8652 | SRCC = 0.8656 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2201 | LCC = 0.8672 | SRCC = 0.8666 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..c6509dee738657741e5b4dcaafc582ff849d431b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d88d930c3cd51e211dc536507bd0810d60d8ec7e1d78aa1e2ae8a4b59bb1bc3 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..14a279d74e939bc931a182ae643f6c373054e081 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4735e16390fb899541490402a31fa869827942d3a70110da13d8124b7ef55bd9 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4a4301e7d32c38d849aa4e65435927e569076b41 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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=0.38328731862398296 +INFO:root: +[Valid][0][UTT][ MSE = 0.3021 | LCC = 0.8147 | SRCC = 0.8277 ] +INFO:root:Average loss=0.23432502284263954 +INFO:root: +[Valid][1][UTT][ MSE = 0.2792 | LCC = 0.8447 | SRCC = 0.8536 ] +INFO:root:Average loss=0.208887215417165 +INFO:root: +[Valid][2][UTT][ MSE = 0.2509 | LCC = 0.8564 | SRCC = 0.8582 ] +INFO:root:Average loss=0.19294263747258064 +INFO:root: +[Valid][3][UTT][ MSE = 0.2615 | LCC = 0.8618 | SRCC = 0.8637 ] +INFO:root:Average loss=0.17943988740444183 +INFO:root: +[Valid][4][UTT][ MSE = 0.2393 | LCC = 0.8675 | SRCC = 0.8663 ] +INFO:root:Average loss=0.16703817668633583 +INFO:root: +[Valid][5][UTT][ MSE = 0.2198 | LCC = 0.8684 | SRCC = 0.8673 ] +INFO:root:Average loss=0.15110275388146058 +INFO:root: +[Valid][6][UTT][ MSE = 0.2481 | LCC = 0.8680 | SRCC = 0.8685 ] +INFO:root:Average loss=0.14555328788283545 +INFO:root: +[Valid][7][UTT][ MSE = 0.2241 | LCC = 0.8698 | SRCC = 0.8697 ] +INFO:root:Average loss=0.1353987710407147 +INFO:root: +[Valid][8][UTT][ MSE = 0.2885 | LCC = 0.8633 | SRCC = 0.8680 ] +INFO:root:Average loss=0.12879409115666 +INFO:root: +[Valid][9][UTT][ MSE = 0.2026 | LCC = 0.8726 | SRCC = 0.8718 ] +INFO:root:Average loss=0.11862330635388692 +INFO:root: +[Valid][10][UTT][ MSE = 0.2191 | LCC = 0.8705 | SRCC = 0.8693 ] +INFO:root:Average loss=0.12293148642549148 +INFO:root: +[Valid][11][UTT][ MSE = 0.2170 | LCC = 0.8697 | SRCC = 0.8690 ] +INFO:root:Average loss=0.10669159717284717 +INFO:root: +[Valid][12][UTT][ MSE = 0.2128 | LCC = 0.8734 | SRCC = 0.8735 ] +INFO:root:Average loss=0.1060151242865966 +INFO:root: +[Valid][13][UTT][ MSE = 0.2332 | LCC = 0.8688 | SRCC = 0.8712 ] +INFO:root:Average loss=0.10426698109278312 +INFO:root: +[Valid][14][UTT][ MSE = 0.2063 | LCC = 0.8723 | SRCC = 0.8723 ] +INFO:root:Average loss=0.09724632392709072 +INFO:root: +[Valid][15][UTT][ MSE = 0.2047 | LCC = 0.8743 | SRCC = 0.8741 ] +INFO:root:Average loss=0.08629898378291191 +INFO:root: +[Valid][16][UTT][ MSE = 0.2073 | LCC = 0.8705 | SRCC = 0.8712 ] +INFO:root:Average loss=0.08502250456084044 +INFO:root: +[Valid][17][UTT][ MSE = 0.2194 | LCC = 0.8680 | SRCC = 0.8677 ] +INFO:root:Average loss=0.07938827846485835 +INFO:root: +[Valid][18][UTT][ MSE = 0.2105 | LCC = 0.8707 | SRCC = 0.8707 ] +INFO:root:Average loss=0.07428172455193141 +INFO:root: +[Valid][19][UTT][ MSE = 0.2072 | LCC = 0.8694 | SRCC = 0.8693 ] +INFO:root:Average loss=0.07891535873596485 +INFO:root: +[Valid][20][UTT][ MSE = 0.2103 | LCC = 0.8715 | SRCC = 0.8711 ] +INFO:root:Average loss=0.06837612624542835 +INFO:root: +[Valid][21][UTT][ MSE = 0.2044 | LCC = 0.8702 | SRCC = 0.8713 ] +INFO:root:Average loss=0.06809659875356235 +INFO:root: +[Valid][22][UTT][ MSE = 0.2236 | LCC = 0.8692 | SRCC = 0.8691 ] +INFO:root:Average loss=0.0656798486239635 +INFO:root: +[Valid][23][UTT][ MSE = 0.2194 | LCC = 0.8675 | SRCC = 0.8675 ] +INFO:root:Average loss=0.06154260430962612 +INFO:root: +[Valid][24][UTT][ MSE = 0.2135 | LCC = 0.8674 | SRCC = 0.8676 ] +INFO:root:Average loss=0.06367634184276447 +INFO:root: +[Valid][25][UTT][ MSE = 0.2051 | LCC = 0.8694 | SRCC = 0.8693 ] +INFO:root:Average loss=0.05841111449094919 +INFO:root: +[Valid][26][UTT][ MSE = 0.2346 | LCC = 0.8669 | SRCC = 0.8664 ] +INFO:root:Average loss=0.058681668188327395 +INFO:root: +[Valid][27][UTT][ MSE = 0.2185 | LCC = 0.8689 | SRCC = 0.8691 ] +INFO:root:Average loss=0.053664946665939614 +INFO:root: +[Valid][28][UTT][ MSE = 0.2084 | LCC = 0.8698 | SRCC = 0.8710 ] +INFO:root:Average loss=0.05334222417038221 +INFO:root: +[Valid][29][UTT][ MSE = 0.2182 | LCC = 0.8692 | SRCC = 0.8695 ] +INFO:root:Average loss=0.053594930813862726 +INFO:root: +[Valid][30][UTT][ MSE = 0.2126 | LCC = 0.8699 | SRCC = 0.8704 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2138 | LCC = 0.8668 | SRCC = 0.8675 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..77b3775f2162888ec6d2249f08625fe02a01042b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b96894cb313602ddb02afbe09ded9dc529c7c1b5d8a72973783dde96129beb6 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..d3a6cebb595181c2f0abf89e9c2cf92793439c96 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e2c50110bfa8d5a92f158b05d9e5dc4fe7eb483ae7e7522b77530f5848d2cb7f +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..95c023fabf7b85984903447f23e10cdcf0a5a4d8 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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=0.4346654407488994 +INFO:root: +[Valid][0][UTT][ MSE = 0.3215 | LCC = 0.8051 | SRCC = 0.8099 ] +INFO:root:Average loss=0.2456913051696924 +INFO:root: +[Valid][1][UTT][ MSE = 0.2750 | LCC = 0.8436 | SRCC = 0.8444 ] +INFO:root:Average loss=0.2107911264667144 +INFO:root: +[Valid][2][UTT][ MSE = 0.2633 | LCC = 0.8520 | SRCC = 0.8509 ] +INFO:root:Average loss=0.19485732244375426 +INFO:root: +[Valid][3][UTT][ MSE = 0.2524 | LCC = 0.8548 | SRCC = 0.8562 ] +INFO:root:Average loss=0.18107353103084442 +INFO:root: +[Valid][4][UTT][ MSE = 0.2408 | LCC = 0.8586 | SRCC = 0.8584 ] +INFO:root:Average loss=0.17112600688750929 +INFO:root: +[Valid][5][UTT][ MSE = 0.2378 | LCC = 0.8611 | SRCC = 0.8605 ] +INFO:root:Average loss=0.15482544832122633 +INFO:root: +[Valid][6][UTT][ MSE = 0.2391 | LCC = 0.8577 | SRCC = 0.8578 ] +INFO:root:Average loss=0.15738575351543915 +INFO:root: +[Valid][7][UTT][ MSE = 0.2511 | LCC = 0.8604 | SRCC = 0.8599 ] +INFO:root:Average loss=0.14151906986267138 +INFO:root: +[Valid][8][UTT][ MSE = 0.2501 | LCC = 0.8554 | SRCC = 0.8584 ] +INFO:root:Average loss=0.1434709623647042 +INFO:root: +[Valid][9][UTT][ MSE = 0.2243 | LCC = 0.8607 | SRCC = 0.8627 ] +INFO:root:Average loss=0.1320912510347672 +INFO:root: +[Valid][10][UTT][ MSE = 0.2320 | LCC = 0.8603 | SRCC = 0.8612 ] +INFO:root:Average loss=0.12326153109853084 +INFO:root: +[Valid][11][UTT][ MSE = 0.2312 | LCC = 0.8605 | SRCC = 0.8607 ] +INFO:root:Average loss=0.11110419254654492 +INFO:root: +[Valid][12][UTT][ MSE = 0.2523 | LCC = 0.8574 | SRCC = 0.8545 ] +INFO:root:Average loss=0.10876463821683174 +INFO:root: +[Valid][13][UTT][ MSE = 0.2265 | LCC = 0.8613 | SRCC = 0.8600 ] +INFO:root:Average loss=0.09852305532265933 +INFO:root: +[Valid][14][UTT][ MSE = 0.2406 | LCC = 0.8605 | SRCC = 0.8590 ] +INFO:root:Average loss=0.09830836655619817 +INFO:root: +[Valid][15][UTT][ MSE = 0.2331 | LCC = 0.8572 | SRCC = 0.8574 ] +INFO:root:Average loss=0.08950293064117432 +INFO:root: +[Valid][16][UTT][ MSE = 0.2285 | LCC = 0.8575 | SRCC = 0.8557 ] +INFO:root:Average loss=0.0911178643313738 +INFO:root: +[Valid][17][UTT][ MSE = 0.2454 | LCC = 0.8557 | SRCC = 0.8557 ] +INFO:root:Average loss=0.08585754337792213 +INFO:root: +[Valid][18][UTT][ MSE = 0.2583 | LCC = 0.8534 | SRCC = 0.8524 ] +INFO:root:Average loss=0.07923929913876912 +INFO:root: +[Valid][19][UTT][ MSE = 0.2432 | LCC = 0.8532 | SRCC = 0.8542 ] +INFO:root:Average loss=0.07608519098124443 +INFO:root: +[Valid][20][UTT][ MSE = 0.2412 | LCC = 0.8532 | SRCC = 0.8533 ] +INFO:root:Average loss=0.0765822785309492 +INFO:root: +[Valid][21][UTT][ MSE = 0.2300 | LCC = 0.8535 | SRCC = 0.8537 ] +INFO:root:Average loss=0.07356323593128951 +INFO:root: +[Valid][22][UTT][ MSE = 0.2373 | LCC = 0.8538 | SRCC = 0.8539 ] +INFO:root:Average loss=0.06662396408426456 +INFO:root: +[Valid][23][UTT][ MSE = 0.2288 | LCC = 0.8567 | SRCC = 0.8558 ] +INFO:root:Average loss=0.06212254408269356 +INFO:root: +[Valid][24][UTT][ MSE = 0.2360 | LCC = 0.8542 | SRCC = 0.8540 ] +INFO:root:Average loss=0.06563702650750294 +INFO:root: +[Valid][25][UTT][ MSE = 0.2350 | LCC = 0.8551 | SRCC = 0.8532 ] +INFO:root:Average loss=0.061967470277196325 +INFO:root: +[Valid][26][UTT][ MSE = 0.2518 | LCC = 0.8521 | SRCC = 0.8521 ] +INFO:root:Average loss=0.05924618158202905 +INFO:root: +[Valid][27][UTT][ MSE = 0.2387 | LCC = 0.8574 | SRCC = 0.8562 ] +INFO:root:Average loss=0.060254556294053026 +INFO:root: +[Valid][28][UTT][ MSE = 0.2364 | LCC = 0.8563 | SRCC = 0.8562 ] +INFO:root:Average loss=0.05588365962299017 +INFO:root: +[Valid][29][UTT][ MSE = 0.2389 | LCC = 0.8550 | SRCC = 0.8535 ] +INFO:root:Average loss=0.05115199280090821 +INFO:root: +[Valid][30][UTT][ MSE = 0.2489 | LCC = 0.8552 | SRCC = 0.8553 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2259 | LCC = 0.8729 | SRCC = 0.8717 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..c1a0ed9a564dfd8ac7536fb4b4ba6ad5b0ec067f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8fad4eac3c415f6db4ca72323dd4a956dc8e8a21ff818b9499d70d2452a65015 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e4f1f673d1e453b6c94d11932c250e7216e5a39d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1dfe3d62037618d40f550ce5806c311bad76ed4938d326e28534d66c2109f20b +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0f0ebd4fb971b4d2761dc928503a18a4bcd8e9c7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer2/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=0.5183176149924597 +INFO:root: +[Valid][0][UTT][ MSE = 0.3402 | LCC = 0.7983 | SRCC = 0.8072 ] +INFO:root:Average loss=0.25404294771261704 +INFO:root: +[Valid][1][UTT][ MSE = 0.2913 | LCC = 0.8389 | SRCC = 0.8465 ] +INFO:root:Average loss=0.21286605222102922 +INFO:root: +[Valid][2][UTT][ MSE = 0.2467 | LCC = 0.8547 | SRCC = 0.8569 ] +INFO:root:Average loss=0.1956100351153276 +INFO:root: +[Valid][3][UTT][ MSE = 0.2362 | LCC = 0.8636 | SRCC = 0.8637 ] +INFO:root:Average loss=0.17954757628150475 +INFO:root: +[Valid][4][UTT][ MSE = 0.2252 | LCC = 0.8627 | SRCC = 0.8650 ] +INFO:root:Average loss=0.17039882926604685 +INFO:root: +[Valid][5][UTT][ MSE = 0.2194 | LCC = 0.8678 | SRCC = 0.8679 ] +INFO:root:Average loss=0.15473800611037475 +INFO:root: +[Valid][6][UTT][ MSE = 0.2220 | LCC = 0.8681 | SRCC = 0.8693 ] +INFO:root:Average loss=0.1437973303672595 +INFO:root: +[Valid][7][UTT][ MSE = 0.2353 | LCC = 0.8694 | SRCC = 0.8697 ] +INFO:root:Average loss=0.1406598038589343 +INFO:root: +[Valid][8][UTT][ MSE = 0.2229 | LCC = 0.8669 | SRCC = 0.8688 ] +INFO:root:Average loss=0.1255074146275337 +INFO:root: +[Valid][9][UTT][ MSE = 0.2113 | LCC = 0.8718 | SRCC = 0.8725 ] +INFO:root:Average loss=0.12084513386854759 +INFO:root: +[Valid][10][UTT][ MSE = 0.2070 | LCC = 0.8716 | SRCC = 0.8721 ] +INFO:root:Average loss=0.11598542428169495 +INFO:root: +[Valid][11][UTT][ MSE = 0.2023 | LCC = 0.8724 | SRCC = 0.8722 ] +INFO:root:Average loss=0.11516084082615681 +INFO:root: +[Valid][12][UTT][ MSE = 0.2102 | LCC = 0.8701 | SRCC = 0.8695 ] +INFO:root:Average loss=0.10360881509498143 +INFO:root: +[Valid][13][UTT][ MSE = 0.2067 | LCC = 0.8698 | SRCC = 0.8696 ] +INFO:root:Average loss=0.09775147534524783 +INFO:root: +[Valid][14][UTT][ MSE = 0.2139 | LCC = 0.8699 | SRCC = 0.8699 ] +INFO:root:Average loss=0.08792002924168721 +INFO:root: +[Valid][15][UTT][ MSE = 0.2102 | LCC = 0.8658 | SRCC = 0.8670 ] +INFO:root:Average loss=0.09280985803940357 +INFO:root: +[Valid][16][UTT][ MSE = 0.2086 | LCC = 0.8690 | SRCC = 0.8677 ] +INFO:root:Average loss=0.0769442127396663 +INFO:root: +[Valid][17][UTT][ MSE = 0.2203 | LCC = 0.8662 | SRCC = 0.8648 ] +INFO:root:Average loss=0.0760998343810057 +INFO:root: +[Valid][18][UTT][ MSE = 0.2180 | LCC = 0.8679 | SRCC = 0.8674 ] +INFO:root:Average loss=0.07305991191130418 +INFO:root: +[Valid][19][UTT][ MSE = 0.2088 | LCC = 0.8663 | SRCC = 0.8662 ] +INFO:root:Average loss=0.07330811038040198 +INFO:root: +[Valid][20][UTT][ MSE = 0.2053 | LCC = 0.8693 | SRCC = 0.8696 ] +INFO:root:Average loss=0.07110693157674411 +INFO:root: +[Valid][21][UTT][ MSE = 0.2174 | LCC = 0.8674 | SRCC = 0.8671 ] +INFO:root:Average loss=0.06349856010041176 +INFO:root: +[Valid][22][UTT][ MSE = 0.2058 | LCC = 0.8669 | SRCC = 0.8673 ] +INFO:root:Average loss=0.06230842446287473 +INFO:root: +[Valid][23][UTT][ MSE = 0.2318 | LCC = 0.8656 | SRCC = 0.8656 ] +INFO:root:Average loss=0.06118717932930359 +INFO:root: +[Valid][24][UTT][ MSE = 0.2143 | LCC = 0.8660 | SRCC = 0.8656 ] +INFO:root:Average loss=0.057917818713646665 +INFO:root: +[Valid][25][UTT][ MSE = 0.2164 | LCC = 0.8640 | SRCC = 0.8636 ] +INFO:root:Average loss=0.0537607007397291 +INFO:root: +[Valid][26][UTT][ MSE = 0.2160 | LCC = 0.8656 | SRCC = 0.8652 ] +INFO:root:Average loss=0.05216836489928074 +INFO:root: +[Valid][27][UTT][ MSE = 0.2128 | LCC = 0.8658 | SRCC = 0.8649 ] +INFO:root:Average loss=0.052562835650184214 +INFO:root: +[Valid][28][UTT][ MSE = 0.2182 | LCC = 0.8649 | SRCC = 0.8657 ] +INFO:root:Average loss=0.04753824624304588 +INFO:root: +[Valid][29][UTT][ MSE = 0.2154 | LCC = 0.8647 | SRCC = 0.8649 ] +INFO:root:Average loss=0.0468658534093545 +INFO:root: +[Valid][30][UTT][ MSE = 0.2158 | LCC = 0.8626 | SRCC = 0.8630 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2072 | LCC = 0.8715 | SRCC = 0.8725 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..1926e4b2acbfd5b3544ed47330a57636d563cc8c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:09307c7492855fc36267967b6d6f9ce65c761cb137bb09bdd6a5e4502dafd960 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..9ae563ccb16cd9e6c457417479bd107f7000d171 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a462812ea2be0340c37694309fb1ccaa12cdb0d9f64bcdfb1d02c5a813995f4e +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..a0fddc8bd4aceb183ee165fccffeeead3de01875 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/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=0.355921008456976 +INFO:root: +[Valid][0][UTT][ MSE = 0.3116 | LCC = 0.8235 | SRCC = 0.8300 ] +INFO:root:Average loss=0.23465218051121786 +INFO:root: +[Valid][1][UTT][ MSE = 0.2659 | LCC = 0.8522 | SRCC = 0.8531 ] +INFO:root:Average loss=0.20662097336772162 +INFO:root: +[Valid][2][UTT][ MSE = 0.2437 | LCC = 0.8622 | SRCC = 0.8612 ] +INFO:root:Average loss=0.18646154858362982 +INFO:root: +[Valid][3][UTT][ MSE = 0.2489 | LCC = 0.8613 | SRCC = 0.8602 ] +INFO:root:Average loss=0.17167662657224214 +INFO:root: +[Valid][4][UTT][ MSE = 0.2663 | LCC = 0.8625 | SRCC = 0.8636 ] +INFO:root:Average loss=0.15713384269903868 +INFO:root: +[Valid][5][UTT][ MSE = 0.2535 | LCC = 0.8657 | SRCC = 0.8647 ] +INFO:root:Average loss=0.14907896174834326 +INFO:root: +[Valid][6][UTT][ MSE = 0.2452 | LCC = 0.8667 | SRCC = 0.8652 ] +INFO:root:Average loss=0.14637647788876143 +INFO:root: +[Valid][7][UTT][ MSE = 0.2284 | LCC = 0.8663 | SRCC = 0.8659 ] +INFO:root:Average loss=0.1365797471923706 +INFO:root: +[Valid][8][UTT][ MSE = 0.2479 | LCC = 0.8678 | SRCC = 0.8668 ] +INFO:root:Average loss=0.12663178203197625 +INFO:root: +[Valid][9][UTT][ MSE = 0.2330 | LCC = 0.8639 | SRCC = 0.8653 ] +INFO:root:Average loss=0.11436240188777447 +INFO:root: +[Valid][10][UTT][ MSE = 0.2493 | LCC = 0.8655 | SRCC = 0.8670 ] +INFO:root:Average loss=0.1074899262151657 +INFO:root: +[Valid][11][UTT][ MSE = 0.2397 | LCC = 0.8656 | SRCC = 0.8651 ] +INFO:root:Average loss=0.10815433059365322 +INFO:root: +[Valid][12][UTT][ MSE = 0.2451 | LCC = 0.8667 | SRCC = 0.8660 ] +INFO:root:Average loss=0.10078255497874358 +INFO:root: +[Valid][13][UTT][ MSE = 0.2472 | LCC = 0.8662 | SRCC = 0.8648 ] +INFO:root:Average loss=0.09455075473166429 +INFO:root: +[Valid][14][UTT][ MSE = 0.2397 | LCC = 0.8651 | SRCC = 0.8634 ] +INFO:root:Average loss=0.08734359052509834 +INFO:root: +[Valid][15][UTT][ MSE = 0.2286 | LCC = 0.8681 | SRCC = 0.8683 ] +INFO:root:Average loss=0.09527109950207747 +INFO:root: +[Valid][16][UTT][ MSE = 0.2171 | LCC = 0.8693 | SRCC = 0.8698 ] +INFO:root:Average loss=0.07834398049192551 +INFO:root: +[Valid][17][UTT][ MSE = 0.2163 | LCC = 0.8652 | SRCC = 0.8649 ] +INFO:root:Average loss=0.078211426257323 +INFO:root: +[Valid][18][UTT][ MSE = 0.2434 | LCC = 0.8634 | SRCC = 0.8627 ] +INFO:root:Average loss=0.07094016494468236 +INFO:root: +[Valid][19][UTT][ MSE = 0.3029 | LCC = 0.8638 | SRCC = 0.8643 ] +INFO:root:Average loss=0.07918375787826684 +INFO:root: +[Valid][20][UTT][ MSE = 0.2641 | LCC = 0.8691 | SRCC = 0.8691 ] +INFO:root:Average loss=0.06755127041385724 +INFO:root: +[Valid][21][UTT][ MSE = 0.2452 | LCC = 0.8659 | SRCC = 0.8653 ] +INFO:root:Average loss=0.06105798697815491 +INFO:root: +[Valid][22][UTT][ MSE = 0.2302 | LCC = 0.8674 | SRCC = 0.8674 ] +INFO:root:Average loss=0.0610471043544702 +INFO:root: +[Valid][23][UTT][ MSE = 0.2314 | LCC = 0.8637 | SRCC = 0.8634 ] +INFO:root:Average loss=0.05914214530434364 +INFO:root: +[Valid][24][UTT][ MSE = 0.2339 | LCC = 0.8653 | SRCC = 0.8645 ] +INFO:root:Average loss=0.056957785708782 +INFO:root: +[Valid][25][UTT][ MSE = 0.2428 | LCC = 0.8685 | SRCC = 0.8681 ] +INFO:root:Average loss=0.055446218412656054 +INFO:root: +[Valid][26][UTT][ MSE = 0.2156 | LCC = 0.8692 | SRCC = 0.8686 ] +INFO:root:Average loss=0.05086506478106364 +INFO:root: +[Valid][27][UTT][ MSE = 0.2534 | LCC = 0.8653 | SRCC = 0.8652 ] +INFO:root:Average loss=0.050315822259737894 +INFO:root: +[Valid][28][UTT][ MSE = 0.2240 | LCC = 0.8681 | SRCC = 0.8685 ] +INFO:root:Average loss=0.04749790569528555 +INFO:root: +[Valid][29][UTT][ MSE = 0.2154 | LCC = 0.8699 | SRCC = 0.8714 ] +INFO:root:Average loss=0.0473342350182625 +INFO:root: +[Valid][30][UTT][ MSE = 0.2129 | LCC = 0.8697 | SRCC = 0.8707 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2166 | LCC = 0.8662 | SRCC = 0.8667 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..60532bd6dfb2b362696134315abb1764f48a003c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85dc9d16531e62c7ba07ec8723fffe70f3c672ee22ed2509ff2c721af59a7820 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4853dd8782e87ca7e76e39b6221a64364b3b3e80 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a1e447b79a6dc3aaa89bc6b624cbf147b17ea4ee6b52c4fd6f83f1f4002d1607 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..840f7816ee1feb0a093eddf800fc93866dd7ce90 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/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=0.5025304895945084 +INFO:root: +[Valid][0][UTT][ MSE = 0.3337 | LCC = 0.8154 | SRCC = 0.8218 ] +INFO:root:Average loss=0.23945961949916986 +INFO:root: +[Valid][1][UTT][ MSE = 0.2751 | LCC = 0.8498 | SRCC = 0.8507 ] +INFO:root:Average loss=0.203442133581027 +INFO:root: +[Valid][2][UTT][ MSE = 0.2605 | LCC = 0.8619 | SRCC = 0.8607 ] +INFO:root:Average loss=0.18557090608355326 +INFO:root: +[Valid][3][UTT][ MSE = 0.2951 | LCC = 0.8647 | SRCC = 0.8652 ] +INFO:root:Average loss=0.18428826255676073 +INFO:root: +[Valid][4][UTT][ MSE = 0.2562 | LCC = 0.8702 | SRCC = 0.8675 ] +INFO:root:Average loss=0.16778464605792975 +INFO:root: +[Valid][5][UTT][ MSE = 0.2322 | LCC = 0.8719 | SRCC = 0.8697 ] +INFO:root:Average loss=0.15184546605898783 +INFO:root: +[Valid][6][UTT][ MSE = 0.2355 | LCC = 0.8715 | SRCC = 0.8712 ] +INFO:root:Average loss=0.1429723268136 +INFO:root: +[Valid][7][UTT][ MSE = 0.2450 | LCC = 0.8672 | SRCC = 0.8679 ] +INFO:root:Average loss=0.13668850503670862 +INFO:root: +[Valid][8][UTT][ MSE = 0.2352 | LCC = 0.8669 | SRCC = 0.8668 ] +INFO:root:Average loss=0.12841153565125588 +INFO:root: +[Valid][9][UTT][ MSE = 0.2416 | LCC = 0.8650 | SRCC = 0.8639 ] +INFO:root:Average loss=0.11567004091846637 +INFO:root: +[Valid][10][UTT][ MSE = 0.2517 | LCC = 0.8680 | SRCC = 0.8688 ] +INFO:root:Average loss=0.1121637424788414 +INFO:root: +[Valid][11][UTT][ MSE = 0.2260 | LCC = 0.8648 | SRCC = 0.8646 ] +INFO:root:Average loss=0.10263121481507252 +INFO:root: +[Valid][12][UTT][ MSE = 0.2240 | LCC = 0.8647 | SRCC = 0.8653 ] +INFO:root:Average loss=0.09912849231981315 +INFO:root: +[Valid][13][UTT][ MSE = 0.2396 | LCC = 0.8645 | SRCC = 0.8639 ] +INFO:root:Average loss=0.09542528406167641 +INFO:root: +[Valid][14][UTT][ MSE = 0.2355 | LCC = 0.8631 | SRCC = 0.8625 ] +INFO:root:Average loss=0.08612546157569458 +INFO:root: +[Valid][15][UTT][ MSE = 0.2212 | LCC = 0.8616 | SRCC = 0.8618 ] +INFO:root:Average loss=0.08596612928578487 +INFO:root: +[Valid][16][UTT][ MSE = 0.2834 | LCC = 0.8628 | SRCC = 0.8672 ] +INFO:root:Average loss=0.08133437455846713 +INFO:root: +[Valid][17][UTT][ MSE = 0.2189 | LCC = 0.8641 | SRCC = 0.8657 ] +INFO:root:Average loss=0.0783611667366364 +INFO:root: +[Valid][18][UTT][ MSE = 0.2247 | LCC = 0.8670 | SRCC = 0.8675 ] +INFO:root:Average loss=0.07355772885374534 +INFO:root: +[Valid][19][UTT][ MSE = 0.2316 | LCC = 0.8649 | SRCC = 0.8675 ] +INFO:root:Average loss=0.06778604045319252 +INFO:root: +[Valid][20][UTT][ MSE = 0.2305 | LCC = 0.8635 | SRCC = 0.8650 ] +INFO:root:Average loss=0.06819725576310585 +INFO:root: +[Valid][21][UTT][ MSE = 0.2568 | LCC = 0.8630 | SRCC = 0.8634 ] +INFO:root:Average loss=0.0660460367321204 +INFO:root: +[Valid][22][UTT][ MSE = 0.2193 | LCC = 0.8671 | SRCC = 0.8684 ] +INFO:root:Average loss=0.06399313608805339 +INFO:root: +[Valid][23][UTT][ MSE = 0.2161 | LCC = 0.8673 | SRCC = 0.8686 ] +INFO:root:Average loss=0.05797218416745846 +INFO:root: +[Valid][24][UTT][ MSE = 0.2337 | LCC = 0.8638 | SRCC = 0.8646 ] +INFO:root:Average loss=0.05678480099409054 +INFO:root: +[Valid][25][UTT][ MSE = 0.2321 | LCC = 0.8597 | SRCC = 0.8608 ] +INFO:root:Average loss=0.057263969276578 +INFO:root: +[Valid][26][UTT][ MSE = 0.2311 | LCC = 0.8621 | SRCC = 0.8631 ] +INFO:root:Average loss=0.051537036609191164 +INFO:root: +[Valid][27][UTT][ MSE = 0.2452 | LCC = 0.8591 | SRCC = 0.8603 ] +INFO:root:Average loss=0.05617561062368063 +INFO:root: +[Valid][28][UTT][ MSE = 0.2237 | LCC = 0.8630 | SRCC = 0.8647 ] +INFO:root:Average loss=0.048178997272864364 +INFO:root: +[Valid][29][UTT][ MSE = 0.2369 | LCC = 0.8595 | SRCC = 0.8597 ] +INFO:root:Average loss=0.04873237102173078 +INFO:root: +[Valid][30][UTT][ MSE = 0.2339 | LCC = 0.8600 | SRCC = 0.8609 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2202 | LCC = 0.8792 | SRCC = 0.8797 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..31c162e0ff58d6b096676f729765352489d4ef39 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc74baaac29d814a325c9a341fb3074ec665111a5db4609e46ccbb7f34075a6a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..a7fbb8e2e693b64736cf390d7b518582c2ee3978 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ddb25c0d48d33759e137500913a297e3e953f131beb7e5f5a0386d173662654b +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..37eaee1985f34736f20d5ad6b29edf6fb88a1491 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.37918334339673704 +INFO:root: +[Valid][0][UTT][ MSE = 0.3115 | LCC = 0.8309 | SRCC = 0.8300 ] +INFO:root:Average loss=0.23216740710613054 +INFO:root: +[Valid][1][UTT][ MSE = 0.2706 | LCC = 0.8540 | SRCC = 0.8522 ] +INFO:root:Average loss=0.20096515921445993 +INFO:root: +[Valid][2][UTT][ MSE = 0.2744 | LCC = 0.8611 | SRCC = 0.8587 ] +INFO:root:Average loss=0.18695878409422362 +INFO:root: +[Valid][3][UTT][ MSE = 0.2353 | LCC = 0.8649 | SRCC = 0.8628 ] +INFO:root:Average loss=0.17145024106288567 +INFO:root: +[Valid][4][UTT][ MSE = 0.2335 | LCC = 0.8668 | SRCC = 0.8653 ] +INFO:root:Average loss=0.16853160477983645 +INFO:root: +[Valid][5][UTT][ MSE = 0.2220 | LCC = 0.8679 | SRCC = 0.8664 ] +INFO:root:Average loss=0.15639599173879012 +INFO:root: +[Valid][6][UTT][ MSE = 0.2455 | LCC = 0.8676 | SRCC = 0.8655 ] +INFO:root:Average loss=0.13937201159886825 +INFO:root: +[Valid][7][UTT][ MSE = 0.2199 | LCC = 0.8693 | SRCC = 0.8676 ] +INFO:root:Average loss=0.13266589014958113 +INFO:root: +[Valid][8][UTT][ MSE = 0.2180 | LCC = 0.8694 | SRCC = 0.8689 ] +INFO:root:Average loss=0.1252513831624618 +INFO:root: +[Valid][9][UTT][ MSE = 0.2429 | LCC = 0.8660 | SRCC = 0.8651 ] +INFO:root:Average loss=0.11720862363775571 +INFO:root: +[Valid][10][UTT][ MSE = 0.2268 | LCC = 0.8660 | SRCC = 0.8662 ] +INFO:root:Average loss=0.10985581634136346 +INFO:root: +[Valid][11][UTT][ MSE = 0.2381 | LCC = 0.8665 | SRCC = 0.8656 ] +INFO:root:Average loss=0.10320697891979645 +INFO:root: +[Valid][12][UTT][ MSE = 0.2338 | LCC = 0.8659 | SRCC = 0.8664 ] +INFO:root:Average loss=0.0981099304671471 +INFO:root: +[Valid][13][UTT][ MSE = 0.2132 | LCC = 0.8678 | SRCC = 0.8678 ] +INFO:root:Average loss=0.10195509277475186 +INFO:root: +[Valid][14][UTT][ MSE = 0.2378 | LCC = 0.8664 | SRCC = 0.8661 ] +INFO:root:Average loss=0.08716951329738666 +INFO:root: +[Valid][15][UTT][ MSE = 0.2387 | LCC = 0.8638 | SRCC = 0.8631 ] +INFO:root:Average loss=0.08335897388557593 +INFO:root: +[Valid][16][UTT][ MSE = 0.2414 | LCC = 0.8617 | SRCC = 0.8632 ] +INFO:root:Average loss=0.07943778762068504 +INFO:root: +[Valid][17][UTT][ MSE = 0.2268 | LCC = 0.8641 | SRCC = 0.8646 ] +INFO:root:Average loss=0.07526580521311516 +INFO:root: +[Valid][18][UTT][ MSE = 0.2288 | LCC = 0.8652 | SRCC = 0.8645 ] +INFO:root:Average loss=0.0707269218774178 +INFO:root: +[Valid][19][UTT][ MSE = 0.2548 | LCC = 0.8617 | SRCC = 0.8620 ] +INFO:root:Average loss=0.07202759346900842 +INFO:root: +[Valid][20][UTT][ MSE = 0.2333 | LCC = 0.8640 | SRCC = 0.8645 ] +INFO:root:Average loss=0.06550765634538272 +INFO:root: +[Valid][21][UTT][ MSE = 0.2437 | LCC = 0.8628 | SRCC = 0.8634 ] +INFO:root:Average loss=0.06607571091407384 +INFO:root: +[Valid][22][UTT][ MSE = 0.2472 | LCC = 0.8605 | SRCC = 0.8613 ] +INFO:root:Average loss=0.06283538439908089 +INFO:root: +[Valid][23][UTT][ MSE = 0.2479 | LCC = 0.8610 | SRCC = 0.8617 ] +INFO:root:Average loss=0.05970505505609207 +INFO:root: +[Valid][24][UTT][ MSE = 0.2486 | LCC = 0.8601 | SRCC = 0.8602 ] +INFO:root:Average loss=0.059083288917556785 +INFO:root: +[Valid][25][UTT][ MSE = 0.2366 | LCC = 0.8622 | SRCC = 0.8629 ] +INFO:root:Average loss=0.05438483081375941 +INFO:root: +[Valid][26][UTT][ MSE = 0.2360 | LCC = 0.8618 | SRCC = 0.8625 ] +INFO:root:Average loss=0.054352012247993395 +INFO:root: +[Valid][27][UTT][ MSE = 0.2212 | LCC = 0.8646 | SRCC = 0.8652 ] +INFO:root:Average loss=0.05147410769206591 +INFO:root: +[Valid][28][UTT][ MSE = 0.2534 | LCC = 0.8623 | SRCC = 0.8640 ] +INFO:root:Average loss=0.05266793848325809 +INFO:root: +[Valid][29][UTT][ MSE = 0.2256 | LCC = 0.8629 | SRCC = 0.8650 ] +INFO:root:Average loss=0.04933074897585007 +INFO:root: +[Valid][30][UTT][ MSE = 0.2541 | LCC = 0.8583 | SRCC = 0.8594 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2186 | LCC = 0.8738 | SRCC = 0.8739 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..bbafb798846b8f050c2bb758bee7bfbe7137db4a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d97f497dcefe747b02d92601d0dc5b263101571df4c9f1d59919e74299563c13 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..2dd066547c858a8b3c2bdd059c4f571f30647c07 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:af309f425a2154b5a38970c2c7a9efbe3ad2d48c57e1ce4c65dc83adbc84bdc4 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..44bad11ced3d48c6c31396a79a18a3d983fadb36 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.3590706300276976 +INFO:root: +[Valid][0][UTT][ MSE = 0.3129 | LCC = 0.8274 | SRCC = 0.8301 ] +INFO:root:Average loss=0.23191000750431648 +INFO:root: +[Valid][1][UTT][ MSE = 0.2791 | LCC = 0.8519 | SRCC = 0.8524 ] +INFO:root:Average loss=0.19896224179329017 +INFO:root: +[Valid][2][UTT][ MSE = 0.2394 | LCC = 0.8612 | SRCC = 0.8600 ] +INFO:root:Average loss=0.18205310251468268 +INFO:root: +[Valid][3][UTT][ MSE = 0.2391 | LCC = 0.8652 | SRCC = 0.8648 ] +INFO:root:Average loss=0.16897123268781564 +INFO:root: +[Valid][4][UTT][ MSE = 0.2559 | LCC = 0.8658 | SRCC = 0.8633 ] +INFO:root:Average loss=0.15552050543901247 +INFO:root: +[Valid][5][UTT][ MSE = 0.2421 | LCC = 0.8660 | SRCC = 0.8639 ] +INFO:root:Average loss=0.15548177283161727 +INFO:root: +[Valid][6][UTT][ MSE = 0.2447 | LCC = 0.8660 | SRCC = 0.8649 ] +INFO:root:Average loss=0.1406933306119381 +INFO:root: +[Valid][7][UTT][ MSE = 0.2266 | LCC = 0.8647 | SRCC = 0.8643 ] +INFO:root:Average loss=0.1300000962920678 +INFO:root: +[Valid][8][UTT][ MSE = 0.2419 | LCC = 0.8650 | SRCC = 0.8640 ] +INFO:root:Average loss=0.13083975676160592 +INFO:root: +[Valid][9][UTT][ MSE = 0.2599 | LCC = 0.8677 | SRCC = 0.8675 ] +INFO:root:Average loss=0.11546836401789616 +INFO:root: +[Valid][10][UTT][ MSE = 0.2389 | LCC = 0.8649 | SRCC = 0.8643 ] +INFO:root:Average loss=0.1116529327745621 +INFO:root: +[Valid][11][UTT][ MSE = 0.2308 | LCC = 0.8654 | SRCC = 0.8655 ] +INFO:root:Average loss=0.10159963316833362 +INFO:root: +[Valid][12][UTT][ MSE = 0.2385 | LCC = 0.8652 | SRCC = 0.8639 ] +INFO:root:Average loss=0.10378883845913105 +INFO:root: +[Valid][13][UTT][ MSE = 0.2419 | LCC = 0.8657 | SRCC = 0.8636 ] +INFO:root:Average loss=0.09466034861711356 +INFO:root: +[Valid][14][UTT][ MSE = 0.2214 | LCC = 0.8651 | SRCC = 0.8633 ] +INFO:root:Average loss=0.0895018013050923 +INFO:root: +[Valid][15][UTT][ MSE = 0.2338 | LCC = 0.8636 | SRCC = 0.8614 ] +INFO:root:Average loss=0.08225275681186946 +INFO:root: +[Valid][16][UTT][ MSE = 0.2400 | LCC = 0.8597 | SRCC = 0.8584 ] +INFO:root:Average loss=0.08201290394824284 +INFO:root: +[Valid][17][UTT][ MSE = 0.2352 | LCC = 0.8603 | SRCC = 0.8580 ] +INFO:root:Average loss=0.07868557604841697 +INFO:root: +[Valid][18][UTT][ MSE = 0.2277 | LCC = 0.8633 | SRCC = 0.8627 ] +INFO:root:Average loss=0.07021642509752359 +INFO:root: +[Valid][19][UTT][ MSE = 0.2428 | LCC = 0.8567 | SRCC = 0.8559 ] +INFO:root:Average loss=0.07020132453777851 +INFO:root: +[Valid][20][UTT][ MSE = 0.2247 | LCC = 0.8637 | SRCC = 0.8629 ] +INFO:root:Average loss=0.06655501077572505 +INFO:root: +[Valid][21][UTT][ MSE = 0.2336 | LCC = 0.8600 | SRCC = 0.8589 ] +INFO:root:Average loss=0.06706953722123916 +INFO:root: +[Valid][22][UTT][ MSE = 0.2318 | LCC = 0.8583 | SRCC = 0.8570 ] +INFO:root:Average loss=0.0602065232128669 +INFO:root: +[Valid][23][UTT][ MSE = 0.2560 | LCC = 0.8591 | SRCC = 0.8582 ] +INFO:root:Average loss=0.06156270702679952 +INFO:root: +[Valid][24][UTT][ MSE = 0.2327 | LCC = 0.8600 | SRCC = 0.8589 ] +INFO:root:Average loss=0.05567158925800751 +INFO:root: +[Valid][25][UTT][ MSE = 0.2524 | LCC = 0.8548 | SRCC = 0.8535 ] +INFO:root:Average loss=0.05586656417029026 +INFO:root: +[Valid][26][UTT][ MSE = 0.2504 | LCC = 0.8579 | SRCC = 0.8573 ] +INFO:root:Average loss=0.05574753870948767 +INFO:root: +[Valid][27][UTT][ MSE = 0.2401 | LCC = 0.8597 | SRCC = 0.8586 ] +INFO:root:Average loss=0.05329497464192219 +INFO:root: +[Valid][28][UTT][ MSE = 0.2228 | LCC = 0.8625 | SRCC = 0.8618 ] +INFO:root:Average loss=0.050464585614510074 +INFO:root: +[Valid][29][UTT][ MSE = 0.2278 | LCC = 0.8581 | SRCC = 0.8573 ] +INFO:root:Average loss=0.0481343948497222 +INFO:root: +[Valid][30][UTT][ MSE = 0.2404 | LCC = 0.8610 | SRCC = 0.8599 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2302 | LCC = 0.8745 | SRCC = 0.8745 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d84cf23f7d2218016e3c37e6e6ff35df3a349b20 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a8573d24e2c3e4e45b74ef3b82cd12f1f33a5d50e450252cb680bbe49d5185c +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..d122f6792149e94df1f67c32a7a838f59f50a634 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ecd78dccadd4dff03847b6b11b6a9c1572f96a976fd144c6cc28cd5e95a4b8b5 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer3/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..761f1e2d24ecc11e0e3c36d4f3b93ffcc07830d2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.39831789258198863 +INFO:root: +[Valid][0][UTT][ MSE = 0.3218 | LCC = 0.8175 | SRCC = 0.8209 ] +INFO:root:Average loss=0.23598485440015793 +INFO:root: +[Valid][1][UTT][ MSE = 0.2667 | LCC = 0.8436 | SRCC = 0.8464 ] +INFO:root:Average loss=0.2059344490751242 +INFO:root: +[Valid][2][UTT][ MSE = 0.2478 | LCC = 0.8570 | SRCC = 0.8557 ] +INFO:root:Average loss=0.1857431798409193 +INFO:root: +[Valid][3][UTT][ MSE = 0.2570 | LCC = 0.8645 | SRCC = 0.8617 ] +INFO:root:Average loss=0.17408048619444555 +INFO:root: +[Valid][4][UTT][ MSE = 0.2538 | LCC = 0.8662 | SRCC = 0.8638 ] +INFO:root:Average loss=0.16331855389170158 +INFO:root: +[Valid][5][UTT][ MSE = 0.2346 | LCC = 0.8629 | SRCC = 0.8606 ] +INFO:root:Average loss=0.1491958118784122 +INFO:root: +[Valid][6][UTT][ MSE = 0.2296 | LCC = 0.8648 | SRCC = 0.8638 ] +INFO:root:Average loss=0.13898966623804507 +INFO:root: +[Valid][7][UTT][ MSE = 0.2301 | LCC = 0.8656 | SRCC = 0.8645 ] +INFO:root:Average loss=0.13551122399094778 +INFO:root: +[Valid][8][UTT][ MSE = 0.2250 | LCC = 0.8625 | SRCC = 0.8618 ] +INFO:root:Average loss=0.12292055623271526 +INFO:root: +[Valid][9][UTT][ MSE = 0.2429 | LCC = 0.8586 | SRCC = 0.8575 ] +INFO:root:Average loss=0.11465249392084587 +INFO:root: +[Valid][10][UTT][ MSE = 0.2367 | LCC = 0.8570 | SRCC = 0.8553 ] +INFO:root:Average loss=0.11036657427365963 +INFO:root: +[Valid][11][UTT][ MSE = 0.2251 | LCC = 0.8612 | SRCC = 0.8598 ] +INFO:root:Average loss=0.1015956872739853 +INFO:root: +[Valid][12][UTT][ MSE = 0.2417 | LCC = 0.8578 | SRCC = 0.8559 ] +INFO:root:Average loss=0.10083628636904252 +INFO:root: +[Valid][13][UTT][ MSE = 0.2327 | LCC = 0.8612 | SRCC = 0.8595 ] +INFO:root:Average loss=0.09846726933924052 +INFO:root: +[Valid][14][UTT][ MSE = 0.2315 | LCC = 0.8607 | SRCC = 0.8588 ] +INFO:root:Average loss=0.08995841715771419 +INFO:root: +[Valid][15][UTT][ MSE = 0.2312 | LCC = 0.8606 | SRCC = 0.8595 ] +INFO:root:Average loss=0.09141965415806343 +INFO:root: +[Valid][16][UTT][ MSE = 0.2328 | LCC = 0.8612 | SRCC = 0.8583 ] +INFO:root:Average loss=0.07975989546722327 +INFO:root: +[Valid][17][UTT][ MSE = 0.2514 | LCC = 0.8517 | SRCC = 0.8506 ] +INFO:root:Average loss=0.07547295972322807 +INFO:root: +[Valid][18][UTT][ MSE = 0.2309 | LCC = 0.8558 | SRCC = 0.8537 ] +INFO:root:Average loss=0.07636507409505355 +INFO:root: +[Valid][19][UTT][ MSE = 0.2350 | LCC = 0.8554 | SRCC = 0.8539 ] +INFO:root:Average loss=0.07275251604807682 +INFO:root: +[Valid][20][UTT][ MSE = 0.2303 | LCC = 0.8553 | SRCC = 0.8529 ] +INFO:root:Average loss=0.06842406237354645 +INFO:root: +[Valid][21][UTT][ MSE = 0.2409 | LCC = 0.8548 | SRCC = 0.8523 ] +INFO:root:Average loss=0.07009032879693386 +INFO:root: +[Valid][22][UTT][ MSE = 0.2378 | LCC = 0.8560 | SRCC = 0.8558 ] +INFO:root:Average loss=0.05991220202010412 +INFO:root: +[Valid][23][UTT][ MSE = 0.2234 | LCC = 0.8573 | SRCC = 0.8561 ] +INFO:root:Average loss=0.06220455465312952 +INFO:root: +[Valid][24][UTT][ MSE = 0.2287 | LCC = 0.8571 | SRCC = 0.8556 ] +INFO:root:Average loss=0.05659589863931521 +INFO:root: +[Valid][25][UTT][ MSE = 0.2301 | LCC = 0.8556 | SRCC = 0.8537 ] +INFO:root:Average loss=0.05587436153720587 +INFO:root: +[Valid][26][UTT][ MSE = 0.2317 | LCC = 0.8544 | SRCC = 0.8529 ] +INFO:root:Average loss=0.054268971348229125 +INFO:root: +[Valid][27][UTT][ MSE = 0.2297 | LCC = 0.8535 | SRCC = 0.8509 ] +INFO:root:Average loss=0.051173943118789256 +INFO:root: +[Valid][28][UTT][ MSE = 0.2373 | LCC = 0.8540 | SRCC = 0.8550 ] +INFO:root:Average loss=0.050943831292291485 +INFO:root: +[Valid][29][UTT][ MSE = 0.2376 | LCC = 0.8542 | SRCC = 0.8515 ] +INFO:root:Average loss=0.04662393315289265 +INFO:root: +[Valid][30][UTT][ MSE = 0.2390 | LCC = 0.8527 | SRCC = 0.8512 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2641 | LCC = 0.8711 | SRCC = 0.8741 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..17d515d057ebe398bf88deb68d28e733d929ad56 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f58e64a42d47ccb1ac0548178e440b9c99b7717dea6a39093afe79e7426ff972 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e04845981614b678cb4f7e79f4dc0eae4af08bae --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4dc2699490131275a2a36752d56d3c663e3182845f9ed09bacec84e2d87b3093 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..3859955d5ebd01efd7864ead974315c598746492 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.39435059558122587 +INFO:root: +[Valid][0][UTT][ MSE = 0.3066 | LCC = 0.8277 | SRCC = 0.8301 ] +INFO:root:Average loss=0.23765110472838083 +INFO:root: +[Valid][1][UTT][ MSE = 0.2999 | LCC = 0.8513 | SRCC = 0.8486 ] +INFO:root:Average loss=0.21071421440977317 +INFO:root: +[Valid][2][UTT][ MSE = 0.2482 | LCC = 0.8578 | SRCC = 0.8552 ] +INFO:root:Average loss=0.19078739665639707 +INFO:root: +[Valid][3][UTT][ MSE = 0.2442 | LCC = 0.8611 | SRCC = 0.8582 ] +INFO:root:Average loss=0.17423691591009116 +INFO:root: +[Valid][4][UTT][ MSE = 0.2439 | LCC = 0.8667 | SRCC = 0.8647 ] +INFO:root:Average loss=0.16446519595308182 +INFO:root: +[Valid][5][UTT][ MSE = 0.2419 | LCC = 0.8644 | SRCC = 0.8617 ] +INFO:root:Average loss=0.14842827637226153 +INFO:root: +[Valid][6][UTT][ MSE = 0.2573 | LCC = 0.8623 | SRCC = 0.8588 ] +INFO:root:Average loss=0.1397505650917689 +INFO:root: +[Valid][7][UTT][ MSE = 0.2287 | LCC = 0.8616 | SRCC = 0.8608 ] +INFO:root:Average loss=0.13285439108044672 +INFO:root: +[Valid][8][UTT][ MSE = 0.2374 | LCC = 0.8616 | SRCC = 0.8593 ] +INFO:root:Average loss=0.1256566497568901 +INFO:root: +[Valid][9][UTT][ MSE = 0.2668 | LCC = 0.8597 | SRCC = 0.8589 ] +INFO:root:Average loss=0.11590990548332532 +INFO:root: +[Valid][10][UTT][ MSE = 0.2274 | LCC = 0.8607 | SRCC = 0.8594 ] +INFO:root:Average loss=0.10724844535191853 +INFO:root: +[Valid][11][UTT][ MSE = 0.2528 | LCC = 0.8598 | SRCC = 0.8582 ] +INFO:root:Average loss=0.10622178218685664 +INFO:root: +[Valid][12][UTT][ MSE = 0.2448 | LCC = 0.8559 | SRCC = 0.8551 ] +INFO:root:Average loss=0.09664766595531733 +INFO:root: +[Valid][13][UTT][ MSE = 0.2309 | LCC = 0.8562 | SRCC = 0.8552 ] +INFO:root:Average loss=0.09320676317199683 +INFO:root: +[Valid][14][UTT][ MSE = 0.2866 | LCC = 0.8544 | SRCC = 0.8537 ] +INFO:root:Average loss=0.08490500470193532 +INFO:root: +[Valid][15][UTT][ MSE = 0.2356 | LCC = 0.8554 | SRCC = 0.8547 ] +INFO:root:Average loss=0.07902185833798005 +INFO:root: +[Valid][16][UTT][ MSE = 0.2376 | LCC = 0.8535 | SRCC = 0.8537 ] +INFO:root:Average loss=0.07390067702493607 +INFO:root: +[Valid][17][UTT][ MSE = 0.2590 | LCC = 0.8490 | SRCC = 0.8482 ] +INFO:root:Average loss=0.0740708032479653 +INFO:root: +[Valid][18][UTT][ MSE = 0.2528 | LCC = 0.8498 | SRCC = 0.8490 ] +INFO:root:Average loss=0.069276903445522 +INFO:root: +[Valid][19][UTT][ MSE = 0.2522 | LCC = 0.8492 | SRCC = 0.8482 ] +INFO:root:Average loss=0.06429070024154125 +INFO:root: +[Valid][20][UTT][ MSE = 0.2891 | LCC = 0.8467 | SRCC = 0.8458 ] +INFO:root:Average loss=0.06924001237329765 +INFO:root: +[Valid][21][UTT][ MSE = 0.2601 | LCC = 0.8491 | SRCC = 0.8466 ] +INFO:root:Average loss=0.06305448338389397 +INFO:root: +[Valid][22][UTT][ MSE = 0.2545 | LCC = 0.8453 | SRCC = 0.8434 ] +INFO:root:Average loss=0.061337730154777184 +INFO:root: +[Valid][23][UTT][ MSE = 0.2712 | LCC = 0.8487 | SRCC = 0.8488 ] +INFO:root:Average loss=0.057803276782998673 +INFO:root: +[Valid][24][UTT][ MSE = 0.2569 | LCC = 0.8497 | SRCC = 0.8484 ] +INFO:root:Average loss=0.05474208768170614 +INFO:root: +[Valid][25][UTT][ MSE = 0.2449 | LCC = 0.8490 | SRCC = 0.8474 ] +INFO:root:Average loss=0.052038299588438794 +INFO:root: +[Valid][26][UTT][ MSE = 0.2428 | LCC = 0.8497 | SRCC = 0.8486 ] +INFO:root:Average loss=0.051307837216135785 +INFO:root: +[Valid][27][UTT][ MSE = 0.2636 | LCC = 0.8443 | SRCC = 0.8432 ] +INFO:root:Average loss=0.049370952714712195 +INFO:root: +[Valid][28][UTT][ MSE = 0.2489 | LCC = 0.8450 | SRCC = 0.8436 ] +INFO:root:Average loss=0.048703427283236615 +INFO:root: +[Valid][29][UTT][ MSE = 0.2678 | LCC = 0.8474 | SRCC = 0.8459 ] +INFO:root:Average loss=0.0452030405927545 +INFO:root: +[Valid][30][UTT][ MSE = 0.2683 | LCC = 0.8441 | SRCC = 0.8428 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2653 | LCC = 0.8652 | SRCC = 0.8665 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..98ce8513af8f4ffc10ff6cacb7b1621ea6b536a1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d3a32f90f3e4b15b3c14f903dcb7fa6028bb1a0853d4a3e9c60b1e74b9bfffd1 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..bb608dcebe9b4fc20f637155744c85e91e2c6c11 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82532f3e22871f223648762e39df62e2a075e402b4ce6e3991fe95e8cde0059f +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..810374158e712aa58b097d958e605556240b2e27 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.4095694103684181 +INFO:root: +[Valid][0][UTT][ MSE = 0.3029 | LCC = 0.8179 | SRCC = 0.8280 ] +INFO:root:Average loss=0.2392322805065375 +INFO:root: +[Valid][1][UTT][ MSE = 0.2682 | LCC = 0.8490 | SRCC = 0.8516 ] +INFO:root:Average loss=0.2094162970017164 +INFO:root: +[Valid][2][UTT][ MSE = 0.2577 | LCC = 0.8645 | SRCC = 0.8617 ] +INFO:root:Average loss=0.19564510872348762 +INFO:root: +[Valid][3][UTT][ MSE = 0.2384 | LCC = 0.8662 | SRCC = 0.8651 ] +INFO:root:Average loss=0.182640946828402 +INFO:root: +[Valid][4][UTT][ MSE = 0.2685 | LCC = 0.8688 | SRCC = 0.8663 ] +INFO:root:Average loss=0.16906185963979134 +INFO:root: +[Valid][5][UTT][ MSE = 0.2315 | LCC = 0.8668 | SRCC = 0.8675 ] +INFO:root:Average loss=0.15499364489164108 +INFO:root: +[Valid][6][UTT][ MSE = 0.2269 | LCC = 0.8699 | SRCC = 0.8680 ] +INFO:root:Average loss=0.14717014592427474 +INFO:root: +[Valid][7][UTT][ MSE = 0.2228 | LCC = 0.8697 | SRCC = 0.8693 ] +INFO:root:Average loss=0.13433195440432963 +INFO:root: +[Valid][8][UTT][ MSE = 0.2191 | LCC = 0.8680 | SRCC = 0.8671 ] +INFO:root:Average loss=0.13560700903718287 +INFO:root: +[Valid][9][UTT][ MSE = 0.2308 | LCC = 0.8666 | SRCC = 0.8658 ] +INFO:root:Average loss=0.12245020193931384 +INFO:root: +[Valid][10][UTT][ MSE = 0.2370 | LCC = 0.8650 | SRCC = 0.8644 ] +INFO:root:Average loss=0.11046277827177292 +INFO:root: +[Valid][11][UTT][ MSE = 0.2394 | LCC = 0.8640 | SRCC = 0.8639 ] +INFO:root:Average loss=0.10817245145638783 +INFO:root: +[Valid][12][UTT][ MSE = 0.2259 | LCC = 0.8643 | SRCC = 0.8633 ] +INFO:root:Average loss=0.09937273118740474 +INFO:root: +[Valid][13][UTT][ MSE = 0.2228 | LCC = 0.8626 | SRCC = 0.8615 ] +INFO:root:Average loss=0.09153014770111977 +INFO:root: +[Valid][14][UTT][ MSE = 0.2388 | LCC = 0.8611 | SRCC = 0.8603 ] +INFO:root:Average loss=0.0875754663004325 +INFO:root: +[Valid][15][UTT][ MSE = 0.2198 | LCC = 0.8636 | SRCC = 0.8635 ] +INFO:root:Average loss=0.0796750030743006 +INFO:root: +[Valid][16][UTT][ MSE = 0.2313 | LCC = 0.8573 | SRCC = 0.8563 ] +INFO:root:Average loss=0.07748165282492454 +INFO:root: +[Valid][17][UTT][ MSE = 0.2390 | LCC = 0.8558 | SRCC = 0.8559 ] +INFO:root:Average loss=0.07156296466023494 +INFO:root: +[Valid][18][UTT][ MSE = 0.2489 | LCC = 0.8531 | SRCC = 0.8527 ] +INFO:root:Average loss=0.06793571745929046 +INFO:root: +[Valid][19][UTT][ MSE = 0.2372 | LCC = 0.8551 | SRCC = 0.8543 ] +INFO:root:Average loss=0.06610348696509998 +INFO:root: +[Valid][20][UTT][ MSE = 0.2367 | LCC = 0.8561 | SRCC = 0.8543 ] +INFO:root:Average loss=0.06588510653147331 +INFO:root: +[Valid][21][UTT][ MSE = 0.2457 | LCC = 0.8530 | SRCC = 0.8518 ] +INFO:root:Average loss=0.05911055708733889 +INFO:root: +[Valid][22][UTT][ MSE = 0.2349 | LCC = 0.8551 | SRCC = 0.8539 ] +INFO:root:Average loss=0.05905954929020924 +INFO:root: +[Valid][23][UTT][ MSE = 0.2538 | LCC = 0.8512 | SRCC = 0.8515 ] +INFO:root:Average loss=0.05660654785923469 +INFO:root: +[Valid][24][UTT][ MSE = 0.2422 | LCC = 0.8536 | SRCC = 0.8544 ] +INFO:root:Average loss=0.053133594493071236 +INFO:root: +[Valid][25][UTT][ MSE = 0.2365 | LCC = 0.8545 | SRCC = 0.8544 ] +INFO:root:Average loss=0.05366006921021602 +INFO:root: +[Valid][26][UTT][ MSE = 0.2337 | LCC = 0.8530 | SRCC = 0.8525 ] +INFO:root:Average loss=0.04683946639012832 +INFO:root: +[Valid][27][UTT][ MSE = 0.2359 | LCC = 0.8536 | SRCC = 0.8526 ] +INFO:root:Average loss=0.05109266368433451 +INFO:root: +[Valid][28][UTT][ MSE = 0.2477 | LCC = 0.8509 | SRCC = 0.8496 ] +INFO:root:Average loss=0.045890964734821744 +INFO:root: +[Valid][29][UTT][ MSE = 0.2347 | LCC = 0.8539 | SRCC = 0.8531 ] +INFO:root:Average loss=0.04638395986209313 +INFO:root: +[Valid][30][UTT][ MSE = 0.2335 | LCC = 0.8542 | SRCC = 0.8531 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2282 | LCC = 0.8721 | SRCC = 0.8724 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..1c399b7ccb1a3c106b1b85b1c61953087ed63af1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64214a52e452cbc9cc239ef4b0860bd02a0e545146a4eba27ccea10a3c86936a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7a9c143061dde11022c8e8cfd136ad5b6c92c52e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16ed153b6a494d63a189d0bf5e9eedb95a73d25db49381555b54f06d1f6b11e9 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..992406b7dfdd739f1ed633c28cda548f27ab4936 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.3659914980332057 +INFO:root: +[Valid][0][UTT][ MSE = 0.2998 | LCC = 0.8345 | SRCC = 0.8377 ] +INFO:root:Average loss=0.22928612125225556 +INFO:root: +[Valid][1][UTT][ MSE = 0.2662 | LCC = 0.8608 | SRCC = 0.8589 ] +INFO:root:Average loss=0.19641570154673013 +INFO:root: +[Valid][2][UTT][ MSE = 0.2361 | LCC = 0.8661 | SRCC = 0.8665 ] +INFO:root:Average loss=0.18497672390479308 +INFO:root: +[Valid][3][UTT][ MSE = 0.2219 | LCC = 0.8686 | SRCC = 0.8677 ] +INFO:root:Average loss=0.16534704925158086 +INFO:root: +[Valid][4][UTT][ MSE = 0.2349 | LCC = 0.8694 | SRCC = 0.8673 ] +INFO:root:Average loss=0.14937491858234772 +INFO:root: +[Valid][5][UTT][ MSE = 0.2389 | LCC = 0.8674 | SRCC = 0.8668 ] +INFO:root:Average loss=0.14304460384524786 +INFO:root: +[Valid][6][UTT][ MSE = 0.2504 | LCC = 0.8667 | SRCC = 0.8652 ] +INFO:root:Average loss=0.13660920296724027 +INFO:root: +[Valid][7][UTT][ MSE = 0.2258 | LCC = 0.8669 | SRCC = 0.8664 ] +INFO:root:Average loss=0.1261036444741946 +INFO:root: +[Valid][8][UTT][ MSE = 0.2229 | LCC = 0.8646 | SRCC = 0.8631 ] +INFO:root:Average loss=0.11727634616769277 +INFO:root: +[Valid][9][UTT][ MSE = 0.2304 | LCC = 0.8644 | SRCC = 0.8620 ] +INFO:root:Average loss=0.10704743613799413 +INFO:root: +[Valid][10][UTT][ MSE = 0.2472 | LCC = 0.8568 | SRCC = 0.8551 ] +INFO:root:Average loss=0.09857524511141655 +INFO:root: +[Valid][11][UTT][ MSE = 0.2295 | LCC = 0.8590 | SRCC = 0.8565 ] +INFO:root:Average loss=0.0980349471553778 +INFO:root: +[Valid][12][UTT][ MSE = 0.2506 | LCC = 0.8593 | SRCC = 0.8585 ] +INFO:root:Average loss=0.09180777113980208 +INFO:root: +[Valid][13][UTT][ MSE = 0.2343 | LCC = 0.8603 | SRCC = 0.8579 ] +INFO:root:Average loss=0.08385325638720623 +INFO:root: +[Valid][14][UTT][ MSE = 0.2401 | LCC = 0.8560 | SRCC = 0.8525 ] +INFO:root:Average loss=0.08133407185475032 +INFO:root: +[Valid][15][UTT][ MSE = 0.2374 | LCC = 0.8539 | SRCC = 0.8513 ] +INFO:root:Average loss=0.07515602635267453 +INFO:root: +[Valid][16][UTT][ MSE = 0.2366 | LCC = 0.8564 | SRCC = 0.8528 ] +INFO:root:Average loss=0.07018295537011746 +INFO:root: +[Valid][17][UTT][ MSE = 0.2335 | LCC = 0.8509 | SRCC = 0.8494 ] +INFO:root:Average loss=0.06806864541692612 +INFO:root: +[Valid][18][UTT][ MSE = 0.2301 | LCC = 0.8556 | SRCC = 0.8542 ] +INFO:root:Average loss=0.07020889193965839 +INFO:root: +[Valid][19][UTT][ MSE = 0.2254 | LCC = 0.8582 | SRCC = 0.8566 ] +INFO:root:Average loss=0.06749248624039002 +INFO:root: +[Valid][20][UTT][ MSE = 0.2309 | LCC = 0.8586 | SRCC = 0.8556 ] +INFO:root:Average loss=0.06857287783462268 +INFO:root: +[Valid][21][UTT][ MSE = 0.2482 | LCC = 0.8549 | SRCC = 0.8510 ] +INFO:root:Average loss=0.05783846807212402 +INFO:root: +[Valid][22][UTT][ MSE = 0.2359 | LCC = 0.8568 | SRCC = 0.8530 ] +INFO:root:Average loss=0.057213118920723595 +INFO:root: +[Valid][23][UTT][ MSE = 0.2310 | LCC = 0.8550 | SRCC = 0.8507 ] +INFO:root:Average loss=0.05433301479579546 +INFO:root: +[Valid][24][UTT][ MSE = 0.2351 | LCC = 0.8580 | SRCC = 0.8553 ] +INFO:root:Average loss=0.049569650433766536 +INFO:root: +[Valid][25][UTT][ MSE = 0.2387 | LCC = 0.8538 | SRCC = 0.8505 ] +INFO:root:Average loss=0.05372804417633093 +INFO:root: +[Valid][26][UTT][ MSE = 0.2301 | LCC = 0.8566 | SRCC = 0.8537 ] +INFO:root:Average loss=0.04767193955679735 +INFO:root: +[Valid][27][UTT][ MSE = 0.2315 | LCC = 0.8531 | SRCC = 0.8500 ] +INFO:root:Average loss=0.04885196561614672 +INFO:root: +[Valid][28][UTT][ MSE = 0.2590 | LCC = 0.8487 | SRCC = 0.8459 ] +INFO:root:Average loss=0.048121544341437325 +INFO:root: +[Valid][29][UTT][ MSE = 0.2497 | LCC = 0.8493 | SRCC = 0.8463 ] +INFO:root:Average loss=0.042226726141495585 +INFO:root: +[Valid][30][UTT][ MSE = 0.2470 | LCC = 0.8516 | SRCC = 0.8493 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2406 | LCC = 0.8668 | SRCC = 0.8681 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e16cae896e9012397b34fd960cbb63f82e81001d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:407662e2316d6e7ddbf1d67862f6a4598564f3512c79cc325abe5bdf7dc11595 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..275e517bef9b75682cc25f4603f2ce800c8fcfdb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0963d0421de0d291acc7962c714f1fe9a58e7bdd568490f5b97dc54ac73495ea +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..887b095ab1822e69ecd7fee1df9fe181a60df52a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.40926807469282395 +INFO:root: +[Valid][0][UTT][ MSE = 0.3165 | LCC = 0.8198 | SRCC = 0.8291 ] +INFO:root:Average loss=0.23892891044035935 +INFO:root: +[Valid][1][UTT][ MSE = 0.2777 | LCC = 0.8475 | SRCC = 0.8499 ] +INFO:root:Average loss=0.2094550841511824 +INFO:root: +[Valid][2][UTT][ MSE = 0.2381 | LCC = 0.8605 | SRCC = 0.8595 ] +INFO:root:Average loss=0.19359203256093538 +INFO:root: +[Valid][3][UTT][ MSE = 0.2334 | LCC = 0.8623 | SRCC = 0.8612 ] +INFO:root:Average loss=0.17102863658697176 +INFO:root: +[Valid][4][UTT][ MSE = 0.2306 | LCC = 0.8638 | SRCC = 0.8631 ] +INFO:root:Average loss=0.15707170342405638 +INFO:root: +[Valid][5][UTT][ MSE = 0.2365 | LCC = 0.8655 | SRCC = 0.8638 ] +INFO:root:Average loss=0.14312714720383668 +INFO:root: +[Valid][6][UTT][ MSE = 0.2349 | LCC = 0.8639 | SRCC = 0.8616 ] +INFO:root:Average loss=0.14059604226778716 +INFO:root: +[Valid][7][UTT][ MSE = 0.2385 | LCC = 0.8650 | SRCC = 0.8627 ] +INFO:root:Average loss=0.1305957446113611 +INFO:root: +[Valid][8][UTT][ MSE = 0.2182 | LCC = 0.8649 | SRCC = 0.8636 ] +INFO:root:Average loss=0.11795706941913335 +INFO:root: +[Valid][9][UTT][ MSE = 0.2309 | LCC = 0.8643 | SRCC = 0.8611 ] +INFO:root:Average loss=0.11058120209819232 +INFO:root: +[Valid][10][UTT][ MSE = 0.2282 | LCC = 0.8626 | SRCC = 0.8622 ] +INFO:root:Average loss=0.1025885470593587 +INFO:root: +[Valid][11][UTT][ MSE = 0.2346 | LCC = 0.8626 | SRCC = 0.8597 ] +INFO:root:Average loss=0.10252026315682973 +INFO:root: +[Valid][12][UTT][ MSE = 0.2384 | LCC = 0.8631 | SRCC = 0.8599 ] +INFO:root:Average loss=0.08762408157762809 +INFO:root: +[Valid][13][UTT][ MSE = 0.2213 | LCC = 0.8600 | SRCC = 0.8583 ] +INFO:root:Average loss=0.09069082852548514 +INFO:root: +[Valid][14][UTT][ MSE = 0.2282 | LCC = 0.8593 | SRCC = 0.8574 ] +INFO:root:Average loss=0.08163948935002853 +INFO:root: +[Valid][15][UTT][ MSE = 0.2399 | LCC = 0.8552 | SRCC = 0.8533 ] +INFO:root:Average loss=0.07628913066135003 +INFO:root: +[Valid][16][UTT][ MSE = 0.2351 | LCC = 0.8556 | SRCC = 0.8526 ] +INFO:root:Average loss=0.07240708253513543 +INFO:root: +[Valid][17][UTT][ MSE = 0.2427 | LCC = 0.8541 | SRCC = 0.8518 ] +INFO:root:Average loss=0.06755199154409078 +INFO:root: +[Valid][18][UTT][ MSE = 0.2370 | LCC = 0.8542 | SRCC = 0.8520 ] +INFO:root:Average loss=0.06781378593773414 +INFO:root: +[Valid][19][UTT][ MSE = 0.2453 | LCC = 0.8521 | SRCC = 0.8493 ] +INFO:root:Average loss=0.06499317102134228 +INFO:root: +[Valid][20][UTT][ MSE = 0.2413 | LCC = 0.8515 | SRCC = 0.8489 ] +INFO:root:Average loss=0.06298336071463731 +INFO:root: +[Valid][21][UTT][ MSE = 0.2466 | LCC = 0.8515 | SRCC = 0.8477 ] +INFO:root:Average loss=0.05724823919053261 +INFO:root: +[Valid][22][UTT][ MSE = 0.2424 | LCC = 0.8530 | SRCC = 0.8496 ] +INFO:root:Average loss=0.05665779763307327 +INFO:root: +[Valid][23][UTT][ MSE = 0.2392 | LCC = 0.8536 | SRCC = 0.8522 ] +INFO:root:Average loss=0.05314639201148962 +INFO:root: +[Valid][24][UTT][ MSE = 0.2474 | LCC = 0.8512 | SRCC = 0.8483 ] +INFO:root:Average loss=0.04893563530192925 +INFO:root: +[Valid][25][UTT][ MSE = 0.2391 | LCC = 0.8538 | SRCC = 0.8507 ] +INFO:root:Average loss=0.05303625375605547 +INFO:root: +[Valid][26][UTT][ MSE = 0.2389 | LCC = 0.8519 | SRCC = 0.8480 ] +INFO:root:Average loss=0.05004637096172724 +INFO:root: +[Valid][27][UTT][ MSE = 0.2337 | LCC = 0.8531 | SRCC = 0.8502 ] +INFO:root:Average loss=0.046557775483681604 +INFO:root: +[Valid][28][UTT][ MSE = 0.2423 | LCC = 0.8520 | SRCC = 0.8501 ] +INFO:root:Average loss=0.042887080962268204 +INFO:root: +[Valid][29][UTT][ MSE = 0.2562 | LCC = 0.8476 | SRCC = 0.8433 ] +INFO:root:Average loss=0.04172271533081165 +INFO:root: +[Valid][30][UTT][ MSE = 0.2373 | LCC = 0.8528 | SRCC = 0.8489 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2547 | LCC = 0.8633 | SRCC = 0.8683 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..88a809d538d42ccc8486337700d9418ad20daaeb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73cd560c036d8f20368beef6017b069d1014e96457016de9751b70930bad4e50 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..eafece5bd66b4c018e0380c8a2a96c5b1518b2a0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2736636f1e8189932a5edda6e41dbb57e6966fc1ff6db202d56283495568fc76 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer4/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..be45d1a608b23569e3c8e4d2448f4fe44ed1e462 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.35447478886598194 +INFO:root: +[Valid][0][UTT][ MSE = 0.3057 | LCC = 0.8274 | SRCC = 0.8309 ] +INFO:root:Average loss=0.23605371782412896 +INFO:root: +[Valid][1][UTT][ MSE = 0.2987 | LCC = 0.8511 | SRCC = 0.8521 ] +INFO:root:Average loss=0.2063313283217259 +INFO:root: +[Valid][2][UTT][ MSE = 0.2444 | LCC = 0.8580 | SRCC = 0.8576 ] +INFO:root:Average loss=0.18858067452525482 +INFO:root: +[Valid][3][UTT][ MSE = 0.2406 | LCC = 0.8659 | SRCC = 0.8629 ] +INFO:root:Average loss=0.17365217743775782 +INFO:root: +[Valid][4][UTT][ MSE = 0.2489 | LCC = 0.8646 | SRCC = 0.8619 ] +INFO:root:Average loss=0.16020698071672365 +INFO:root: +[Valid][5][UTT][ MSE = 0.2579 | LCC = 0.8658 | SRCC = 0.8646 ] +INFO:root:Average loss=0.15111384684076676 +INFO:root: +[Valid][6][UTT][ MSE = 0.2415 | LCC = 0.8654 | SRCC = 0.8637 ] +INFO:root:Average loss=0.14244573601545432 +INFO:root: +[Valid][7][UTT][ MSE = 0.2630 | LCC = 0.8644 | SRCC = 0.8628 ] +INFO:root:Average loss=0.1307863905452765 +INFO:root: +[Valid][8][UTT][ MSE = 0.2563 | LCC = 0.8616 | SRCC = 0.8613 ] +INFO:root:Average loss=0.12051759994564912 +INFO:root: +[Valid][9][UTT][ MSE = 0.2460 | LCC = 0.8621 | SRCC = 0.8606 ] +INFO:root:Average loss=0.11061557659353966 +INFO:root: +[Valid][10][UTT][ MSE = 0.2745 | LCC = 0.8611 | SRCC = 0.8621 ] +INFO:root:Average loss=0.1077811296742696 +INFO:root: +[Valid][11][UTT][ MSE = 0.2345 | LCC = 0.8631 | SRCC = 0.8612 ] +INFO:root:Average loss=0.10053943441464351 +INFO:root: +[Valid][12][UTT][ MSE = 0.2203 | LCC = 0.8617 | SRCC = 0.8614 ] +INFO:root:Average loss=0.09619324033459027 +INFO:root: +[Valid][13][UTT][ MSE = 0.2383 | LCC = 0.8591 | SRCC = 0.8565 ] +INFO:root:Average loss=0.08979493990922585 +INFO:root: +[Valid][14][UTT][ MSE = 0.2516 | LCC = 0.8609 | SRCC = 0.8584 ] +INFO:root:Average loss=0.08993569852258915 +INFO:root: +[Valid][15][UTT][ MSE = 0.2706 | LCC = 0.8594 | SRCC = 0.8572 ] +INFO:root:Average loss=0.08439612231002404 +INFO:root: +[Valid][16][UTT][ MSE = 0.2573 | LCC = 0.8595 | SRCC = 0.8575 ] +INFO:root:Average loss=0.0760320071130991 +INFO:root: +[Valid][17][UTT][ MSE = 0.2412 | LCC = 0.8572 | SRCC = 0.8544 ] +INFO:root:Average loss=0.07526649859471199 +INFO:root: +[Valid][18][UTT][ MSE = 0.2588 | LCC = 0.8563 | SRCC = 0.8556 ] +INFO:root:Average loss=0.06760775679961228 +INFO:root: +[Valid][19][UTT][ MSE = 0.2307 | LCC = 0.8596 | SRCC = 0.8556 ] +INFO:root:Average loss=0.06895673274993896 +INFO:root: +[Valid][20][UTT][ MSE = 0.2378 | LCC = 0.8605 | SRCC = 0.8589 ] +INFO:root:Average loss=0.059871731325984 +INFO:root: +[Valid][21][UTT][ MSE = 0.2601 | LCC = 0.8602 | SRCC = 0.8574 ] +INFO:root:Average loss=0.0660487930648602 +INFO:root: +[Valid][22][UTT][ MSE = 0.2271 | LCC = 0.8600 | SRCC = 0.8567 ] +INFO:root:Average loss=0.057395887250701584 +INFO:root: +[Valid][23][UTT][ MSE = 0.2451 | LCC = 0.8623 | SRCC = 0.8602 ] +INFO:root:Average loss=0.05702023105457043 +INFO:root: +[Valid][24][UTT][ MSE = 0.2757 | LCC = 0.8563 | SRCC = 0.8534 ] +INFO:root:Average loss=0.051732671447098255 +INFO:root: +[Valid][25][UTT][ MSE = 0.2623 | LCC = 0.8569 | SRCC = 0.8548 ] +INFO:root:Average loss=0.050937341836591564 +INFO:root: +[Valid][26][UTT][ MSE = 0.2463 | LCC = 0.8589 | SRCC = 0.8573 ] +INFO:root:Average loss=0.05131654064051616 +INFO:root: +[Valid][27][UTT][ MSE = 0.2859 | LCC = 0.8572 | SRCC = 0.8561 ] +INFO:root:Average loss=0.047640154544168554 +INFO:root: +[Valid][28][UTT][ MSE = 0.3015 | LCC = 0.8566 | SRCC = 0.8562 ] +INFO:root:Average loss=0.04731529920051495 +INFO:root: +[Valid][29][UTT][ MSE = 0.2746 | LCC = 0.8562 | SRCC = 0.8532 ] +INFO:root:Average loss=0.04522312210442928 +INFO:root: +[Valid][30][UTT][ MSE = 0.2400 | LCC = 0.8540 | SRCC = 0.8514 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2364 | LCC = 0.8686 | SRCC = 0.8708 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..0aaf9b7d9cca86990ebc6e8e67bd6f548186fe67 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bde78208b3f06f6c13f4d2278e69c8b9a2d22a8f79139911c1c482f39367218a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..af77a1666818a9758ef66aa1e05527fe5593718b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94c6275d84eb64d2c8aa8ac11b48e401bfab97a657874a59439e02214afc0773 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..a6b39cfbe23912ddf89fc9badd42a94b62b1cd9a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.44423634119522876 +INFO:root: +[Valid][0][UTT][ MSE = 0.2990 | LCC = 0.8239 | SRCC = 0.8317 ] +INFO:root:Average loss=0.24600252031515807 +INFO:root: +[Valid][1][UTT][ MSE = 0.2776 | LCC = 0.8504 | SRCC = 0.8515 ] +INFO:root:Average loss=0.2120879366993904 +INFO:root: +[Valid][2][UTT][ MSE = 0.2514 | LCC = 0.8604 | SRCC = 0.8599 ] +INFO:root:Average loss=0.19321685389448434 +INFO:root: +[Valid][3][UTT][ MSE = 0.2436 | LCC = 0.8586 | SRCC = 0.8589 ] +INFO:root:Average loss=0.17314277598873162 +INFO:root: +[Valid][4][UTT][ MSE = 0.2387 | LCC = 0.8619 | SRCC = 0.8617 ] +INFO:root:Average loss=0.16573860610906893 +INFO:root: +[Valid][5][UTT][ MSE = 0.2286 | LCC = 0.8651 | SRCC = 0.8630 ] +INFO:root:Average loss=0.14895567880609098 +INFO:root: +[Valid][6][UTT][ MSE = 0.2282 | LCC = 0.8605 | SRCC = 0.8593 ] +INFO:root:Average loss=0.14125682270297638 +INFO:root: +[Valid][7][UTT][ MSE = 0.2271 | LCC = 0.8593 | SRCC = 0.8581 ] +INFO:root:Average loss=0.1352340905712201 +INFO:root: +[Valid][8][UTT][ MSE = 0.2353 | LCC = 0.8591 | SRCC = 0.8591 ] +INFO:root:Average loss=0.12643945618317678 +INFO:root: +[Valid][9][UTT][ MSE = 0.2387 | LCC = 0.8550 | SRCC = 0.8522 ] +INFO:root:Average loss=0.11226332187652588 +INFO:root: +[Valid][10][UTT][ MSE = 0.2293 | LCC = 0.8574 | SRCC = 0.8553 ] +INFO:root:Average loss=0.10661670193076134 +INFO:root: +[Valid][11][UTT][ MSE = 0.2288 | LCC = 0.8556 | SRCC = 0.8556 ] +INFO:root:Average loss=0.09771254524970666 +INFO:root: +[Valid][12][UTT][ MSE = 0.2304 | LCC = 0.8552 | SRCC = 0.8531 ] +INFO:root:Average loss=0.08979310954992588 +INFO:root: +[Valid][13][UTT][ MSE = 0.2322 | LCC = 0.8533 | SRCC = 0.8523 ] +INFO:root:Average loss=0.08373629120297921 +INFO:root: +[Valid][14][UTT][ MSE = 0.2290 | LCC = 0.8532 | SRCC = 0.8524 ] +INFO:root:Average loss=0.08023282890327466 +INFO:root: +[Valid][15][UTT][ MSE = 0.2380 | LCC = 0.8480 | SRCC = 0.8472 ] +INFO:root:Average loss=0.07495343226652879 +INFO:root: +[Valid][16][UTT][ MSE = 0.2325 | LCC = 0.8521 | SRCC = 0.8504 ] +INFO:root:Average loss=0.07115861539466259 +INFO:root: +[Valid][17][UTT][ MSE = 0.2468 | LCC = 0.8506 | SRCC = 0.8499 ] +INFO:root:Average loss=0.07123042332629363 +INFO:root: +[Valid][18][UTT][ MSE = 0.2413 | LCC = 0.8460 | SRCC = 0.8446 ] +INFO:root:Average loss=0.06582355375091235 +INFO:root: +[Valid][19][UTT][ MSE = 0.2394 | LCC = 0.8470 | SRCC = 0.8456 ] +INFO:root:Average loss=0.06005764747850406 +INFO:root: +[Valid][20][UTT][ MSE = 0.2377 | LCC = 0.8461 | SRCC = 0.8465 ] +INFO:root:Average loss=0.061252878453487 +INFO:root: +[Valid][21][UTT][ MSE = 0.2481 | LCC = 0.8449 | SRCC = 0.8440 ] +INFO:root:Average loss=0.053401183910094775 +INFO:root: +[Valid][22][UTT][ MSE = 0.2395 | LCC = 0.8457 | SRCC = 0.8447 ] +INFO:root:Average loss=0.05163371202368767 +INFO:root: +[Valid][23][UTT][ MSE = 0.2467 | LCC = 0.8465 | SRCC = 0.8465 ] +INFO:root:Average loss=0.05174460381460495 +INFO:root: +[Valid][24][UTT][ MSE = 0.2374 | LCC = 0.8465 | SRCC = 0.8462 ] +INFO:root:Average loss=0.04781143923696035 +INFO:root: +[Valid][25][UTT][ MSE = 0.2407 | LCC = 0.8442 | SRCC = 0.8424 ] +INFO:root:Average loss=0.04681229560325543 +INFO:root: +[Valid][26][UTT][ MSE = 0.2445 | LCC = 0.8479 | SRCC = 0.8467 ] +INFO:root:Average loss=0.04537919256836176 +INFO:root: +[Valid][27][UTT][ MSE = 0.2401 | LCC = 0.8451 | SRCC = 0.8447 ] +INFO:root:Average loss=0.046311221204889126 +INFO:root: +[Valid][28][UTT][ MSE = 0.2377 | LCC = 0.8472 | SRCC = 0.8472 ] +INFO:root:Average loss=0.042300977457601294 +INFO:root: +[Valid][29][UTT][ MSE = 0.2395 | LCC = 0.8446 | SRCC = 0.8444 ] +INFO:root:Average loss=0.041690496942744806 +INFO:root: +[Valid][30][UTT][ MSE = 0.2426 | LCC = 0.8438 | SRCC = 0.8426 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2267 | LCC = 0.8676 | SRCC = 0.8658 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..99fea062906b0291f99aad4b7a4be5bc2e1617b6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ae834ddcc412adc0226781585a96c2a6dbc7f75b44363f675fb82c02c7753c7 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..b5ad7a1ea466c79c3705a644b6ff7077e689ecd8 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34ec85ec576ef6b4ccf883cdf8bdd56a3d65df6d0a06ec27d79a14e0d023977f +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..4a29f858dcf580d24ecc712a6bf332e4bcc43739 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.3532790440397385 +INFO:root: +[Valid][0][UTT][ MSE = 0.2934 | LCC = 0.8278 | SRCC = 0.8305 ] +INFO:root:Average loss=0.23542056805812395 +INFO:root: +[Valid][1][UTT][ MSE = 0.2883 | LCC = 0.8505 | SRCC = 0.8502 ] +INFO:root:Average loss=0.21428402742514244 +INFO:root: +[Valid][2][UTT][ MSE = 0.2664 | LCC = 0.8623 | SRCC = 0.8621 ] +INFO:root:Average loss=0.1941398564630594 +INFO:root: +[Valid][3][UTT][ MSE = 0.2332 | LCC = 0.8660 | SRCC = 0.8656 ] +INFO:root:Average loss=0.17796254215332177 +INFO:root: +[Valid][4][UTT][ MSE = 0.2561 | LCC = 0.8691 | SRCC = 0.8673 ] +INFO:root:Average loss=0.15791213063475412 +INFO:root: +[Valid][5][UTT][ MSE = 0.2239 | LCC = 0.8685 | SRCC = 0.8671 ] +INFO:root:Average loss=0.15667983354666296 +INFO:root: +[Valid][6][UTT][ MSE = 0.2277 | LCC = 0.8689 | SRCC = 0.8691 ] +INFO:root:Average loss=0.13750377555306142 +INFO:root: +[Valid][7][UTT][ MSE = 0.2325 | LCC = 0.8633 | SRCC = 0.8609 ] +INFO:root:Average loss=0.13369866938163072 +INFO:root: +[Valid][8][UTT][ MSE = 0.2235 | LCC = 0.8586 | SRCC = 0.8613 ] +INFO:root:Average loss=0.12291351963694279 +INFO:root: +[Valid][9][UTT][ MSE = 0.2270 | LCC = 0.8621 | SRCC = 0.8622 ] +INFO:root:Average loss=0.11415767430877075 +INFO:root: +[Valid][10][UTT][ MSE = 0.2307 | LCC = 0.8643 | SRCC = 0.8627 ] +INFO:root:Average loss=0.10583419228593509 +INFO:root: +[Valid][11][UTT][ MSE = 0.2333 | LCC = 0.8599 | SRCC = 0.8578 ] +INFO:root:Average loss=0.09753793845765102 +INFO:root: +[Valid][12][UTT][ MSE = 0.2438 | LCC = 0.8615 | SRCC = 0.8602 ] +INFO:root:Average loss=0.09199596979679206 +INFO:root: +[Valid][13][UTT][ MSE = 0.2560 | LCC = 0.8585 | SRCC = 0.8589 ] +INFO:root:Average loss=0.09039602528970975 +INFO:root: +[Valid][14][UTT][ MSE = 0.2320 | LCC = 0.8575 | SRCC = 0.8577 ] +INFO:root:Average loss=0.08855613650610814 +INFO:root: +[Valid][15][UTT][ MSE = 0.2427 | LCC = 0.8549 | SRCC = 0.8536 ] +INFO:root:Average loss=0.07628511202832063 +INFO:root: +[Valid][16][UTT][ MSE = 0.2407 | LCC = 0.8567 | SRCC = 0.8567 ] +INFO:root:Average loss=0.07328453340018407 +INFO:root: +[Valid][17][UTT][ MSE = 0.2453 | LCC = 0.8509 | SRCC = 0.8515 ] +INFO:root:Average loss=0.07109743853410085 +INFO:root: +[Valid][18][UTT][ MSE = 0.2395 | LCC = 0.8545 | SRCC = 0.8528 ] +INFO:root:Average loss=0.0644948690270002 +INFO:root: +[Valid][19][UTT][ MSE = 0.2458 | LCC = 0.8537 | SRCC = 0.8531 ] +INFO:root:Average loss=0.06405318065140492 +INFO:root: +[Valid][20][UTT][ MSE = 0.2496 | LCC = 0.8513 | SRCC = 0.8511 ] +INFO:root:Average loss=0.06173017826408912 +INFO:root: +[Valid][21][UTT][ MSE = 0.2305 | LCC = 0.8546 | SRCC = 0.8545 ] +INFO:root:Average loss=0.053901965801532455 +INFO:root: +[Valid][22][UTT][ MSE = 0.2373 | LCC = 0.8525 | SRCC = 0.8521 ] +INFO:root:Average loss=0.055815434704224266 +INFO:root: +[Valid][23][UTT][ MSE = 0.2350 | LCC = 0.8504 | SRCC = 0.8503 ] +INFO:root:Average loss=0.0539230795051807 +INFO:root: +[Valid][24][UTT][ MSE = 0.2398 | LCC = 0.8503 | SRCC = 0.8503 ] +INFO:root:Average loss=0.05106512972941765 +INFO:root: +[Valid][25][UTT][ MSE = 0.2410 | LCC = 0.8499 | SRCC = 0.8511 ] +INFO:root:Average loss=0.05006508751270863 +INFO:root: +[Valid][26][UTT][ MSE = 0.2334 | LCC = 0.8507 | SRCC = 0.8504 ] +INFO:root:Average loss=0.04767979695819891 +INFO:root: +[Valid][27][UTT][ MSE = 0.2328 | LCC = 0.8512 | SRCC = 0.8507 ] +INFO:root:Average loss=0.04681019492160816 +INFO:root: +[Valid][28][UTT][ MSE = 0.2404 | LCC = 0.8535 | SRCC = 0.8527 ] +INFO:root:Average loss=0.04501170376076912 +INFO:root: +[Valid][29][UTT][ MSE = 0.2451 | LCC = 0.8522 | SRCC = 0.8516 ] +INFO:root:Average loss=0.04231249581640347 +INFO:root: +[Valid][30][UTT][ MSE = 0.2419 | LCC = 0.8498 | SRCC = 0.8489 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2760 | LCC = 0.8626 | SRCC = 0.8641 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..f045f054c409c7eeb2d935cef85dcb405cf71b60 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9cf51f3a3eead36f0b7009b143bed606e9ebaec4f2813b047a5412bc9f278edd +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7f2a0b8a9cbf5d172205156316d809963a9d689a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5be37790b9a09dae4f464107570b304b4caa397f96eba3ba8c823f26dd58e20b +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..b9261dce888d5dbc5cdb9749107a73c5d520e971 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.4027602838782164 +INFO:root: +[Valid][0][UTT][ MSE = 0.3175 | LCC = 0.8213 | SRCC = 0.8276 ] +INFO:root:Average loss=0.24248880835679862 +INFO:root: +[Valid][1][UTT][ MSE = 0.2649 | LCC = 0.8436 | SRCC = 0.8458 ] +INFO:root:Average loss=0.21423527006155405 +INFO:root: +[Valid][2][UTT][ MSE = 0.2488 | LCC = 0.8589 | SRCC = 0.8569 ] +INFO:root:Average loss=0.1917875418678308 +INFO:root: +[Valid][3][UTT][ MSE = 0.2348 | LCC = 0.8582 | SRCC = 0.8570 ] +INFO:root:Average loss=0.180630062157527 +INFO:root: +[Valid][4][UTT][ MSE = 0.2286 | LCC = 0.8612 | SRCC = 0.8596 ] +INFO:root:Average loss=0.16909473685499948 +INFO:root: +[Valid][5][UTT][ MSE = 0.2367 | LCC = 0.8630 | SRCC = 0.8615 ] +INFO:root:Average loss=0.15024232482298827 +INFO:root: +[Valid][6][UTT][ MSE = 0.2219 | LCC = 0.8601 | SRCC = 0.8588 ] +INFO:root:Average loss=0.14234386394039178 +INFO:root: +[Valid][7][UTT][ MSE = 0.2279 | LCC = 0.8623 | SRCC = 0.8598 ] +INFO:root:Average loss=0.13202622284491858 +INFO:root: +[Valid][8][UTT][ MSE = 0.2416 | LCC = 0.8599 | SRCC = 0.8580 ] +INFO:root:Average loss=0.12129659535220036 +INFO:root: +[Valid][9][UTT][ MSE = 0.2364 | LCC = 0.8593 | SRCC = 0.8577 ] +INFO:root:Average loss=0.11854073663170521 +INFO:root: +[Valid][10][UTT][ MSE = 0.2296 | LCC = 0.8553 | SRCC = 0.8527 ] +INFO:root:Average loss=0.10408478889327782 +INFO:root: +[Valid][11][UTT][ MSE = 0.2416 | LCC = 0.8574 | SRCC = 0.8552 ] +INFO:root:Average loss=0.10343983291815488 +INFO:root: +[Valid][12][UTT][ MSE = 0.2492 | LCC = 0.8564 | SRCC = 0.8559 ] +INFO:root:Average loss=0.10006399032397148 +INFO:root: +[Valid][13][UTT][ MSE = 0.2303 | LCC = 0.8574 | SRCC = 0.8551 ] +INFO:root:Average loss=0.08611657881201842 +INFO:root: +[Valid][14][UTT][ MSE = 0.2307 | LCC = 0.8556 | SRCC = 0.8547 ] +INFO:root:Average loss=0.08110168127295299 +INFO:root: +[Valid][15][UTT][ MSE = 0.2338 | LCC = 0.8533 | SRCC = 0.8518 ] +INFO:root:Average loss=0.07647321058007386 +INFO:root: +[Valid][16][UTT][ MSE = 0.2276 | LCC = 0.8535 | SRCC = 0.8528 ] +INFO:root:Average loss=0.07355992304973114 +INFO:root: +[Valid][17][UTT][ MSE = 0.2549 | LCC = 0.8524 | SRCC = 0.8511 ] +INFO:root:Average loss=0.06579209854587531 +INFO:root: +[Valid][18][UTT][ MSE = 0.2302 | LCC = 0.8512 | SRCC = 0.8507 ] +INFO:root:Average loss=0.06732475451934032 +INFO:root: +[Valid][19][UTT][ MSE = 0.2357 | LCC = 0.8495 | SRCC = 0.8478 ] +INFO:root:Average loss=0.06826886716179359 +INFO:root: +[Valid][20][UTT][ MSE = 0.2332 | LCC = 0.8518 | SRCC = 0.8494 ] +INFO:root:Average loss=0.058669458549374186 +INFO:root: +[Valid][21][UTT][ MSE = 0.2392 | LCC = 0.8509 | SRCC = 0.8491 ] +INFO:root:Average loss=0.05678849576566464 +INFO:root: +[Valid][22][UTT][ MSE = 0.2530 | LCC = 0.8485 | SRCC = 0.8462 ] +INFO:root:Average loss=0.05577952519823343 +INFO:root: +[Valid][23][UTT][ MSE = 0.2455 | LCC = 0.8506 | SRCC = 0.8484 ] +INFO:root:Average loss=0.05084414116274088 +INFO:root: +[Valid][24][UTT][ MSE = 0.2363 | LCC = 0.8491 | SRCC = 0.8472 ] +INFO:root:Average loss=0.05187613454957803 +INFO:root: +[Valid][25][UTT][ MSE = 0.2633 | LCC = 0.8459 | SRCC = 0.8445 ] +INFO:root:Average loss=0.05268848532189926 +INFO:root: +[Valid][26][UTT][ MSE = 0.2394 | LCC = 0.8472 | SRCC = 0.8456 ] +INFO:root:Average loss=0.048698993877340585 +INFO:root: +[Valid][27][UTT][ MSE = 0.2399 | LCC = 0.8467 | SRCC = 0.8459 ] +INFO:root:Average loss=0.04616852369732582 +INFO:root: +[Valid][28][UTT][ MSE = 0.2365 | LCC = 0.8485 | SRCC = 0.8462 ] +INFO:root:Average loss=0.04534845875623899 +INFO:root: +[Valid][29][UTT][ MSE = 0.2486 | LCC = 0.8510 | SRCC = 0.8499 ] +INFO:root:Average loss=0.044269719471534096 +INFO:root: +[Valid][30][UTT][ MSE = 0.2517 | LCC = 0.8502 | SRCC = 0.8499 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2600 | LCC = 0.8584 | SRCC = 0.8611 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..0eee4939ad00221e9bd79147204e3a4d7f69321e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:872bb65a71c7013acb7bfa01dffe85f7a7946e27454e2bd36831a0cb4292bbc5 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..8f7129a15a3cdcec569ecf5c0cdbfbf4946f6b5d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94094a0a5725510200f08fe8b9c3bcb245f5fd10e291cd9cd3734f4e2542550d +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..00d3f236e26464c8c2f57befed6730bf70c2626a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.390387865404288 +INFO:root: +[Valid][0][UTT][ MSE = 0.3325 | LCC = 0.8312 | SRCC = 0.8322 ] +INFO:root:Average loss=0.24299630560936072 +INFO:root: +[Valid][1][UTT][ MSE = 0.2783 | LCC = 0.8501 | SRCC = 0.8485 ] +INFO:root:Average loss=0.211145974504642 +INFO:root: +[Valid][2][UTT][ MSE = 0.2853 | LCC = 0.8590 | SRCC = 0.8554 ] +INFO:root:Average loss=0.1905519438859744 +INFO:root: +[Valid][3][UTT][ MSE = 0.2613 | LCC = 0.8611 | SRCC = 0.8579 ] +INFO:root:Average loss=0.17283148280321023 +INFO:root: +[Valid][4][UTT][ MSE = 0.2459 | LCC = 0.8606 | SRCC = 0.8592 ] +INFO:root:Average loss=0.15754488492623353 +INFO:root: +[Valid][5][UTT][ MSE = 0.2418 | LCC = 0.8614 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1468845687997647 +INFO:root: +[Valid][6][UTT][ MSE = 0.2556 | LCC = 0.8576 | SRCC = 0.8537 ] +INFO:root:Average loss=0.13577705201430199 +INFO:root: +[Valid][7][UTT][ MSE = 0.2799 | LCC = 0.8554 | SRCC = 0.8527 ] +INFO:root:Average loss=0.12693976572690865 +INFO:root: +[Valid][8][UTT][ MSE = 0.2357 | LCC = 0.8568 | SRCC = 0.8547 ] +INFO:root:Average loss=0.11799867842824031 +INFO:root: +[Valid][9][UTT][ MSE = 0.2393 | LCC = 0.8575 | SRCC = 0.8547 ] +INFO:root:Average loss=0.10472278019938713 +INFO:root: +[Valid][10][UTT][ MSE = 0.2408 | LCC = 0.8561 | SRCC = 0.8528 ] +INFO:root:Average loss=0.09823061525821686 +INFO:root: +[Valid][11][UTT][ MSE = 0.2496 | LCC = 0.8518 | SRCC = 0.8500 ] +INFO:root:Average loss=0.09376707706504907 +INFO:root: +[Valid][12][UTT][ MSE = 0.2390 | LCC = 0.8562 | SRCC = 0.8544 ] +INFO:root:Average loss=0.09031981053069615 +INFO:root: +[Valid][13][UTT][ MSE = 0.2698 | LCC = 0.8488 | SRCC = 0.8466 ] +INFO:root:Average loss=0.0860510080670699 +INFO:root: +[Valid][14][UTT][ MSE = 0.2456 | LCC = 0.8530 | SRCC = 0.8515 ] +INFO:root:Average loss=0.08128327317535877 +INFO:root: +[Valid][15][UTT][ MSE = 0.2428 | LCC = 0.8513 | SRCC = 0.8488 ] +INFO:root:Average loss=0.07254672227188563 +INFO:root: +[Valid][16][UTT][ MSE = 0.2470 | LCC = 0.8494 | SRCC = 0.8466 ] +INFO:root:Average loss=0.07539081836167054 +INFO:root: +[Valid][17][UTT][ MSE = 0.2708 | LCC = 0.8478 | SRCC = 0.8456 ] +INFO:root:Average loss=0.06739270319350255 +INFO:root: +[Valid][18][UTT][ MSE = 0.2478 | LCC = 0.8478 | SRCC = 0.8458 ] +INFO:root:Average loss=0.06240894650228512 +INFO:root: +[Valid][19][UTT][ MSE = 0.2623 | LCC = 0.8472 | SRCC = 0.8447 ] +INFO:root:Average loss=0.06134763445991736 +INFO:root: +[Valid][20][UTT][ MSE = 0.2409 | LCC = 0.8493 | SRCC = 0.8466 ] +INFO:root:Average loss=0.05671720681950832 +INFO:root: +[Valid][21][UTT][ MSE = 0.2600 | LCC = 0.8456 | SRCC = 0.8433 ] +INFO:root:Average loss=0.05585585759045222 +INFO:root: +[Valid][22][UTT][ MSE = 0.2624 | LCC = 0.8429 | SRCC = 0.8410 ] +INFO:root:Average loss=0.05149386407664189 +INFO:root: +[Valid][23][UTT][ MSE = 0.2472 | LCC = 0.8465 | SRCC = 0.8455 ] +INFO:root:Average loss=0.05721375546776331 +INFO:root: +[Valid][24][UTT][ MSE = 0.2588 | LCC = 0.8434 | SRCC = 0.8419 ] +INFO:root:Average loss=0.05000851413187308 +INFO:root: +[Valid][25][UTT][ MSE = 0.2621 | LCC = 0.8415 | SRCC = 0.8397 ] +INFO:root:Average loss=0.04677107400045945 +INFO:root: +[Valid][26][UTT][ MSE = 0.2593 | LCC = 0.8392 | SRCC = 0.8372 ] +INFO:root:Average loss=0.04444294145856148 +INFO:root: +[Valid][27][UTT][ MSE = 0.2728 | LCC = 0.8383 | SRCC = 0.8362 ] +INFO:root:Average loss=0.04362839450820898 +INFO:root: +[Valid][28][UTT][ MSE = 0.2587 | LCC = 0.8388 | SRCC = 0.8361 ] +INFO:root:Average loss=0.04268149138452151 +INFO:root: +[Valid][29][UTT][ MSE = 0.2612 | LCC = 0.8372 | SRCC = 0.8342 ] +INFO:root:Average loss=0.042681077805658184 +INFO:root: +[Valid][30][UTT][ MSE = 0.2534 | LCC = 0.8405 | SRCC = 0.8388 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2285 | LCC = 0.8669 | SRCC = 0.8648 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..83d3d20b350020902324f304ccdf6af0b883b47f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85794d523ba3d13215b2b4913eeceb5dc8ec262df97b6a6371a374d77e537ac0 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4575ea9682a8e1c9634a11ec54f3034ddada4105 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61dc1938b5234767afe427ac28803dc9911cc353070b2f0240d6c3fbf59c0c1d +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer5/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e760f95b5177dc8452a7538caa9bdc84b18ad1d1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.4032483144830435 +INFO:root: +[Valid][0][UTT][ MSE = 0.3211 | LCC = 0.8338 | SRCC = 0.8352 ] +INFO:root:Average loss=0.2348736129128016 +INFO:root: +[Valid][1][UTT][ MSE = 0.2914 | LCC = 0.8510 | SRCC = 0.8495 ] +INFO:root:Average loss=0.2102135672019078 +INFO:root: +[Valid][2][UTT][ MSE = 0.2564 | LCC = 0.8600 | SRCC = 0.8567 ] +INFO:root:Average loss=0.18792994492329085 +INFO:root: +[Valid][3][UTT][ MSE = 0.2597 | LCC = 0.8609 | SRCC = 0.8584 ] +INFO:root:Average loss=0.176579239181219 +INFO:root: +[Valid][4][UTT][ MSE = 0.2774 | LCC = 0.8602 | SRCC = 0.8586 ] +INFO:root:Average loss=0.16118293112287155 +INFO:root: +[Valid][5][UTT][ MSE = 0.2622 | LCC = 0.8617 | SRCC = 0.8591 ] +INFO:root:Average loss=0.14973078075891885 +INFO:root: +[Valid][6][UTT][ MSE = 0.2731 | LCC = 0.8589 | SRCC = 0.8572 ] +INFO:root:Average loss=0.13891522729626068 +INFO:root: +[Valid][7][UTT][ MSE = 0.2656 | LCC = 0.8591 | SRCC = 0.8566 ] +INFO:root:Average loss=0.12982434196731982 +INFO:root: +[Valid][8][UTT][ MSE = 0.3466 | LCC = 0.8547 | SRCC = 0.8553 ] +INFO:root:Average loss=0.1187166609825232 +INFO:root: +[Valid][9][UTT][ MSE = 0.2461 | LCC = 0.8576 | SRCC = 0.8560 ] +INFO:root:Average loss=0.11086661225328079 +INFO:root: +[Valid][10][UTT][ MSE = 0.2552 | LCC = 0.8549 | SRCC = 0.8539 ] +INFO:root:Average loss=0.10810093982861592 +INFO:root: +[Valid][11][UTT][ MSE = 0.2698 | LCC = 0.8506 | SRCC = 0.8493 ] +INFO:root:Average loss=0.0963942218476381 +INFO:root: +[Valid][12][UTT][ MSE = 0.2698 | LCC = 0.8499 | SRCC = 0.8472 ] +INFO:root:Average loss=0.0908628606165831 +INFO:root: +[Valid][13][UTT][ MSE = 0.2642 | LCC = 0.8471 | SRCC = 0.8444 ] +INFO:root:Average loss=0.08656715095425263 +INFO:root: +[Valid][14][UTT][ MSE = 0.2712 | LCC = 0.8473 | SRCC = 0.8449 ] +INFO:root:Average loss=0.08374684514143528 +INFO:root: +[Valid][15][UTT][ MSE = 0.2676 | LCC = 0.8447 | SRCC = 0.8415 ] +INFO:root:Average loss=0.07634989606837432 +INFO:root: +[Valid][16][UTT][ MSE = 0.2672 | LCC = 0.8415 | SRCC = 0.8387 ] +INFO:root:Average loss=0.0709335132000538 +INFO:root: +[Valid][17][UTT][ MSE = 0.2577 | LCC = 0.8450 | SRCC = 0.8435 ] +INFO:root:Average loss=0.06986614135213387 +INFO:root: +[Valid][18][UTT][ MSE = 0.2601 | LCC = 0.8407 | SRCC = 0.8372 ] +INFO:root:Average loss=0.07277473869423072 +INFO:root: +[Valid][19][UTT][ MSE = 0.2738 | LCC = 0.8437 | SRCC = 0.8435 ] +INFO:root:Average loss=0.06414768694398496 +INFO:root: +[Valid][20][UTT][ MSE = 0.2595 | LCC = 0.8425 | SRCC = 0.8404 ] +INFO:root:Average loss=0.0602857620956806 +INFO:root: +[Valid][21][UTT][ MSE = 0.2573 | LCC = 0.8455 | SRCC = 0.8430 ] +INFO:root:Average loss=0.06395931102526493 +INFO:root: +[Valid][22][UTT][ MSE = 0.2818 | LCC = 0.8451 | SRCC = 0.8435 ] +INFO:root:Average loss=0.05494058055755419 +INFO:root: +[Valid][23][UTT][ MSE = 0.2690 | LCC = 0.8414 | SRCC = 0.8383 ] +INFO:root:Average loss=0.052682448751651324 +INFO:root: +[Valid][24][UTT][ MSE = 0.2695 | LCC = 0.8426 | SRCC = 0.8389 ] +INFO:root:Average loss=0.05400323691085363 +INFO:root: +[Valid][25][UTT][ MSE = 0.2599 | LCC = 0.8412 | SRCC = 0.8380 ] +INFO:root:Average loss=0.05037222516078215 +INFO:root: +[Valid][26][UTT][ MSE = 0.2630 | LCC = 0.8444 | SRCC = 0.8426 ] +INFO:root:Average loss=0.049636176142555014 +INFO:root: +[Valid][27][UTT][ MSE = 0.2599 | LCC = 0.8426 | SRCC = 0.8387 ] +INFO:root:Average loss=0.05030002062901472 +INFO:root: +[Valid][28][UTT][ MSE = 0.2650 | LCC = 0.8455 | SRCC = 0.8425 ] +INFO:root:Average loss=0.0447976511831467 +INFO:root: +[Valid][29][UTT][ MSE = 0.2674 | LCC = 0.8435 | SRCC = 0.8401 ] +INFO:root:Average loss=0.04598891976265571 +INFO:root: +[Valid][30][UTT][ MSE = 0.2646 | LCC = 0.8432 | SRCC = 0.8400 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2561 | LCC = 0.8614 | SRCC = 0.8590 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..f55061a0155a4f8ae05c82d0b8e98d25e68b54f1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efe2ab1cdbfb0b5a6886ff03c4982db399c2f7af03604365560accd89b34d77f +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..fb3322d75fef647a23af903838c2c107976804e6 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f087619f78d2c35a56bc9f233c3b6c482c8468cec1fa5a7cf6f7a7ea19ead75 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..e5d36dcb2f629619f615946d0da90dc72a456e5c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.35967685473270905 +INFO:root: +[Valid][0][UTT][ MSE = 0.3316 | LCC = 0.8282 | SRCC = 0.8308 ] +INFO:root:Average loss=0.23711970352973694 +INFO:root: +[Valid][1][UTT][ MSE = 0.2918 | LCC = 0.8471 | SRCC = 0.8465 ] +INFO:root:Average loss=0.21093845329223535 +INFO:root: +[Valid][2][UTT][ MSE = 0.2880 | LCC = 0.8558 | SRCC = 0.8544 ] +INFO:root:Average loss=0.19452034978148264 +INFO:root: +[Valid][3][UTT][ MSE = 0.3761 | LCC = 0.8506 | SRCC = 0.8541 ] +INFO:root:Average loss=0.17138560527028182 +INFO:root: +[Valid][4][UTT][ MSE = 0.2969 | LCC = 0.8504 | SRCC = 0.8499 ] +INFO:root:Average loss=0.15878390453946897 +INFO:root: +[Valid][5][UTT][ MSE = 0.2688 | LCC = 0.8505 | SRCC = 0.8468 ] +INFO:root:Average loss=0.14624023189147314 +INFO:root: +[Valid][6][UTT][ MSE = 0.2970 | LCC = 0.8477 | SRCC = 0.8457 ] +INFO:root:Average loss=0.13760267857175607 +INFO:root: +[Valid][7][UTT][ MSE = 0.2630 | LCC = 0.8486 | SRCC = 0.8455 ] +INFO:root:Average loss=0.1259141712425611 +INFO:root: +[Valid][8][UTT][ MSE = 0.2989 | LCC = 0.8472 | SRCC = 0.8458 ] +INFO:root:Average loss=0.12264146455205403 +INFO:root: +[Valid][9][UTT][ MSE = 0.3421 | LCC = 0.8415 | SRCC = 0.8386 ] +INFO:root:Average loss=0.11746781434004124 +INFO:root: +[Valid][10][UTT][ MSE = 0.2892 | LCC = 0.8410 | SRCC = 0.8388 ] +INFO:root:Average loss=0.10772516081730525 +INFO:root: +[Valid][11][UTT][ MSE = 0.2524 | LCC = 0.8418 | SRCC = 0.8376 ] +INFO:root:Average loss=0.09865013452676627 +INFO:root: +[Valid][12][UTT][ MSE = 0.2793 | LCC = 0.8406 | SRCC = 0.8368 ] +INFO:root:Average loss=0.09415292601363781 +INFO:root: +[Valid][13][UTT][ MSE = 0.2870 | LCC = 0.8404 | SRCC = 0.8383 ] +INFO:root:Average loss=0.09736601320596841 +INFO:root: +[Valid][14][UTT][ MSE = 0.2647 | LCC = 0.8390 | SRCC = 0.8345 ] +INFO:root:Average loss=0.07905932811972423 +INFO:root: +[Valid][15][UTT][ MSE = 0.2657 | LCC = 0.8400 | SRCC = 0.8357 ] +INFO:root:Average loss=0.0782618826876084 +INFO:root: +[Valid][16][UTT][ MSE = 0.2818 | LCC = 0.8376 | SRCC = 0.8324 ] +INFO:root:Average loss=0.0708045453692858 +INFO:root: +[Valid][17][UTT][ MSE = 0.2773 | LCC = 0.8354 | SRCC = 0.8310 ] +INFO:root:Average loss=0.06795135818612881 +INFO:root: +[Valid][18][UTT][ MSE = 0.2882 | LCC = 0.8360 | SRCC = 0.8331 ] +INFO:root:Average loss=0.06667435522644947 +INFO:root: +[Valid][19][UTT][ MSE = 0.3332 | LCC = 0.8321 | SRCC = 0.8294 ] +INFO:root:Average loss=0.0671768439694857 +INFO:root: +[Valid][20][UTT][ MSE = 0.2952 | LCC = 0.8328 | SRCC = 0.8295 ] +INFO:root:Average loss=0.06286245742096351 +INFO:root: +[Valid][21][UTT][ MSE = 0.2908 | LCC = 0.8349 | SRCC = 0.8322 ] +INFO:root:Average loss=0.06274988129734993 +INFO:root: +[Valid][22][UTT][ MSE = 0.2872 | LCC = 0.8363 | SRCC = 0.8333 ] +INFO:root:Average loss=0.0573060365441518 +INFO:root: +[Valid][23][UTT][ MSE = 0.2804 | LCC = 0.8378 | SRCC = 0.8345 ] +INFO:root:Average loss=0.05414001185160417 +INFO:root: +[Valid][24][UTT][ MSE = 0.3222 | LCC = 0.8364 | SRCC = 0.8337 ] +INFO:root:Average loss=0.051952354347285554 +INFO:root: +[Valid][25][UTT][ MSE = 0.2925 | LCC = 0.8326 | SRCC = 0.8293 ] +INFO:root:Average loss=0.050910882030924164 +INFO:root: +[Valid][26][UTT][ MSE = 0.2516 | LCC = 0.8398 | SRCC = 0.8359 ] +INFO:root:Average loss=0.05010906876757359 +INFO:root: +[Valid][27][UTT][ MSE = 0.2696 | LCC = 0.8389 | SRCC = 0.8361 ] +INFO:root:Average loss=0.050320616278510824 +INFO:root: +[Valid][28][UTT][ MSE = 0.2939 | LCC = 0.8338 | SRCC = 0.8316 ] +INFO:root:Average loss=0.044204892948842965 +INFO:root: +[Valid][29][UTT][ MSE = 0.2789 | LCC = 0.8310 | SRCC = 0.8272 ] +INFO:root:Average loss=0.044112035646461524 +INFO:root: +[Valid][30][UTT][ MSE = 0.2950 | LCC = 0.8317 | SRCC = 0.8292 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2565 | LCC = 0.8646 | SRCC = 0.8650 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..0da6b592b3be04a83e7c3600e7a310bd9a4d971c --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:29353ce518c8b1c62cfe54ccb579f9265a0c14edddbf5916ab89a4f88b8d0893 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..990de750dbe109a95d585376b03308e4aae8ddbf --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c8490136e15080cf77f82781347895bd99d013a2332cca11a2dc3b82b60ad06 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..1b645926e335117b0a1a0cb7d3d8a208b7b9a863 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.34992452558034504 +INFO:root: +[Valid][0][UTT][ MSE = 0.3138 | LCC = 0.8356 | SRCC = 0.8366 ] +INFO:root:Average loss=0.23610629504307723 +INFO:root: +[Valid][1][UTT][ MSE = 0.2573 | LCC = 0.8537 | SRCC = 0.8533 ] +INFO:root:Average loss=0.2204148165690593 +INFO:root: +[Valid][2][UTT][ MSE = 0.2613 | LCC = 0.8614 | SRCC = 0.8596 ] +INFO:root:Average loss=0.19698775339967164 +INFO:root: +[Valid][3][UTT][ MSE = 0.2450 | LCC = 0.8629 | SRCC = 0.8610 ] +INFO:root:Average loss=0.17850503392326525 +INFO:root: +[Valid][4][UTT][ MSE = 0.2596 | LCC = 0.8628 | SRCC = 0.8603 ] +INFO:root:Average loss=0.17331499625474978 +INFO:root: +[Valid][5][UTT][ MSE = 0.2598 | LCC = 0.8607 | SRCC = 0.8592 ] +INFO:root:Average loss=0.15550493706877416 +INFO:root: +[Valid][6][UTT][ MSE = 0.2535 | LCC = 0.8591 | SRCC = 0.8583 ] +INFO:root:Average loss=0.1490496658743956 +INFO:root: +[Valid][7][UTT][ MSE = 0.2457 | LCC = 0.8575 | SRCC = 0.8570 ] +INFO:root:Average loss=0.13337358660422838 +INFO:root: +[Valid][8][UTT][ MSE = 0.2516 | LCC = 0.8560 | SRCC = 0.8555 ] +INFO:root:Average loss=0.12512784604078683 +INFO:root: +[Valid][9][UTT][ MSE = 0.2824 | LCC = 0.8491 | SRCC = 0.8493 ] +INFO:root:Average loss=0.11923328385903285 +INFO:root: +[Valid][10][UTT][ MSE = 0.2465 | LCC = 0.8525 | SRCC = 0.8521 ] +INFO:root:Average loss=0.11249589743331456 +INFO:root: +[Valid][11][UTT][ MSE = 0.2427 | LCC = 0.8504 | SRCC = 0.8501 ] +INFO:root:Average loss=0.10388663936501895 +INFO:root: +[Valid][12][UTT][ MSE = 0.2977 | LCC = 0.8465 | SRCC = 0.8474 ] +INFO:root:Average loss=0.09730450293192497 +INFO:root: +[Valid][13][UTT][ MSE = 0.2579 | LCC = 0.8482 | SRCC = 0.8481 ] +INFO:root:Average loss=0.10021176513953087 +INFO:root: +[Valid][14][UTT][ MSE = 0.2569 | LCC = 0.8454 | SRCC = 0.8458 ] +INFO:root:Average loss=0.08863390643054093 +INFO:root: +[Valid][15][UTT][ MSE = 0.2721 | LCC = 0.8492 | SRCC = 0.8509 ] +INFO:root:Average loss=0.0814015305099579 +INFO:root: +[Valid][16][UTT][ MSE = 0.2807 | LCC = 0.8458 | SRCC = 0.8461 ] +INFO:root:Average loss=0.07591542563377282 +INFO:root: +[Valid][17][UTT][ MSE = 0.2435 | LCC = 0.8503 | SRCC = 0.8502 ] +INFO:root:Average loss=0.07123773301450106 +INFO:root: +[Valid][18][UTT][ MSE = 0.2499 | LCC = 0.8480 | SRCC = 0.8484 ] +INFO:root:Average loss=0.07220442215792644 +INFO:root: +[Valid][19][UTT][ MSE = 0.2596 | LCC = 0.8444 | SRCC = 0.8437 ] +INFO:root:Average loss=0.0652000264575084 +INFO:root: +[Valid][20][UTT][ MSE = 0.2746 | LCC = 0.8409 | SRCC = 0.8407 ] +INFO:root:Average loss=0.06492882317457443 +INFO:root: +[Valid][21][UTT][ MSE = 0.2613 | LCC = 0.8428 | SRCC = 0.8430 ] +INFO:root:Average loss=0.059209355081503205 +INFO:root: +[Valid][22][UTT][ MSE = 0.2665 | LCC = 0.8410 | SRCC = 0.8418 ] +INFO:root:Average loss=0.05710396839257998 +INFO:root: +[Valid][23][UTT][ MSE = 0.2527 | LCC = 0.8460 | SRCC = 0.8463 ] +INFO:root:Average loss=0.05515124238072298 +INFO:root: +[Valid][24][UTT][ MSE = 0.2691 | LCC = 0.8355 | SRCC = 0.8348 ] +INFO:root:Average loss=0.05341970548033714 +INFO:root: +[Valid][25][UTT][ MSE = 0.2562 | LCC = 0.8387 | SRCC = 0.8380 ] +INFO:root:Average loss=0.05050580924711166 +INFO:root: +[Valid][26][UTT][ MSE = 0.2574 | LCC = 0.8414 | SRCC = 0.8417 ] +INFO:root:Average loss=0.04950651545555164 +INFO:root: +[Valid][27][UTT][ MSE = 0.2551 | LCC = 0.8394 | SRCC = 0.8387 ] +INFO:root:Average loss=0.04619490820914507 +INFO:root: +[Valid][28][UTT][ MSE = 0.2629 | LCC = 0.8392 | SRCC = 0.8394 ] +INFO:root:Average loss=0.0459886768546242 +INFO:root: +[Valid][29][UTT][ MSE = 0.2656 | LCC = 0.8385 | SRCC = 0.8388 ] +INFO:root:Average loss=0.04635453527458967 +INFO:root: +[Valid][30][UTT][ MSE = 0.2724 | LCC = 0.8362 | SRCC = 0.8359 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2593 | LCC = 0.8607 | SRCC = 0.8628 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..0d4c1edb5c7217b193ece13d84a0a96470c5e4f1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:849de156bc2f3c76fddd07fab1c7d721232238f22ebbedfd36826095c315dd30 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..5317fa0f5010ac28c424fa324d2132fa200d3351 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8181119eef4a0b220b35e5a1d0092d4ea125eee59d20a66438cda81fa452cec4 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..256f0608e6f9cd44e7a9947ed5151682d9f760de --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.36013838667900133 +INFO:root: +[Valid][0][UTT][ MSE = 0.3168 | LCC = 0.8310 | SRCC = 0.8355 ] +INFO:root:Average loss=0.23723947677092674 +INFO:root: +[Valid][1][UTT][ MSE = 0.2616 | LCC = 0.8521 | SRCC = 0.8527 ] +INFO:root:Average loss=0.21054338396359712 +INFO:root: +[Valid][2][UTT][ MSE = 0.2684 | LCC = 0.8595 | SRCC = 0.8575 ] +INFO:root:Average loss=0.19509602615084404 +INFO:root: +[Valid][3][UTT][ MSE = 0.2587 | LCC = 0.8649 | SRCC = 0.8637 ] +INFO:root:Average loss=0.182223535883121 +INFO:root: +[Valid][4][UTT][ MSE = 0.2713 | LCC = 0.8610 | SRCC = 0.8617 ] +INFO:root:Average loss=0.17700555968361023 +INFO:root: +[Valid][5][UTT][ MSE = 0.2566 | LCC = 0.8632 | SRCC = 0.8633 ] +INFO:root:Average loss=0.15644565835977212 +INFO:root: +[Valid][6][UTT][ MSE = 0.2603 | LCC = 0.8596 | SRCC = 0.8585 ] +INFO:root:Average loss=0.15452930990319985 +INFO:root: +[Valid][7][UTT][ MSE = 0.2642 | LCC = 0.8624 | SRCC = 0.8617 ] +INFO:root:Average loss=0.13644647884827393 +INFO:root: +[Valid][8][UTT][ MSE = 0.2597 | LCC = 0.8571 | SRCC = 0.8560 ] +INFO:root:Average loss=0.12365600887017372 +INFO:root: +[Valid][9][UTT][ MSE = 0.2406 | LCC = 0.8526 | SRCC = 0.8536 ] +INFO:root:Average loss=0.11982880236628728 +INFO:root: +[Valid][10][UTT][ MSE = 0.2580 | LCC = 0.8473 | SRCC = 0.8499 ] +INFO:root:Average loss=0.11170673141112694 +INFO:root: +[Valid][11][UTT][ MSE = 0.2845 | LCC = 0.8461 | SRCC = 0.8462 ] +INFO:root:Average loss=0.11101976591042984 +INFO:root: +[Valid][12][UTT][ MSE = 0.2933 | LCC = 0.8459 | SRCC = 0.8461 ] +INFO:root:Average loss=0.10371038045447606 +INFO:root: +[Valid][13][UTT][ MSE = 0.2563 | LCC = 0.8497 | SRCC = 0.8504 ] +INFO:root:Average loss=0.09338421011582398 +INFO:root: +[Valid][14][UTT][ MSE = 0.2626 | LCC = 0.8422 | SRCC = 0.8442 ] +INFO:root:Average loss=0.08728231040713115 +INFO:root: +[Valid][15][UTT][ MSE = 0.2725 | LCC = 0.8436 | SRCC = 0.8440 ] +INFO:root:Average loss=0.08433714805123134 +INFO:root: +[Valid][16][UTT][ MSE = 0.2839 | LCC = 0.8404 | SRCC = 0.8417 ] +INFO:root:Average loss=0.07610477468906304 +INFO:root: +[Valid][17][UTT][ MSE = 0.2672 | LCC = 0.8402 | SRCC = 0.8414 ] +INFO:root:Average loss=0.07050252753572586 +INFO:root: +[Valid][18][UTT][ MSE = 0.2615 | LCC = 0.8434 | SRCC = 0.8448 ] +INFO:root:Average loss=0.07019792578350274 +INFO:root: +[Valid][19][UTT][ MSE = 0.2635 | LCC = 0.8426 | SRCC = 0.8428 ] +INFO:root:Average loss=0.0661021151030675 +INFO:root: +[Valid][20][UTT][ MSE = 0.2571 | LCC = 0.8411 | SRCC = 0.8427 ] +INFO:root:Average loss=0.06218768995350752 +INFO:root: +[Valid][21][UTT][ MSE = 0.2622 | LCC = 0.8404 | SRCC = 0.8409 ] +INFO:root:Average loss=0.06014617179066707 +INFO:root: +[Valid][22][UTT][ MSE = 0.2728 | LCC = 0.8366 | SRCC = 0.8370 ] +INFO:root:Average loss=0.05750192123918961 +INFO:root: +[Valid][23][UTT][ MSE = 0.2671 | LCC = 0.8405 | SRCC = 0.8411 ] +INFO:root:Average loss=0.05556542999469317 +INFO:root: +[Valid][24][UTT][ MSE = 0.2599 | LCC = 0.8401 | SRCC = 0.8410 ] +INFO:root:Average loss=0.05658941274174513 +INFO:root: +[Valid][25][UTT][ MSE = 0.2666 | LCC = 0.8387 | SRCC = 0.8393 ] +INFO:root:Average loss=0.05353813981398558 +INFO:root: +[Valid][26][UTT][ MSE = 0.2724 | LCC = 0.8370 | SRCC = 0.8381 ] +INFO:root:Average loss=0.04724407955431021 +INFO:root: +[Valid][27][UTT][ MSE = 0.2733 | LCC = 0.8363 | SRCC = 0.8359 ] +INFO:root:Average loss=0.04795557055144738 +INFO:root: +[Valid][28][UTT][ MSE = 0.2588 | LCC = 0.8421 | SRCC = 0.8437 ] +INFO:root:Average loss=0.04644946061456815 +INFO:root: +[Valid][29][UTT][ MSE = 0.2652 | LCC = 0.8371 | SRCC = 0.8385 ] +INFO:root:Average loss=0.04362446996263969 +INFO:root: +[Valid][30][UTT][ MSE = 0.2759 | LCC = 0.8367 | SRCC = 0.8384 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2550 | LCC = 0.8693 | SRCC = 0.8686 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..9aeb1a2b1cca95ce05f7439f448f07b88fb815cd --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:918911d06bcc98c9eaa2ae307c1ca5e0d5bbc92439aa12fab1249cd778292c9a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..2620e1fcb195f3f07f94ff71677dae206873a4f1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dce1147ab8924680ed6cb68b050e6507ba77c304ab1e4ad62dae6b2a14a258aa +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..495520fafb529cdb7da4f1e16c0f5c2d6bc38e64 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.3985204266814085 +INFO:root: +[Valid][0][UTT][ MSE = 0.3723 | LCC = 0.8221 | SRCC = 0.8265 ] +INFO:root:Average loss=0.24721316248178482 +INFO:root: +[Valid][1][UTT][ MSE = 0.3335 | LCC = 0.8462 | SRCC = 0.8468 ] +INFO:root:Average loss=0.21697327628349647 +INFO:root: +[Valid][2][UTT][ MSE = 0.2969 | LCC = 0.8581 | SRCC = 0.8574 ] +INFO:root:Average loss=0.19414820321477377 +INFO:root: +[Valid][3][UTT][ MSE = 0.2474 | LCC = 0.8597 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1832635156237162 +INFO:root: +[Valid][4][UTT][ MSE = 0.2599 | LCC = 0.8627 | SRCC = 0.8617 ] +INFO:root:Average loss=0.16699392386736014 +INFO:root: +[Valid][5][UTT][ MSE = 0.2890 | LCC = 0.8594 | SRCC = 0.8615 ] +INFO:root:Average loss=0.1549624311618316 +INFO:root: +[Valid][6][UTT][ MSE = 0.2620 | LCC = 0.8618 | SRCC = 0.8618 ] +INFO:root:Average loss=0.1448875208122608 +INFO:root: +[Valid][7][UTT][ MSE = 0.2803 | LCC = 0.8575 | SRCC = 0.8588 ] +INFO:root:Average loss=0.1338488265680961 +INFO:root: +[Valid][8][UTT][ MSE = 0.2564 | LCC = 0.8566 | SRCC = 0.8560 ] +INFO:root:Average loss=0.12721320652426818 +INFO:root: +[Valid][9][UTT][ MSE = 0.2636 | LCC = 0.8547 | SRCC = 0.8533 ] +INFO:root:Average loss=0.11536911903665616 +INFO:root: +[Valid][10][UTT][ MSE = 0.2719 | LCC = 0.8547 | SRCC = 0.8543 ] +INFO:root:Average loss=0.10727411360503772 +INFO:root: +[Valid][11][UTT][ MSE = 0.2642 | LCC = 0.8522 | SRCC = 0.8526 ] +INFO:root:Average loss=0.10773733048102795 +INFO:root: +[Valid][12][UTT][ MSE = 0.2621 | LCC = 0.8538 | SRCC = 0.8521 ] +INFO:root:Average loss=0.09891653801195133 +INFO:root: +[Valid][13][UTT][ MSE = 0.2865 | LCC = 0.8534 | SRCC = 0.8540 ] +INFO:root:Average loss=0.09033119774017578 +INFO:root: +[Valid][14][UTT][ MSE = 0.2635 | LCC = 0.8504 | SRCC = 0.8487 ] +INFO:root:Average loss=0.08526179977716544 +INFO:root: +[Valid][15][UTT][ MSE = 0.2562 | LCC = 0.8527 | SRCC = 0.8507 ] +INFO:root:Average loss=0.08129143504760204 +INFO:root: +[Valid][16][UTT][ MSE = 0.2533 | LCC = 0.8507 | SRCC = 0.8511 ] +INFO:root:Average loss=0.0748566631705333 +INFO:root: +[Valid][17][UTT][ MSE = 0.2902 | LCC = 0.8466 | SRCC = 0.8479 ] +INFO:root:Average loss=0.07164051331197605 +INFO:root: +[Valid][18][UTT][ MSE = 0.2698 | LCC = 0.8471 | SRCC = 0.8455 ] +INFO:root:Average loss=0.06858830636319442 +INFO:root: +[Valid][19][UTT][ MSE = 0.2794 | LCC = 0.8455 | SRCC = 0.8444 ] +INFO:root:Average loss=0.06381108640478207 +INFO:root: +[Valid][20][UTT][ MSE = 0.2758 | LCC = 0.8507 | SRCC = 0.8510 ] +INFO:root:Average loss=0.06705740557458156 +INFO:root: +[Valid][21][UTT][ MSE = 0.2632 | LCC = 0.8497 | SRCC = 0.8497 ] +INFO:root:Average loss=0.06321678902858342 +INFO:root: +[Valid][22][UTT][ MSE = 0.2820 | LCC = 0.8480 | SRCC = 0.8482 ] +INFO:root:Average loss=0.0566813223159466 +INFO:root: +[Valid][23][UTT][ MSE = 0.2807 | LCC = 0.8451 | SRCC = 0.8457 ] +INFO:root:Average loss=0.05743397853504389 +INFO:root: +[Valid][24][UTT][ MSE = 0.2707 | LCC = 0.8444 | SRCC = 0.8431 ] +INFO:root:Average loss=0.05141855945858436 +INFO:root: +[Valid][25][UTT][ MSE = 0.2827 | LCC = 0.8455 | SRCC = 0.8446 ] +INFO:root:Average loss=0.05098641940798515 +INFO:root: +[Valid][26][UTT][ MSE = 0.2666 | LCC = 0.8469 | SRCC = 0.8468 ] +INFO:root:Average loss=0.046258388087153435 +INFO:root: +[Valid][27][UTT][ MSE = 0.2562 | LCC = 0.8422 | SRCC = 0.8399 ] +INFO:root:Average loss=0.050062847490876146 +INFO:root: +[Valid][28][UTT][ MSE = 0.2579 | LCC = 0.8425 | SRCC = 0.8403 ] +INFO:root:Average loss=0.04746131560741327 +INFO:root: +[Valid][29][UTT][ MSE = 0.2843 | LCC = 0.8445 | SRCC = 0.8454 ] +INFO:root:Average loss=0.04402158703081883 +INFO:root: +[Valid][30][UTT][ MSE = 0.2638 | LCC = 0.8411 | SRCC = 0.8401 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2522 | LCC = 0.8632 | SRCC = 0.8640 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..4e2be860b7a5ccbef34e2d5d1924eecfd93ee4c4 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b57b81418285a588e873c94590d04c8bf5f240d697c8cae48d98bf477bc428e9 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..11aa581eea98ae9b4d758f0fbc61ec4bbd0ad53e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:818470bf7b150ce74f402130c4fafc7378f792f919239e64940a6fc33e46dbd2 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer6/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..fe394bbb689ae5e23edcbbdd1611721b83328948 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.36821127319947267 +INFO:root: +[Valid][0][UTT][ MSE = 0.3602 | LCC = 0.8310 | SRCC = 0.8343 ] +INFO:root:Average loss=0.24302268754213285 +INFO:root: +[Valid][1][UTT][ MSE = 0.3001 | LCC = 0.8535 | SRCC = 0.8548 ] +INFO:root:Average loss=0.20963786790768305 +INFO:root: +[Valid][2][UTT][ MSE = 0.2732 | LCC = 0.8656 | SRCC = 0.8643 ] +INFO:root:Average loss=0.20486899694571128 +INFO:root: +[Valid][3][UTT][ MSE = 0.3117 | LCC = 0.8667 | SRCC = 0.8658 ] +INFO:root:Average loss=0.18760078725142357 +INFO:root: +[Valid][4][UTT][ MSE = 0.2849 | LCC = 0.8638 | SRCC = 0.8648 ] +INFO:root:Average loss=0.17118957170691246 +INFO:root: +[Valid][5][UTT][ MSE = 0.2655 | LCC = 0.8649 | SRCC = 0.8633 ] +INFO:root:Average loss=0.16321835131981435 +INFO:root: +[Valid][6][UTT][ MSE = 0.2471 | LCC = 0.8687 | SRCC = 0.8672 ] +INFO:root:Average loss=0.14019734345567533 +INFO:root: +[Valid][7][UTT][ MSE = 0.2495 | LCC = 0.8665 | SRCC = 0.8664 ] +INFO:root:Average loss=0.13409451958842766 +INFO:root: +[Valid][8][UTT][ MSE = 0.2383 | LCC = 0.8615 | SRCC = 0.8619 ] +INFO:root:Average loss=0.12116928866658455 +INFO:root: +[Valid][9][UTT][ MSE = 0.2383 | LCC = 0.8656 | SRCC = 0.8655 ] +INFO:root:Average loss=0.11557567816896316 +INFO:root: +[Valid][10][UTT][ MSE = 0.2509 | LCC = 0.8569 | SRCC = 0.8563 ] +INFO:root:Average loss=0.1102052740752697 +INFO:root: +[Valid][11][UTT][ MSE = 0.2524 | LCC = 0.8590 | SRCC = 0.8580 ] +INFO:root:Average loss=0.10095256213576366 +INFO:root: +[Valid][12][UTT][ MSE = 0.2298 | LCC = 0.8619 | SRCC = 0.8622 ] +INFO:root:Average loss=0.09183010478050281 +INFO:root: +[Valid][13][UTT][ MSE = 0.2380 | LCC = 0.8565 | SRCC = 0.8558 ] +INFO:root:Average loss=0.08986802623631099 +INFO:root: +[Valid][14][UTT][ MSE = 0.2943 | LCC = 0.8594 | SRCC = 0.8583 ] +INFO:root:Average loss=0.08365005889955239 +INFO:root: +[Valid][15][UTT][ MSE = 0.2479 | LCC = 0.8569 | SRCC = 0.8554 ] +INFO:root:Average loss=0.07860091003852013 +INFO:root: +[Valid][16][UTT][ MSE = 0.2918 | LCC = 0.8552 | SRCC = 0.8536 ] +INFO:root:Average loss=0.07497164330039269 +INFO:root: +[Valid][17][UTT][ MSE = 0.2472 | LCC = 0.8571 | SRCC = 0.8568 ] +INFO:root:Average loss=0.07313460703843679 +INFO:root: +[Valid][18][UTT][ MSE = 0.2659 | LCC = 0.8534 | SRCC = 0.8534 ] +INFO:root:Average loss=0.06358573471124356 +INFO:root: +[Valid][19][UTT][ MSE = 0.2456 | LCC = 0.8508 | SRCC = 0.8504 ] +INFO:root:Average loss=0.06191529452991791 +INFO:root: +[Valid][20][UTT][ MSE = 0.2694 | LCC = 0.8512 | SRCC = 0.8512 ] +INFO:root:Average loss=0.06594719580159737 +INFO:root: +[Valid][21][UTT][ MSE = 0.2665 | LCC = 0.8485 | SRCC = 0.8478 ] +INFO:root:Average loss=0.05777014070787491 +INFO:root: +[Valid][22][UTT][ MSE = 0.2801 | LCC = 0.8520 | SRCC = 0.8517 ] +INFO:root:Average loss=0.05697945935221819 +INFO:root: +[Valid][23][UTT][ MSE = 0.2793 | LCC = 0.8542 | SRCC = 0.8542 ] +INFO:root:Average loss=0.04914052192217264 +INFO:root: +[Valid][24][UTT][ MSE = 0.2440 | LCC = 0.8544 | SRCC = 0.8535 ] +INFO:root:Average loss=0.04874551043105431 +INFO:root: +[Valid][25][UTT][ MSE = 0.2487 | LCC = 0.8546 | SRCC = 0.8536 ] +INFO:root:Average loss=0.045773990476169646 +INFO:root: +[Valid][26][UTT][ MSE = 0.2629 | LCC = 0.8521 | SRCC = 0.8510 ] +INFO:root:Average loss=0.04854452134802555 +INFO:root: +[Valid][27][UTT][ MSE = 0.2659 | LCC = 0.8538 | SRCC = 0.8549 ] +INFO:root:Average loss=0.04436725709969417 +INFO:root: +[Valid][28][UTT][ MSE = 0.2810 | LCC = 0.8493 | SRCC = 0.8491 ] +INFO:root:Average loss=0.04403507996063966 +INFO:root: +[Valid][29][UTT][ MSE = 0.2439 | LCC = 0.8501 | SRCC = 0.8491 ] +INFO:root:Average loss=0.0424503291455599 +INFO:root: +[Valid][30][UTT][ MSE = 0.2479 | LCC = 0.8527 | SRCC = 0.8520 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2422 | LCC = 0.8628 | SRCC = 0.8637 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..60cecb55851e55f6170c0f2829d7aa543ee2e0f0 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:15cc09d282e0733b7826730f0035b160855c4076f0b08a94c704b7797e2e2b83 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..bb763a6cef7528e982a2a050851989a366f5d829 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aceade7e270f973a8a297dd9d9e619aa8d9db44a1152e9c4fc383033e969b1c9 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..f2c879d772e313db65e3fe27f2522f36e24cf6f2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.38936794167145705 +INFO:root: +[Valid][0][UTT][ MSE = 0.3785 | LCC = 0.8266 | SRCC = 0.8316 ] +INFO:root:Average loss=0.24305556408869913 +INFO:root: +[Valid][1][UTT][ MSE = 0.3095 | LCC = 0.8520 | SRCC = 0.8494 ] +INFO:root:Average loss=0.21885436582259643 +INFO:root: +[Valid][2][UTT][ MSE = 0.3015 | LCC = 0.8610 | SRCC = 0.8590 ] +INFO:root:Average loss=0.19094685254952845 +INFO:root: +[Valid][3][UTT][ MSE = 0.2864 | LCC = 0.8638 | SRCC = 0.8615 ] +INFO:root:Average loss=0.17939216032242164 +INFO:root: +[Valid][4][UTT][ MSE = 0.2467 | LCC = 0.8625 | SRCC = 0.8610 ] +INFO:root:Average loss=0.16792545715967813 +INFO:root: +[Valid][5][UTT][ MSE = 0.2602 | LCC = 0.8626 | SRCC = 0.8598 ] +INFO:root:Average loss=0.15360649885275426 +INFO:root: +[Valid][6][UTT][ MSE = 0.2801 | LCC = 0.8624 | SRCC = 0.8576 ] +INFO:root:Average loss=0.14738893317870605 +INFO:root: +[Valid][7][UTT][ MSE = 0.2472 | LCC = 0.8634 | SRCC = 0.8624 ] +INFO:root:Average loss=0.1408664783797203 +INFO:root: +[Valid][8][UTT][ MSE = 0.2917 | LCC = 0.8596 | SRCC = 0.8580 ] +INFO:root:Average loss=0.12898144956964713 +INFO:root: +[Valid][9][UTT][ MSE = 0.2485 | LCC = 0.8580 | SRCC = 0.8555 ] +INFO:root:Average loss=0.11740485597879459 +INFO:root: +[Valid][10][UTT][ MSE = 0.2553 | LCC = 0.8507 | SRCC = 0.8470 ] +INFO:root:Average loss=0.11333150053635621 +INFO:root: +[Valid][11][UTT][ MSE = 0.2428 | LCC = 0.8542 | SRCC = 0.8506 ] +INFO:root:Average loss=0.1045435161735767 +INFO:root: +[Valid][12][UTT][ MSE = 0.2501 | LCC = 0.8553 | SRCC = 0.8539 ] +INFO:root:Average loss=0.09854458659314193 +INFO:root: +[Valid][13][UTT][ MSE = 0.2534 | LCC = 0.8572 | SRCC = 0.8545 ] +INFO:root:Average loss=0.0874477053204408 +INFO:root: +[Valid][14][UTT][ MSE = 0.2494 | LCC = 0.8537 | SRCC = 0.8514 ] +INFO:root:Average loss=0.084876321351681 +INFO:root: +[Valid][15][UTT][ MSE = 0.2309 | LCC = 0.8531 | SRCC = 0.8510 ] +INFO:root:Average loss=0.08370339631652221 +INFO:root: +[Valid][16][UTT][ MSE = 0.2743 | LCC = 0.8519 | SRCC = 0.8488 ] +INFO:root:Average loss=0.0741450122724741 +INFO:root: +[Valid][17][UTT][ MSE = 0.2614 | LCC = 0.8529 | SRCC = 0.8495 ] +INFO:root:Average loss=0.07035192575019139 +INFO:root: +[Valid][18][UTT][ MSE = 0.2486 | LCC = 0.8485 | SRCC = 0.8459 ] +INFO:root:Average loss=0.06658251979985298 +INFO:root: +[Valid][19][UTT][ MSE = 0.2708 | LCC = 0.8463 | SRCC = 0.8440 ] +INFO:root:Average loss=0.062188576286037765 +INFO:root: +[Valid][20][UTT][ MSE = 0.2617 | LCC = 0.8414 | SRCC = 0.8381 ] +INFO:root:Average loss=0.06166656587559443 +INFO:root: +[Valid][21][UTT][ MSE = 0.2561 | LCC = 0.8467 | SRCC = 0.8442 ] +INFO:root:Average loss=0.06397376539042363 +INFO:root: +[Valid][22][UTT][ MSE = 0.2664 | LCC = 0.8457 | SRCC = 0.8434 ] +INFO:root:Average loss=0.06153976845626648 +INFO:root: +[Valid][23][UTT][ MSE = 0.2637 | LCC = 0.8457 | SRCC = 0.8438 ] +INFO:root:Average loss=0.05711865472869995 +INFO:root: +[Valid][24][UTT][ MSE = 0.2702 | LCC = 0.8385 | SRCC = 0.8359 ] +INFO:root:Average loss=0.05448780982540204 +INFO:root: +[Valid][25][UTT][ MSE = 0.2919 | LCC = 0.8446 | SRCC = 0.8422 ] +INFO:root:Average loss=0.050270472247249044 +INFO:root: +[Valid][26][UTT][ MSE = 0.2519 | LCC = 0.8491 | SRCC = 0.8467 ] +INFO:root:Average loss=0.048808894955959074 +INFO:root: +[Valid][27][UTT][ MSE = 0.2504 | LCC = 0.8458 | SRCC = 0.8432 ] +INFO:root:Average loss=0.049361278517888144 +INFO:root: +[Valid][28][UTT][ MSE = 0.2598 | LCC = 0.8473 | SRCC = 0.8440 ] +INFO:root:Average loss=0.05051465203555731 +INFO:root: +[Valid][29][UTT][ MSE = 0.2741 | LCC = 0.8413 | SRCC = 0.8379 ] +INFO:root:Average loss=0.04536973059368439 +INFO:root: +[Valid][30][UTT][ MSE = 0.2678 | LCC = 0.8453 | SRCC = 0.8421 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2570 | LCC = 0.8660 | SRCC = 0.8647 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..5f1f3a4192cfefb1a6d78f05180c140c109736a5 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94c4c1ce491acede1ce77d2d90a92ed35f0df4044006ca2b5238c75ea8a36823 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..23b0b20b433b21c6cf07393e7e171b4da871bbcf --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea77a34f98fa0ca94dce645034804bf53b55eb0067e08052ba38efa58738b29a +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..2d805fb7f72c01c03aeb811fbc1ac45e2b7bfeee --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.382133974860876 +INFO:root: +[Valid][0][UTT][ MSE = 0.3085 | LCC = 0.8286 | SRCC = 0.8309 ] +INFO:root:Average loss=0.24140969721170571 +INFO:root: +[Valid][1][UTT][ MSE = 0.3029 | LCC = 0.8513 | SRCC = 0.8489 ] +INFO:root:Average loss=0.21385551798037994 +INFO:root: +[Valid][2][UTT][ MSE = 0.2423 | LCC = 0.8575 | SRCC = 0.8586 ] +INFO:root:Average loss=0.19693021590893084 +INFO:root: +[Valid][3][UTT][ MSE = 0.2308 | LCC = 0.8611 | SRCC = 0.8617 ] +INFO:root:Average loss=0.19514649189435518 +INFO:root: +[Valid][4][UTT][ MSE = 0.2290 | LCC = 0.8627 | SRCC = 0.8624 ] +INFO:root:Average loss=0.17847114992447388 +INFO:root: +[Valid][5][UTT][ MSE = 0.2314 | LCC = 0.8639 | SRCC = 0.8626 ] +INFO:root:Average loss=0.16318614417925859 +INFO:root: +[Valid][6][UTT][ MSE = 0.2380 | LCC = 0.8627 | SRCC = 0.8614 ] +INFO:root:Average loss=0.15126552146214706 +INFO:root: +[Valid][7][UTT][ MSE = 0.2384 | LCC = 0.8594 | SRCC = 0.8592 ] +INFO:root:Average loss=0.14128512870042753 +INFO:root: +[Valid][8][UTT][ MSE = 0.2430 | LCC = 0.8550 | SRCC = 0.8552 ] +INFO:root:Average loss=0.13365667504377854 +INFO:root: +[Valid][9][UTT][ MSE = 0.2419 | LCC = 0.8603 | SRCC = 0.8584 ] +INFO:root:Average loss=0.12213804123875423 +INFO:root: +[Valid][10][UTT][ MSE = 0.2696 | LCC = 0.8582 | SRCC = 0.8569 ] +INFO:root:Average loss=0.12221542153602992 +INFO:root: +[Valid][11][UTT][ MSE = 0.2493 | LCC = 0.8582 | SRCC = 0.8566 ] +INFO:root:Average loss=0.10835590557410167 +INFO:root: +[Valid][12][UTT][ MSE = 0.2366 | LCC = 0.8548 | SRCC = 0.8530 ] +INFO:root:Average loss=0.1007872332747166 +INFO:root: +[Valid][13][UTT][ MSE = 0.2333 | LCC = 0.8533 | SRCC = 0.8515 ] +INFO:root:Average loss=0.09416908961840165 +INFO:root: +[Valid][14][UTT][ MSE = 0.2297 | LCC = 0.8540 | SRCC = 0.8531 ] +INFO:root:Average loss=0.08897640766241612 +INFO:root: +[Valid][15][UTT][ MSE = 0.2564 | LCC = 0.8480 | SRCC = 0.8462 ] +INFO:root:Average loss=0.08237178910237092 +INFO:root: +[Valid][16][UTT][ MSE = 0.2509 | LCC = 0.8506 | SRCC = 0.8484 ] +INFO:root:Average loss=0.07729462104348037 +INFO:root: +[Valid][17][UTT][ MSE = 0.2427 | LCC = 0.8503 | SRCC = 0.8487 ] +INFO:root:Average loss=0.07307877749777757 +INFO:root: +[Valid][18][UTT][ MSE = 0.2372 | LCC = 0.8486 | SRCC = 0.8473 ] +INFO:root:Average loss=0.07307783538141312 +INFO:root: +[Valid][19][UTT][ MSE = 0.2549 | LCC = 0.8485 | SRCC = 0.8489 ] +INFO:root:Average loss=0.0750117252747982 +INFO:root: +[Valid][20][UTT][ MSE = 0.2421 | LCC = 0.8511 | SRCC = 0.8517 ] +INFO:root:Average loss=0.0674340163763517 +INFO:root: +[Valid][21][UTT][ MSE = 0.2390 | LCC = 0.8490 | SRCC = 0.8492 ] +INFO:root:Average loss=0.06681850972848061 +INFO:root: +[Valid][22][UTT][ MSE = 0.2360 | LCC = 0.8503 | SRCC = 0.8500 ] +INFO:root:Average loss=0.059631408932499394 +INFO:root: +[Valid][23][UTT][ MSE = 0.2477 | LCC = 0.8485 | SRCC = 0.8472 ] +INFO:root:Average loss=0.05515147946201838 +INFO:root: +[Valid][24][UTT][ MSE = 0.2441 | LCC = 0.8491 | SRCC = 0.8473 ] +INFO:root:Average loss=0.05369805310590145 +INFO:root: +[Valid][25][UTT][ MSE = 0.2496 | LCC = 0.8443 | SRCC = 0.8427 ] +INFO:root:Average loss=0.051728918145482354 +INFO:root: +[Valid][26][UTT][ MSE = 0.2654 | LCC = 0.8437 | SRCC = 0.8429 ] +INFO:root:Average loss=0.052002511703624174 +INFO:root: +[Valid][27][UTT][ MSE = 0.2413 | LCC = 0.8472 | SRCC = 0.8463 ] +INFO:root:Average loss=0.04777215478511957 +INFO:root: +[Valid][28][UTT][ MSE = 0.2485 | LCC = 0.8450 | SRCC = 0.8449 ] +INFO:root:Average loss=0.0486344373665559 +INFO:root: +[Valid][29][UTT][ MSE = 0.2549 | LCC = 0.8437 | SRCC = 0.8413 ] +INFO:root:Average loss=0.04901128997787451 +INFO:root: +[Valid][30][UTT][ MSE = 0.2463 | LCC = 0.8464 | SRCC = 0.8461 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2547 | LCC = 0.8595 | SRCC = 0.8621 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ac97c25bf284fd5316f6427f67cc43bdf7a7606d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7123e6bcae05e41bc18b5eac7d9f37177e8530de59bf4047ad74f5497d4343d +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..bcfeaef0839810eed038d2cf87ad7a147b546988 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77a935b34e2d8fdc72eec0eb14aa9b9113c9581867462bf8d246cd54695c8aa0 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..ec6fedb2cd4a75f314334a691d2d944ef14c91d2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.3602311683770938 +INFO:root: +[Valid][0][UTT][ MSE = 0.3013 | LCC = 0.8311 | SRCC = 0.8377 ] +INFO:root:Average loss=0.2314279081347661 +INFO:root: +[Valid][1][UTT][ MSE = 0.2948 | LCC = 0.8537 | SRCC = 0.8565 ] +INFO:root:Average loss=0.2146023859580358 +INFO:root: +[Valid][2][UTT][ MSE = 0.2697 | LCC = 0.8586 | SRCC = 0.8572 ] +INFO:root:Average loss=0.18873514827245322 +INFO:root: +[Valid][3][UTT][ MSE = 0.2453 | LCC = 0.8627 | SRCC = 0.8587 ] +INFO:root:Average loss=0.1790798407716629 +INFO:root: +[Valid][4][UTT][ MSE = 0.2473 | LCC = 0.8637 | SRCC = 0.8602 ] +INFO:root:Average loss=0.16489050957636955 +INFO:root: +[Valid][5][UTT][ MSE = 0.2408 | LCC = 0.8642 | SRCC = 0.8619 ] +INFO:root:Average loss=0.1626514409406063 +INFO:root: +[Valid][6][UTT][ MSE = 0.2563 | LCC = 0.8607 | SRCC = 0.8615 ] +INFO:root:Average loss=0.14616460486864433 +INFO:root: +[Valid][7][UTT][ MSE = 0.2488 | LCC = 0.8646 | SRCC = 0.8616 ] +INFO:root:Average loss=0.13987669195884314 +INFO:root: +[Valid][8][UTT][ MSE = 0.2387 | LCC = 0.8601 | SRCC = 0.8578 ] +INFO:root:Average loss=0.12660786222953063 +INFO:root: +[Valid][9][UTT][ MSE = 0.2582 | LCC = 0.8550 | SRCC = 0.8524 ] +INFO:root:Average loss=0.12211758576524563 +INFO:root: +[Valid][10][UTT][ MSE = 0.2438 | LCC = 0.8586 | SRCC = 0.8560 ] +INFO:root:Average loss=0.11142660285799931 +INFO:root: +[Valid][11][UTT][ MSE = 0.2330 | LCC = 0.8555 | SRCC = 0.8540 ] +INFO:root:Average loss=0.11101482942318305 +INFO:root: +[Valid][12][UTT][ MSE = 0.2470 | LCC = 0.8569 | SRCC = 0.8537 ] +INFO:root:Average loss=0.10078286408231808 +INFO:root: +[Valid][13][UTT][ MSE = 0.2326 | LCC = 0.8551 | SRCC = 0.8532 ] +INFO:root:Average loss=0.09744389431598859 +INFO:root: +[Valid][14][UTT][ MSE = 0.2429 | LCC = 0.8516 | SRCC = 0.8490 ] +INFO:root:Average loss=0.08517993007523891 +INFO:root: +[Valid][15][UTT][ MSE = 0.2498 | LCC = 0.8470 | SRCC = 0.8450 ] +INFO:root:Average loss=0.08592507333900684 +INFO:root: +[Valid][16][UTT][ MSE = 0.2448 | LCC = 0.8500 | SRCC = 0.8472 ] +INFO:root:Average loss=0.07910298049832001 +INFO:root: +[Valid][17][UTT][ MSE = 0.2404 | LCC = 0.8490 | SRCC = 0.8470 ] +INFO:root:Average loss=0.0754692969031823 +INFO:root: +[Valid][18][UTT][ MSE = 0.2505 | LCC = 0.8487 | SRCC = 0.8459 ] +INFO:root:Average loss=0.0730399110664924 +INFO:root: +[Valid][19][UTT][ MSE = 0.2468 | LCC = 0.8464 | SRCC = 0.8444 ] +INFO:root:Average loss=0.07024951451099835 +INFO:root: +[Valid][20][UTT][ MSE = 0.2546 | LCC = 0.8494 | SRCC = 0.8469 ] +INFO:root:Average loss=0.06742030816773574 +INFO:root: +[Valid][21][UTT][ MSE = 0.2499 | LCC = 0.8450 | SRCC = 0.8422 ] +INFO:root:Average loss=0.06090548820793629 +INFO:root: +[Valid][22][UTT][ MSE = 0.2402 | LCC = 0.8478 | SRCC = 0.8459 ] +INFO:root:Average loss=0.05983414309911239 +INFO:root: +[Valid][23][UTT][ MSE = 0.2777 | LCC = 0.8453 | SRCC = 0.8440 ] +INFO:root:Average loss=0.05721822935037124 +INFO:root: +[Valid][24][UTT][ MSE = 0.2774 | LCC = 0.8415 | SRCC = 0.8405 ] +INFO:root:Average loss=0.055587698538333945 +INFO:root: +[Valid][25][UTT][ MSE = 0.2510 | LCC = 0.8507 | SRCC = 0.8475 ] +INFO:root:Average loss=0.05475469609388174 +INFO:root: +[Valid][26][UTT][ MSE = 0.2497 | LCC = 0.8474 | SRCC = 0.8458 ] +INFO:root:Average loss=0.05216233709301704 +INFO:root: +[Valid][27][UTT][ MSE = 0.2588 | LCC = 0.8461 | SRCC = 0.8438 ] +INFO:root:Average loss=0.05129785520526079 +INFO:root: +[Valid][28][UTT][ MSE = 0.2428 | LCC = 0.8496 | SRCC = 0.8467 ] +INFO:root:Average loss=0.04872809844807936 +INFO:root: +[Valid][29][UTT][ MSE = 0.2593 | LCC = 0.8428 | SRCC = 0.8422 ] +INFO:root:Average loss=0.04772496127929443 +INFO:root: +[Valid][30][UTT][ MSE = 0.2447 | LCC = 0.8486 | SRCC = 0.8460 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2652 | LCC = 0.8530 | SRCC = 0.8550 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..9f0b512fc2284b885e3210bdf1611e8ab240e2ac --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6ef6edde42761c5ab5a6f83ed7bbd2db0fcaeb664004a36ce643db84566e5ec +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..aed8147b8f11303d7b8ed7802f3df779821dee48 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8b6629544120ec5fba0fbfe87b16b17a7b21acfef2a821e5c110add4ab8086de +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..666d4cb5ff7443798432a570fc5fcb4d7606c8c2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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=0.36397342804150706 +INFO:root: +[Valid][0][UTT][ MSE = 0.3046 | LCC = 0.8251 | SRCC = 0.8322 ] +INFO:root:Average loss=0.245444923830338 +INFO:root: +[Valid][1][UTT][ MSE = 0.2605 | LCC = 0.8443 | SRCC = 0.8500 ] +INFO:root:Average loss=0.21586048297393015 +INFO:root: +[Valid][2][UTT][ MSE = 0.2450 | LCC = 0.8562 | SRCC = 0.8590 ] +INFO:root:Average loss=0.19770498077074686 +INFO:root: +[Valid][3][UTT][ MSE = 0.2452 | LCC = 0.8596 | SRCC = 0.8598 ] +INFO:root:Average loss=0.17895091936374322 +INFO:root: +[Valid][4][UTT][ MSE = 0.2505 | LCC = 0.8623 | SRCC = 0.8620 ] +INFO:root:Average loss=0.1671215909986924 +INFO:root: +[Valid][5][UTT][ MSE = 0.2399 | LCC = 0.8608 | SRCC = 0.8596 ] +INFO:root:Average loss=0.15898744666423553 +INFO:root: +[Valid][6][UTT][ MSE = 0.2825 | LCC = 0.8615 | SRCC = 0.8623 ] +INFO:root:Average loss=0.15407248309407479 +INFO:root: +[Valid][7][UTT][ MSE = 0.2384 | LCC = 0.8615 | SRCC = 0.8609 ] +INFO:root:Average loss=0.13730604860645074 +INFO:root: +[Valid][8][UTT][ MSE = 0.2331 | LCC = 0.8609 | SRCC = 0.8605 ] +INFO:root:Average loss=0.12922947815595529 +INFO:root: +[Valid][9][UTT][ MSE = 0.2345 | LCC = 0.8607 | SRCC = 0.8595 ] +INFO:root:Average loss=0.11929303579605542 +INFO:root: +[Valid][10][UTT][ MSE = 0.2268 | LCC = 0.8589 | SRCC = 0.8590 ] +INFO:root:Average loss=0.1182055145693131 +INFO:root: +[Valid][11][UTT][ MSE = 0.2319 | LCC = 0.8622 | SRCC = 0.8608 ] +INFO:root:Average loss=0.10349101773821391 +INFO:root: +[Valid][12][UTT][ MSE = 0.2437 | LCC = 0.8586 | SRCC = 0.8561 ] +INFO:root:Average loss=0.09732067967072511 +INFO:root: +[Valid][13][UTT][ MSE = 0.2348 | LCC = 0.8580 | SRCC = 0.8560 ] +INFO:root:Average loss=0.0948904394530333 +INFO:root: +[Valid][14][UTT][ MSE = 0.2472 | LCC = 0.8559 | SRCC = 0.8533 ] +INFO:root:Average loss=0.08575341659478652 +INFO:root: +[Valid][15][UTT][ MSE = 0.2566 | LCC = 0.8558 | SRCC = 0.8556 ] +INFO:root:Average loss=0.08504044985732971 +INFO:root: +[Valid][16][UTT][ MSE = 0.2453 | LCC = 0.8572 | SRCC = 0.8550 ] +INFO:root:Average loss=0.07672872661780088 +INFO:root: +[Valid][17][UTT][ MSE = 0.2420 | LCC = 0.8538 | SRCC = 0.8536 ] +INFO:root:Average loss=0.0779275038303473 +INFO:root: +[Valid][18][UTT][ MSE = 0.2523 | LCC = 0.8528 | SRCC = 0.8542 ] +INFO:root:Average loss=0.06673648877021594 +INFO:root: +[Valid][19][UTT][ MSE = 0.2386 | LCC = 0.8509 | SRCC = 0.8511 ] +INFO:root:Average loss=0.06712445110464707 +INFO:root: +[Valid][20][UTT][ MSE = 0.2366 | LCC = 0.8492 | SRCC = 0.8495 ] +INFO:root:Average loss=0.06233649624463839 +INFO:root: +[Valid][21][UTT][ MSE = 0.2372 | LCC = 0.8500 | SRCC = 0.8504 ] +INFO:root:Average loss=0.06075268434599424 +INFO:root: +[Valid][22][UTT][ MSE = 0.2413 | LCC = 0.8474 | SRCC = 0.8470 ] +INFO:root:Average loss=0.06079782927647615 +INFO:root: +[Valid][23][UTT][ MSE = 0.2393 | LCC = 0.8506 | SRCC = 0.8492 ] +INFO:root:Average loss=0.05743453283913624 +INFO:root: +[Valid][24][UTT][ MSE = 0.2459 | LCC = 0.8470 | SRCC = 0.8459 ] +INFO:root:Average loss=0.05254620913034066 +INFO:root: +[Valid][25][UTT][ MSE = 0.2495 | LCC = 0.8464 | SRCC = 0.8468 ] +INFO:root:Average loss=0.052903206063768804 +INFO:root: +[Valid][26][UTT][ MSE = 0.2467 | LCC = 0.8506 | SRCC = 0.8501 ] +INFO:root:Average loss=0.050381544046103954 +INFO:root: +[Valid][27][UTT][ MSE = 0.2414 | LCC = 0.8496 | SRCC = 0.8488 ] +INFO:root:Average loss=0.048478107660626754 +INFO:root: +[Valid][28][UTT][ MSE = 0.2453 | LCC = 0.8501 | SRCC = 0.8489 ] +INFO:root:Average loss=0.04521266791300896 +INFO:root: +[Valid][29][UTT][ MSE = 0.2698 | LCC = 0.8472 | SRCC = 0.8471 ] +INFO:root:Average loss=0.04534978928187719 +INFO:root: +[Valid][30][UTT][ MSE = 0.2578 | LCC = 0.8485 | SRCC = 0.8479 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2740 | LCC = 0.8580 | SRCC = 0.8623 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d26c599c1eec7d09bb9f6b7563483164b2fb5e8b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be219f13e81164ebc153e0e08315c396bc41e64a4c8959e9768bc580aa4bd960 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..0f850d3e62455c7809120ad4d48c48703d712fab --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1bc6cec96155d59d6ddf111bf638e4fd50a97d80c313a5b1ef0b7754a2fc8a89 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..d89d86474576e0621aa3f13e20759f397de7fe8a --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer7/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=0.38335184084299284 +INFO:root: +[Valid][0][UTT][ MSE = 0.3168 | LCC = 0.8372 | SRCC = 0.8387 ] +INFO:root:Average loss=0.23597935396127212 +INFO:root: +[Valid][1][UTT][ MSE = 0.2796 | LCC = 0.8555 | SRCC = 0.8565 ] +INFO:root:Average loss=0.20632159136808836 +INFO:root: +[Valid][2][UTT][ MSE = 0.2955 | LCC = 0.8632 | SRCC = 0.8613 ] +INFO:root:Average loss=0.19342476922350052 +INFO:root: +[Valid][3][UTT][ MSE = 0.2712 | LCC = 0.8681 | SRCC = 0.8651 ] +INFO:root:Average loss=0.18052131023544532 +INFO:root: +[Valid][4][UTT][ MSE = 0.2671 | LCC = 0.8661 | SRCC = 0.8632 ] +INFO:root:Average loss=0.1681719010648055 +INFO:root: +[Valid][5][UTT][ MSE = 0.2500 | LCC = 0.8656 | SRCC = 0.8628 ] +INFO:root:Average loss=0.16178731171366495 +INFO:root: +[Valid][6][UTT][ MSE = 0.2392 | LCC = 0.8684 | SRCC = 0.8663 ] +INFO:root:Average loss=0.14304526627827913 +INFO:root: +[Valid][7][UTT][ MSE = 0.2393 | LCC = 0.8658 | SRCC = 0.8638 ] +INFO:root:Average loss=0.13700010914068955 +INFO:root: +[Valid][8][UTT][ MSE = 0.2513 | LCC = 0.8623 | SRCC = 0.8597 ] +INFO:root:Average loss=0.12617877574685293 +INFO:root: +[Valid][9][UTT][ MSE = 0.2428 | LCC = 0.8607 | SRCC = 0.8597 ] +INFO:root:Average loss=0.11835498190843143 +INFO:root: +[Valid][10][UTT][ MSE = 0.2382 | LCC = 0.8633 | SRCC = 0.8611 ] +INFO:root:Average loss=0.11623024453337376 +INFO:root: +[Valid][11][UTT][ MSE = 0.2586 | LCC = 0.8587 | SRCC = 0.8563 ] +INFO:root:Average loss=0.10344805444280307 +INFO:root: +[Valid][12][UTT][ MSE = 0.2733 | LCC = 0.8560 | SRCC = 0.8544 ] +INFO:root:Average loss=0.09296522805323967 +INFO:root: +[Valid][13][UTT][ MSE = 0.2577 | LCC = 0.8543 | SRCC = 0.8535 ] +INFO:root:Average loss=0.09691680394686185 +INFO:root: +[Valid][14][UTT][ MSE = 0.2663 | LCC = 0.8536 | SRCC = 0.8499 ] +INFO:root:Average loss=0.08777036259953792 +INFO:root: +[Valid][15][UTT][ MSE = 0.2873 | LCC = 0.8523 | SRCC = 0.8493 ] +INFO:root:Average loss=0.08127651831660515 +INFO:root: +[Valid][16][UTT][ MSE = 0.2755 | LCC = 0.8482 | SRCC = 0.8453 ] +INFO:root:Average loss=0.08566597123176624 +INFO:root: +[Valid][17][UTT][ MSE = 0.2486 | LCC = 0.8511 | SRCC = 0.8499 ] +INFO:root:Average loss=0.07266660373753463 +INFO:root: +[Valid][18][UTT][ MSE = 0.2895 | LCC = 0.8512 | SRCC = 0.8493 ] +INFO:root:Average loss=0.06931361670677479 +INFO:root: +[Valid][19][UTT][ MSE = 0.2471 | LCC = 0.8534 | SRCC = 0.8509 ] +INFO:root:Average loss=0.06731925771022454 +INFO:root: +[Valid][20][UTT][ MSE = 0.2771 | LCC = 0.8454 | SRCC = 0.8430 ] +INFO:root:Average loss=0.0686812514009384 +INFO:root: +[Valid][21][UTT][ MSE = 0.2696 | LCC = 0.8446 | SRCC = 0.8409 ] +INFO:root:Average loss=0.06004914273627293 +INFO:root: +[Valid][22][UTT][ MSE = 0.2741 | LCC = 0.8501 | SRCC = 0.8483 ] +INFO:root:Average loss=0.05724196532406868 +INFO:root: +[Valid][23][UTT][ MSE = 0.3476 | LCC = 0.8452 | SRCC = 0.8442 ] +INFO:root:Average loss=0.06277956632085335 +INFO:root: +[Valid][24][UTT][ MSE = 0.2449 | LCC = 0.8477 | SRCC = 0.8442 ] +INFO:root:Average loss=0.0521102187056572 +INFO:root: +[Valid][25][UTT][ MSE = 0.2496 | LCC = 0.8472 | SRCC = 0.8452 ] +INFO:root:Average loss=0.051256921428900495 +INFO:root: +[Valid][26][UTT][ MSE = 0.2628 | LCC = 0.8404 | SRCC = 0.8383 ] +INFO:root:Average loss=0.04894598922095238 +INFO:root: +[Valid][27][UTT][ MSE = 0.2803 | LCC = 0.8455 | SRCC = 0.8432 ] +INFO:root:Average loss=0.048654796866079174 +INFO:root: +[Valid][28][UTT][ MSE = 0.2659 | LCC = 0.8487 | SRCC = 0.8461 ] +INFO:root:Average loss=0.048401727747076594 +INFO:root: +[Valid][29][UTT][ MSE = 0.2509 | LCC = 0.8481 | SRCC = 0.8454 ] +INFO:root:Average loss=0.045584204439551405 +INFO:root: +[Valid][30][UTT][ MSE = 0.2641 | LCC = 0.8493 | SRCC = 0.8464 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2431 | LCC = 0.8589 | SRCC = 0.8602 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..3c132af0cd8b433bc85250e4450e7328756f770e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fd3a9ac8cba002fd643b27767017dc03ca69e42a6b3cf4d72f343a25e3e63832 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..577834d7114adaebab5bec63d624f59647d2577d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2a2fef5c282eb620d614008d63792ed060a75305bcfaf3052307be910a1b2cc5 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..aac10edc2c34fcc7e1c1e894e780294c1fdd8cb1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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=0.3402170141537984 +INFO:root: +[Valid][0][UTT][ MSE = 0.3559 | LCC = 0.8319 | SRCC = 0.8368 ] +INFO:root:Average loss=0.2441949643767797 +INFO:root: +[Valid][1][UTT][ MSE = 0.3137 | LCC = 0.8527 | SRCC = 0.8492 ] +INFO:root:Average loss=0.2187048899821746 +INFO:root: +[Valid][2][UTT][ MSE = 0.2588 | LCC = 0.8596 | SRCC = 0.8577 ] +INFO:root:Average loss=0.20523394300387457 +INFO:root: +[Valid][3][UTT][ MSE = 0.2502 | LCC = 0.8623 | SRCC = 0.8599 ] +INFO:root:Average loss=0.1897807207245093 +INFO:root: +[Valid][4][UTT][ MSE = 0.2421 | LCC = 0.8643 | SRCC = 0.8636 ] +INFO:root:Average loss=0.17880811943457678 +INFO:root: +[Valid][5][UTT][ MSE = 0.2608 | LCC = 0.8652 | SRCC = 0.8634 ] +INFO:root:Average loss=0.1703132263934001 +INFO:root: +[Valid][6][UTT][ MSE = 0.2758 | LCC = 0.8608 | SRCC = 0.8615 ] +INFO:root:Average loss=0.15961933365234962 +INFO:root: +[Valid][7][UTT][ MSE = 0.2521 | LCC = 0.8639 | SRCC = 0.8617 ] +INFO:root:Average loss=0.1645943918862404 +INFO:root: +[Valid][8][UTT][ MSE = 0.3135 | LCC = 0.8607 | SRCC = 0.8609 ] +INFO:root:Average loss=0.14518306127343422 +INFO:root: +[Valid][9][UTT][ MSE = 0.2863 | LCC = 0.8640 | SRCC = 0.8622 ] +INFO:root:Average loss=0.13368538910379776 +INFO:root: +[Valid][10][UTT][ MSE = 0.2335 | LCC = 0.8594 | SRCC = 0.8602 ] +INFO:root:Average loss=0.12934706455621964 +INFO:root: +[Valid][11][UTT][ MSE = 0.2665 | LCC = 0.8578 | SRCC = 0.8562 ] +INFO:root:Average loss=0.12665524505651915 +INFO:root: +[Valid][12][UTT][ MSE = 0.2430 | LCC = 0.8609 | SRCC = 0.8580 ] +INFO:root:Average loss=0.11458899539250594 +INFO:root: +[Valid][13][UTT][ MSE = 0.2824 | LCC = 0.8516 | SRCC = 0.8510 ] +INFO:root:Average loss=0.10889020781868543 +INFO:root: +[Valid][14][UTT][ MSE = 0.2587 | LCC = 0.8535 | SRCC = 0.8522 ] +INFO:root:Average loss=0.10381360170550835 +INFO:root: +[Valid][15][UTT][ MSE = 0.2625 | LCC = 0.8535 | SRCC = 0.8525 ] +INFO:root:Average loss=0.09656045838999443 +INFO:root: +[Valid][16][UTT][ MSE = 0.2769 | LCC = 0.8492 | SRCC = 0.8497 ] +INFO:root:Average loss=0.09219557051666272 +INFO:root: +[Valid][17][UTT][ MSE = 0.2724 | LCC = 0.8528 | SRCC = 0.8529 ] +INFO:root:Average loss=0.08918967900367883 +INFO:root: +[Valid][18][UTT][ MSE = 0.2803 | LCC = 0.8511 | SRCC = 0.8491 ] +INFO:root:Average loss=0.08707221941306041 +INFO:root: +[Valid][19][UTT][ MSE = 0.2497 | LCC = 0.8556 | SRCC = 0.8552 ] +INFO:root:Average loss=0.08079892416030933 +INFO:root: +[Valid][20][UTT][ MSE = 0.2648 | LCC = 0.8504 | SRCC = 0.8502 ] +INFO:root:Average loss=0.07577006366008367 +INFO:root: +[Valid][21][UTT][ MSE = 0.2539 | LCC = 0.8535 | SRCC = 0.8535 ] +INFO:root:Average loss=0.07495671964417665 +INFO:root: +[Valid][22][UTT][ MSE = 0.2477 | LCC = 0.8521 | SRCC = 0.8503 ] +INFO:root:Average loss=0.07002059341623233 +INFO:root: +[Valid][23][UTT][ MSE = 0.2462 | LCC = 0.8482 | SRCC = 0.8484 ] +INFO:root:Average loss=0.06676036563630287 +INFO:root: +[Valid][24][UTT][ MSE = 0.2989 | LCC = 0.8435 | SRCC = 0.8424 ] +INFO:root:Average loss=0.06812790723947379 +INFO:root: +[Valid][25][UTT][ MSE = 0.2472 | LCC = 0.8477 | SRCC = 0.8463 ] +INFO:root:Average loss=0.061988590810543455 +INFO:root: +[Valid][26][UTT][ MSE = 0.2600 | LCC = 0.8506 | SRCC = 0.8474 ] +INFO:root:Average loss=0.05794303343655207 +INFO:root: +[Valid][27][UTT][ MSE = 0.2461 | LCC = 0.8527 | SRCC = 0.8505 ] +INFO:root:Average loss=0.05631502082523627 +INFO:root: +[Valid][28][UTT][ MSE = 0.2637 | LCC = 0.8463 | SRCC = 0.8468 ] +INFO:root:Average loss=0.060104798477811694 +INFO:root: +[Valid][29][UTT][ MSE = 0.2571 | LCC = 0.8484 | SRCC = 0.8469 ] +INFO:root:Average loss=0.054433862320505656 +INFO:root: +[Valid][30][UTT][ MSE = 0.2608 | LCC = 0.8462 | SRCC = 0.8452 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2462 | LCC = 0.8669 | SRCC = 0.8665 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..673519e9f029d5302b8a30cb4eebd5ea1efa6932 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f36b52fe3428938dcf70491b47b7c36f80e21ec7d0f171868a7288b5215394c6 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..f89ac09974f4c34f76f7528eb0767041dbeee006 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a4a9c3143850081d0eb7700876aa2d06a9b7abacbf4121df86197361ad0bada +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..7ea7b9108e03ae82ab6040a507fc0885037604b8 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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=0.32820918200871885 +INFO:root: +[Valid][0][UTT][ MSE = 0.2957 | LCC = 0.8287 | SRCC = 0.8332 ] +INFO:root:Average loss=0.23871814268521774 +INFO:root: +[Valid][1][UTT][ MSE = 0.2856 | LCC = 0.8430 | SRCC = 0.8463 ] +INFO:root:Average loss=0.2165304571390152 +INFO:root: +[Valid][2][UTT][ MSE = 0.2508 | LCC = 0.8557 | SRCC = 0.8547 ] +INFO:root:Average loss=0.2010145808259646 +INFO:root: +[Valid][3][UTT][ MSE = 0.2509 | LCC = 0.8595 | SRCC = 0.8583 ] +INFO:root:Average loss=0.19355477641026178 +INFO:root: +[Valid][4][UTT][ MSE = 0.2756 | LCC = 0.8620 | SRCC = 0.8596 ] +INFO:root:Average loss=0.18434447069198656 +INFO:root: +[Valid][5][UTT][ MSE = 0.2403 | LCC = 0.8579 | SRCC = 0.8645 ] +INFO:root:Average loss=0.17679273270261595 +INFO:root: +[Valid][6][UTT][ MSE = 0.2495 | LCC = 0.8616 | SRCC = 0.8606 ] +INFO:root:Average loss=0.1632066828509172 +INFO:root: +[Valid][7][UTT][ MSE = 0.2328 | LCC = 0.8629 | SRCC = 0.8624 ] +INFO:root:Average loss=0.15165668085981637 +INFO:root: +[Valid][8][UTT][ MSE = 0.2540 | LCC = 0.8605 | SRCC = 0.8605 ] +INFO:root:Average loss=0.1479603885075985 +INFO:root: +[Valid][9][UTT][ MSE = 0.2329 | LCC = 0.8615 | SRCC = 0.8610 ] +INFO:root:Average loss=0.13531134793391594 +INFO:root: +[Valid][10][UTT][ MSE = 0.2380 | LCC = 0.8598 | SRCC = 0.8583 ] +INFO:root:Average loss=0.13345116338668725 +INFO:root: +[Valid][11][UTT][ MSE = 0.2404 | LCC = 0.8570 | SRCC = 0.8563 ] +INFO:root:Average loss=0.12506322486278337 +INFO:root: +[Valid][12][UTT][ MSE = 0.2543 | LCC = 0.8554 | SRCC = 0.8547 ] +INFO:root:Average loss=0.12052316677111846 +INFO:root: +[Valid][13][UTT][ MSE = 0.2597 | LCC = 0.8536 | SRCC = 0.8512 ] +INFO:root:Average loss=0.11571305350233348 +INFO:root: +[Valid][14][UTT][ MSE = 0.2346 | LCC = 0.8630 | SRCC = 0.8616 ] +INFO:root:Average loss=0.10625300097924012 +INFO:root: +[Valid][15][UTT][ MSE = 0.2309 | LCC = 0.8573 | SRCC = 0.8566 ] +INFO:root:Average loss=0.10383581312803122 +INFO:root: +[Valid][16][UTT][ MSE = 0.2447 | LCC = 0.8540 | SRCC = 0.8544 ] +INFO:root:Average loss=0.0937476628101789 +INFO:root: +[Valid][17][UTT][ MSE = 0.2495 | LCC = 0.8507 | SRCC = 0.8495 ] +INFO:root:Average loss=0.09246135025452344 +INFO:root: +[Valid][18][UTT][ MSE = 0.2435 | LCC = 0.8512 | SRCC = 0.8499 ] +INFO:root:Average loss=0.08594100970106247 +INFO:root: +[Valid][19][UTT][ MSE = 0.2435 | LCC = 0.8505 | SRCC = 0.8490 ] +INFO:root:Average loss=0.08692024896542232 +INFO:root: +[Valid][20][UTT][ MSE = 0.2528 | LCC = 0.8462 | SRCC = 0.8448 ] +INFO:root:Average loss=0.07682577940898064 +INFO:root: +[Valid][21][UTT][ MSE = 0.2327 | LCC = 0.8505 | SRCC = 0.8497 ] +INFO:root:Average loss=0.07345408678818972 +INFO:root: +[Valid][22][UTT][ MSE = 0.2559 | LCC = 0.8467 | SRCC = 0.8451 ] +INFO:root:Average loss=0.07295294368687348 +INFO:root: +[Valid][23][UTT][ MSE = 0.2642 | LCC = 0.8440 | SRCC = 0.8433 ] +INFO:root:Average loss=0.07139696839910287 +INFO:root: +[Valid][24][UTT][ MSE = 0.2481 | LCC = 0.8471 | SRCC = 0.8471 ] +INFO:root:Average loss=0.06768394887256317 +INFO:root: +[Valid][25][UTT][ MSE = 0.2585 | LCC = 0.8434 | SRCC = 0.8429 ] +INFO:root:Average loss=0.0641708282323984 +INFO:root: +[Valid][26][UTT][ MSE = 0.2439 | LCC = 0.8472 | SRCC = 0.8464 ] +INFO:root:Average loss=0.06205814432066221 +INFO:root: +[Valid][27][UTT][ MSE = 0.2493 | LCC = 0.8446 | SRCC = 0.8437 ] +INFO:root:Average loss=0.060754102583114915 +INFO:root: +[Valid][28][UTT][ MSE = 0.2512 | LCC = 0.8470 | SRCC = 0.8459 ] +INFO:root:Average loss=0.05440507891277472 +INFO:root: +[Valid][29][UTT][ MSE = 0.2476 | LCC = 0.8452 | SRCC = 0.8442 ] +INFO:root:Average loss=0.05309845936986116 +INFO:root: +[Valid][30][UTT][ MSE = 0.2461 | LCC = 0.8464 | SRCC = 0.8454 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2395 | LCC = 0.8588 | SRCC = 0.8564 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ecd9f7c294b19a979ce464cf9c3e11b9b3de8ad1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:587bf3ce032fb1778853dcf33d6ed181a2687495eab4f20befec97ab448266c7 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..e768ffe77e369255642b206784e19282e1ae0e25 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68edb032b1cd02c84160054cfe2910f9bbbe38a01b212b07f4a5a52a3d31c174 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..98b9e3d476eadf4420cc3b9f7af03c8741dab237 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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=0.3641928759140846 +INFO:root: +[Valid][0][UTT][ MSE = 0.3306 | LCC = 0.8360 | SRCC = 0.8384 ] +INFO:root:Average loss=0.2331586921444306 +INFO:root: +[Valid][1][UTT][ MSE = 0.2768 | LCC = 0.8569 | SRCC = 0.8576 ] +INFO:root:Average loss=0.21501163144906363 +INFO:root: +[Valid][2][UTT][ MSE = 0.2693 | LCC = 0.8607 | SRCC = 0.8611 ] +INFO:root:Average loss=0.20393017717661002 +INFO:root: +[Valid][3][UTT][ MSE = 0.2714 | LCC = 0.8651 | SRCC = 0.8667 ] +INFO:root:Average loss=0.1931002248938267 +INFO:root: +[Valid][4][UTT][ MSE = 0.2970 | LCC = 0.8624 | SRCC = 0.8646 ] +INFO:root:Average loss=0.17608074308970037 +INFO:root: +[Valid][5][UTT][ MSE = 0.2578 | LCC = 0.8638 | SRCC = 0.8670 ] +INFO:root:Average loss=0.16923117981507227 +INFO:root: +[Valid][6][UTT][ MSE = 0.3133 | LCC = 0.8568 | SRCC = 0.8645 ] +INFO:root:Average loss=0.1721971003481975 +INFO:root: +[Valid][7][UTT][ MSE = 0.2862 | LCC = 0.8630 | SRCC = 0.8653 ] +INFO:root:Average loss=0.15376667458659563 +INFO:root: +[Valid][8][UTT][ MSE = 0.2670 | LCC = 0.8594 | SRCC = 0.8598 ] +INFO:root:Average loss=0.14586045411534798 +INFO:root: +[Valid][9][UTT][ MSE = 0.2698 | LCC = 0.8597 | SRCC = 0.8604 ] +INFO:root:Average loss=0.13061381064546415 +INFO:root: +[Valid][10][UTT][ MSE = 0.2811 | LCC = 0.8596 | SRCC = 0.8592 ] +INFO:root:Average loss=0.13266851055698517 +INFO:root: +[Valid][11][UTT][ MSE = 0.2667 | LCC = 0.8597 | SRCC = 0.8592 ] +INFO:root:Average loss=0.12182425115353022 +INFO:root: +[Valid][12][UTT][ MSE = 0.2390 | LCC = 0.8582 | SRCC = 0.8568 ] +INFO:root:Average loss=0.1193983979905263 +INFO:root: +[Valid][13][UTT][ MSE = 0.2788 | LCC = 0.8518 | SRCC = 0.8519 ] +INFO:root:Average loss=0.10864970680230703 +INFO:root: +[Valid][14][UTT][ MSE = 0.2584 | LCC = 0.8524 | SRCC = 0.8523 ] +INFO:root:Average loss=0.10901646129786968 +INFO:root: +[Valid][15][UTT][ MSE = 0.2562 | LCC = 0.8586 | SRCC = 0.8564 ] +INFO:root:Average loss=0.09716410103898782 +INFO:root: +[Valid][16][UTT][ MSE = 0.2701 | LCC = 0.8490 | SRCC = 0.8478 ] +INFO:root:Average loss=0.10246286369287051 +INFO:root: +[Valid][17][UTT][ MSE = 0.2726 | LCC = 0.8532 | SRCC = 0.8535 ] +INFO:root:Average loss=0.09045895924552894 +INFO:root: +[Valid][18][UTT][ MSE = 0.2919 | LCC = 0.8458 | SRCC = 0.8453 ] +INFO:root:Average loss=0.08376105650304219 +INFO:root: +[Valid][19][UTT][ MSE = 0.2542 | LCC = 0.8515 | SRCC = 0.8506 ] +INFO:root:Average loss=0.0823517799950563 +INFO:root: +[Valid][20][UTT][ MSE = 0.2468 | LCC = 0.8538 | SRCC = 0.8537 ] +INFO:root:Average loss=0.07916218954592179 +INFO:root: +[Valid][21][UTT][ MSE = 0.2722 | LCC = 0.8495 | SRCC = 0.8497 ] +INFO:root:Average loss=0.07589805976320536 +INFO:root: +[Valid][22][UTT][ MSE = 0.2448 | LCC = 0.8523 | SRCC = 0.8497 ] +INFO:root:Average loss=0.07477361818727775 +INFO:root: +[Valid][23][UTT][ MSE = 0.2578 | LCC = 0.8523 | SRCC = 0.8532 ] +INFO:root:Average loss=0.07197199886043866 +INFO:root: +[Valid][24][UTT][ MSE = 0.2550 | LCC = 0.8448 | SRCC = 0.8433 ] +INFO:root:Average loss=0.0676754981948015 +INFO:root: +[Valid][25][UTT][ MSE = 0.2533 | LCC = 0.8464 | SRCC = 0.8479 ] +INFO:root:Average loss=0.0615634930630525 +INFO:root: +[Valid][26][UTT][ MSE = 0.2605 | LCC = 0.8442 | SRCC = 0.8440 ] +INFO:root:Average loss=0.061427938345915235 +INFO:root: +[Valid][27][UTT][ MSE = 0.2717 | LCC = 0.8482 | SRCC = 0.8479 ] +INFO:root:Average loss=0.05824610299597948 +INFO:root: +[Valid][28][UTT][ MSE = 0.2782 | LCC = 0.8456 | SRCC = 0.8455 ] +INFO:root:Average loss=0.057789162708780706 +INFO:root: +[Valid][29][UTT][ MSE = 0.2500 | LCC = 0.8495 | SRCC = 0.8496 ] +INFO:root:Average loss=0.05593977118722904 +INFO:root: +[Valid][30][UTT][ MSE = 0.2678 | LCC = 0.8440 | SRCC = 0.8425 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2532 | LCC = 0.8675 | SRCC = 0.8678 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..e7c406fa003445992b56e8dd05d590e93f3b5205 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7e18882b0dbe90b40c32c43d804431e58f9f80bdaca7ce4b40eb31f6f3f927e +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..d6478d1b3152e610b62acdc2c6eeca5a757929ff --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47f0020fc39d7a8ab28986fdd664cdae76eb4935376b0a2cce87b5b26d189942 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..429d9ab9089a4c983a194b18571b6c0f2479c7df --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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=0.3410821631550789 +INFO:root: +[Valid][0][UTT][ MSE = 0.2989 | LCC = 0.8345 | SRCC = 0.8335 ] +INFO:root:Average loss=0.2439834089615406 +INFO:root: +[Valid][1][UTT][ MSE = 0.2530 | LCC = 0.8485 | SRCC = 0.8514 ] +INFO:root:Average loss=0.21876975607413512 +INFO:root: +[Valid][2][UTT][ MSE = 0.2499 | LCC = 0.8620 | SRCC = 0.8604 ] +INFO:root:Average loss=0.20447931792109442 +INFO:root: +[Valid][3][UTT][ MSE = 0.2521 | LCC = 0.8661 | SRCC = 0.8649 ] +INFO:root:Average loss=0.19490099144287598 +INFO:root: +[Valid][4][UTT][ MSE = 0.2326 | LCC = 0.8669 | SRCC = 0.8664 ] +INFO:root:Average loss=0.17930671429404846 +INFO:root: +[Valid][5][UTT][ MSE = 0.2311 | LCC = 0.8659 | SRCC = 0.8662 ] +INFO:root:Average loss=0.17265549731942323 +INFO:root: +[Valid][6][UTT][ MSE = 0.2540 | LCC = 0.8662 | SRCC = 0.8640 ] +INFO:root:Average loss=0.16309759116325623 +INFO:root: +[Valid][7][UTT][ MSE = 0.2629 | LCC = 0.8614 | SRCC = 0.8601 ] +INFO:root:Average loss=0.16369123336596367 +INFO:root: +[Valid][8][UTT][ MSE = 0.2371 | LCC = 0.8634 | SRCC = 0.8634 ] +INFO:root:Average loss=0.14593265482630485 +INFO:root: +[Valid][9][UTT][ MSE = 0.2837 | LCC = 0.8577 | SRCC = 0.8577 ] +INFO:root:Average loss=0.13991680359229064 +INFO:root: +[Valid][10][UTT][ MSE = 0.2248 | LCC = 0.8595 | SRCC = 0.8585 ] +INFO:root:Average loss=0.13096408049265543 +INFO:root: +[Valid][11][UTT][ MSE = 0.2440 | LCC = 0.8577 | SRCC = 0.8578 ] +INFO:root:Average loss=0.12200968254071015 +INFO:root: +[Valid][12][UTT][ MSE = 0.2564 | LCC = 0.8582 | SRCC = 0.8572 ] +INFO:root:Average loss=0.11561161785935745 +INFO:root: +[Valid][13][UTT][ MSE = 0.2439 | LCC = 0.8566 | SRCC = 0.8530 ] +INFO:root:Average loss=0.10790994553229748 +INFO:root: +[Valid][14][UTT][ MSE = 0.2291 | LCC = 0.8583 | SRCC = 0.8574 ] +INFO:root:Average loss=0.10573387117340015 +INFO:root: +[Valid][15][UTT][ MSE = 0.2823 | LCC = 0.8489 | SRCC = 0.8491 ] +INFO:root:Average loss=0.0981198213994503 +INFO:root: +[Valid][16][UTT][ MSE = 0.2500 | LCC = 0.8529 | SRCC = 0.8544 ] +INFO:root:Average loss=0.10147282786858387 +INFO:root: +[Valid][17][UTT][ MSE = 0.2324 | LCC = 0.8531 | SRCC = 0.8533 ] +INFO:root:Average loss=0.0896896202212725 +INFO:root: +[Valid][18][UTT][ MSE = 0.2411 | LCC = 0.8534 | SRCC = 0.8516 ] +INFO:root:Average loss=0.08966096949118835 +INFO:root: +[Valid][19][UTT][ MSE = 0.2353 | LCC = 0.8511 | SRCC = 0.8488 ] +INFO:root:Average loss=0.07674624140446003 +INFO:root: +[Valid][20][UTT][ MSE = 0.2631 | LCC = 0.8456 | SRCC = 0.8434 ] +INFO:root:Average loss=0.07811522049208482 +INFO:root: +[Valid][21][UTT][ MSE = 0.2371 | LCC = 0.8530 | SRCC = 0.8512 ] +INFO:root:Average loss=0.07460661948873447 +INFO:root: +[Valid][22][UTT][ MSE = 0.2365 | LCC = 0.8552 | SRCC = 0.8531 ] +INFO:root:Average loss=0.07118728541984008 +INFO:root: +[Valid][23][UTT][ MSE = 0.2374 | LCC = 0.8513 | SRCC = 0.8507 ] +INFO:root:Average loss=0.06571866380862701 +INFO:root: +[Valid][24][UTT][ MSE = 0.2366 | LCC = 0.8526 | SRCC = 0.8491 ] +INFO:root:Average loss=0.06625225141835518 +INFO:root: +[Valid][25][UTT][ MSE = 0.2425 | LCC = 0.8533 | SRCC = 0.8499 ] +INFO:root:Average loss=0.06517252722420754 +INFO:root: +[Valid][26][UTT][ MSE = 0.2557 | LCC = 0.8473 | SRCC = 0.8438 ] +INFO:root:Average loss=0.062970264790914 +INFO:root: +[Valid][27][UTT][ MSE = 0.2455 | LCC = 0.8510 | SRCC = 0.8488 ] +INFO:root:Average loss=0.060132404550527915 +INFO:root: +[Valid][28][UTT][ MSE = 0.2422 | LCC = 0.8478 | SRCC = 0.8467 ] +INFO:root:Average loss=0.05758569212869192 +INFO:root: +[Valid][29][UTT][ MSE = 0.2355 | LCC = 0.8513 | SRCC = 0.8497 ] +INFO:root:Average loss=0.05529439138869444 +INFO:root: +[Valid][30][UTT][ MSE = 0.2482 | LCC = 0.8446 | SRCC = 0.8433 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2315 | LCC = 0.8658 | SRCC = 0.8673 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..170eb06c6382387781b34d408420712e1a501423 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f369c0ade4d01762f221ca0179f313b1b1b2df5fc16aab5b912d17ed212dac9 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..1fc3eca66ba511caf31f01f9ea0487840ed469fb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0665a53a9c1028cf8d0abca6a22067d723f87f387f39f37942f5d800136de7d9 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..0895b287ca0b3be50f1c497d53865f658c4827ff --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer8/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=0.3549042284870759 +INFO:root: +[Valid][0][UTT][ MSE = 0.3317 | LCC = 0.8343 | SRCC = 0.8351 ] +INFO:root:Average loss=0.2486132501791685 +INFO:root: +[Valid][1][UTT][ MSE = 0.2892 | LCC = 0.8528 | SRCC = 0.8506 ] +INFO:root:Average loss=0.2186838046480448 +INFO:root: +[Valid][2][UTT][ MSE = 0.2810 | LCC = 0.8594 | SRCC = 0.8562 ] +INFO:root:Average loss=0.20482413757305878 +INFO:root: +[Valid][3][UTT][ MSE = 0.3353 | LCC = 0.8635 | SRCC = 0.8617 ] +INFO:root:Average loss=0.19687385092943144 +INFO:root: +[Valid][4][UTT][ MSE = 0.2518 | LCC = 0.8674 | SRCC = 0.8671 ] +INFO:root:Average loss=0.19050699539291552 +INFO:root: +[Valid][5][UTT][ MSE = 0.2975 | LCC = 0.8631 | SRCC = 0.8649 ] +INFO:root:Average loss=0.17280711873601645 +INFO:root: +[Valid][6][UTT][ MSE = 0.2582 | LCC = 0.8676 | SRCC = 0.8685 ] +INFO:root:Average loss=0.16970804047126037 +INFO:root: +[Valid][7][UTT][ MSE = 0.2672 | LCC = 0.8717 | SRCC = 0.8702 ] +INFO:root:Average loss=0.15834723584927046 +INFO:root: +[Valid][8][UTT][ MSE = 0.2639 | LCC = 0.8662 | SRCC = 0.8669 ] +INFO:root:Average loss=0.14880975259420198 +INFO:root: +[Valid][9][UTT][ MSE = 0.2473 | LCC = 0.8709 | SRCC = 0.8672 ] +INFO:root:Average loss=0.13862347182555076 +INFO:root: +[Valid][10][UTT][ MSE = 0.2421 | LCC = 0.8718 | SRCC = 0.8691 ] +INFO:root:Average loss=0.1296871766830102 +INFO:root: +[Valid][11][UTT][ MSE = 0.2493 | LCC = 0.8652 | SRCC = 0.8619 ] +INFO:root:Average loss=0.13181034494669008 +INFO:root: +[Valid][12][UTT][ MSE = 0.2393 | LCC = 0.8681 | SRCC = 0.8671 ] +INFO:root:Average loss=0.1178916827417337 +INFO:root: +[Valid][13][UTT][ MSE = 0.2523 | LCC = 0.8666 | SRCC = 0.8636 ] +INFO:root:Average loss=0.11287180458505948 +INFO:root: +[Valid][14][UTT][ MSE = 0.2575 | LCC = 0.8678 | SRCC = 0.8640 ] +INFO:root:Average loss=0.1068801703170324 +INFO:root: +[Valid][15][UTT][ MSE = 0.2388 | LCC = 0.8634 | SRCC = 0.8621 ] +INFO:root:Average loss=0.10364608032008012 +INFO:root: +[Valid][16][UTT][ MSE = 0.2543 | LCC = 0.8636 | SRCC = 0.8611 ] +INFO:root:Average loss=0.09660870844546036 +INFO:root: +[Valid][17][UTT][ MSE = 0.2765 | LCC = 0.8655 | SRCC = 0.8629 ] +INFO:root:Average loss=0.09448756812474667 +INFO:root: +[Valid][18][UTT][ MSE = 0.2555 | LCC = 0.8634 | SRCC = 0.8599 ] +INFO:root:Average loss=0.09033383204577825 +INFO:root: +[Valid][19][UTT][ MSE = 0.2367 | LCC = 0.8650 | SRCC = 0.8614 ] +INFO:root:Average loss=0.0823800103404583 +INFO:root: +[Valid][20][UTT][ MSE = 0.2703 | LCC = 0.8599 | SRCC = 0.8587 ] +INFO:root:Average loss=0.07647293252058518 +INFO:root: +[Valid][21][UTT][ MSE = 0.2687 | LCC = 0.8546 | SRCC = 0.8510 ] +INFO:root:Average loss=0.07521691760764672 +INFO:root: +[Valid][22][UTT][ MSE = 0.2720 | LCC = 0.8586 | SRCC = 0.8551 ] +INFO:root:Average loss=0.07102653537041102 +INFO:root: +[Valid][23][UTT][ MSE = 0.2813 | LCC = 0.8593 | SRCC = 0.8582 ] +INFO:root:Average loss=0.07068108323101814 +INFO:root: +[Valid][24][UTT][ MSE = 0.2416 | LCC = 0.8562 | SRCC = 0.8545 ] +INFO:root:Average loss=0.06536077679349826 +INFO:root: +[Valid][25][UTT][ MSE = 0.2596 | LCC = 0.8597 | SRCC = 0.8573 ] +INFO:root:Average loss=0.06467219064824092 +INFO:root: +[Valid][26][UTT][ MSE = 0.2410 | LCC = 0.8615 | SRCC = 0.8593 ] +INFO:root:Average loss=0.05911202184282816 +INFO:root: +[Valid][27][UTT][ MSE = 0.2413 | LCC = 0.8568 | SRCC = 0.8544 ] +INFO:root:Average loss=0.05879734547283405 +INFO:root: +[Valid][28][UTT][ MSE = 0.2604 | LCC = 0.8527 | SRCC = 0.8501 ] +INFO:root:Average loss=0.05641559893504167 +INFO:root: +[Valid][29][UTT][ MSE = 0.2418 | LCC = 0.8573 | SRCC = 0.8575 ] +INFO:root:Average loss=0.05939881283885393 +INFO:root: +[Valid][30][UTT][ MSE = 0.2605 | LCC = 0.8550 | SRCC = 0.8513 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2471 | LCC = 0.8601 | SRCC = 0.8631 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..7b7ed2083b838bf8e881c869ce322ec046af951d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:523dbb8eb1ccd7bd991b3e22644c1a38043057b8da21eaa6e41ec0272c265683 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..bd26858a5a533a6badddd81b409f102b2538ce42 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e153a9ef192b669cde6421eb34046397e6a702ba0ac09c7172418d4e0996688d +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..012b8465b849d208596389a60850cc4ec0f1cbcb --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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=0.3261918459947293 +INFO:root: +[Valid][0][UTT][ MSE = 0.3213 | LCC = 0.8348 | SRCC = 0.8355 ] +INFO:root:Average loss=0.25515777751421315 +INFO:root: +[Valid][1][UTT][ MSE = 0.2871 | LCC = 0.8437 | SRCC = 0.8439 ] +INFO:root:Average loss=0.23329516213673812 +INFO:root: +[Valid][2][UTT][ MSE = 0.2900 | LCC = 0.8590 | SRCC = 0.8555 ] +INFO:root:Average loss=0.22188410764703384 +INFO:root: +[Valid][3][UTT][ MSE = 0.2564 | LCC = 0.8625 | SRCC = 0.8609 ] +INFO:root:Average loss=0.21450309455394745 +INFO:root: +[Valid][4][UTT][ MSE = 0.2692 | LCC = 0.8601 | SRCC = 0.8599 ] +INFO:root:Average loss=0.20372846722602844 +INFO:root: +[Valid][5][UTT][ MSE = 0.2903 | LCC = 0.8715 | SRCC = 0.8702 ] +INFO:root:Average loss=0.2026046568002456 +INFO:root: +[Valid][6][UTT][ MSE = 0.2472 | LCC = 0.8634 | SRCC = 0.8634 ] +INFO:root:Average loss=0.19371828064322472 +INFO:root: +[Valid][7][UTT][ MSE = 0.2513 | LCC = 0.8692 | SRCC = 0.8679 ] +INFO:root:Average loss=0.18351634993002966 +INFO:root: +[Valid][8][UTT][ MSE = 0.2553 | LCC = 0.8664 | SRCC = 0.8655 ] +INFO:root:Average loss=0.18054162548520628 +INFO:root: +[Valid][9][UTT][ MSE = 0.2545 | LCC = 0.8714 | SRCC = 0.8700 ] +INFO:root:Average loss=0.17154379246326593 +INFO:root: +[Valid][10][UTT][ MSE = 0.2387 | LCC = 0.8671 | SRCC = 0.8666 ] +INFO:root:Average loss=0.1681134074162214 +INFO:root: +[Valid][11][UTT][ MSE = 0.2372 | LCC = 0.8680 | SRCC = 0.8680 ] +INFO:root:Average loss=0.1677925441509638 +INFO:root: +[Valid][12][UTT][ MSE = 0.2433 | LCC = 0.8728 | SRCC = 0.8721 ] +INFO:root:Average loss=0.1612604457216385 +INFO:root: +[Valid][13][UTT][ MSE = 0.2387 | LCC = 0.8615 | SRCC = 0.8627 ] +INFO:root:Average loss=0.15542279623257807 +INFO:root: +[Valid][14][UTT][ MSE = 0.2577 | LCC = 0.8640 | SRCC = 0.8633 ] +INFO:root:Average loss=0.1434595342247914 +INFO:root: +[Valid][15][UTT][ MSE = 0.2502 | LCC = 0.8588 | SRCC = 0.8612 ] +INFO:root:Average loss=0.14975747103110337 +INFO:root: +[Valid][16][UTT][ MSE = 0.2607 | LCC = 0.8619 | SRCC = 0.8612 ] +INFO:root:Average loss=0.14272125772176644 +INFO:root: +[Valid][17][UTT][ MSE = 0.2466 | LCC = 0.8665 | SRCC = 0.8661 ] +INFO:root:Average loss=0.1347831928004057 +INFO:root: +[Valid][18][UTT][ MSE = 0.2530 | LCC = 0.8638 | SRCC = 0.8637 ] +INFO:root:Average loss=0.12926727160811424 +INFO:root: +[Valid][19][UTT][ MSE = 0.2689 | LCC = 0.8595 | SRCC = 0.8589 ] +INFO:root:Average loss=0.1272902661600174 +INFO:root: +[Valid][20][UTT][ MSE = 0.2492 | LCC = 0.8615 | SRCC = 0.8615 ] +INFO:root:Average loss=0.11998188447875854 +INFO:root: +[Valid][21][UTT][ MSE = 0.2545 | LCC = 0.8581 | SRCC = 0.8576 ] +INFO:root:Average loss=0.12197175583778283 +INFO:root: +[Valid][22][UTT][ MSE = 0.3032 | LCC = 0.8556 | SRCC = 0.8569 ] +INFO:root:Average loss=0.11725875716178845 +INFO:root: +[Valid][23][UTT][ MSE = 0.2595 | LCC = 0.8518 | SRCC = 0.8509 ] +INFO:root:Average loss=0.11218050943735318 +INFO:root: +[Valid][24][UTT][ MSE = 0.3017 | LCC = 0.8500 | SRCC = 0.8492 ] +INFO:root:Average loss=0.11172519423640691 +INFO:root: +[Valid][25][UTT][ MSE = 0.2685 | LCC = 0.8545 | SRCC = 0.8549 ] +INFO:root:Average loss=0.10116524058274734 +INFO:root: +[Valid][26][UTT][ MSE = 0.2580 | LCC = 0.8558 | SRCC = 0.8533 ] +INFO:root:Average loss=0.09833318233872071 +INFO:root: +[Valid][27][UTT][ MSE = 0.2632 | LCC = 0.8535 | SRCC = 0.8531 ] +INFO:root:Average loss=0.09665097926671688 +INFO:root: +[Valid][28][UTT][ MSE = 0.2638 | LCC = 0.8514 | SRCC = 0.8505 ] +INFO:root:Average loss=0.09867233587190127 +INFO:root: +[Valid][29][UTT][ MSE = 0.2619 | LCC = 0.8483 | SRCC = 0.8478 ] +INFO:root:Average loss=0.09017600754323678 +INFO:root: +[Valid][30][UTT][ MSE = 0.2436 | LCC = 0.8544 | SRCC = 0.8548 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2621 | LCC = 0.8641 | SRCC = 0.8628 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..efc73da1bd2da293639d7c63e0d5e95e2c5c84fd --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efe47631ed048fe047841a5ac67c5735b9f9028efb60c2cf5bb8d55aa1ba631a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..3a2f299fd2196f9f7a2e1b57cf7180cb12e84670 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27f438563668678fdc7a3304914bda0095d9cb0660629dd94960172901372e25 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..45afed37deeffe5ff04331295af626c67949a256 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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=0.30171274202756393 +INFO:root: +[Valid][0][UTT][ MSE = 0.3062 | LCC = 0.8404 | SRCC = 0.8395 ] +INFO:root:Average loss=0.24735527810377952 +INFO:root: +[Valid][1][UTT][ MSE = 0.2630 | LCC = 0.8491 | SRCC = 0.8521 ] +INFO:root:Average loss=0.22958607226610184 +INFO:root: +[Valid][2][UTT][ MSE = 0.2569 | LCC = 0.8602 | SRCC = 0.8592 ] +INFO:root:Average loss=0.21939270886091086 +INFO:root: +[Valid][3][UTT][ MSE = 0.3440 | LCC = 0.8617 | SRCC = 0.8619 ] +INFO:root:Average loss=0.22150632643546814 +INFO:root: +[Valid][4][UTT][ MSE = 0.2650 | LCC = 0.8673 | SRCC = 0.8657 ] +INFO:root:Average loss=0.2058237183552522 +INFO:root: +[Valid][5][UTT][ MSE = 0.2678 | LCC = 0.8675 | SRCC = 0.8671 ] +INFO:root:Average loss=0.20126484277156684 +INFO:root: +[Valid][6][UTT][ MSE = 0.2729 | LCC = 0.8666 | SRCC = 0.8662 ] +INFO:root:Average loss=0.18987009731622842 +INFO:root: +[Valid][7][UTT][ MSE = 0.2519 | LCC = 0.8671 | SRCC = 0.8675 ] +INFO:root:Average loss=0.19595674874308783 +INFO:root: +[Valid][8][UTT][ MSE = 0.2370 | LCC = 0.8683 | SRCC = 0.8675 ] +INFO:root:Average loss=0.1773128178066168 +INFO:root: +[Valid][9][UTT][ MSE = 0.2447 | LCC = 0.8680 | SRCC = 0.8688 ] +INFO:root:Average loss=0.17701911037931076 +INFO:root: +[Valid][10][UTT][ MSE = 0.2694 | LCC = 0.8690 | SRCC = 0.8675 ] +INFO:root:Average loss=0.1711610347414628 +INFO:root: +[Valid][11][UTT][ MSE = 0.2454 | LCC = 0.8669 | SRCC = 0.8659 ] +INFO:root:Average loss=0.1753987399622416 +INFO:root: +[Valid][12][UTT][ MSE = 0.2415 | LCC = 0.8679 | SRCC = 0.8668 ] +INFO:root:Average loss=0.15757432465369886 +INFO:root: +[Valid][13][UTT][ MSE = 0.2359 | LCC = 0.8671 | SRCC = 0.8669 ] +INFO:root:Average loss=0.1616271364574249 +INFO:root: +[Valid][14][UTT][ MSE = 0.2541 | LCC = 0.8605 | SRCC = 0.8612 ] +INFO:root:Average loss=0.15946402085515168 +INFO:root: +[Valid][15][UTT][ MSE = 0.2475 | LCC = 0.8641 | SRCC = 0.8639 ] +INFO:root:Average loss=0.14789404614995688 +INFO:root: +[Valid][16][UTT][ MSE = 0.2494 | LCC = 0.8595 | SRCC = 0.8604 ] +INFO:root:Average loss=0.14191924771055198 +INFO:root: +[Valid][17][UTT][ MSE = 0.2570 | LCC = 0.8589 | SRCC = 0.8588 ] +INFO:root:Average loss=0.13748258208999267 +INFO:root: +[Valid][18][UTT][ MSE = 0.2574 | LCC = 0.8577 | SRCC = 0.8565 ] +INFO:root:Average loss=0.13515741874774298 +INFO:root: +[Valid][19][UTT][ MSE = 0.2548 | LCC = 0.8598 | SRCC = 0.8584 ] +INFO:root:Average loss=0.1336734610108229 +INFO:root: +[Valid][20][UTT][ MSE = 0.2387 | LCC = 0.8602 | SRCC = 0.8612 ] +INFO:root:Average loss=0.12295146343799737 +INFO:root: +[Valid][21][UTT][ MSE = 0.2647 | LCC = 0.8535 | SRCC = 0.8514 ] +INFO:root:Average loss=0.1233148826047396 +INFO:root: +[Valid][22][UTT][ MSE = 0.2493 | LCC = 0.8554 | SRCC = 0.8529 ] +INFO:root:Average loss=0.1204994352391133 +INFO:root: +[Valid][23][UTT][ MSE = 0.2371 | LCC = 0.8580 | SRCC = 0.8575 ] +INFO:root:Average loss=0.12187462500654735 +INFO:root: +[Valid][24][UTT][ MSE = 0.2364 | LCC = 0.8591 | SRCC = 0.8586 ] +INFO:root:Average loss=0.11454739574438487 +INFO:root: +[Valid][25][UTT][ MSE = 0.2459 | LCC = 0.8568 | SRCC = 0.8563 ] +INFO:root:Average loss=0.11087185364121045 +INFO:root: +[Valid][26][UTT][ MSE = 0.2446 | LCC = 0.8563 | SRCC = 0.8555 ] +INFO:root:Average loss=0.10655014598980928 +INFO:root: +[Valid][27][UTT][ MSE = 0.2541 | LCC = 0.8510 | SRCC = 0.8507 ] +INFO:root:Average loss=0.10380384650750038 +INFO:root: +[Valid][28][UTT][ MSE = 0.2443 | LCC = 0.8537 | SRCC = 0.8537 ] +INFO:root:Average loss=0.0980947325244928 +INFO:root: +[Valid][29][UTT][ MSE = 0.2598 | LCC = 0.8485 | SRCC = 0.8455 ] +INFO:root:Average loss=0.10046223216714004 +INFO:root: +[Valid][30][UTT][ MSE = 0.2421 | LCC = 0.8517 | SRCC = 0.8516 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2767 | LCC = 0.8596 | SRCC = 0.8591 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..7c9d3f146187361a875137bc18a18cb008603765 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:93021fd28957eabd4a3b40e34a9250e2e0f93e233c5b1443f4be022dedf97244 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..3a967dea857c9110540fe63ec5b6cdfddbe7ef5f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bf4900c3ce582ca6b93edc0491a420cfbb08d7dddc54faf963a682d846a4a201 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..89d68adceb3f26f89184d27ad0c19c0cd7bbaa65 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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=0.34623666699880207 +INFO:root: +[Valid][0][UTT][ MSE = 0.3029 | LCC = 0.8377 | SRCC = 0.8369 ] +INFO:root:Average loss=0.2503905529395128 +INFO:root: +[Valid][1][UTT][ MSE = 0.3102 | LCC = 0.8493 | SRCC = 0.8483 ] +INFO:root:Average loss=0.2321739049676137 +INFO:root: +[Valid][2][UTT][ MSE = 0.2837 | LCC = 0.8590 | SRCC = 0.8561 ] +INFO:root:Average loss=0.22460153297736093 +INFO:root: +[Valid][3][UTT][ MSE = 0.2725 | LCC = 0.8607 | SRCC = 0.8571 ] +INFO:root:Average loss=0.20963521664723372 +INFO:root: +[Valid][4][UTT][ MSE = 0.2725 | LCC = 0.8625 | SRCC = 0.8609 ] +INFO:root:Average loss=0.2049122840548173 +INFO:root: +[Valid][5][UTT][ MSE = 0.2521 | LCC = 0.8648 | SRCC = 0.8639 ] +INFO:root:Average loss=0.20137875116406342 +INFO:root: +[Valid][6][UTT][ MSE = 0.2760 | LCC = 0.8652 | SRCC = 0.8654 ] +INFO:root:Average loss=0.19566338910506323 +INFO:root: +[Valid][7][UTT][ MSE = 0.2410 | LCC = 0.8657 | SRCC = 0.8649 ] +INFO:root:Average loss=0.19298111360806686 +INFO:root: +[Valid][8][UTT][ MSE = 0.2279 | LCC = 0.8638 | SRCC = 0.8666 ] +INFO:root:Average loss=0.18066376762894484 +INFO:root: +[Valid][9][UTT][ MSE = 0.2620 | LCC = 0.8646 | SRCC = 0.8646 ] +INFO:root:Average loss=0.17702935569179365 +INFO:root: +[Valid][10][UTT][ MSE = 0.2543 | LCC = 0.8667 | SRCC = 0.8637 ] +INFO:root:Average loss=0.16982431241717094 +INFO:root: +[Valid][11][UTT][ MSE = 0.2362 | LCC = 0.8648 | SRCC = 0.8647 ] +INFO:root:Average loss=0.165732079018385 +INFO:root: +[Valid][12][UTT][ MSE = 0.2400 | LCC = 0.8654 | SRCC = 0.8636 ] +INFO:root:Average loss=0.15780565820825407 +INFO:root: +[Valid][13][UTT][ MSE = 0.2935 | LCC = 0.8594 | SRCC = 0.8580 ] +INFO:root:Average loss=0.14708234856908137 +INFO:root: +[Valid][14][UTT][ MSE = 0.2328 | LCC = 0.8639 | SRCC = 0.8632 ] +INFO:root:Average loss=0.14839457835142428 +INFO:root: +[Valid][15][UTT][ MSE = 0.2577 | LCC = 0.8650 | SRCC = 0.8631 ] +INFO:root:Average loss=0.14691603241058496 +INFO:root: +[Valid][16][UTT][ MSE = 0.2618 | LCC = 0.8629 | SRCC = 0.8617 ] +INFO:root:Average loss=0.13483756952560866 +INFO:root: +[Valid][17][UTT][ MSE = 0.2430 | LCC = 0.8621 | SRCC = 0.8611 ] +INFO:root:Average loss=0.13496750965714455 +INFO:root: +[Valid][18][UTT][ MSE = 0.2623 | LCC = 0.8630 | SRCC = 0.8618 ] +INFO:root:Average loss=0.12899398736846754 +INFO:root: +[Valid][19][UTT][ MSE = 0.2570 | LCC = 0.8576 | SRCC = 0.8565 ] +INFO:root:Average loss=0.12485129787371708 +INFO:root: +[Valid][20][UTT][ MSE = 0.2419 | LCC = 0.8566 | SRCC = 0.8558 ] +INFO:root:Average loss=0.11753995248522514 +INFO:root: +[Valid][21][UTT][ MSE = 0.2332 | LCC = 0.8582 | SRCC = 0.8587 ] +INFO:root:Average loss=0.1163296556721131 +INFO:root: +[Valid][22][UTT][ MSE = 0.2403 | LCC = 0.8577 | SRCC = 0.8542 ] +INFO:root:Average loss=0.11199496008264713 +INFO:root: +[Valid][23][UTT][ MSE = 0.2456 | LCC = 0.8503 | SRCC = 0.8537 ] +INFO:root:Average loss=0.11529060510488656 +INFO:root: +[Valid][24][UTT][ MSE = 0.2692 | LCC = 0.8566 | SRCC = 0.8543 ] +INFO:root:Average loss=0.10844204106774086 +INFO:root: +[Valid][25][UTT][ MSE = 0.2606 | LCC = 0.8514 | SRCC = 0.8495 ] +INFO:root:Average loss=0.10392422620684673 +INFO:root: +[Valid][26][UTT][ MSE = 0.2699 | LCC = 0.8503 | SRCC = 0.8514 ] +INFO:root:Average loss=0.10299247722977246 +INFO:root: +[Valid][27][UTT][ MSE = 0.2433 | LCC = 0.8535 | SRCC = 0.8516 ] +INFO:root:Average loss=0.0967668191744731 +INFO:root: +[Valid][28][UTT][ MSE = 0.2827 | LCC = 0.8508 | SRCC = 0.8475 ] +INFO:root:Average loss=0.09791986099802531 +INFO:root: +[Valid][29][UTT][ MSE = 0.2631 | LCC = 0.8558 | SRCC = 0.8537 ] +INFO:root:Average loss=0.08882993903870766 +INFO:root: +[Valid][30][UTT][ MSE = 0.2534 | LCC = 0.8530 | SRCC = 0.8518 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2588 | LCC = 0.8667 | SRCC = 0.8661 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..4076454f601d7c99c31621bbb7c19dc5811d960e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:664a4a5fab0dc4d41ee16b20448a7ff95b71da7a4137ec5733186bbe75693978 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..acf89b411f46ab4244cd1372c78f91cfe54da714 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d59b3c6a1390ffa8afe65dd5ea7b456f4ded17cd5d1c5893651c747b3b30f01c +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..c551a15d5bca0fab0cd9ef8c9baeb9cdd8e69cbe --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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=0.3534568190956727 +INFO:root: +[Valid][0][UTT][ MSE = 0.3257 | LCC = 0.8367 | SRCC = 0.8364 ] +INFO:root:Average loss=0.2427778931764456 +INFO:root: +[Valid][1][UTT][ MSE = 0.3433 | LCC = 0.8524 | SRCC = 0.8519 ] +INFO:root:Average loss=0.23261064587113184 +INFO:root: +[Valid][2][UTT][ MSE = 0.2922 | LCC = 0.8566 | SRCC = 0.8549 ] +INFO:root:Average loss=0.21961841713159513 +INFO:root: +[Valid][3][UTT][ MSE = 0.2770 | LCC = 0.8573 | SRCC = 0.8571 ] +INFO:root:Average loss=0.2098760914344054 +INFO:root: +[Valid][4][UTT][ MSE = 0.2480 | LCC = 0.8602 | SRCC = 0.8594 ] +INFO:root:Average loss=0.20804087225443277 +INFO:root: +[Valid][5][UTT][ MSE = 0.3113 | LCC = 0.8652 | SRCC = 0.8639 ] +INFO:root:Average loss=0.20307454973077163 +INFO:root: +[Valid][6][UTT][ MSE = 0.2725 | LCC = 0.8642 | SRCC = 0.8646 ] +INFO:root:Average loss=0.19156524844658682 +INFO:root: +[Valid][7][UTT][ MSE = 0.2741 | LCC = 0.8670 | SRCC = 0.8650 ] +INFO:root:Average loss=0.18446782011634263 +INFO:root: +[Valid][8][UTT][ MSE = 0.2676 | LCC = 0.8620 | SRCC = 0.8631 ] +INFO:root:Average loss=0.18522583158352438 +INFO:root: +[Valid][9][UTT][ MSE = 0.3291 | LCC = 0.8653 | SRCC = 0.8671 ] +INFO:root:Average loss=0.1776545613239973 +INFO:root: +[Valid][10][UTT][ MSE = 0.2680 | LCC = 0.8665 | SRCC = 0.8647 ] +INFO:root:Average loss=0.17123926765261552 +INFO:root: +[Valid][11][UTT][ MSE = 0.2585 | LCC = 0.8687 | SRCC = 0.8670 ] +INFO:root:Average loss=0.17468309154113135 +INFO:root: +[Valid][12][UTT][ MSE = 0.2691 | LCC = 0.8608 | SRCC = 0.8570 ] +INFO:root:Average loss=0.1606554428163247 +INFO:root: +[Valid][13][UTT][ MSE = 0.2644 | LCC = 0.8663 | SRCC = 0.8638 ] +INFO:root:Average loss=0.15611121488305238 +INFO:root: +[Valid][14][UTT][ MSE = 0.2857 | LCC = 0.8593 | SRCC = 0.8571 ] +INFO:root:Average loss=0.14958299963902205 +INFO:root: +[Valid][15][UTT][ MSE = 0.2885 | LCC = 0.8630 | SRCC = 0.8614 ] +INFO:root:Average loss=0.1477968968833104 +INFO:root: +[Valid][16][UTT][ MSE = 0.2824 | LCC = 0.8623 | SRCC = 0.8639 ] +INFO:root:Average loss=0.15105268578880873 +INFO:root: +[Valid][17][UTT][ MSE = 0.2727 | LCC = 0.8603 | SRCC = 0.8600 ] +INFO:root:Average loss=0.1319523223508627 +INFO:root: +[Valid][18][UTT][ MSE = 0.2786 | LCC = 0.8564 | SRCC = 0.8532 ] +INFO:root:Average loss=0.1356138918453302 +INFO:root: +[Valid][19][UTT][ MSE = 0.2491 | LCC = 0.8618 | SRCC = 0.8603 ] +INFO:root:Average loss=0.13028881173485365 +INFO:root: +[Valid][20][UTT][ MSE = 0.2880 | LCC = 0.8505 | SRCC = 0.8487 ] +INFO:root:Average loss=0.12020434448734307 +INFO:root: +[Valid][21][UTT][ MSE = 0.2473 | LCC = 0.8535 | SRCC = 0.8520 ] +INFO:root:Average loss=0.11786855155458817 +INFO:root: +[Valid][22][UTT][ MSE = 0.2861 | LCC = 0.8523 | SRCC = 0.8495 ] +INFO:root:Average loss=0.11762338160322262 +INFO:root: +[Valid][23][UTT][ MSE = 0.2511 | LCC = 0.8537 | SRCC = 0.8538 ] +INFO:root:Average loss=0.11585306615019456 +INFO:root: +[Valid][24][UTT][ MSE = 0.2644 | LCC = 0.8531 | SRCC = 0.8525 ] +INFO:root:Average loss=0.11555874185302319 +INFO:root: +[Valid][25][UTT][ MSE = 0.2603 | LCC = 0.8546 | SRCC = 0.8525 ] +INFO:root:Average loss=0.11148543073198734 +INFO:root: +[Valid][26][UTT][ MSE = 0.2563 | LCC = 0.8529 | SRCC = 0.8522 ] +INFO:root:Average loss=0.10565162727083915 +INFO:root: +[Valid][27][UTT][ MSE = 0.3012 | LCC = 0.8510 | SRCC = 0.8495 ] +INFO:root:Average loss=0.10118713134374374 +INFO:root: +[Valid][28][UTT][ MSE = 0.2587 | LCC = 0.8484 | SRCC = 0.8491 ] +INFO:root:Average loss=0.09346406964155343 +INFO:root: +[Valid][29][UTT][ MSE = 0.2692 | LCC = 0.8503 | SRCC = 0.8491 ] +INFO:root:Average loss=0.09560236435096997 +INFO:root: +[Valid][30][UTT][ MSE = 0.2751 | LCC = 0.8492 | SRCC = 0.8492 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2642 | LCC = 0.8664 | SRCC = 0.8704 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..d8012d772d937d39d8de9bd11ee8a517e0213df3 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19067dc0a7e86cb2bff5b2f3d167be2dd7912f999b118d18eea97000f963acbd +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..9b3bcd4f7eef41cd79ac4d7e948ec2a0a36943db --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f40a33d11edd82b66457974638ea4c084f358c4b7655c8589857b942b4203f7 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..8bf8d4826c929aaaf74e50435dbfa9620bb36af8 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_5epoch/layer9/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=0.36598835102258587 +INFO:root: +[Valid][0][UTT][ MSE = 0.3402 | LCC = 0.8281 | SRCC = 0.8285 ] +INFO:root:Average loss=0.2577115595340729 +INFO:root: +[Valid][1][UTT][ MSE = 0.3128 | LCC = 0.8405 | SRCC = 0.8407 ] +INFO:root:Average loss=0.23510586451261473 +INFO:root: +[Valid][2][UTT][ MSE = 0.3129 | LCC = 0.8549 | SRCC = 0.8522 ] +INFO:root:Average loss=0.22438102100904173 +INFO:root: +[Valid][3][UTT][ MSE = 0.3265 | LCC = 0.8516 | SRCC = 0.8524 ] +INFO:root:Average loss=0.21832740345062354 +INFO:root: +[Valid][4][UTT][ MSE = 0.2682 | LCC = 0.8579 | SRCC = 0.8583 ] +INFO:root:Average loss=0.20762086955782694 +INFO:root: +[Valid][5][UTT][ MSE = 0.2779 | LCC = 0.8650 | SRCC = 0.8632 ] +INFO:root:Average loss=0.20112017188698816 +INFO:root: +[Valid][6][UTT][ MSE = 0.3349 | LCC = 0.8660 | SRCC = 0.8642 ] +INFO:root:Average loss=0.19899039314343378 +INFO:root: +[Valid][7][UTT][ MSE = 0.2896 | LCC = 0.8638 | SRCC = 0.8660 ] +INFO:root:Average loss=0.19322490157225194 +INFO:root: +[Valid][8][UTT][ MSE = 0.2800 | LCC = 0.8678 | SRCC = 0.8654 ] +INFO:root:Average loss=0.1823662038988028 +INFO:root: +[Valid][9][UTT][ MSE = 0.2708 | LCC = 0.8689 | SRCC = 0.8682 ] +INFO:root:Average loss=0.17253085034780014 +INFO:root: +[Valid][10][UTT][ MSE = 0.2736 | LCC = 0.8708 | SRCC = 0.8684 ] +INFO:root:Average loss=0.1728742661384436 +INFO:root: +[Valid][11][UTT][ MSE = 0.2898 | LCC = 0.8662 | SRCC = 0.8649 ] +INFO:root:Average loss=0.17207205075866136 +INFO:root: +[Valid][12][UTT][ MSE = 0.2505 | LCC = 0.8696 | SRCC = 0.8689 ] +INFO:root:Average loss=0.16310476588133055 +INFO:root: +[Valid][13][UTT][ MSE = 0.2438 | LCC = 0.8719 | SRCC = 0.8707 ] +INFO:root:Average loss=0.15626410146554312 +INFO:root: +[Valid][14][UTT][ MSE = 0.2456 | LCC = 0.8629 | SRCC = 0.8648 ] +INFO:root:Average loss=0.15256070765929344 +INFO:root: +[Valid][15][UTT][ MSE = 0.2783 | LCC = 0.8661 | SRCC = 0.8662 ] +INFO:root:Average loss=0.14740048635464448 +INFO:root: +[Valid][16][UTT][ MSE = 0.2611 | LCC = 0.8625 | SRCC = 0.8620 ] +INFO:root:Average loss=0.1393830921405401 +INFO:root: +[Valid][17][UTT][ MSE = 0.2524 | LCC = 0.8657 | SRCC = 0.8655 ] +INFO:root:Average loss=0.1372955349775461 +INFO:root: +[Valid][18][UTT][ MSE = 0.2498 | LCC = 0.8652 | SRCC = 0.8661 ] +INFO:root:Average loss=0.13281875122816134 +INFO:root: +[Valid][19][UTT][ MSE = 0.2263 | LCC = 0.8622 | SRCC = 0.8665 ] +INFO:root:Average loss=0.1279822941391896 +INFO:root: +[Valid][20][UTT][ MSE = 0.2224 | LCC = 0.8628 | SRCC = 0.8633 ] +INFO:root:Average loss=0.12368151650596888 +INFO:root: +[Valid][21][UTT][ MSE = 0.2393 | LCC = 0.8601 | SRCC = 0.8605 ] +INFO:root:Average loss=0.11652135228117307 +INFO:root: +[Valid][22][UTT][ MSE = 0.2669 | LCC = 0.8601 | SRCC = 0.8585 ] +INFO:root:Average loss=0.11459748838574459 +INFO:root: +[Valid][23][UTT][ MSE = 0.2578 | LCC = 0.8547 | SRCC = 0.8537 ] +INFO:root:Average loss=0.11453868792607234 +INFO:root: +[Valid][24][UTT][ MSE = 0.2560 | LCC = 0.8587 | SRCC = 0.8568 ] +INFO:root:Average loss=0.11497864093727027 +INFO:root: +[Valid][25][UTT][ MSE = 0.2560 | LCC = 0.8581 | SRCC = 0.8578 ] +INFO:root:Average loss=0.10567407224040765 +INFO:root: +[Valid][26][UTT][ MSE = 0.2414 | LCC = 0.8556 | SRCC = 0.8563 ] +INFO:root:Average loss=0.1003360669964399 +INFO:root: +[Valid][27][UTT][ MSE = 0.2477 | LCC = 0.8542 | SRCC = 0.8536 ] +INFO:root:Average loss=0.10069668741944508 +INFO:root: +[Valid][28][UTT][ MSE = 0.2553 | LCC = 0.8530 | SRCC = 0.8520 ] +INFO:root:Average loss=0.092719852590026 +INFO:root: +[Valid][29][UTT][ MSE = 0.2620 | LCC = 0.8513 | SRCC = 0.8518 ] +INFO:root:Average loss=0.09085672969619434 +INFO:root: +[Valid][30][UTT][ MSE = 0.2737 | LCC = 0.8510 | SRCC = 0.8500 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2427 | LCC = 0.8671 | SRCC = 0.8681 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..9106e84e55a648f05ec7382c58fa7b83040b5779 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:347c949e3d26f7c93dc9b7932aad896a9f4c657f92a40b545ccf3632a4aa188f +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7d344b8ebaee368232ccb65614e02fb039b29c35 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a81f31ee1cc1825a9990c5a64a220b5bd2e5710f5b16d11929237564a812eb86 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..31bf9ac6a1f34134af64fc6839861ad73b59b765 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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=0.5163642080166401 +INFO:root: +[Valid][0][UTT][ MSE = 0.3789 | LCC = 0.7599 | SRCC = 0.7626 ] +INFO:root:Average loss=0.29631015524650234 +INFO:root: +[Valid][1][UTT][ MSE = 0.3292 | LCC = 0.8027 | SRCC = 0.8052 ] +INFO:root:Average loss=0.2633215198532129 +INFO:root: +[Valid][2][UTT][ MSE = 0.2984 | LCC = 0.8146 | SRCC = 0.8156 ] +INFO:root:Average loss=0.24295559773842493 +INFO:root: +[Valid][3][UTT][ MSE = 0.2860 | LCC = 0.8286 | SRCC = 0.8275 ] +INFO:root:Average loss=0.2210187533727059 +INFO:root: +[Valid][4][UTT][ MSE = 0.2708 | LCC = 0.8353 | SRCC = 0.8345 ] +INFO:root:Average loss=0.20853095253308615 +INFO:root: +[Valid][5][UTT][ MSE = 0.2637 | LCC = 0.8373 | SRCC = 0.8393 ] +INFO:root:Average loss=0.19494426976411772 +INFO:root: +[Valid][6][UTT][ MSE = 0.2607 | LCC = 0.8374 | SRCC = 0.8388 ] +INFO:root:Average loss=0.18538387454090974 +INFO:root: +[Valid][7][UTT][ MSE = 0.2620 | LCC = 0.8362 | SRCC = 0.8407 ] +INFO:root:Average loss=0.1829232286948424 +INFO:root: +[Valid][8][UTT][ MSE = 0.2563 | LCC = 0.8365 | SRCC = 0.8400 ] +INFO:root:Average loss=0.16535514870133156 +INFO:root: +[Valid][9][UTT][ MSE = 0.2441 | LCC = 0.8432 | SRCC = 0.8452 ] +INFO:root:Average loss=0.15480385978634542 +INFO:root: +[Valid][10][UTT][ MSE = 0.2863 | LCC = 0.8416 | SRCC = 0.8419 ] +INFO:root:Average loss=0.15312999439163086 +INFO:root: +[Valid][11][UTT][ MSE = 0.2548 | LCC = 0.8367 | SRCC = 0.8392 ] +INFO:root:Average loss=0.14609907691677412 +INFO:root: +[Valid][12][UTT][ MSE = 0.2462 | LCC = 0.8478 | SRCC = 0.8475 ] +INFO:root:Average loss=0.15136885050779733 +INFO:root: +[Valid][13][UTT][ MSE = 0.2472 | LCC = 0.8436 | SRCC = 0.8482 ] +INFO:root:Average loss=0.127789434332114 +INFO:root: +[Valid][14][UTT][ MSE = 0.2396 | LCC = 0.8496 | SRCC = 0.8512 ] +INFO:root:Average loss=0.1302586677364814 +INFO:root: +[Valid][15][UTT][ MSE = 0.2488 | LCC = 0.8483 | SRCC = 0.8478 ] +INFO:root:Average loss=0.12196954282430503 +INFO:root: +[Valid][16][UTT][ MSE = 0.2363 | LCC = 0.8499 | SRCC = 0.8502 ] +INFO:root:Average loss=0.1143175673026305 +INFO:root: +[Valid][17][UTT][ MSE = 0.2689 | LCC = 0.8425 | SRCC = 0.8485 ] +INFO:root:Average loss=0.11305760076412788 +INFO:root: +[Valid][18][UTT][ MSE = 0.2358 | LCC = 0.8485 | SRCC = 0.8487 ] +INFO:root:Average loss=0.10982131795623364 +INFO:root: +[Valid][19][UTT][ MSE = 0.2338 | LCC = 0.8484 | SRCC = 0.8495 ] +INFO:root:Average loss=0.10525415054498574 +INFO:root: +[Valid][20][UTT][ MSE = 0.2365 | LCC = 0.8544 | SRCC = 0.8543 ] +INFO:root:Average loss=0.10402038349555089 +INFO:root: +[Valid][21][UTT][ MSE = 0.2263 | LCC = 0.8531 | SRCC = 0.8551 ] +INFO:root:Average loss=0.09755289521163855 +INFO:root: +[Valid][22][UTT][ MSE = 0.2250 | LCC = 0.8569 | SRCC = 0.8576 ] +INFO:root:Average loss=0.09399899563346154 +INFO:root: +[Valid][23][UTT][ MSE = 0.2388 | LCC = 0.8498 | SRCC = 0.8504 ] +INFO:root:Average loss=0.08815563622957621 +INFO:root: +[Valid][24][UTT][ MSE = 0.2375 | LCC = 0.8481 | SRCC = 0.8499 ] +INFO:root:Average loss=0.09276632042840505 +INFO:root: +[Valid][25][UTT][ MSE = 0.2671 | LCC = 0.8454 | SRCC = 0.8508 ] +INFO:root:Average loss=0.10284659596016774 +INFO:root: +[Valid][26][UTT][ MSE = 0.2425 | LCC = 0.8514 | SRCC = 0.8514 ] +INFO:root:Average loss=0.08485921782751878 +INFO:root: +[Valid][27][UTT][ MSE = 0.2425 | LCC = 0.8489 | SRCC = 0.8505 ] +INFO:root:Average loss=0.08429477072488038 +INFO:root: +[Valid][28][UTT][ MSE = 0.2315 | LCC = 0.8520 | SRCC = 0.8521 ] +INFO:root:Average loss=0.07732289174619393 +INFO:root: +[Valid][29][UTT][ MSE = 0.2301 | LCC = 0.8508 | SRCC = 0.8525 ] +INFO:root:Average loss=0.07949110349783531 +INFO:root: +[Valid][30][UTT][ MSE = 0.2293 | LCC = 0.8515 | SRCC = 0.8513 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2357 | LCC = 0.8607 | SRCC = 0.8614 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..c0323e73cd687326ebd786f28683fbad566900de --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3072ea1ccbc96a9d8d1d9fe99826f5a854c6d61444ac3330bacdb8a4d55670a +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..2ab29b5a2cd54625453f1a573a2ff260efc91184 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b8fd0da8151c4793f061e3d38677d7e090b97896cdb9bd1234e203892e36122e +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..6529096aa0d51e62c3f2bb7adf87bcfe22d3334f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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=0.5086980330256315 +INFO:root: +[Valid][0][UTT][ MSE = 0.3715 | LCC = 0.7512 | SRCC = 0.7558 ] +INFO:root:Average loss=0.30535538093401837 +INFO:root: +[Valid][1][UTT][ MSE = 0.3323 | LCC = 0.7985 | SRCC = 0.7979 ] +INFO:root:Average loss=0.2591889821566068 +INFO:root: +[Valid][2][UTT][ MSE = 0.2868 | LCC = 0.8167 | SRCC = 0.8162 ] +INFO:root:Average loss=0.2439753206876608 +INFO:root: +[Valid][3][UTT][ MSE = 0.2877 | LCC = 0.8224 | SRCC = 0.8206 ] +INFO:root:Average loss=0.2269977730436203 +INFO:root: +[Valid][4][UTT][ MSE = 0.2759 | LCC = 0.8303 | SRCC = 0.8288 ] +INFO:root:Average loss=0.21136175917509276 +INFO:root: +[Valid][5][UTT][ MSE = 0.3028 | LCC = 0.8281 | SRCC = 0.8290 ] +INFO:root:Average loss=0.21182122501807335 +INFO:root: +[Valid][6][UTT][ MSE = 0.2902 | LCC = 0.8281 | SRCC = 0.8310 ] +INFO:root:Average loss=0.190667424733058 +INFO:root: +[Valid][7][UTT][ MSE = 0.2587 | LCC = 0.8383 | SRCC = 0.8367 ] +INFO:root:Average loss=0.18095173868231285 +INFO:root: +[Valid][8][UTT][ MSE = 0.2629 | LCC = 0.8384 | SRCC = 0.8376 ] +INFO:root:Average loss=0.17146639076945108 +INFO:root: +[Valid][9][UTT][ MSE = 0.2616 | LCC = 0.8371 | SRCC = 0.8333 ] +INFO:root:Average loss=0.17104634451560485 +INFO:root: +[Valid][10][UTT][ MSE = 0.2555 | LCC = 0.8391 | SRCC = 0.8396 ] +INFO:root:Average loss=0.16366322758870247 +INFO:root: +[Valid][11][UTT][ MSE = 0.2484 | LCC = 0.8428 | SRCC = 0.8420 ] +INFO:root:Average loss=0.15247840099991897 +INFO:root: +[Valid][12][UTT][ MSE = 0.2526 | LCC = 0.8451 | SRCC = 0.8434 ] +INFO:root:Average loss=0.1455710330643715 +INFO:root: +[Valid][13][UTT][ MSE = 0.2700 | LCC = 0.8399 | SRCC = 0.8393 ] +INFO:root:Average loss=0.1419865953234526 +INFO:root: +[Valid][14][UTT][ MSE = 0.2523 | LCC = 0.8442 | SRCC = 0.8434 ] +INFO:root:Average loss=0.1380590945482254 +INFO:root: +[Valid][15][UTT][ MSE = 0.2476 | LCC = 0.8401 | SRCC = 0.8406 ] +INFO:root:Average loss=0.12693194767985588 +INFO:root: +[Valid][16][UTT][ MSE = 0.2470 | LCC = 0.8443 | SRCC = 0.8404 ] +INFO:root:Average loss=0.12659801810215682 +INFO:root: +[Valid][17][UTT][ MSE = 0.2392 | LCC = 0.8439 | SRCC = 0.8425 ] +INFO:root:Average loss=0.12102289707996906 +INFO:root: +[Valid][18][UTT][ MSE = 0.2523 | LCC = 0.8437 | SRCC = 0.8437 ] +INFO:root:Average loss=0.11881916048244023 +INFO:root: +[Valid][19][UTT][ MSE = 0.2668 | LCC = 0.8372 | SRCC = 0.8380 ] +INFO:root:Average loss=0.11221206933259964 +INFO:root: +[Valid][20][UTT][ MSE = 0.2895 | LCC = 0.8365 | SRCC = 0.8381 ] +INFO:root:Average loss=0.11057525739455834 +INFO:root: +[Valid][21][UTT][ MSE = 0.2410 | LCC = 0.8466 | SRCC = 0.8455 ] +INFO:root:Average loss=0.10402307955500407 +INFO:root: +[Valid][22][UTT][ MSE = 0.2445 | LCC = 0.8407 | SRCC = 0.8400 ] +INFO:root:Average loss=0.09566213868749447 +INFO:root: +[Valid][23][UTT][ MSE = 0.2373 | LCC = 0.8483 | SRCC = 0.8476 ] +INFO:root:Average loss=0.09483862381715041 +INFO:root: +[Valid][24][UTT][ MSE = 0.2382 | LCC = 0.8488 | SRCC = 0.8483 ] +INFO:root:Average loss=0.0921592310262032 +INFO:root: +[Valid][25][UTT][ MSE = 0.2391 | LCC = 0.8483 | SRCC = 0.8481 ] +INFO:root:Average loss=0.09137299422843334 +INFO:root: +[Valid][26][UTT][ MSE = 0.2425 | LCC = 0.8483 | SRCC = 0.8472 ] +INFO:root:Average loss=0.09923075469067463 +INFO:root: +[Valid][27][UTT][ MSE = 0.2423 | LCC = 0.8479 | SRCC = 0.8468 ] +INFO:root:Average loss=0.0957482553158815 +INFO:root: +[Valid][28][UTT][ MSE = 0.2342 | LCC = 0.8486 | SRCC = 0.8469 ] +INFO:root:Average loss=0.09040134471769516 +INFO:root: +[Valid][29][UTT][ MSE = 0.2424 | LCC = 0.8489 | SRCC = 0.8475 ] +INFO:root:Average loss=0.09167999673921329 +INFO:root: +[Valid][30][UTT][ MSE = 0.2373 | LCC = 0.8456 | SRCC = 0.8455 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2541 | LCC = 0.8581 | SRCC = 0.8569 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..1f9f568751a72d147e21de2c4930a6dad94413ae --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdf14a03699c301bfe6c54934da4076676c8933b5202cd4b1f56ee79dc35576f +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..b393c48ebdd8177103104c0bb5d2f34ed46f04e1 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d4b26ed1e6b94224ca7709ea4d16793b9740f08a6dcb583ee556160762d68d89 +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run2/train.log new file mode 100644 index 0000000000000000000000000000000000000000..428f7b14f9148587cbb0fcd5d6ffcf2b755444a7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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=0.5801382030432041 +INFO:root: +[Valid][0][UTT][ MSE = 0.4031 | LCC = 0.7398 | SRCC = 0.7462 ] +INFO:root:Average loss=0.3125178557940019 +INFO:root: +[Valid][1][UTT][ MSE = 0.3130 | LCC = 0.8073 | SRCC = 0.8076 ] +INFO:root:Average loss=0.2592305208627994 +INFO:root: +[Valid][2][UTT][ MSE = 0.2871 | LCC = 0.8197 | SRCC = 0.8193 ] +INFO:root:Average loss=0.2390029869782619 +INFO:root: +[Valid][3][UTT][ MSE = 0.2951 | LCC = 0.8321 | SRCC = 0.8308 ] +INFO:root:Average loss=0.22948138893414766 +INFO:root: +[Valid][4][UTT][ MSE = 0.3179 | LCC = 0.8267 | SRCC = 0.8274 ] +INFO:root:Average loss=0.22635673445004684 +INFO:root: +[Valid][5][UTT][ MSE = 0.2763 | LCC = 0.8351 | SRCC = 0.8355 ] +INFO:root:Average loss=0.20311911022051787 +INFO:root: +[Valid][6][UTT][ MSE = 0.2957 | LCC = 0.8371 | SRCC = 0.8388 ] +INFO:root:Average loss=0.1956417094438504 +INFO:root: +[Valid][7][UTT][ MSE = 0.2596 | LCC = 0.8397 | SRCC = 0.8398 ] +INFO:root:Average loss=0.1830768769559188 +INFO:root: +[Valid][8][UTT][ MSE = 0.2782 | LCC = 0.8408 | SRCC = 0.8411 ] +INFO:root:Average loss=0.17033110635402876 +INFO:root: +[Valid][9][UTT][ MSE = 0.2535 | LCC = 0.8415 | SRCC = 0.8424 ] +INFO:root:Average loss=0.16839494288731843 +INFO:root: +[Valid][10][UTT][ MSE = 0.2465 | LCC = 0.8443 | SRCC = 0.8445 ] +INFO:root:Average loss=0.15932525894962823 +INFO:root: +[Valid][11][UTT][ MSE = 0.2481 | LCC = 0.8459 | SRCC = 0.8470 ] +INFO:root:Average loss=0.15391197017370126 +INFO:root: +[Valid][12][UTT][ MSE = 0.2463 | LCC = 0.8500 | SRCC = 0.8499 ] +INFO:root:Average loss=0.14536149551471075 +INFO:root: +[Valid][13][UTT][ MSE = 0.2373 | LCC = 0.8493 | SRCC = 0.8497 ] +INFO:root:Average loss=0.13988079369450226 +INFO:root: +[Valid][14][UTT][ MSE = 0.2391 | LCC = 0.8487 | SRCC = 0.8478 ] +INFO:root:Average loss=0.13332303642080381 +INFO:root: +[Valid][15][UTT][ MSE = 0.2512 | LCC = 0.8487 | SRCC = 0.8480 ] +INFO:root:Average loss=0.1234085535009702 +INFO:root: +[Valid][16][UTT][ MSE = 0.2438 | LCC = 0.8512 | SRCC = 0.8507 ] +INFO:root:Average loss=0.12454267285573177 +INFO:root: +[Valid][17][UTT][ MSE = 0.2333 | LCC = 0.8494 | SRCC = 0.8498 ] +INFO:root:Average loss=0.1190307033367646 +INFO:root: +[Valid][18][UTT][ MSE = 0.2345 | LCC = 0.8505 | SRCC = 0.8499 ] +INFO:root:Average loss=0.10688479225604962 +INFO:root: +[Valid][19][UTT][ MSE = 0.2448 | LCC = 0.8498 | SRCC = 0.8483 ] +INFO:root:Average loss=0.11009916070944224 +INFO:root: +[Valid][20][UTT][ MSE = 0.2508 | LCC = 0.8455 | SRCC = 0.8459 ] +INFO:root:Average loss=0.10798309991757075 +INFO:root: +[Valid][21][UTT][ MSE = 0.2485 | LCC = 0.8480 | SRCC = 0.8504 ] +INFO:root:Average loss=0.09596960767148396 +INFO:root: +[Valid][22][UTT][ MSE = 0.2411 | LCC = 0.8492 | SRCC = 0.8476 ] +INFO:root:Average loss=0.09674872419773003 +INFO:root: +[Valid][23][UTT][ MSE = 0.2342 | LCC = 0.8504 | SRCC = 0.8496 ] +INFO:root:Average loss=0.09684574422545922 +INFO:root: +[Valid][24][UTT][ MSE = 0.2359 | LCC = 0.8510 | SRCC = 0.8493 ] +INFO:root:Average loss=0.08903854903884423 +INFO:root: +[Valid][25][UTT][ MSE = 0.2493 | LCC = 0.8474 | SRCC = 0.8464 ] +INFO:root:Average loss=0.0900917380857162 +INFO:root: +[Valid][26][UTT][ MSE = 0.2377 | LCC = 0.8497 | SRCC = 0.8485 ] +INFO:root:Average loss=0.08162814350082324 +INFO:root: +[Valid][27][UTT][ MSE = 0.2729 | LCC = 0.8466 | SRCC = 0.8442 ] +INFO:root:Average loss=0.07885252412122029 +INFO:root: +[Valid][28][UTT][ MSE = 0.2479 | LCC = 0.8449 | SRCC = 0.8443 ] +INFO:root:Average loss=0.08032448733082184 +INFO:root: +[Valid][29][UTT][ MSE = 0.2594 | LCC = 0.8503 | SRCC = 0.8495 ] +INFO:root:Average loss=0.08011147542259632 +INFO:root: +[Valid][30][UTT][ MSE = 0.2363 | LCC = 0.8482 | SRCC = 0.8472 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2476 | LCC = 0.8582 | SRCC = 0.8621 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..68d1f64ced4da3e24ad9fc1262513d804b72d31b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bbe735d6a4b7fca3105ba7e80c5c769f95ac64034dab2046d21fd48ae6ebd659 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..4def705843fba4283ee8ce591f04ad7c66f565a7 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66194f399bc2f97e3d78d1ac83587a344f3131d1150774c65e555e442cd2a5ad +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run3/train.log new file mode 100644 index 0000000000000000000000000000000000000000..df7f14d1a0ae0f577a678c348a324515dbd7aaf2 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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=0.5343722368662174 +INFO:root: +[Valid][0][UTT][ MSE = 0.3944 | LCC = 0.7558 | SRCC = 0.7579 ] +INFO:root:Average loss=0.30632266306724304 +INFO:root: +[Valid][1][UTT][ MSE = 0.3285 | LCC = 0.7863 | SRCC = 0.7914 ] +INFO:root:Average loss=0.26929831351989353 +INFO:root: +[Valid][2][UTT][ MSE = 0.3017 | LCC = 0.8045 | SRCC = 0.8096 ] +INFO:root:Average loss=0.2491036044099392 +INFO:root: +[Valid][3][UTT][ MSE = 0.2993 | LCC = 0.8130 | SRCC = 0.8162 ] +INFO:root:Average loss=0.22121785294551116 +INFO:root: +[Valid][4][UTT][ MSE = 0.2884 | LCC = 0.8218 | SRCC = 0.8214 ] +INFO:root:Average loss=0.21524314200266814 +INFO:root: +[Valid][5][UTT][ MSE = 0.2738 | LCC = 0.8227 | SRCC = 0.8265 ] +INFO:root:Average loss=0.20220068756204385 +INFO:root: +[Valid][6][UTT][ MSE = 0.2931 | LCC = 0.8289 | SRCC = 0.8302 ] +INFO:root:Average loss=0.19279855699875417 +INFO:root: +[Valid][7][UTT][ MSE = 0.2739 | LCC = 0.8306 | SRCC = 0.8314 ] +INFO:root:Average loss=0.18699805161509758 +INFO:root: +[Valid][8][UTT][ MSE = 0.2735 | LCC = 0.8305 | SRCC = 0.8327 ] +INFO:root:Average loss=0.181230313789386 +INFO:root: +[Valid][9][UTT][ MSE = 0.2690 | LCC = 0.8294 | SRCC = 0.8308 ] +INFO:root:Average loss=0.1706711153189341 +INFO:root: +[Valid][10][UTT][ MSE = 0.2723 | LCC = 0.8268 | SRCC = 0.8311 ] +INFO:root:Average loss=0.16383771569683 +INFO:root: +[Valid][11][UTT][ MSE = 0.2603 | LCC = 0.8343 | SRCC = 0.8353 ] +INFO:root:Average loss=0.14910378689185166 +INFO:root: +[Valid][12][UTT][ MSE = 0.2568 | LCC = 0.8317 | SRCC = 0.8326 ] +INFO:root:Average loss=0.14156104691135576 +INFO:root: +[Valid][13][UTT][ MSE = 0.2588 | LCC = 0.8329 | SRCC = 0.8347 ] +INFO:root:Average loss=0.13447169329111391 +INFO:root: +[Valid][14][UTT][ MSE = 0.2612 | LCC = 0.8341 | SRCC = 0.8343 ] +INFO:root:Average loss=0.14698495228703207 +INFO:root: +[Valid][15][UTT][ MSE = 0.2984 | LCC = 0.8294 | SRCC = 0.8330 ] +INFO:root:Average loss=0.1355760684953286 +INFO:root: +[Valid][16][UTT][ MSE = 0.2586 | LCC = 0.8364 | SRCC = 0.8377 ] +INFO:root:Average loss=0.12135763246661578 +INFO:root: +[Valid][17][UTT][ MSE = 0.2597 | LCC = 0.8372 | SRCC = 0.8369 ] +INFO:root:Average loss=0.1142003303154921 +INFO:root: +[Valid][18][UTT][ MSE = 0.2550 | LCC = 0.8441 | SRCC = 0.8422 ] +INFO:root:Average loss=0.11831635313156323 +INFO:root: +[Valid][19][UTT][ MSE = 0.2660 | LCC = 0.8449 | SRCC = 0.8427 ] +INFO:root:Average loss=0.1100625577263343 +INFO:root: +[Valid][20][UTT][ MSE = 0.2693 | LCC = 0.8362 | SRCC = 0.8354 ] +INFO:root:Average loss=0.11267179632798219 +INFO:root: +[Valid][21][UTT][ MSE = 0.2557 | LCC = 0.8369 | SRCC = 0.8361 ] +INFO:root:Average loss=0.10265023347276908 +INFO:root: +[Valid][22][UTT][ MSE = 0.2532 | LCC = 0.8380 | SRCC = 0.8369 ] +INFO:root:Average loss=0.09883993481978393 +INFO:root: +[Valid][23][UTT][ MSE = 0.2447 | LCC = 0.8443 | SRCC = 0.8432 ] +INFO:root:Average loss=0.10375880440458274 +INFO:root: +[Valid][24][UTT][ MSE = 0.2484 | LCC = 0.8436 | SRCC = 0.8428 ] +INFO:root:Average loss=0.09179503843188286 +INFO:root: +[Valid][25][UTT][ MSE = 0.2444 | LCC = 0.8444 | SRCC = 0.8403 ] +INFO:root:Average loss=0.09916901187254833 +INFO:root: +[Valid][26][UTT][ MSE = 0.2416 | LCC = 0.8446 | SRCC = 0.8435 ] +INFO:root:Average loss=0.09760114555366528 +INFO:root: +[Valid][27][UTT][ MSE = 0.2414 | LCC = 0.8426 | SRCC = 0.8427 ] +INFO:root:Average loss=0.08982251369609283 +INFO:root: +[Valid][28][UTT][ MSE = 0.2463 | LCC = 0.8414 | SRCC = 0.8416 ] +INFO:root:Average loss=0.08102281496692927 +INFO:root: +[Valid][29][UTT][ MSE = 0.2449 | LCC = 0.8439 | SRCC = 0.8444 ] +INFO:root:Average loss=0.08595061794114418 +INFO:root: +[Valid][30][UTT][ MSE = 0.2390 | LCC = 0.8462 | SRCC = 0.8454 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2393 | LCC = 0.8592 | SRCC = 0.8610 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..427734d9232d14bb2e23743ca400b0aff696b10f --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dfda1bbebb856c7b623ccbfccbe7277caf0756a78c14d47cf0b266229b95b116 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..7c71b73b21ad4acbea1e2ddeb9c4df5b35e49e27 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e11771852bf1476d153f654901b62a342e8ab284ef7c828484180f920c9508f +size 2749638 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/run4/train.log new file mode 100644 index 0000000000000000000000000000000000000000..7f9bc893efb85aa54a7b661cfe2efc4a981dcde9 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer0/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=0.5699333479771247 +INFO:root: +[Valid][0][UTT][ MSE = 0.4396 | LCC = 0.7247 | SRCC = 0.7323 ] +INFO:root:Average loss=0.31989315954538494 +INFO:root: +[Valid][1][UTT][ MSE = 0.3160 | LCC = 0.7908 | SRCC = 0.7985 ] +INFO:root:Average loss=0.25459787593438077 +INFO:root: +[Valid][2][UTT][ MSE = 0.2846 | LCC = 0.8139 | SRCC = 0.8183 ] +INFO:root:Average loss=0.23338301613544807 +INFO:root: +[Valid][3][UTT][ MSE = 0.2885 | LCC = 0.8289 | SRCC = 0.8303 ] +INFO:root:Average loss=0.2304530443671422 +INFO:root: +[Valid][4][UTT][ MSE = 0.2752 | LCC = 0.8326 | SRCC = 0.8333 ] +INFO:root:Average loss=0.21530579049617815 +INFO:root: +[Valid][5][UTT][ MSE = 0.2637 | LCC = 0.8355 | SRCC = 0.8350 ] +INFO:root:Average loss=0.19580000505233422 +INFO:root: +[Valid][6][UTT][ MSE = 0.2538 | LCC = 0.8425 | SRCC = 0.8410 ] +INFO:root:Average loss=0.18688765779519692 +INFO:root: +[Valid][7][UTT][ MSE = 0.2699 | LCC = 0.8360 | SRCC = 0.8383 ] +INFO:root:Average loss=0.17805368386400053 +INFO:root: +[Valid][8][UTT][ MSE = 0.2481 | LCC = 0.8442 | SRCC = 0.8433 ] +INFO:root:Average loss=0.17730239855173308 +INFO:root: +[Valid][9][UTT][ MSE = 0.2559 | LCC = 0.8394 | SRCC = 0.8403 ] +INFO:root:Average loss=0.1598843300762849 +INFO:root: +[Valid][10][UTT][ MSE = 0.2627 | LCC = 0.8420 | SRCC = 0.8430 ] +INFO:root:Average loss=0.1591207649654303 +INFO:root: +[Valid][11][UTT][ MSE = 0.2912 | LCC = 0.8372 | SRCC = 0.8413 ] +INFO:root:Average loss=0.14937095678387544 +INFO:root: +[Valid][12][UTT][ MSE = 0.2451 | LCC = 0.8427 | SRCC = 0.8432 ] +INFO:root:Average loss=0.14519709062117797 +INFO:root: +[Valid][13][UTT][ MSE = 0.2507 | LCC = 0.8444 | SRCC = 0.8436 ] +INFO:root:Average loss=0.1409055613554441 +INFO:root: +[Valid][14][UTT][ MSE = 0.2655 | LCC = 0.8433 | SRCC = 0.8425 ] +INFO:root:Average loss=0.13867197185754776 +INFO:root: +[Valid][15][UTT][ MSE = 0.2485 | LCC = 0.8447 | SRCC = 0.8435 ] +INFO:root:Average loss=0.12636633838216463 +INFO:root: +[Valid][16][UTT][ MSE = 0.2650 | LCC = 0.8417 | SRCC = 0.8435 ] +INFO:root:Average loss=0.11577655909917293 +INFO:root: +[Valid][17][UTT][ MSE = 0.2615 | LCC = 0.8429 | SRCC = 0.8445 ] +INFO:root:Average loss=0.11223116068121715 +INFO:root: +[Valid][18][UTT][ MSE = 0.2395 | LCC = 0.8463 | SRCC = 0.8460 ] +INFO:root:Average loss=0.11161293920416099 +INFO:root: +[Valid][19][UTT][ MSE = 0.2509 | LCC = 0.8452 | SRCC = 0.8451 ] +INFO:root:Average loss=0.10704566137148784 +INFO:root: +[Valid][20][UTT][ MSE = 0.2490 | LCC = 0.8452 | SRCC = 0.8421 ] +INFO:root:Average loss=0.11756933423189017 +INFO:root: +[Valid][21][UTT][ MSE = 0.2371 | LCC = 0.8497 | SRCC = 0.8490 ] +INFO:root:Average loss=0.09858589098812678 +INFO:root: +[Valid][22][UTT][ MSE = 0.2350 | LCC = 0.8476 | SRCC = 0.8468 ] +INFO:root:Average loss=0.092626216272131 +INFO:root: +[Valid][23][UTT][ MSE = 0.2670 | LCC = 0.8425 | SRCC = 0.8431 ] +INFO:root:Average loss=0.09267581894229619 +INFO:root: +[Valid][24][UTT][ MSE = 0.2660 | LCC = 0.8426 | SRCC = 0.8438 ] +INFO:root:Average loss=0.08651518234266685 +INFO:root: +[Valid][25][UTT][ MSE = 0.2470 | LCC = 0.8403 | SRCC = 0.8398 ] +INFO:root:Average loss=0.08597099594771862 +INFO:root: +[Valid][26][UTT][ MSE = 0.2401 | LCC = 0.8474 | SRCC = 0.8463 ] +INFO:root:Average loss=0.0857331481977151 +INFO:root: +[Valid][27][UTT][ MSE = 0.2406 | LCC = 0.8487 | SRCC = 0.8480 ] +INFO:root:Average loss=0.09086344892588946 +INFO:root: +[Valid][28][UTT][ MSE = 0.2427 | LCC = 0.8502 | SRCC = 0.8502 ] +INFO:root:Average loss=0.08142291190914619 +INFO:root: +[Valid][29][UTT][ MSE = 0.2363 | LCC = 0.8476 | SRCC = 0.8461 ] +INFO:root:Average loss=0.07683732432241623 +INFO:root: +[Valid][30][UTT][ MSE = 0.2593 | LCC = 0.8449 | SRCC = 0.8440 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2325 | LCC = 0.8559 | SRCC = 0.8537 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..0a45711234431c4f928b3bd7917483c2a56d5c57 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1b83aee33791967d08ed4c9caa1c8ea353907f31a66b4f99068a3b7237fc66bd +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..a4e2b557beb3a14efc934efeb42a945e67748d1e --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f27efa24e010d1b975f41b1b0cc03509888c7c5e02ddcdc47142e68c2d81ba1 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run0/train.log new file mode 100644 index 0000000000000000000000000000000000000000..48c912c3164000e90da99f5550353619a46d2e02 --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/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=0.48730611915771777 +INFO:root: +[Valid][0][UTT][ MSE = 0.3429 | LCC = 0.7868 | SRCC = 0.7904 ] +INFO:root:Average loss=0.25775410597905135 +INFO:root: +[Valid][1][UTT][ MSE = 0.2936 | LCC = 0.8297 | SRCC = 0.8328 ] +INFO:root:Average loss=0.21891280894096082 +INFO:root: +[Valid][2][UTT][ MSE = 0.2754 | LCC = 0.8467 | SRCC = 0.8473 ] +INFO:root:Average loss=0.20493962997809434 +INFO:root: +[Valid][3][UTT][ MSE = 0.2730 | LCC = 0.8504 | SRCC = 0.8517 ] +INFO:root:Average loss=0.18952996637194586 +INFO:root: +[Valid][4][UTT][ MSE = 0.2421 | LCC = 0.8577 | SRCC = 0.8592 ] +INFO:root:Average loss=0.18309211883789453 +INFO:root: +[Valid][5][UTT][ MSE = 0.2458 | LCC = 0.8602 | SRCC = 0.8604 ] +INFO:root:Average loss=0.1614677145695075 +INFO:root: +[Valid][6][UTT][ MSE = 0.2810 | LCC = 0.8557 | SRCC = 0.8568 ] +INFO:root:Average loss=0.15303127515392426 +INFO:root: +[Valid][7][UTT][ MSE = 0.2536 | LCC = 0.8607 | SRCC = 0.8621 ] +INFO:root:Average loss=0.13929454762584123 +INFO:root: +[Valid][8][UTT][ MSE = 0.2409 | LCC = 0.8638 | SRCC = 0.8641 ] +INFO:root:Average loss=0.13784939060226464 +INFO:root: +[Valid][9][UTT][ MSE = 0.2254 | LCC = 0.8664 | SRCC = 0.8666 ] +INFO:root:Average loss=0.12484071623438443 +INFO:root: +[Valid][10][UTT][ MSE = 0.2430 | LCC = 0.8623 | SRCC = 0.8634 ] +INFO:root:Average loss=0.11916246246068905 +INFO:root: +[Valid][11][UTT][ MSE = 0.2344 | LCC = 0.8654 | SRCC = 0.8646 ] +INFO:root:Average loss=0.11184625346691181 +INFO:root: +[Valid][12][UTT][ MSE = 0.2229 | LCC = 0.8645 | SRCC = 0.8651 ] +INFO:root:Average loss=0.11221431893033859 +INFO:root: +[Valid][13][UTT][ MSE = 0.2438 | LCC = 0.8633 | SRCC = 0.8628 ] +INFO:root:Average loss=0.10638893099549489 +INFO:root: +[Valid][14][UTT][ MSE = 0.2273 | LCC = 0.8667 | SRCC = 0.8660 ] +INFO:root:Average loss=0.09682618769315574 +INFO:root: +[Valid][15][UTT][ MSE = 0.2283 | LCC = 0.8623 | SRCC = 0.8628 ] +INFO:root:Average loss=0.09853018877597955 +INFO:root: +[Valid][16][UTT][ MSE = 0.2167 | LCC = 0.8677 | SRCC = 0.8674 ] +INFO:root:Average loss=0.08587414795198502 +INFO:root: +[Valid][17][UTT][ MSE = 0.2220 | LCC = 0.8661 | SRCC = 0.8664 ] +INFO:root:Average loss=0.08369163748545524 +INFO:root: +[Valid][18][UTT][ MSE = 0.2343 | LCC = 0.8641 | SRCC = 0.8632 ] +INFO:root:Average loss=0.08181805908679962 +INFO:root: +[Valid][19][UTT][ MSE = 0.2369 | LCC = 0.8617 | SRCC = 0.8626 ] +INFO:root:Average loss=0.07560732401907444 +INFO:root: +[Valid][20][UTT][ MSE = 0.2449 | LCC = 0.8631 | SRCC = 0.8632 ] +INFO:root:Average loss=0.07348189899363579 +INFO:root: +[Valid][21][UTT][ MSE = 0.2205 | LCC = 0.8648 | SRCC = 0.8635 ] +INFO:root:Average loss=0.07080196030437946 +INFO:root: +[Valid][22][UTT][ MSE = 0.2129 | LCC = 0.8649 | SRCC = 0.8651 ] +INFO:root:Average loss=0.07112420455385478 +INFO:root: +[Valid][23][UTT][ MSE = 0.2148 | LCC = 0.8643 | SRCC = 0.8627 ] +INFO:root:Average loss=0.07127113344195561 +INFO:root: +[Valid][24][UTT][ MSE = 0.2303 | LCC = 0.8662 | SRCC = 0.8644 ] +INFO:root:Average loss=0.06404577338924775 +INFO:root: +[Valid][25][UTT][ MSE = 0.2234 | LCC = 0.8636 | SRCC = 0.8641 ] +INFO:root:Average loss=0.06434260814999923 +INFO:root: +[Valid][26][UTT][ MSE = 0.2182 | LCC = 0.8654 | SRCC = 0.8644 ] +INFO:root:Average loss=0.060632982124120764 +INFO:root: +[Valid][27][UTT][ MSE = 0.2401 | LCC = 0.8633 | SRCC = 0.8625 ] +INFO:root:Average loss=0.056822738299767174 +INFO:root: +[Valid][28][UTT][ MSE = 0.2306 | LCC = 0.8631 | SRCC = 0.8623 ] +INFO:root:Average loss=0.05506541919058714 +INFO:root: +[Valid][29][UTT][ MSE = 0.2206 | LCC = 0.8676 | SRCC = 0.8670 ] +INFO:root:Average loss=0.056702594965314254 +INFO:root: +[Valid][30][UTT][ MSE = 0.2159 | LCC = 0.8669 | SRCC = 0.8677 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2269 | LCC = 0.8596 | SRCC = 0.8601 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/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 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model.pt new file mode 100644 index 0000000000000000000000000000000000000000..ab5d184272b62577a17068be433b12c4dff9c98d --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:80fc10368c87ab4c1943fa69a0da80faf973768264d81ef8f3a17dc536c902c0 +size 2747979 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model_best.pt b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model_best.pt new file mode 100644 index 0000000000000000000000000000000000000000..62bf6c9eed30f049941e890641837abc9668d5ab --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/model_best.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b34ffa03d0fef267efc5a03e4b48b3ad34850e36c58a1d33f969b4453f9421d2 +size 2749702 diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/train.log b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run1/train.log new file mode 100644 index 0000000000000000000000000000000000000000..5cf20235a812e583c033002e8157ed2d4496b85b --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/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=0.403356230029693 +INFO:root: +[Valid][0][UTT][ MSE = 0.3458 | LCC = 0.8019 | SRCC = 0.8079 ] +INFO:root:Average loss=0.24504654147686103 +INFO:root: +[Valid][1][UTT][ MSE = 0.3074 | LCC = 0.8392 | SRCC = 0.8392 ] +INFO:root:Average loss=0.2212523239163252 +INFO:root: +[Valid][2][UTT][ MSE = 0.2726 | LCC = 0.8517 | SRCC = 0.8513 ] +INFO:root:Average loss=0.1919358273347219 +INFO:root: +[Valid][3][UTT][ MSE = 0.2663 | LCC = 0.8574 | SRCC = 0.8555 ] +INFO:root:Average loss=0.18206220702865186 +INFO:root: +[Valid][4][UTT][ MSE = 0.2564 | LCC = 0.8603 | SRCC = 0.8603 ] +INFO:root:Average loss=0.16380356882627195 +INFO:root: +[Valid][5][UTT][ MSE = 0.2320 | LCC = 0.8540 | SRCC = 0.8554 ] +INFO:root:Average loss=0.16085966294392562 +INFO:root: +[Valid][6][UTT][ MSE = 0.2394 | LCC = 0.8651 | SRCC = 0.8647 ] +INFO:root:Average loss=0.1436474424524185 +INFO:root: +[Valid][7][UTT][ MSE = 0.2280 | LCC = 0.8665 | SRCC = 0.8658 ] +INFO:root:Average loss=0.13650570857601288 +INFO:root: +[Valid][8][UTT][ MSE = 0.2187 | LCC = 0.8632 | SRCC = 0.8630 ] +INFO:root:Average loss=0.1265812458900305 +INFO:root: +[Valid][9][UTT][ MSE = 0.2271 | LCC = 0.8659 | SRCC = 0.8659 ] +INFO:root:Average loss=0.12167770673449223 +INFO:root: +[Valid][10][UTT][ MSE = 0.2265 | LCC = 0.8644 | SRCC = 0.8617 ] +INFO:root:Average loss=0.10690382180305627 +INFO:root: +[Valid][11][UTT][ MSE = 0.2306 | LCC = 0.8630 | SRCC = 0.8636 ] +INFO:root:Average loss=0.10681357177404258 +INFO:root: +[Valid][12][UTT][ MSE = 0.2147 | LCC = 0.8689 | SRCC = 0.8693 ] +INFO:root:Average loss=0.09842756686684413 +INFO:root: +[Valid][13][UTT][ MSE = 0.2134 | LCC = 0.8661 | SRCC = 0.8651 ] +INFO:root:Average loss=0.09208398469938682 +INFO:root: +[Valid][14][UTT][ MSE = 0.2364 | LCC = 0.8658 | SRCC = 0.8663 ] +INFO:root:Average loss=0.08977004651648876 +INFO:root: +[Valid][15][UTT][ MSE = 0.2188 | LCC = 0.8675 | SRCC = 0.8671 ] +INFO:root:Average loss=0.08824189088474481 +INFO:root: +[Valid][16][UTT][ MSE = 0.2103 | LCC = 0.8672 | SRCC = 0.8663 ] +INFO:root:Average loss=0.08159601664504944 +INFO:root: +[Valid][17][UTT][ MSE = 0.2136 | LCC = 0.8668 | SRCC = 0.8663 ] +INFO:root:Average loss=0.08364848324503654 +INFO:root: +[Valid][18][UTT][ MSE = 0.2188 | LCC = 0.8679 | SRCC = 0.8672 ] +INFO:root:Average loss=0.07345642732122006 +INFO:root: +[Valid][19][UTT][ MSE = 0.2215 | LCC = 0.8674 | SRCC = 0.8669 ] +INFO:root:Average loss=0.06951262395924483 +INFO:root: +[Valid][20][UTT][ MSE = 0.2171 | LCC = 0.8685 | SRCC = 0.8672 ] +INFO:root:Average loss=0.0685470396031936 +INFO:root: +[Valid][21][UTT][ MSE = 0.2104 | LCC = 0.8664 | SRCC = 0.8656 ] +INFO:root:Average loss=0.06710382741995347 +INFO:root: +[Valid][22][UTT][ MSE = 0.2155 | LCC = 0.8654 | SRCC = 0.8655 ] +INFO:root:Average loss=0.06758066519903831 +INFO:root: +[Valid][23][UTT][ MSE = 0.2049 | LCC = 0.8690 | SRCC = 0.8684 ] +INFO:root:Average loss=0.061953793255946576 +INFO:root: +[Valid][24][UTT][ MSE = 0.2184 | LCC = 0.8698 | SRCC = 0.8687 ] +INFO:root:Average loss=0.06258088641632827 +INFO:root: +[Valid][25][UTT][ MSE = 0.2269 | LCC = 0.8646 | SRCC = 0.8636 ] +INFO:root:Average loss=0.05940822564447538 +INFO:root: +[Valid][26][UTT][ MSE = 0.2072 | LCC = 0.8677 | SRCC = 0.8661 ] +INFO:root:Average loss=0.055898898639358006 +INFO:root: +[Valid][27][UTT][ MSE = 0.2153 | LCC = 0.8674 | SRCC = 0.8671 ] +INFO:root:Average loss=0.0554384434930025 +INFO:root: +[Valid][28][UTT][ MSE = 0.2156 | LCC = 0.8651 | SRCC = 0.8646 ] +INFO:root:Average loss=0.05663360583667572 +INFO:root: +[Valid][29][UTT][ MSE = 0.2206 | LCC = 0.8684 | SRCC = 0.8675 ] +INFO:root:Average loss=0.05260227128672294 +INFO:root: +[Valid][30][UTT][ MSE = 0.2750 | LCC = 0.8664 | SRCC = 0.8668 ] +INFO:root: +[Test][31][UTT][ MSE = 0.2660 | LCC = 0.8548 | SRCC = 0.8542 ] diff --git a/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run2/config.gin b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/run2/config.gin new file mode 100644 index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa --- /dev/null +++ b/results_bvcc/w2v2_base_bvcc_finetuned_6epoch/layer1/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