diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer10/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f25c2a754b5c4e60e4d9ad76350be4633e2423fd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17c954b27d304f577e9fd998036c8afa372b2fb1bb1ecb167af33267fe87ebd6
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer10/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a03da0b2b2bbce08897556fdc24e120cb4e82750
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b982e79fea04f796dbed29dabf4b3392178d6f9164e38998293ad06c9387d387
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer10/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..4de38edadbe779a38e42cde04046ab44c068fc93
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.6746184119047263
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4778 | LCC = 0.7159 | SRCC = 0.7216 ]
+INFO:root:Average loss=0.3387069020133752
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3338 | LCC = 0.8053 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.2847767711067811
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3085 | LCC = 0.8174 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.26873792765232235
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3536 | LCC = 0.8174 | SRCC = 0.8175 ]
+INFO:root:Average loss=0.23792060521932748
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3200 | LCC = 0.8162 | SRCC = 0.8128 ]
+INFO:root:Average loss=0.2266076065790959
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3025 | LCC = 0.8133 | SRCC = 0.8105 ]
+INFO:root:Average loss=0.21253541111946106
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3221 | LCC = 0.8146 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.19001203135420114
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3453 | LCC = 0.8117 | SRCC = 0.8131 ]
+INFO:root:Average loss=0.17686052477130523
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2949 | LCC = 0.8136 | SRCC = 0.8142 ]
+INFO:root:Average loss=0.16311080323962066
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3003 | LCC = 0.8100 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.1606394347663109
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3156 | LCC = 0.8080 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.14482228400615546
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3110 | LCC = 0.8137 | SRCC = 0.8113 ]
+INFO:root:Average loss=0.14077927630681258
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2824 | LCC = 0.8181 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.1233119808901579
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2920 | LCC = 0.8107 | SRCC = 0.8101 ]
+INFO:root:Average loss=0.12160828308417247
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2835 | LCC = 0.8124 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.11040873868534198
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2900 | LCC = 0.8126 | SRCC = 0.8101 ]
+INFO:root:Average loss=0.10938038973089977
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2914 | LCC = 0.8137 | SRCC = 0.8119 ]
+INFO:root:Average loss=0.10012815907024421
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3084 | LCC = 0.8083 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.10388120755744286
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2943 | LCC = 0.8107 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.08978882174079235
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3115 | LCC = 0.8074 | SRCC = 0.8049 ]
+INFO:root:Average loss=0.08635260518162678
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3136 | LCC = 0.8057 | SRCC = 0.8069 ]
+INFO:root:Average loss=0.07857741916982028
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2885 | LCC = 0.8124 | SRCC = 0.8121 ]
+INFO:root:Average loss=0.08343375440782462
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3734 | LCC = 0.8007 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.08682268590499194
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3107 | LCC = 0.8072 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.07026724560329548
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2972 | LCC = 0.8077 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.0699538896099115
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3158 | LCC = 0.8078 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.0652948336628003
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2908 | LCC = 0.8137 | SRCC = 0.8124 ]
+INFO:root:Average loss=0.06554413371934341
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2928 | LCC = 0.8084 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.06051074364819588
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3119 | LCC = 0.8020 | SRCC = 0.8009 ]
+INFO:root:Average loss=0.0663526530067126
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3253 | LCC = 0.8010 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.06058197826720201
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3325 | LCC = 0.8038 | SRCC = 0.8048 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2816 | LCC = 0.8313 | SRCC = 0.8339 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer10/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer10/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer10/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0f62c365ab1641cafc99d494c4087b713587a617
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a9677dbfb8f2aad402e8378478a102d4599225892a987b760e678b5ed4ccc42
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer10/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0ca70e82213728e46ffc7fec52abb524c728ad8f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bef1cdb12deabd38fd53fc24ae27e1f98b2f9318a55b33d24b2f0916d0fb429c
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer10/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..fc4adc103102a1d1bd4f98fc7e8f4f3410bdbb25
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5988704699736375
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4256 | LCC = 0.7561 | SRCC = 0.7585 ]
+INFO:root:Average loss=0.3351873986614056
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3306 | LCC = 0.8009 | SRCC = 0.8010 ]
+INFO:root:Average loss=0.2886387335184293
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3175 | LCC = 0.8080 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.2622057657975417
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3165 | LCC = 0.8089 | SRCC = 0.8082 ]
+INFO:root:Average loss=0.2366573931888128
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2980 | LCC = 0.8144 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.22310085451373687
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3147 | LCC = 0.8087 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.23775893927384645
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2993 | LCC = 0.8173 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.19699565645975944
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2875 | LCC = 0.8165 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.17750538675448832
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3055 | LCC = 0.8035 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.18402894739157113
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2866 | LCC = 0.8115 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.15437620715835157
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3052 | LCC = 0.8091 | SRCC = 0.8077 ]
+INFO:root:Average loss=0.15082182897589144
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3023 | LCC = 0.8042 | SRCC = 0.8035 ]
+INFO:root:Average loss=0.14588640639797235
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3007 | LCC = 0.8087 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.13715299237997103
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3086 | LCC = 0.8106 | SRCC = 0.8090 ]
+INFO:root:Average loss=0.12994531336694193
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2964 | LCC = 0.8083 | SRCC = 0.8081 ]
+INFO:root:Average loss=0.11570002855016635
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2924 | LCC = 0.8070 | SRCC = 0.8052 ]
+INFO:root:Average loss=0.10925124208323467
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2987 | LCC = 0.8022 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.10947629188497861
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2971 | LCC = 0.8047 | SRCC = 0.8042 ]
+INFO:root:Average loss=0.11201404579556905
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3016 | LCC = 0.8077 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.09224934384035759
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2914 | LCC = 0.8080 | SRCC = 0.8071 ]
+INFO:root:Average loss=0.08938650796428704
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3218 | LCC = 0.8049 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.09217797339153595
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3163 | LCC = 0.8044 | SRCC = 0.8039 ]
+INFO:root:Average loss=0.08819882896466133
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3013 | LCC = 0.8039 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.07946886528187837
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2983 | LCC = 0.8056 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.07416030201010215
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3147 | LCC = 0.8069 | SRCC = 0.8071 ]
+INFO:root:Average loss=0.07000843511942105
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3102 | LCC = 0.8058 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.07103851313392322
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3075 | LCC = 0.8059 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.07397129558599912
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3003 | LCC = 0.8026 | SRCC = 0.8048 ]
+INFO:root:Average loss=0.0688283399034005
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2973 | LCC = 0.8031 | SRCC = 0.8037 ]
+INFO:root:Average loss=0.06074103789451795
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3053 | LCC = 0.8039 | SRCC = 0.8037 ]
+INFO:root:Average loss=0.05725609396512692
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3043 | LCC = 0.8031 | SRCC = 0.8028 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2946 | LCC = 0.8287 | SRCC = 0.8330 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer10/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer10/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer10/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8ae4092d026c6c8135266026b299e2453917a201
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43b280e478e0984619e04a636aa21c3ca58b59766f3a1b907a3e1a28a2bf5628
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer10/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..de4dbd49e36d342bf4eb1e3df56c845c2557ea06
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer10/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0908d121f460742f853bf06173266c352b2723d6d316379565c5b860f3781a2
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer10/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer10/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..38e0688f49d31076c54bc20a08699ec312df0114
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.6203618072546445
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4357 | LCC = 0.7424 | SRCC = 0.7445 ]
+INFO:root:Average loss=0.32978219195054126
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3305 | LCC = 0.8015 | SRCC = 0.8003 ]
+INFO:root:Average loss=0.2830068669640101
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3209 | LCC = 0.8173 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.2473831717402507
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3117 | LCC = 0.8149 | SRCC = 0.8109 ]
+INFO:root:Average loss=0.2436703187532914
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3327 | LCC = 0.8152 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.22828983591917235
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3274 | LCC = 0.8206 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.19720400258516654
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2941 | LCC = 0.8195 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.199496636310449
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2920 | LCC = 0.8163 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.17800079887876144
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2984 | LCC = 0.8118 | SRCC = 0.8102 ]
+INFO:root:Average loss=0.16389273231228194
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2954 | LCC = 0.8107 | SRCC = 0.8117 ]
+INFO:root:Average loss=0.15842864261223719
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3438 | LCC = 0.8145 | SRCC = 0.8152 ]
+INFO:root:Average loss=0.1473681323039226
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3195 | LCC = 0.8119 | SRCC = 0.8122 ]
+INFO:root:Average loss=0.14767956275206345
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2986 | LCC = 0.8129 | SRCC = 0.8106 ]
+INFO:root:Average loss=0.12661838818054932
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3090 | LCC = 0.8052 | SRCC = 0.8046 ]
+INFO:root:Average loss=0.12022751426467529
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2973 | LCC = 0.8087 | SRCC = 0.8068 ]
+INFO:root:Average loss=0.11529687218941174
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2894 | LCC = 0.8115 | SRCC = 0.8105 ]
+INFO:root:Average loss=0.11008151047504865
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3617 | LCC = 0.8013 | SRCC = 0.8013 ]
+INFO:root:Average loss=0.10258800918475175
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3020 | LCC = 0.8053 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.09492622760052864
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2955 | LCC = 0.8096 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.09255884143595512
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3150 | LCC = 0.8041 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.0843467849951524
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2996 | LCC = 0.8023 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.0875364528157008
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3082 | LCC = 0.8034 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.08333408980606458
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3099 | LCC = 0.8018 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.0776115395128727
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3091 | LCC = 0.7993 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.07417628866357681
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3151 | LCC = 0.7950 | SRCC = 0.7930 ]
+INFO:root:Average loss=0.07346136371294658
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3040 | LCC = 0.8005 | SRCC = 0.7995 ]
+INFO:root:Average loss=0.06646742208454853
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3142 | LCC = 0.8004 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.06499218501341648
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3218 | LCC = 0.8002 | SRCC = 0.8004 ]
+INFO:root:Average loss=0.06706518684633267
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3453 | LCC = 0.7962 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.06496112478467134
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3409 | LCC = 0.8005 | SRCC = 0.8014 ]
+INFO:root:Average loss=0.05908839113246172
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3119 | LCC = 0.7975 | SRCC = 0.7962 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3395 | LCC = 0.8158 | SRCC = 0.8257 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer12/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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_large_asr_en/layer12/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer12/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ee4e44b3e2dafe53860fab012eace8676c227bc2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9627174152b51416edf38ab0aaeb16341f92d0a3eae0b4a5fea045918080aa2
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer12/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c5360dd4d09c13b254079e6fdb50b19cff1afb55
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b3b06a62fff74787b637a935ffe19ad02e0bfb8a2f69a6353632c73c4a258b3
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer12/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..afc49b086fdf334f7980abea8031b04497c36af6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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=0.5967773099740347
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5083 | LCC = 0.7323 | SRCC = 0.7392 ]
+INFO:root:Average loss=0.34689008731108445
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3440 | LCC = 0.7944 | SRCC = 0.7930 ]
+INFO:root:Average loss=0.3308933778451039
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3955 | LCC = 0.7990 | SRCC = 0.8004 ]
+INFO:root:Average loss=0.2722178880984967
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3819 | LCC = 0.8036 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.2590073903019612
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3180 | LCC = 0.8129 | SRCC = 0.8098 ]
+INFO:root:Average loss=0.23053675374159446
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3261 | LCC = 0.8118 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.21345278983696914
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3180 | LCC = 0.8190 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.20548403683381203
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3376 | LCC = 0.8126 | SRCC = 0.8108 ]
+INFO:root:Average loss=0.20480541846691033
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3139 | LCC = 0.8094 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.17525660494963327
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2926 | LCC = 0.8121 | SRCC = 0.8099 ]
+INFO:root:Average loss=0.17095294269995812
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2990 | LCC = 0.8079 | SRCC = 0.8063 ]
+INFO:root:Average loss=0.15322914127355966
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3106 | LCC = 0.8120 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.14001862017007974
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3499 | LCC = 0.8112 | SRCC = 0.8095 ]
+INFO:root:Average loss=0.15172623604153976
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3013 | LCC = 0.8063 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.14074587439879394
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3060 | LCC = 0.8110 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.12415926702893697
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2993 | LCC = 0.8074 | SRCC = 0.8067 ]
+INFO:root:Average loss=0.1146744348299809
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3094 | LCC = 0.8033 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.11406851951510479
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3587 | LCC = 0.7996 | SRCC = 0.7986 ]
+INFO:root:Average loss=0.10659033136490063
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3100 | LCC = 0.8036 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.09453476110521035
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3329 | LCC = 0.8067 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.08874362950714734
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2957 | LCC = 0.8076 | SRCC = 0.8051 ]
+INFO:root:Average loss=0.09242981037077232
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3054 | LCC = 0.8065 | SRCC = 0.8043 ]
+INFO:root:Average loss=0.07949087367607997
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3520 | LCC = 0.8035 | SRCC = 0.8004 ]
+INFO:root:Average loss=0.07928717561448231
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2984 | LCC = 0.8094 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.08720068447291851
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3126 | LCC = 0.8090 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.07371282224089672
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3219 | LCC = 0.8051 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.07177434422266789
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2913 | LCC = 0.8100 | SRCC = 0.8077 ]
+INFO:root:Average loss=0.06913609372881743
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3401 | LCC = 0.8019 | SRCC = 0.7991 ]
+INFO:root:Average loss=0.0761683588513197
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3065 | LCC = 0.8055 | SRCC = 0.8031 ]
+INFO:root:Average loss=0.0654251968058256
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3148 | LCC = 0.8042 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.06390051720425105
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3053 | LCC = 0.8099 | SRCC = 0.8086 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2847 | LCC = 0.8197 | SRCC = 0.8195 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer12/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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_large_asr_en/layer12/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer12/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3c989a91500c9c746608134130ff67bc6de92191
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf84ab06f58f2a54803a6a19b3781544137bdae1ecba533d2c3d8b94ea169253
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer12/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..18bc8f67a57a09b8171bd11703f30b62da8689ac
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3196ed7c16941f06cf9acdbc5fae9e8942902b9695a99be04a78db14bdab4d47
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer12/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..86cf7ac6b986ad889eb63a38130555144d0c2335
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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=0.6111231052722687
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4574 | LCC = 0.7163 | SRCC = 0.7186 ]
+INFO:root:Average loss=0.3580839347380858
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4009 | LCC = 0.7808 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.30512720556595385
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3525 | LCC = 0.8025 | SRCC = 0.8021 ]
+INFO:root:Average loss=0.2711172887148001
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3829 | LCC = 0.8011 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.2620871812105179
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4072 | LCC = 0.8052 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.2431554207817102
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3058 | LCC = 0.8136 | SRCC = 0.8097 ]
+INFO:root:Average loss=0.2137223243331298
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3034 | LCC = 0.8146 | SRCC = 0.8108 ]
+INFO:root:Average loss=0.2017128905042624
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3873 | LCC = 0.8064 | SRCC = 0.8097 ]
+INFO:root:Average loss=0.18478004701244524
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2929 | LCC = 0.8128 | SRCC = 0.8109 ]
+INFO:root:Average loss=0.17394910303828043
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2902 | LCC = 0.8135 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.1696313469646833
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3878 | LCC = 0.8061 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.17139079574591073
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3245 | LCC = 0.8067 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.13853010191367224
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3798 | LCC = 0.7954 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.13903944729230341
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3053 | LCC = 0.8020 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.12900117956675017
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2957 | LCC = 0.8049 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.12103551416060863
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3212 | LCC = 0.8053 | SRCC = 0.8034 ]
+INFO:root:Average loss=0.11534078113543682
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3003 | LCC = 0.8066 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.10717813364970379
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3113 | LCC = 0.8017 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.10493040657960452
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2933 | LCC = 0.8079 | SRCC = 0.8048 ]
+INFO:root:Average loss=0.0925721837541996
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3083 | LCC = 0.8038 | SRCC = 0.7998 ]
+INFO:root:Average loss=0.08859112156698337
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3068 | LCC = 0.7973 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.09077216226321
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3093 | LCC = 0.8010 | SRCC = 0.7955 ]
+INFO:root:Average loss=0.08821797834183925
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3173 | LCC = 0.8019 | SRCC = 0.8002 ]
+INFO:root:Average loss=0.07885123287829068
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3267 | LCC = 0.7952 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.07573024904689728
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3180 | LCC = 0.7988 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.06840286609262992
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3259 | LCC = 0.7977 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.06431310022106537
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3008 | LCC = 0.8013 | SRCC = 0.7996 ]
+INFO:root:Average loss=0.06694942450103088
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3120 | LCC = 0.7936 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.061848485173705295
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2987 | LCC = 0.8031 | SRCC = 0.8002 ]
+INFO:root:Average loss=0.05977314920761646
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3274 | LCC = 0.7991 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.05841148921694511
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3039 | LCC = 0.8013 | SRCC = 0.7976 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2870 | LCC = 0.8221 | SRCC = 0.8224 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer12/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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_large_asr_en/layer12/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer12/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a393830fc28e1bf6e96fcf1e62b6535fa7b61612
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04f9c1c017313ab3328b7a6db815c9180fbf9bd7746903a73d48991660515a71
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer12/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6a797b49ca366cb1bcb87e47a02ff9948c40c27a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e50e5dd0821a67cae239020a0d55a9bd863f2b5c1b7db8d8f732ccf9dd190188
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer12/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..d76709f0ce349dde011c9411f98bf13927d2d6a9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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=0.6102118430993496
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4486 | LCC = 0.7210 | SRCC = 0.7226 ]
+INFO:root:Average loss=0.3598540024115489
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3737 | LCC = 0.7838 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.29358902401649034
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3682 | LCC = 0.7965 | SRCC = 0.7936 ]
+INFO:root:Average loss=0.2608952222344203
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3139 | LCC = 0.8086 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.24057510724434486
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3437 | LCC = 0.8072 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.22202213547932795
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3303 | LCC = 0.8044 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.22276107336466128
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3215 | LCC = 0.8071 | SRCC = 0.8033 ]
+INFO:root:Average loss=0.19395450770090789
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2991 | LCC = 0.8101 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.18139779625030664
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3013 | LCC = 0.8105 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.16541293788796815
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3383 | LCC = 0.8089 | SRCC = 0.8058 ]
+INFO:root:Average loss=0.15609066302959734
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3098 | LCC = 0.8048 | SRCC = 0.8012 ]
+INFO:root:Average loss=0.14942565073187536
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3018 | LCC = 0.8046 | SRCC = 0.8013 ]
+INFO:root:Average loss=0.1444778493963755
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3218 | LCC = 0.8104 | SRCC = 0.8074 ]
+INFO:root:Average loss=0.12929204392891663
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3533 | LCC = 0.8064 | SRCC = 0.8051 ]
+INFO:root:Average loss=0.12945508059018698
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3040 | LCC = 0.8066 | SRCC = 0.8034 ]
+INFO:root:Average loss=0.1215602186245796
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3055 | LCC = 0.8009 | SRCC = 0.7961 ]
+INFO:root:Average loss=0.10900326655843319
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2961 | LCC = 0.8056 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.1064690011434066
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3581 | LCC = 0.8007 | SRCC = 0.8006 ]
+INFO:root:Average loss=0.10853447392582893
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3901 | LCC = 0.8002 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.08927819065940686
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3061 | LCC = 0.8089 | SRCC = 0.8049 ]
+INFO:root:Average loss=0.08615353049185032
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3125 | LCC = 0.8016 | SRCC = 0.7982 ]
+INFO:root:Average loss=0.0913308357867675
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2913 | LCC = 0.8074 | SRCC = 0.8046 ]
+INFO:root:Average loss=0.07851233495733677
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3270 | LCC = 0.7990 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.07828337899767436
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3264 | LCC = 0.8047 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.07540841166598675
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2951 | LCC = 0.8084 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.0709887038056667
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3058 | LCC = 0.8052 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.07145582717389633
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3043 | LCC = 0.8051 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.06620600533026916
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3009 | LCC = 0.8102 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.06880018468468617
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3437 | LCC = 0.8032 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.06038604270762358
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3027 | LCC = 0.8078 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.05491556771672689
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3178 | LCC = 0.8009 | SRCC = 0.8005 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2999 | LCC = 0.8180 | SRCC = 0.8209 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer12/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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_large_asr_en/layer12/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer12/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5aa1d6ad224d039e25c49ab1f84413db2bc52349
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ead20be6dba04f1b5410d68166d64d5cff49fbe9a78bbfda959b188c7f6222b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer12/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9680243f0f95e367f09874e996f1d5825d6435b7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b2d9689b85d9f33cec017b4875d26aa8b002a174bb54cf0aa34e88f48d904b9
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer12/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..46f6c342a9ece8d63e8e3ae4f7d946f387d8bebe
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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=0.7038338230206416
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4931 | LCC = 0.6788 | SRCC = 0.6801 ]
+INFO:root:Average loss=0.37384560455878574
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3693 | LCC = 0.7838 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.2999933856802109
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3384 | LCC = 0.8002 | SRCC = 0.7987 ]
+INFO:root:Average loss=0.26500188291837007
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3288 | LCC = 0.8070 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.23847780701441643
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3540 | LCC = 0.8071 | SRCC = 0.8052 ]
+INFO:root:Average loss=0.22249199602848443
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3009 | LCC = 0.8112 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.20372035392583945
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3045 | LCC = 0.8111 | SRCC = 0.8069 ]
+INFO:root:Average loss=0.19250235897608292
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2994 | LCC = 0.8121 | SRCC = 0.8074 ]
+INFO:root:Average loss=0.17996701588615394
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2776 | LCC = 0.8202 | SRCC = 0.8168 ]
+INFO:root:Average loss=0.1668266461063654
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3254 | LCC = 0.8164 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.15905942366673395
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3296 | LCC = 0.8108 | SRCC = 0.8096 ]
+INFO:root:Average loss=0.15147841702669096
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3526 | LCC = 0.8057 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.1517454889149238
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3294 | LCC = 0.8116 | SRCC = 0.8085 ]
+INFO:root:Average loss=0.14102680484453836
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3441 | LCC = 0.8147 | SRCC = 0.8140 ]
+INFO:root:Average loss=0.12603258274686643
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2879 | LCC = 0.8125 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.11384060959785412
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3237 | LCC = 0.8094 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.10927917483525398
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3029 | LCC = 0.8081 | SRCC = 0.8050 ]
+INFO:root:Average loss=0.10366906598210335
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3050 | LCC = 0.8042 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.09829619240302306
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3236 | LCC = 0.8103 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.09296508334003963
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3247 | LCC = 0.8074 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.08567066232745464
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3026 | LCC = 0.8062 | SRCC = 0.8014 ]
+INFO:root:Average loss=0.08407249258687863
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3115 | LCC = 0.8053 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.07837912526268226
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2995 | LCC = 0.8018 | SRCC = 0.7982 ]
+INFO:root:Average loss=0.08392994831769894
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3279 | LCC = 0.8054 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.07431051794152993
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2946 | LCC = 0.8070 | SRCC = 0.8038 ]
+INFO:root:Average loss=0.06799868661432694
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3061 | LCC = 0.7993 | SRCC = 0.7971 ]
+INFO:root:Average loss=0.0673973303861343
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3030 | LCC = 0.8015 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.06311425772041847
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3116 | LCC = 0.8084 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.06494387826667382
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3072 | LCC = 0.8091 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.059478260385684475
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3006 | LCC = 0.8058 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.059121602143232636
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2950 | LCC = 0.8070 | SRCC = 0.8054 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3062 | LCC = 0.8157 | SRCC = 0.8164 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer12/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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_large_asr_en/layer12/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer12/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a75b2acac42345f60397b74d3bb924d2e684cb59
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da8afe3d99c6ced5a81feaa639a55c8e1a9b589fbf4b0a8699e50494160966a1
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer12/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6535dd6bd1d2f0d06039b12e28af4503119b9cce
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d0807b5465cb26fcd978057b9d0de1c7cdb5d993e499ad0e61943909c16ea76
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer12/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer12/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..656028f153fcd5f68706e3a5cc6759202b59663e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer12/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=0.629870287500895
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4628 | LCC = 0.7109 | SRCC = 0.7140 ]
+INFO:root:Average loss=0.3486146888671777
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3833 | LCC = 0.7779 | SRCC = 0.7778 ]
+INFO:root:Average loss=0.30247746923795116
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3435 | LCC = 0.7981 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.2903971492480009
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3342 | LCC = 0.8005 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.2524496593918556
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3906 | LCC = 0.8058 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.22911507846453252
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3181 | LCC = 0.8064 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.21391650709586266
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3005 | LCC = 0.8122 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.1907470042888935
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3211 | LCC = 0.8106 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.182094951852774
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3109 | LCC = 0.8112 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.17697418882296637
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3565 | LCC = 0.8053 | SRCC = 0.8055 ]
+INFO:root:Average loss=0.15253328579740646
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2970 | LCC = 0.8096 | SRCC = 0.8092 ]
+INFO:root:Average loss=0.15185313031841546
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3611 | LCC = 0.8006 | SRCC = 0.7988 ]
+INFO:root:Average loss=0.13487681068288973
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3007 | LCC = 0.8081 | SRCC = 0.8063 ]
+INFO:root:Average loss=0.12277116760229453
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3128 | LCC = 0.8080 | SRCC = 0.8051 ]
+INFO:root:Average loss=0.12296248007661258
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3159 | LCC = 0.8034 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.10632928155171566
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3184 | LCC = 0.8101 | SRCC = 0.8085 ]
+INFO:root:Average loss=0.11399256285184468
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2917 | LCC = 0.8103 | SRCC = 0.8100 ]
+INFO:root:Average loss=0.10046655785005826
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3252 | LCC = 0.8092 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.0938771465458931
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3440 | LCC = 0.8056 | SRCC = 0.8035 ]
+INFO:root:Average loss=0.09173524800019386
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3307 | LCC = 0.8088 | SRCC = 0.8072 ]
+INFO:root:Average loss=0.08001724840738834
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3070 | LCC = 0.8075 | SRCC = 0.8050 ]
+INFO:root:Average loss=0.08038300457291114
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3170 | LCC = 0.8071 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.0751321111829617
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3115 | LCC = 0.8099 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.07465068246118534
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3101 | LCC = 0.8038 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.06820170103739469
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3353 | LCC = 0.8043 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.06471494991236772
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2934 | LCC = 0.8094 | SRCC = 0.8074 ]
+INFO:root:Average loss=0.0650003196623845
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2972 | LCC = 0.8107 | SRCC = 0.8071 ]
+INFO:root:Average loss=0.06452292547776149
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3088 | LCC = 0.8027 | SRCC = 0.8000 ]
+INFO:root:Average loss=0.06001786782573431
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3289 | LCC = 0.8055 | SRCC = 0.8035 ]
+INFO:root:Average loss=0.05812227386885729
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3155 | LCC = 0.8026 | SRCC = 0.8003 ]
+INFO:root:Average loss=0.055573841031545244
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3125 | LCC = 0.8098 | SRCC = 0.8081 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3076 | LCC = 0.8107 | SRCC = 0.8140 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer14/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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_large_asr_en/layer14/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer14/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fcd82fc96f0bab5e3d8d32b72620463e8772de66
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7bdb91f00f6807e0aa3c8fb169b2e8a9043ac5a04b0d19b446bd071909cd2f4
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer14/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f6fbfd01478904890cf2cd436ea12c3eb023a4bc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3568c4c623091099662ba2f8dbdaf778a4352b39059e1c3faf36bcfa6578647b
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer14/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..96870f2c12e811a22a59f0f99f7a0dcc040bff84
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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=0.639124198219715
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5824 | LCC = 0.6603 | SRCC = 0.6654 ]
+INFO:root:Average loss=0.40829688004958326
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3725 | LCC = 0.7650 | SRCC = 0.7667 ]
+INFO:root:Average loss=0.3168532964892877
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3455 | LCC = 0.7900 | SRCC = 0.7882 ]
+INFO:root:Average loss=0.2736568292364096
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3320 | LCC = 0.8046 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.2780578547181227
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4968 | LCC = 0.7974 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.24588502771579301
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3295 | LCC = 0.8075 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.2572734610010416
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3514 | LCC = 0.8064 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.21617332186836463
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3010 | LCC = 0.8103 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.19947211979291377
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3068 | LCC = 0.8131 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.18418184152016273
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3275 | LCC = 0.8129 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.18136565587841547
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3026 | LCC = 0.8115 | SRCC = 0.8093 ]
+INFO:root:Average loss=0.17634256021716657
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3040 | LCC = 0.8105 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.16672048908777726
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2918 | LCC = 0.8074 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.1531867396373015
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2858 | LCC = 0.8110 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.13818359909913477
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3116 | LCC = 0.8099 | SRCC = 0.8066 ]
+INFO:root:Average loss=0.13977856437365213
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3141 | LCC = 0.8075 | SRCC = 0.8063 ]
+INFO:root:Average loss=0.12866498329318488
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3146 | LCC = 0.8058 | SRCC = 0.8037 ]
+INFO:root:Average loss=0.12641079996067744
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3072 | LCC = 0.8084 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.11902084497687145
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3461 | LCC = 0.8102 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.10734949982128082
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2973 | LCC = 0.8039 | SRCC = 0.8031 ]
+INFO:root:Average loss=0.10676883839261837
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2959 | LCC = 0.8040 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.09338671265122218
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3056 | LCC = 0.8074 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.11823369424121502
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2920 | LCC = 0.8080 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.09307295757417496
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3335 | LCC = 0.8079 | SRCC = 0.8067 ]
+INFO:root:Average loss=0.08407396402878639
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3107 | LCC = 0.8055 | SRCC = 0.8040 ]
+INFO:root:Average loss=0.08697844788623162
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3272 | LCC = 0.8020 | SRCC = 0.8012 ]
+INFO:root:Average loss=0.0841066506333076
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2910 | LCC = 0.8075 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.0758334105500044
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2999 | LCC = 0.8063 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.07849707659811546
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2987 | LCC = 0.8069 | SRCC = 0.8055 ]
+INFO:root:Average loss=0.07672988012050971
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3049 | LCC = 0.8049 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.06985761306415765
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2994 | LCC = 0.8019 | SRCC = 0.8003 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3380 | LCC = 0.8146 | SRCC = 0.8211 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer14/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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_large_asr_en/layer14/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer14/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6c65cdbd2a590b352b0b262e7dce4fddd2762781
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0334e35d89573d14eff4c4c7292361bddda31360cdb14bf0e31f9f9060bb5063
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer14/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0dacb8c7c41726ebf47fae684123185e5b9d22f5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e1a8ba90d97d5ef442aa7226f11e56b51a1136d1243f8b231f6d2a049dfda46
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer14/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..935eb2752386771aea840df151bdb21264348fde
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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=0.5812411392346407
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4863 | LCC = 0.7401 | SRCC = 0.7450 ]
+INFO:root:Average loss=0.3544054925441742
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3510 | LCC = 0.7938 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.31247991896592653
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3257 | LCC = 0.8021 | SRCC = 0.7997 ]
+INFO:root:Average loss=0.2948195548393788
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3317 | LCC = 0.8104 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.2630130499601364
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3274 | LCC = 0.8069 | SRCC = 0.8040 ]
+INFO:root:Average loss=0.24918348724261308
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3135 | LCC = 0.8088 | SRCC = 0.8055 ]
+INFO:root:Average loss=0.2262976987239642
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2992 | LCC = 0.8145 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.20898243402823424
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2921 | LCC = 0.8181 | SRCC = 0.8129 ]
+INFO:root:Average loss=0.21231674899657568
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3155 | LCC = 0.8145 | SRCC = 0.8144 ]
+INFO:root:Average loss=0.190538444962257
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2829 | LCC = 0.8193 | SRCC = 0.8154 ]
+INFO:root:Average loss=0.18087279672423998
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2871 | LCC = 0.8229 | SRCC = 0.8205 ]
+INFO:root:Average loss=0.16717880161908957
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2927 | LCC = 0.8147 | SRCC = 0.8125 ]
+INFO:root:Average loss=0.15654730997406519
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3133 | LCC = 0.8090 | SRCC = 0.8085 ]
+INFO:root:Average loss=0.15454688276617956
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2922 | LCC = 0.8134 | SRCC = 0.8109 ]
+INFO:root:Average loss=0.14769291667602
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3003 | LCC = 0.8128 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.1449632299825167
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2969 | LCC = 0.8185 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.1303374068095134
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2905 | LCC = 0.8096 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.1257298872447931
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3071 | LCC = 0.8121 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.11752169378674947
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2882 | LCC = 0.8136 | SRCC = 0.8121 ]
+INFO:root:Average loss=0.10992229252289502
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3138 | LCC = 0.8099 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.10096019081389293
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2823 | LCC = 0.8153 | SRCC = 0.8135 ]
+INFO:root:Average loss=0.09729361983063893
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3079 | LCC = 0.8128 | SRCC = 0.8123 ]
+INFO:root:Average loss=0.093237780368863
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3062 | LCC = 0.8121 | SRCC = 0.8110 ]
+INFO:root:Average loss=0.08845462774236997
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2951 | LCC = 0.8120 | SRCC = 0.8110 ]
+INFO:root:Average loss=0.09973976398125672
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2915 | LCC = 0.8118 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.08304178046110348
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3001 | LCC = 0.8034 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.07921191433874461
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2891 | LCC = 0.8084 | SRCC = 0.8081 ]
+INFO:root:Average loss=0.08367712767078327
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3105 | LCC = 0.8090 | SRCC = 0.8105 ]
+INFO:root:Average loss=0.0783615472416083
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3004 | LCC = 0.8082 | SRCC = 0.8082 ]
+INFO:root:Average loss=0.08223956723052722
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2974 | LCC = 0.8095 | SRCC = 0.8100 ]
+INFO:root:Average loss=0.07453711927892306
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3065 | LCC = 0.8109 | SRCC = 0.8105 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2835 | LCC = 0.8207 | SRCC = 0.8208 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer14/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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_large_asr_en/layer14/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer14/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..75f2bfe6b0e4e942fe335e681a681e40fc1d7271
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3dc384c09284d751fa7edb1e270768905ff6dfd7c70bcae7e9f414d6b353c69
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer14/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ca4916f61888dda42a37af70515547878d290266
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cd66cf3e7d05c3b94cfe9b9c72c6beaf7479f1479d85db356f58db841f9dad7
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer14/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..aab44c93640e17c65af4b58f6a62fcd07c0d35a2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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=0.6524099031319985
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4754 | LCC = 0.6931 | SRCC = 0.6955 ]
+INFO:root:Average loss=0.35758627225191164
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3836 | LCC = 0.7808 | SRCC = 0.7829 ]
+INFO:root:Average loss=0.3075551348618972
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3402 | LCC = 0.7947 | SRCC = 0.7927 ]
+INFO:root:Average loss=0.2832821517800673
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3283 | LCC = 0.8025 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.2530775230664473
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3363 | LCC = 0.8030 | SRCC = 0.8020 ]
+INFO:root:Average loss=0.24408865108703956
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3090 | LCC = 0.8028 | SRCC = 0.8009 ]
+INFO:root:Average loss=0.2182787757080335
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3090 | LCC = 0.8111 | SRCC = 0.8079 ]
+INFO:root:Average loss=0.2112214899598024
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2955 | LCC = 0.8129 | SRCC = 0.8092 ]
+INFO:root:Average loss=0.19962961322222
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2955 | LCC = 0.8131 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.1844892527621526
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3150 | LCC = 0.8071 | SRCC = 0.8049 ]
+INFO:root:Average loss=0.18336004553697047
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3011 | LCC = 0.8135 | SRCC = 0.8102 ]
+INFO:root:Average loss=0.18344622048047873
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3021 | LCC = 0.8163 | SRCC = 0.8123 ]
+INFO:root:Average loss=0.163018013135745
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3010 | LCC = 0.8080 | SRCC = 0.8050 ]
+INFO:root:Average loss=0.14758349515688726
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3488 | LCC = 0.7970 | SRCC = 0.7955 ]
+INFO:root:Average loss=0.13971616642979476
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3724 | LCC = 0.8026 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.14740928291128233
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3598 | LCC = 0.8041 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.1275135239538474
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3341 | LCC = 0.8032 | SRCC = 0.8028 ]
+INFO:root:Average loss=0.12222580993786836
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3528 | LCC = 0.7907 | SRCC = 0.7888 ]
+INFO:root:Average loss=0.115394481863731
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3066 | LCC = 0.7961 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.11262325063729897
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3114 | LCC = 0.7992 | SRCC = 0.7955 ]
+INFO:root:Average loss=0.10322136613420951
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3046 | LCC = 0.7978 | SRCC = 0.7954 ]
+INFO:root:Average loss=0.10801306600945118
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3106 | LCC = 0.7995 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.09470349884568116
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3348 | LCC = 0.8007 | SRCC = 0.8006 ]
+INFO:root:Average loss=0.09117928639245339
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3315 | LCC = 0.7981 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.0903872536161007
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3039 | LCC = 0.8068 | SRCC = 0.8055 ]
+INFO:root:Average loss=0.08537686267533363
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3091 | LCC = 0.8008 | SRCC = 0.7980 ]
+INFO:root:Average loss=0.08537523032954106
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3165 | LCC = 0.8010 | SRCC = 0.7998 ]
+INFO:root:Average loss=0.08757533223774189
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2990 | LCC = 0.8046 | SRCC = 0.8021 ]
+INFO:root:Average loss=0.07680801588755387
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3423 | LCC = 0.8014 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.07683007342693134
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3016 | LCC = 0.8037 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.07444462165809594
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3368 | LCC = 0.8025 | SRCC = 0.8021 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3370 | LCC = 0.7984 | SRCC = 0.8052 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer14/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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_large_asr_en/layer14/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer14/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..20030aec09804f13ea4ec13dd9dd53ebed0e4487
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:069389fc67653cde3cc867f7aaba683212efec873d300ef5aee50aa5b581a980
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer14/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..94e2f2122f128b1fab227ae7021060c5f45f0e58
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ef4b777e9e71d9ac8929c8805232b3010a34334d3386463b4f3820a6efc2dfd
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer14/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ea6ccbd83588291f9285644eec924480641f6512
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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=0.6228337616492541
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4639 | LCC = 0.7118 | SRCC = 0.7112 ]
+INFO:root:Average loss=0.36973212048029286
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3787 | LCC = 0.7793 | SRCC = 0.7781 ]
+INFO:root:Average loss=0.31610557245902526
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3625 | LCC = 0.7884 | SRCC = 0.7869 ]
+INFO:root:Average loss=0.2852938834291238
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3119 | LCC = 0.8009 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.2740824589362511
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3700 | LCC = 0.7982 | SRCC = 0.7940 ]
+INFO:root:Average loss=0.24433110864498678
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3307 | LCC = 0.8002 | SRCC = 0.7949 ]
+INFO:root:Average loss=0.23941120295188364
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3721 | LCC = 0.8017 | SRCC = 0.7983 ]
+INFO:root:Average loss=0.2147610577253195
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3154 | LCC = 0.8049 | SRCC = 0.8007 ]
+INFO:root:Average loss=0.1963093800422473
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3047 | LCC = 0.8082 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.19134491806228957
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2998 | LCC = 0.8076 | SRCC = 0.8039 ]
+INFO:root:Average loss=0.17910867795730248
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3235 | LCC = 0.8018 | SRCC = 0.8003 ]
+INFO:root:Average loss=0.1719729864062407
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3029 | LCC = 0.8019 | SRCC = 0.7983 ]
+INFO:root:Average loss=0.15981111025963074
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3317 | LCC = 0.8037 | SRCC = 0.8021 ]
+INFO:root:Average loss=0.1607825325276607
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3118 | LCC = 0.8021 | SRCC = 0.8010 ]
+INFO:root:Average loss=0.15734562745843178
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3096 | LCC = 0.7990 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.1332972825337679
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3301 | LCC = 0.7923 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.1389224115663614
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3122 | LCC = 0.8018 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.13425574289300504
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3329 | LCC = 0.8014 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.11482151206105183
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3155 | LCC = 0.7997 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.11373232744443111
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3101 | LCC = 0.8007 | SRCC = 0.7981 ]
+INFO:root:Average loss=0.10444456740067555
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3060 | LCC = 0.7958 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.09774937857993138
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3169 | LCC = 0.7925 | SRCC = 0.7909 ]
+INFO:root:Average loss=0.09432413472005954
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3318 | LCC = 0.7895 | SRCC = 0.7878 ]
+INFO:root:Average loss=0.09473877371503757
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3360 | LCC = 0.7912 | SRCC = 0.7904 ]
+INFO:root:Average loss=0.09524659310969022
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3147 | LCC = 0.7952 | SRCC = 0.7928 ]
+INFO:root:Average loss=0.08925897604188858
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3078 | LCC = 0.7954 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.08056991270337349
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3018 | LCC = 0.7978 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.07634544687775466
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3187 | LCC = 0.7910 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.07849306880663602
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3045 | LCC = 0.7995 | SRCC = 0.7976 ]
+INFO:root:Average loss=0.07884880671134362
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3241 | LCC = 0.7931 | SRCC = 0.7911 ]
+INFO:root:Average loss=0.0762025519536856
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3172 | LCC = 0.7931 | SRCC = 0.7910 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3233 | LCC = 0.8182 | SRCC = 0.8232 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer14/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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_large_asr_en/layer14/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer14/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6f142188e37ab75eb0184c7dc5e63776c118dac1
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85672245a0b51022641a0d909fb1c8b224bc0e681ed39290f994f2275b15daf4
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer14/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..618b666d1162823de523b2c446a336d63d1f750b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5afadd72191475a9b936a4ff7e854fbfa08c81774a315128368698ed55c638b
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer14/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer14/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..cb595827f1c32a524b10636cc303aa8df76e6996
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer14/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=0.7026129311475998
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5354 | LCC = 0.6535 | SRCC = 0.6544 ]
+INFO:root:Average loss=0.4123027328497324
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4705 | LCC = 0.7603 | SRCC = 0.7674 ]
+INFO:root:Average loss=0.3294103884926209
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3594 | LCC = 0.7968 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.28935185647927797
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3877 | LCC = 0.7997 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.25666792155840457
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3263 | LCC = 0.8053 | SRCC = 0.8014 ]
+INFO:root:Average loss=0.2462401195214345
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3060 | LCC = 0.8127 | SRCC = 0.8098 ]
+INFO:root:Average loss=0.2243162466165347
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3157 | LCC = 0.8073 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.22121054048721606
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3203 | LCC = 0.8085 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.1997496387324272
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3457 | LCC = 0.8069 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.18888872116804123
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3043 | LCC = 0.8117 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.18737222177860063
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3383 | LCC = 0.8053 | SRCC = 0.8043 ]
+INFO:root:Average loss=0.17684167212782761
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3061 | LCC = 0.8093 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.15987134476502737
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3332 | LCC = 0.8037 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.14494832079762068
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3040 | LCC = 0.8045 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.1484078340805494
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3032 | LCC = 0.8076 | SRCC = 0.8033 ]
+INFO:root:Average loss=0.1324362083314321
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2984 | LCC = 0.8059 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.1255868646578911
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3033 | LCC = 0.8057 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.11893585448463757
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2931 | LCC = 0.8106 | SRCC = 0.8079 ]
+INFO:root:Average loss=0.12923317039624238
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2963 | LCC = 0.8076 | SRCC = 0.8048 ]
+INFO:root:Average loss=0.10622644285934094
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3009 | LCC = 0.8092 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.1025756287555664
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3025 | LCC = 0.8018 | SRCC = 0.7982 ]
+INFO:root:Average loss=0.09379468289896464
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3046 | LCC = 0.8036 | SRCC = 0.8012 ]
+INFO:root:Average loss=0.09388349921657489
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3263 | LCC = 0.8060 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.0900142370030666
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3171 | LCC = 0.8026 | SRCC = 0.8014 ]
+INFO:root:Average loss=0.08962828742387967
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3109 | LCC = 0.8086 | SRCC = 0.8070 ]
+INFO:root:Average loss=0.07964161200783192
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3045 | LCC = 0.8023 | SRCC = 0.7981 ]
+INFO:root:Average loss=0.0749060032554926
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2993 | LCC = 0.8044 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.07863314425907074
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3002 | LCC = 0.8070 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.07635377657910188
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3258 | LCC = 0.8039 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.06744844829424834
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3399 | LCC = 0.8048 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.06707029378949067
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3050 | LCC = 0.8076 | SRCC = 0.8053 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3049 | LCC = 0.8079 | SRCC = 0.8130 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer16/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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_large_asr_en/layer16/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer16/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b0d68ac1d06dd18ac3fb73ea7f6e59edaa1a279c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ab263a58bb871a0bbbc0ac0333b72d60632f2548299dcdf0f66171ed652b77c
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer16/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..de4a19407dd02bddab05954e0cda2e4045955134
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b5b74a09ebb88eedbbb6a503c09906c8abeba6d23b638bb312236d08cd1ddc9
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer16/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b06daafb9dc02b9ab931c2a1b205ac3d1b1a1901
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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=0.6045289822877982
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5157 | LCC = 0.6997 | SRCC = 0.7075 ]
+INFO:root:Average loss=0.3855379789303511
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4022 | LCC = 0.7636 | SRCC = 0.7666 ]
+INFO:root:Average loss=0.3314328975020311
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3535 | LCC = 0.7894 | SRCC = 0.7895 ]
+INFO:root:Average loss=0.3095303649703662
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3552 | LCC = 0.7812 | SRCC = 0.7793 ]
+INFO:root:Average loss=0.27557561642084366
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3384 | LCC = 0.7975 | SRCC = 0.7944 ]
+INFO:root:Average loss=0.2570803916225067
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3212 | LCC = 0.8041 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.24496627847353616
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3900 | LCC = 0.7874 | SRCC = 0.7854 ]
+INFO:root:Average loss=0.24344932555388182
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3274 | LCC = 0.7914 | SRCC = 0.7908 ]
+INFO:root:Average loss=0.2289665433076712
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3481 | LCC = 0.7942 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.216319522414452
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3297 | LCC = 0.7968 | SRCC = 0.7951 ]
+INFO:root:Average loss=0.21196157838671636
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3198 | LCC = 0.7959 | SRCC = 0.7958 ]
+INFO:root:Average loss=0.18836556814419916
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3405 | LCC = 0.7873 | SRCC = 0.7834 ]
+INFO:root:Average loss=0.18668227146069208
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3201 | LCC = 0.7958 | SRCC = 0.7931 ]
+INFO:root:Average loss=0.20887977878252664
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3046 | LCC = 0.8064 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.1684195515819085
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2979 | LCC = 0.8052 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.16316595921913782
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3062 | LCC = 0.8002 | SRCC = 0.7986 ]
+INFO:root:Average loss=0.15891984277046645
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3100 | LCC = 0.7999 | SRCC = 0.7988 ]
+INFO:root:Average loss=0.14544746766869837
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2949 | LCC = 0.8021 | SRCC = 0.8008 ]
+INFO:root:Average loss=0.15633754699658126
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3089 | LCC = 0.7949 | SRCC = 0.7925 ]
+INFO:root:Average loss=0.15230250358581543
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3887 | LCC = 0.7941 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.13482884938518205
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3018 | LCC = 0.7990 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.12442011376604056
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3228 | LCC = 0.7954 | SRCC = 0.7933 ]
+INFO:root:Average loss=0.12203077054940738
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3534 | LCC = 0.7989 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.11553829784194629
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3144 | LCC = 0.7932 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.10880390143929383
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3186 | LCC = 0.7921 | SRCC = 0.7900 ]
+INFO:root:Average loss=0.09980233734807907
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3178 | LCC = 0.7941 | SRCC = 0.7928 ]
+INFO:root:Average loss=0.10376477499420826
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3195 | LCC = 0.7928 | SRCC = 0.7925 ]
+INFO:root:Average loss=0.09654545459227684
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3101 | LCC = 0.7942 | SRCC = 0.7936 ]
+INFO:root:Average loss=0.11127876213345772
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3104 | LCC = 0.7926 | SRCC = 0.7923 ]
+INFO:root:Average loss=0.09303435578178136
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3064 | LCC = 0.7964 | SRCC = 0.7959 ]
+INFO:root:Average loss=0.08888796970057182
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3174 | LCC = 0.7934 | SRCC = 0.7914 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2981 | LCC = 0.8126 | SRCC = 0.8138 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer16/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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_large_asr_en/layer16/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer16/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bfa0cdb17d222c85449584915c3ca5836b051d7c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e44fabee7493ceb3863c0347b47d5d689cae269ade3da43e4660a7d79424f9e7
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer16/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f8c62323485d1aa103e2f08aeb6ee36993b11228
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06e285678bd551a3be7fe0b4047f28fc41911fa4b9b84bfb8fc3bc0579c5d79c
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer16/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..bb7c05df0125e1f5cd021629eae6e3a116696c23
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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=0.6340954139446601
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5405 | LCC = 0.6657 | SRCC = 0.6770 ]
+INFO:root:Average loss=0.40024191790666336
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5298 | LCC = 0.7456 | SRCC = 0.7543 ]
+INFO:root:Average loss=0.33920538883942825
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3566 | LCC = 0.7903 | SRCC = 0.7869 ]
+INFO:root:Average loss=0.2973359599709511
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3974 | LCC = 0.7889 | SRCC = 0.7866 ]
+INFO:root:Average loss=0.278354168511354
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3284 | LCC = 0.8039 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.261143011542467
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3325 | LCC = 0.8024 | SRCC = 0.7976 ]
+INFO:root:Average loss=0.25011139191113985
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3486 | LCC = 0.8015 | SRCC = 0.7961 ]
+INFO:root:Average loss=0.23744418643988097
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3097 | LCC = 0.8069 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.22340847781071296
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3335 | LCC = 0.7976 | SRCC = 0.7926 ]
+INFO:root:Average loss=0.22299599609313867
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3571 | LCC = 0.8073 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.20294541139633226
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3277 | LCC = 0.8043 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.20834104067240006
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3325 | LCC = 0.8059 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.18875720667151305
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2975 | LCC = 0.8103 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.17926106974482536
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3237 | LCC = 0.8024 | SRCC = 0.7983 ]
+INFO:root:Average loss=0.1697360057479296
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3041 | LCC = 0.8000 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.16696689889217034
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3085 | LCC = 0.8020 | SRCC = 0.7970 ]
+INFO:root:Average loss=0.16126962999502817
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3159 | LCC = 0.8044 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.1490314886547052
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3258 | LCC = 0.8005 | SRCC = 0.7980 ]
+INFO:root:Average loss=0.15126367352711848
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3342 | LCC = 0.7991 | SRCC = 0.7964 ]
+INFO:root:Average loss=0.14067628368353233
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3366 | LCC = 0.7972 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.15042344108223915
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3526 | LCC = 0.7924 | SRCC = 0.7899 ]
+INFO:root:Average loss=0.13439147776136032
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3076 | LCC = 0.7990 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.12486588859405273
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3320 | LCC = 0.7920 | SRCC = 0.7877 ]
+INFO:root:Average loss=0.12443696583310764
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3285 | LCC = 0.7895 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.11504852943695508
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3609 | LCC = 0.7956 | SRCC = 0.7921 ]
+INFO:root:Average loss=0.11089846071524498
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3203 | LCC = 0.7997 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.10934313038029732
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3112 | LCC = 0.7940 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.09857210975426894
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3010 | LCC = 0.7998 | SRCC = 0.7971 ]
+INFO:root:Average loss=0.10711307885746162
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3464 | LCC = 0.7914 | SRCC = 0.7903 ]
+INFO:root:Average loss=0.10042486912929095
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3374 | LCC = 0.7834 | SRCC = 0.7791 ]
+INFO:root:Average loss=0.09625535682798961
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3253 | LCC = 0.7854 | SRCC = 0.7821 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3149 | LCC = 0.8102 | SRCC = 0.8170 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer16/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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_large_asr_en/layer16/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer16/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2b0a44011e18e79a5104d9ab42800db4502d0354
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f17ea790c9c1f6fc64358b4620cdbb4e400781b031086c24955b7ab063a4ad2e
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer16/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..dfac520f1b9e12d404fbf979d7121077559be3bc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f61bef0c0372660a037bb98b73638f1b9ad4c91de3af23075403b9df5114d5d
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer16/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e3f227a832709db63740445515a3f0fba6af8fdb
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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=0.657288542160621
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5794 | LCC = 0.6114 | SRCC = 0.6162 ]
+INFO:root:Average loss=0.40220736272824115
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4392 | LCC = 0.7524 | SRCC = 0.7593 ]
+INFO:root:Average loss=0.33026147461854494
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3835 | LCC = 0.7787 | SRCC = 0.7764 ]
+INFO:root:Average loss=0.3086832670065073
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4132 | LCC = 0.7784 | SRCC = 0.7784 ]
+INFO:root:Average loss=0.27899152155105883
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3432 | LCC = 0.7913 | SRCC = 0.7870 ]
+INFO:root:Average loss=0.26980655124554265
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3149 | LCC = 0.8007 | SRCC = 0.7949 ]
+INFO:root:Average loss=0.2540996462488786
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3292 | LCC = 0.8037 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.24517427193812835
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3229 | LCC = 0.8049 | SRCC = 0.7996 ]
+INFO:root:Average loss=0.22538406841265848
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3302 | LCC = 0.8026 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.21539975855595025
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3017 | LCC = 0.8098 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.2084357663989067
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3171 | LCC = 0.8074 | SRCC = 0.8038 ]
+INFO:root:Average loss=0.19319197019705406
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3200 | LCC = 0.8045 | SRCC = 0.7990 ]
+INFO:root:Average loss=0.18870560452342033
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3039 | LCC = 0.8072 | SRCC = 0.8010 ]
+INFO:root:Average loss=0.18214081934629342
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2966 | LCC = 0.8052 | SRCC = 0.7990 ]
+INFO:root:Average loss=0.1678267621841186
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3060 | LCC = 0.8097 | SRCC = 0.8042 ]
+INFO:root:Average loss=0.15425458359412658
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3196 | LCC = 0.8078 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.1617930031930789
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3194 | LCC = 0.8037 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.14492820633145478
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3020 | LCC = 0.8064 | SRCC = 0.8003 ]
+INFO:root:Average loss=0.1437309701473285
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3230 | LCC = 0.8017 | SRCC = 0.8004 ]
+INFO:root:Average loss=0.1390629074512384
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3131 | LCC = 0.8008 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.13242085115649763
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3377 | LCC = 0.7981 | SRCC = 0.7950 ]
+INFO:root:Average loss=0.1308021463262729
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3096 | LCC = 0.7995 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.1199426304262418
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3315 | LCC = 0.7930 | SRCC = 0.7866 ]
+INFO:root:Average loss=0.1132065311838419
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3344 | LCC = 0.7993 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.11419395252298085
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3048 | LCC = 0.8014 | SRCC = 0.7948 ]
+INFO:root:Average loss=0.10965073118225122
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3134 | LCC = 0.8014 | SRCC = 0.7959 ]
+INFO:root:Average loss=0.10087619201304057
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3221 | LCC = 0.8027 | SRCC = 0.7997 ]
+INFO:root:Average loss=0.09667974853744873
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3196 | LCC = 0.7987 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.0966238718575392
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3179 | LCC = 0.8022 | SRCC = 0.8007 ]
+INFO:root:Average loss=0.09608871384690969
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3358 | LCC = 0.7988 | SRCC = 0.7987 ]
+INFO:root:Average loss=0.08986696729866359
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3099 | LCC = 0.7999 | SRCC = 0.7974 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2993 | LCC = 0.8203 | SRCC = 0.8232 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer16/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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_large_asr_en/layer16/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer16/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..03a49d1c5e2cacdf532d51259d64d493fcd6f5f8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:026863d6ae639c73a9cf49143ab1fd51197cce22fe04bd41225a0a16769659ba
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer16/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..055af78693f5617711d81faab71dc014bae5abc9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a837d8c118839371a4a6deb2053a76b2e78a8987d34c23191c0663458c0c0e65
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer16/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b7ccbfb4adcd4052b409036eaef9c2efb8e7482b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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=0.6723408076243523
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5557 | LCC = 0.6463 | SRCC = 0.6543 ]
+INFO:root:Average loss=0.4006558996744645
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4246 | LCC = 0.7776 | SRCC = 0.7784 ]
+INFO:root:Average loss=0.34098843141244006
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3222 | LCC = 0.7997 | SRCC = 0.7991 ]
+INFO:root:Average loss=0.3153627287500944
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4108 | LCC = 0.7900 | SRCC = 0.7923 ]
+INFO:root:Average loss=0.27693622807661694
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3903 | LCC = 0.7972 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.2603942215060576
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3260 | LCC = 0.8050 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.2527946873735159
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3533 | LCC = 0.7968 | SRCC = 0.7946 ]
+INFO:root:Average loss=0.25168594737083483
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3339 | LCC = 0.8054 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.2563933768333533
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3102 | LCC = 0.8108 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.21988856544097266
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3165 | LCC = 0.8056 | SRCC = 0.8008 ]
+INFO:root:Average loss=0.1964635105851369
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3415 | LCC = 0.7933 | SRCC = 0.7907 ]
+INFO:root:Average loss=0.19502703530284074
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2998 | LCC = 0.8096 | SRCC = 0.8052 ]
+INFO:root:Average loss=0.18960357199494654
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2964 | LCC = 0.8070 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.18062509825596443
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3320 | LCC = 0.8051 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.16464038976492026
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3908 | LCC = 0.7887 | SRCC = 0.7903 ]
+INFO:root:Average loss=0.15582569554830208
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3137 | LCC = 0.7977 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.1557566132874061
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3050 | LCC = 0.8005 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.1417962185656413
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3331 | LCC = 0.7925 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.14614321129062238
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3137 | LCC = 0.8025 | SRCC = 0.8003 ]
+INFO:root:Average loss=0.1397981749704251
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3050 | LCC = 0.7998 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.13983271844111955
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3238 | LCC = 0.7967 | SRCC = 0.7963 ]
+INFO:root:Average loss=0.12850720855670097
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3531 | LCC = 0.7908 | SRCC = 0.7922 ]
+INFO:root:Average loss=0.11814073100686073
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3241 | LCC = 0.7909 | SRCC = 0.7901 ]
+INFO:root:Average loss=0.1203837201763422
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3116 | LCC = 0.8007 | SRCC = 0.7983 ]
+INFO:root:Average loss=0.11787659455186282
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3124 | LCC = 0.7987 | SRCC = 0.7963 ]
+INFO:root:Average loss=0.11823239325521848
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3059 | LCC = 0.7995 | SRCC = 0.7988 ]
+INFO:root:Average loss=0.11230465979912342
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3116 | LCC = 0.7993 | SRCC = 0.7964 ]
+INFO:root:Average loss=0.10316386570533116
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3088 | LCC = 0.7955 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.09984191402028768
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3064 | LCC = 0.7967 | SRCC = 0.7944 ]
+INFO:root:Average loss=0.09446386391153702
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3199 | LCC = 0.7975 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.09352919736351722
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3027 | LCC = 0.7966 | SRCC = 0.7941 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2919 | LCC = 0.8170 | SRCC = 0.8197 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer16/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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_large_asr_en/layer16/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer16/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..059b49814dcd8aaf4ae34c44627733c43aefd38f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e80feec821fcb3663d2911f35a91d8f9a2bdf8f9fa1bfaec589a4bec3f24c4f
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer16/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e481539c1068558f9caeeb96cb7cef48b082e2ed
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c742a8cde17845310a48d247a90dcea24ea9a56823955eec9fb2832713ea773c
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer16/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer16/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..1a02e78844c22ca12bcf9c7e10a9ee23c67ca288
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer16/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=0.6789450901441085
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6046 | LCC = 0.6282 | SRCC = 0.6400 ]
+INFO:root:Average loss=0.3924536086045779
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5374 | LCC = 0.7519 | SRCC = 0.7650 ]
+INFO:root:Average loss=0.33923711685033947
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3980 | LCC = 0.7845 | SRCC = 0.7843 ]
+INFO:root:Average loss=0.3028064412184251
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3569 | LCC = 0.7910 | SRCC = 0.7891 ]
+INFO:root:Average loss=0.2819619507361681
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4103 | LCC = 0.7917 | SRCC = 0.7897 ]
+INFO:root:Average loss=0.253051302753962
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3777 | LCC = 0.7928 | SRCC = 0.7898 ]
+INFO:root:Average loss=0.2618762203134023
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3135 | LCC = 0.7993 | SRCC = 0.7945 ]
+INFO:root:Average loss=0.23920700412530166
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3354 | LCC = 0.7951 | SRCC = 0.7906 ]
+INFO:root:Average loss=0.2156798541545868
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3594 | LCC = 0.8019 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.20976531830353615
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3858 | LCC = 0.8016 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.20255629469950995
+INFO:root:
+[Valid][10][UTT][ MSE = 0.4731 | LCC = 0.7945 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.18334445166282165
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3012 | LCC = 0.8026 | SRCC = 0.7985 ]
+INFO:root:Average loss=0.1816348512776387
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3071 | LCC = 0.8096 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.1824573727372365
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3031 | LCC = 0.8096 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.17719774635938498
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3844 | LCC = 0.8007 | SRCC = 0.7984 ]
+INFO:root:Average loss=0.15193971733634287
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3707 | LCC = 0.7955 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.15421904604404402
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3131 | LCC = 0.7977 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.14411996371853045
+INFO:root:
+[Valid][17][UTT][ MSE = 0.4548 | LCC = 0.7951 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.1521642519495426
+INFO:root:
+[Valid][18][UTT][ MSE = 0.4206 | LCC = 0.7947 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.134348280250262
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3408 | LCC = 0.7994 | SRCC = 0.7977 ]
+INFO:root:Average loss=0.12350613385056838
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3982 | LCC = 0.7966 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.11615657777740405
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3228 | LCC = 0.7961 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.11736642588407566
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3287 | LCC = 0.7952 | SRCC = 0.7923 ]
+INFO:root:Average loss=0.11491044056721222
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3149 | LCC = 0.7997 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.11043391405389859
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3558 | LCC = 0.7986 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.10436675081459376
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3049 | LCC = 0.7997 | SRCC = 0.7949 ]
+INFO:root:Average loss=0.09988003224134445
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3036 | LCC = 0.8046 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.09782322639456162
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3179 | LCC = 0.7994 | SRCC = 0.7976 ]
+INFO:root:Average loss=0.09417254659227836
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3043 | LCC = 0.7972 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.0951730421720407
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3467 | LCC = 0.7944 | SRCC = 0.7917 ]
+INFO:root:Average loss=0.08588779942156413
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3239 | LCC = 0.7986 | SRCC = 0.7951 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3161 | LCC = 0.7994 | SRCC = 0.8042 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer18/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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_large_asr_en/layer18/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer18/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ca0bee0e4d420510d7a2857962a6a319aff2d230
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54806c1ee8b854af6de8a06d584046d39af2090dcad57715a543a623830e4218
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer18/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..246040c66046c3204450e5f78436133f55a6d8e4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e4484400a487e6b32a76c40f68d70e7fe270c1d88a28806b8371742bbf8c6b3
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer18/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..091d34c1e8fa087246576d3d4de18f319aef7fb2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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=0.7117670296858518
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6106 | LCC = 0.5562 | SRCC = 0.5599 ]
+INFO:root:Average loss=0.4429682099666351
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4165 | LCC = 0.7560 | SRCC = 0.7549 ]
+INFO:root:Average loss=0.34332561913209086
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3795 | LCC = 0.7786 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.33117712518343556
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3859 | LCC = 0.7810 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.3002591692866423
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4364 | LCC = 0.7793 | SRCC = 0.7825 ]
+INFO:root:Average loss=0.29595710241641754
+INFO:root:
+[Valid][5][UTT][ MSE = 0.5978 | LCC = 0.7606 | SRCC = 0.7703 ]
+INFO:root:Average loss=0.2854265845738925
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3992 | LCC = 0.7896 | SRCC = 0.7865 ]
+INFO:root:Average loss=0.2570588980347682
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4801 | LCC = 0.7870 | SRCC = 0.7865 ]
+INFO:root:Average loss=0.24651958029239607
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3699 | LCC = 0.7975 | SRCC = 0.7963 ]
+INFO:root:Average loss=0.23112520155234215
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3400 | LCC = 0.7849 | SRCC = 0.7809 ]
+INFO:root:Average loss=0.23206110823995027
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3191 | LCC = 0.8003 | SRCC = 0.7950 ]
+INFO:root:Average loss=0.22758932363910553
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3340 | LCC = 0.8009 | SRCC = 0.7980 ]
+INFO:root:Average loss=0.23039031487244827
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3134 | LCC = 0.7988 | SRCC = 0.7939 ]
+INFO:root:Average loss=0.20950960396574095
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3124 | LCC = 0.8005 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.20721171194544205
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3363 | LCC = 0.7921 | SRCC = 0.7863 ]
+INFO:root:Average loss=0.21501756459474564
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3767 | LCC = 0.7988 | SRCC = 0.7997 ]
+INFO:root:Average loss=0.19353947922205314
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3524 | LCC = 0.7953 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.2042483474390629
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3973 | LCC = 0.7927 | SRCC = 0.7963 ]
+INFO:root:Average loss=0.19271951503096482
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3450 | LCC = 0.7994 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.1777458359988836
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3268 | LCC = 0.7960 | SRCC = 0.7931 ]
+INFO:root:Average loss=0.17725283920000762
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3130 | LCC = 0.7975 | SRCC = 0.7939 ]
+INFO:root:Average loss=0.16132712880006203
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3239 | LCC = 0.7937 | SRCC = 0.7921 ]
+INFO:root:Average loss=0.1622514983591361
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3745 | LCC = 0.7915 | SRCC = 0.7937 ]
+INFO:root:Average loss=0.1590261387710388
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3361 | LCC = 0.7984 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.15630254884942985
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3237 | LCC = 0.7910 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.1588927767215631
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3202 | LCC = 0.7900 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.13642840326214448
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3803 | LCC = 0.7820 | SRCC = 0.7783 ]
+INFO:root:Average loss=0.1368061573459552
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3244 | LCC = 0.7891 | SRCC = 0.7859 ]
+INFO:root:Average loss=0.13994430072414568
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3420 | LCC = 0.7905 | SRCC = 0.7872 ]
+INFO:root:Average loss=0.13627814902709082
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3053 | LCC = 0.7999 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.12679368515427297
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3321 | LCC = 0.7948 | SRCC = 0.7925 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2898 | LCC = 0.8205 | SRCC = 0.8205 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer18/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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_large_asr_en/layer18/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer18/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1185e60fd13d7c49746fb1c780a056a9456bf544
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c8cd81b7213ff9390c6dc02590a4bbe83c0395bb4099d271a7d8310e3eddc77
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer18/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ed9601a4f689dfd4280aa46675e89ce9bf0e6b4c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:593d3865f42704b007068c431f83e256c9060351f9e98d5ec27ee1acada8845f
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer18/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..0953e0e466089832a98a4869ab6d9b699befa148
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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=0.6100775396976715
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5029 | LCC = 0.6695 | SRCC = 0.6716 ]
+INFO:root:Average loss=0.39632055965753704
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3912 | LCC = 0.7464 | SRCC = 0.7449 ]
+INFO:root:Average loss=0.3389633920712349
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3969 | LCC = 0.7667 | SRCC = 0.7668 ]
+INFO:root:Average loss=0.3145429369730827
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3327 | LCC = 0.7868 | SRCC = 0.7828 ]
+INFO:root:Average loss=0.3081854285720067
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3772 | LCC = 0.7875 | SRCC = 0.7860 ]
+INFO:root:Average loss=0.2819000033613963
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4233 | LCC = 0.7910 | SRCC = 0.7885 ]
+INFO:root:Average loss=0.2781742157844397
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3342 | LCC = 0.7890 | SRCC = 0.7839 ]
+INFO:root:Average loss=0.27238520177511066
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4357 | LCC = 0.7955 | SRCC = 0.7951 ]
+INFO:root:Average loss=0.2578228832437442
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3221 | LCC = 0.7968 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.23971376835535735
+INFO:root:
+[Valid][9][UTT][ MSE = 0.4019 | LCC = 0.7974 | SRCC = 0.7959 ]
+INFO:root:Average loss=0.23505156888411596
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3698 | LCC = 0.7994 | SRCC = 0.7956 ]
+INFO:root:Average loss=0.23671586486773613
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3269 | LCC = 0.7959 | SRCC = 0.7883 ]
+INFO:root:Average loss=0.22069437457964972
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3186 | LCC = 0.7998 | SRCC = 0.7946 ]
+INFO:root:Average loss=0.22055166615889624
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3149 | LCC = 0.7990 | SRCC = 0.7948 ]
+INFO:root:Average loss=0.20877059348500693
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3316 | LCC = 0.8002 | SRCC = 0.7955 ]
+INFO:root:Average loss=0.20384110701389802
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3206 | LCC = 0.8021 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.19095621640101457
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3322 | LCC = 0.8022 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.20597576063412887
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3609 | LCC = 0.7907 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.18552888433138529
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3281 | LCC = 0.7977 | SRCC = 0.7929 ]
+INFO:root:Average loss=0.1806377992989161
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3205 | LCC = 0.7995 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.16829875168892053
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3083 | LCC = 0.7999 | SRCC = 0.7950 ]
+INFO:root:Average loss=0.20723392069339752
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3417 | LCC = 0.7978 | SRCC = 0.7958 ]
+INFO:root:Average loss=0.15616247621484292
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3172 | LCC = 0.7957 | SRCC = 0.7921 ]
+INFO:root:Average loss=0.16115129290100855
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3140 | LCC = 0.8010 | SRCC = 0.7971 ]
+INFO:root:Average loss=0.15326706663920328
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3191 | LCC = 0.7953 | SRCC = 0.7909 ]
+INFO:root:Average loss=0.1477968695645149
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3422 | LCC = 0.7897 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.14118077491338438
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3262 | LCC = 0.7911 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.13815056246060592
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3609 | LCC = 0.7907 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.14220604510643545
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3124 | LCC = 0.7952 | SRCC = 0.7917 ]
+INFO:root:Average loss=0.13270198276791817
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3171 | LCC = 0.7967 | SRCC = 0.7932 ]
+INFO:root:Average loss=0.13343474068320715
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3153 | LCC = 0.7937 | SRCC = 0.7928 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3404 | LCC = 0.8106 | SRCC = 0.8156 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer18/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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_large_asr_en/layer18/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer18/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9c2baf6c5e6a23bd6fa43d87154699b8766b7625
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8991ed34427425a4aee92ab25dc785b3c54f8d17a12b3900cb52d8185dbf70bd
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer18/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..530d32a8e5e82fe25fa1fbfaf043b2fdd52419f9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dff2d14f6f9b5b94ce5dfb80d6fd8fa192b3395d59d1e947a75ec4168df30fcd
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer18/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..0332c0f9db0f3dffe9ccd79069d762aa0c084377
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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=0.6498745316878344
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5672 | LCC = 0.5822 | SRCC = 0.5859 ]
+INFO:root:Average loss=0.4190582212729332
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3937 | LCC = 0.7451 | SRCC = 0.7467 ]
+INFO:root:Average loss=0.34608255823453266
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3709 | LCC = 0.7743 | SRCC = 0.7750 ]
+INFO:root:Average loss=0.3128836504541911
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3989 | LCC = 0.7769 | SRCC = 0.7781 ]
+INFO:root:Average loss=0.303414937777397
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3451 | LCC = 0.7858 | SRCC = 0.7867 ]
+INFO:root:Average loss=0.27943470692023253
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3322 | LCC = 0.7841 | SRCC = 0.7818 ]
+INFO:root:Average loss=0.27902159094810486
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3335 | LCC = 0.7957 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.2726952401109231
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3644 | LCC = 0.7933 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.2502315416932106
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4261 | LCC = 0.7847 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.25346136742677444
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3640 | LCC = 0.7846 | SRCC = 0.7841 ]
+INFO:root:Average loss=0.23119884633865112
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3356 | LCC = 0.7910 | SRCC = 0.7891 ]
+INFO:root:Average loss=0.2504566865853774
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3813 | LCC = 0.7945 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.22866168923867056
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3258 | LCC = 0.7936 | SRCC = 0.7897 ]
+INFO:root:Average loss=0.20235701994254038
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4092 | LCC = 0.7973 | SRCC = 0.7944 ]
+INFO:root:Average loss=0.20871572693188986
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3421 | LCC = 0.7957 | SRCC = 0.7917 ]
+INFO:root:Average loss=0.20271458524541977
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3288 | LCC = 0.7950 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.19426856591151312
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3424 | LCC = 0.7922 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.18068612481539065
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3454 | LCC = 0.7953 | SRCC = 0.7925 ]
+INFO:root:Average loss=0.22155579217733481
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3586 | LCC = 0.7930 | SRCC = 0.7922 ]
+INFO:root:Average loss=0.17281864946469283
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3282 | LCC = 0.7958 | SRCC = 0.7925 ]
+INFO:root:Average loss=0.17149671998161536
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3621 | LCC = 0.7866 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.1633777135075667
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3290 | LCC = 0.7896 | SRCC = 0.7861 ]
+INFO:root:Average loss=0.164691912344633
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3280 | LCC = 0.7917 | SRCC = 0.7899 ]
+INFO:root:Average loss=0.17497703776909754
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3574 | LCC = 0.7909 | SRCC = 0.7895 ]
+INFO:root:Average loss=0.15623905337773836
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3140 | LCC = 0.7905 | SRCC = 0.7877 ]
+INFO:root:Average loss=0.1476696371459044
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3445 | LCC = 0.7858 | SRCC = 0.7819 ]
+INFO:root:Average loss=0.15407639340712473
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3378 | LCC = 0.7939 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.1539198377957711
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3845 | LCC = 0.7792 | SRCC = 0.7785 ]
+INFO:root:Average loss=0.147111481294418
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3298 | LCC = 0.7920 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.12842425016256478
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3434 | LCC = 0.7853 | SRCC = 0.7831 ]
+INFO:root:Average loss=0.12791333777400163
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3210 | LCC = 0.7916 | SRCC = 0.7886 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.4727 | LCC = 0.7961 | SRCC = 0.8036 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer18/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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_large_asr_en/layer18/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer18/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7e2eb3e5e9a36af996ff28ba1e8da2238685ab8f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f3920f38705c74d6d04a7f3fe5ac986c3f57ea71dc6fefee7dca0d931bf93d1
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer18/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ce633d54c0acb438d7b6f72577daf28dc5d91a4b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68ebda3c405c3dc8ce44d47731abc46f29684dc096cf8d9ab123d79b03b7f792
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer18/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..2cd0fdc30be51a08e3015579aa15d5e03999da8c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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=0.6612447439095913
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5898 | LCC = 0.5767 | SRCC = 0.5824 ]
+INFO:root:Average loss=0.4175907266445649
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4072 | LCC = 0.7425 | SRCC = 0.7462 ]
+INFO:root:Average loss=0.3482749811731852
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3744 | LCC = 0.7748 | SRCC = 0.7719 ]
+INFO:root:Average loss=0.315353733033706
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3600 | LCC = 0.7881 | SRCC = 0.7839 ]
+INFO:root:Average loss=0.30815917272598314
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3575 | LCC = 0.7932 | SRCC = 0.7885 ]
+INFO:root:Average loss=0.2883022379798767
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3618 | LCC = 0.7944 | SRCC = 0.7903 ]
+INFO:root:Average loss=0.32269049206605327
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3650 | LCC = 0.7917 | SRCC = 0.7903 ]
+INFO:root:Average loss=0.2622916667889326
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3846 | LCC = 0.7956 | SRCC = 0.7959 ]
+INFO:root:Average loss=0.2615417465567589
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3289 | LCC = 0.7993 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.2784638666571715
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3287 | LCC = 0.7983 | SRCC = 0.7943 ]
+INFO:root:Average loss=0.2506447155506183
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3303 | LCC = 0.7987 | SRCC = 0.7926 ]
+INFO:root:Average loss=0.23278762973271883
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4516 | LCC = 0.7772 | SRCC = 0.7807 ]
+INFO:root:Average loss=0.22371764213610917
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3149 | LCC = 0.7976 | SRCC = 0.7924 ]
+INFO:root:Average loss=0.22959341968481356
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3255 | LCC = 0.7947 | SRCC = 0.7920 ]
+INFO:root:Average loss=0.22521428572825897
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3986 | LCC = 0.7920 | SRCC = 0.7928 ]
+INFO:root:Average loss=0.22304367751647264
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3608 | LCC = 0.7874 | SRCC = 0.7857 ]
+INFO:root:Average loss=0.20353279625758147
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3438 | LCC = 0.7930 | SRCC = 0.7878 ]
+INFO:root:Average loss=0.1941765105495086
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3246 | LCC = 0.7928 | SRCC = 0.7894 ]
+INFO:root:Average loss=0.1868523258047226
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3265 | LCC = 0.7888 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.19702866940926284
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3122 | LCC = 0.7978 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.18590284750247613
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3254 | LCC = 0.7882 | SRCC = 0.7850 ]
+INFO:root:Average loss=0.16575849982790458
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3154 | LCC = 0.7910 | SRCC = 0.7874 ]
+INFO:root:Average loss=0.16042653175118643
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3191 | LCC = 0.7890 | SRCC = 0.7870 ]
+INFO:root:Average loss=0.15388470028455442
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3281 | LCC = 0.7897 | SRCC = 0.7889 ]
+INFO:root:Average loss=0.15991780295585975
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3122 | LCC = 0.7921 | SRCC = 0.7882 ]
+INFO:root:Average loss=0.15555444474403674
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3197 | LCC = 0.7928 | SRCC = 0.7882 ]
+INFO:root:Average loss=0.14707355153484222
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3461 | LCC = 0.7867 | SRCC = 0.7847 ]
+INFO:root:Average loss=0.14318261314661074
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3229 | LCC = 0.7899 | SRCC = 0.7867 ]
+INFO:root:Average loss=0.14612185926391527
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4313 | LCC = 0.7779 | SRCC = 0.7793 ]
+INFO:root:Average loss=0.13033067626066697
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3156 | LCC = 0.7899 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.13584847136949882
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3153 | LCC = 0.7887 | SRCC = 0.7860 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3190 | LCC = 0.8175 | SRCC = 0.8203 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer18/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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_large_asr_en/layer18/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer18/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..84792085e459e71e329feafc887493ea75534e63
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b59664b93c6246b2ed6bc2a694db27ca86381c7168752e10778aa0e0c021b744
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer18/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c4e1a204d2ca2163f0dc211d8c2b4da3d81fc31e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bc58b31845979890f261912a965fe2c2b64f3c6328f6ad6b9b8f801782344de
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer18/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer18/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ecf4903bdad53d30c29f3f65bfbe797f71114830
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer18/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=0.620705016148396
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5143 | LCC = 0.6488 | SRCC = 0.6529 ]
+INFO:root:Average loss=0.4299055742911803
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5073 | LCC = 0.7421 | SRCC = 0.7473 ]
+INFO:root:Average loss=0.37901009982212996
+INFO:root:
+[Valid][2][UTT][ MSE = 0.6384 | LCC = 0.7337 | SRCC = 0.7586 ]
+INFO:root:Average loss=0.35804880410432816
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3679 | LCC = 0.7709 | SRCC = 0.7703 ]
+INFO:root:Average loss=0.317875111523347
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3605 | LCC = 0.7820 | SRCC = 0.7808 ]
+INFO:root:Average loss=0.2855420552003078
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3504 | LCC = 0.7890 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.30648843065286296
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3334 | LCC = 0.7930 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.27025447690334076
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3482 | LCC = 0.7878 | SRCC = 0.7865 ]
+INFO:root:Average loss=0.2733399100028552
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3377 | LCC = 0.7956 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.2472452135422291
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3175 | LCC = 0.7940 | SRCC = 0.7893 ]
+INFO:root:Average loss=0.25132507697129863
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3262 | LCC = 0.7947 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.2326729708375075
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3082 | LCC = 0.8032 | SRCC = 0.7997 ]
+INFO:root:Average loss=0.2201961986720562
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3340 | LCC = 0.7932 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.21561670723633888
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3121 | LCC = 0.7964 | SRCC = 0.7937 ]
+INFO:root:Average loss=0.20966326158780318
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3326 | LCC = 0.8019 | SRCC = 0.7984 ]
+INFO:root:Average loss=0.2019918044217122
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3230 | LCC = 0.7944 | SRCC = 0.7909 ]
+INFO:root:Average loss=0.1892697040278178
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3133 | LCC = 0.7949 | SRCC = 0.7905 ]
+INFO:root:Average loss=0.19909662352158472
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3631 | LCC = 0.7939 | SRCC = 0.7917 ]
+INFO:root:Average loss=0.19228956112876916
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3277 | LCC = 0.7977 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.19329996187335405
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3807 | LCC = 0.7974 | SRCC = 0.7939 ]
+INFO:root:Average loss=0.1746395735595471
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3456 | LCC = 0.7877 | SRCC = 0.7848 ]
+INFO:root:Average loss=0.17260519988261736
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3158 | LCC = 0.7947 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.1811121917114808
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3234 | LCC = 0.7931 | SRCC = 0.7888 ]
+INFO:root:Average loss=0.17454615254432726
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3452 | LCC = 0.7837 | SRCC = 0.7794 ]
+INFO:root:Average loss=0.15564076182169792
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3249 | LCC = 0.7900 | SRCC = 0.7844 ]
+INFO:root:Average loss=0.14205458225348058
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3149 | LCC = 0.7915 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.14475278155161783
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3111 | LCC = 0.7942 | SRCC = 0.7904 ]
+INFO:root:Average loss=0.14409908012319833
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3224 | LCC = 0.7894 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.13968591067271355
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3959 | LCC = 0.7831 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.1425604686523095
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3203 | LCC = 0.7879 | SRCC = 0.7843 ]
+INFO:root:Average loss=0.13342578346148515
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3208 | LCC = 0.7874 | SRCC = 0.7834 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3018 | LCC = 0.8186 | SRCC = 0.8231 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer2/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer2/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer2/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f9926ed36d0100fe1f6bfb823aa611a6972dcad3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07d66af9f240144ed93d30ff583e7087a595590cbf7165438e21d3ec7deb162b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer2/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c507ca78544f089be9af15a434f1f35282a18fef
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b107b4e4aea67e52136bc8d5a0aa13dd979e1fe612803fd2691739fe480bf14a
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer2/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..dc4815e751b7fb58e0e667b1e2d1b696cdcfbb5c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5448629168363718
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4987 | LCC = 0.7278 | SRCC = 0.7424 ]
+INFO:root:Average loss=0.3735619600002582
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3612 | LCC = 0.7646 | SRCC = 0.7681 ]
+INFO:root:Average loss=0.3058084467282662
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3286 | LCC = 0.7982 | SRCC = 0.7957 ]
+INFO:root:Average loss=0.2775260304602293
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3106 | LCC = 0.8104 | SRCC = 0.8090 ]
+INFO:root:Average loss=0.24438482580276635
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2868 | LCC = 0.8313 | SRCC = 0.8281 ]
+INFO:root:Average loss=0.2281216969474768
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2867 | LCC = 0.8348 | SRCC = 0.8320 ]
+INFO:root:Average loss=0.218433854671625
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2778 | LCC = 0.8385 | SRCC = 0.8366 ]
+INFO:root:Average loss=0.2074329944757315
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2526 | LCC = 0.8479 | SRCC = 0.8457 ]
+INFO:root:Average loss=0.19310672628955963
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2644 | LCC = 0.8436 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.185204513466511
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2551 | LCC = 0.8490 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.17728332945933709
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2322 | LCC = 0.8566 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.1709509412638652
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2299 | LCC = 0.8604 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.16004643875818986
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2406 | LCC = 0.8621 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.15109363093208045
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2475 | LCC = 0.8567 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.14788174333098608
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2431 | LCC = 0.8620 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.14261956331439507
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2277 | LCC = 0.8613 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.13571210855092758
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2473 | LCC = 0.8601 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.13766248887166
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2520 | LCC = 0.8557 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.12423945027284133
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2130 | LCC = 0.8661 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.13331230720266318
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2321 | LCC = 0.8618 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.127897886129526
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2200 | LCC = 0.8612 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.11575032866153961
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2197 | LCC = 0.8606 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.1061496696411035
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2168 | LCC = 0.8649 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.11107473896864133
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2213 | LCC = 0.8632 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.12090412546426822
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2437 | LCC = 0.8599 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.09929641713507664
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2167 | LCC = 0.8611 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.09182747892844371
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2185 | LCC = 0.8632 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.08878763597936203
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2248 | LCC = 0.8617 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.08967026351736142
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2498 | LCC = 0.8554 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.08856717840983318
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2335 | LCC = 0.8623 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.08880227106886032
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2238 | LCC = 0.8603 | SRCC = 0.8598 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2243 | LCC = 0.8631 | SRCC = 0.8633 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer2/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer2/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer2/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5b5371dee8db3f755191abbf9c0ac60a2a025201
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b0b45db267061069f82c69d7bc6f08ca2ad9dc88ae1f760c25f9fa1e613b512
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer2/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..dff56a2255520d81a5171af4e57c556fa6ffe062
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:866455218e480a760de998e2d5707dafdafff2f9185b2ba1b13fa9cabc1320cb
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer2/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c59a8fb39230d68562ab04d204c436e5f4b047d2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5009243392791504
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4070 | LCC = 0.7554 | SRCC = 0.7556 ]
+INFO:root:Average loss=0.3270696527682818
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3459 | LCC = 0.7967 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.28988579794382435
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3037 | LCC = 0.8131 | SRCC = 0.8104 ]
+INFO:root:Average loss=0.26783973074112183
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3934 | LCC = 0.8224 | SRCC = 0.8247 ]
+INFO:root:Average loss=0.24383772584872368
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2850 | LCC = 0.8385 | SRCC = 0.8340 ]
+INFO:root:Average loss=0.23403263913515288
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2782 | LCC = 0.8444 | SRCC = 0.8394 ]
+INFO:root:Average loss=0.210208407865885
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2818 | LCC = 0.8425 | SRCC = 0.8404 ]
+INFO:root:Average loss=0.2035445766762281
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2769 | LCC = 0.8482 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.1985910881597262
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2390 | LCC = 0.8521 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.18409389133254686
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2784 | LCC = 0.8517 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.17199970877323395
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2909 | LCC = 0.8576 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.16042996608676055
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2611 | LCC = 0.8599 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.16411867928810608
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2423 | LCC = 0.8589 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.15433048600187668
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2462 | LCC = 0.8597 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.15010429651309282
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2603 | LCC = 0.8638 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.1445619214612704
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2271 | LCC = 0.8625 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.14426322567921418
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3125 | LCC = 0.8566 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.1305345074297526
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2309 | LCC = 0.8632 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.12486533553172381
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2299 | LCC = 0.8673 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.12710205876292327
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3349 | LCC = 0.8595 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.12546413773909593
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2212 | LCC = 0.8680 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.11998425185298309
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2355 | LCC = 0.8668 | SRCC = 0.8661 ]
+INFO:root:Average loss=0.11232606464853653
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2268 | LCC = 0.8671 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.10300317674111097
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2244 | LCC = 0.8669 | SRCC = 0.8667 ]
+INFO:root:Average loss=0.10152096043412502
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2304 | LCC = 0.8631 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.09734077450747673
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2836 | LCC = 0.8630 | SRCC = 0.8652 ]
+INFO:root:Average loss=0.09783914384360497
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2424 | LCC = 0.8664 | SRCC = 0.8677 ]
+INFO:root:Average loss=0.09305632802156302
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2265 | LCC = 0.8663 | SRCC = 0.8670 ]
+INFO:root:Average loss=0.0863001415362725
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2333 | LCC = 0.8654 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.08970705715891643
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2397 | LCC = 0.8653 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.08639181247697426
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2297 | LCC = 0.8635 | SRCC = 0.8629 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2208 | LCC = 0.8634 | SRCC = 0.8635 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer2/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer2/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer2/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f5b391a751d967514acef3a74628528437519c81
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c9e362d80931525bb94efd88c6d671876756c2424b4fc53ec025291a7e77a44
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer2/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1bcb4d0ceffae7dc689b78427eaf08d8d7a43a13
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56798790d8b87d548ae9f87ab70704971fcff6ec4aa687e45189fe97687ee478
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer2/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..897574bf82b1a41e7cc97f7b53309ac442bd9397
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5889513393243154
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4313 | LCC = 0.7317 | SRCC = 0.7332 ]
+INFO:root:Average loss=0.32772406744651306
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3435 | LCC = 0.7835 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.29447240153184306
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3096 | LCC = 0.8132 | SRCC = 0.8119 ]
+INFO:root:Average loss=0.2620118860250864
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3812 | LCC = 0.8126 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.25861442165497023
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2768 | LCC = 0.8286 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.2390630841255188
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3150 | LCC = 0.8334 | SRCC = 0.8366 ]
+INFO:root:Average loss=0.22180476822914222
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2708 | LCC = 0.8347 | SRCC = 0.8349 ]
+INFO:root:Average loss=0.21432542074949312
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2662 | LCC = 0.8417 | SRCC = 0.8405 ]
+INFO:root:Average loss=0.2136356119926159
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2895 | LCC = 0.8394 | SRCC = 0.8437 ]
+INFO:root:Average loss=0.20659936314974076
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2521 | LCC = 0.8535 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.2000135998122203
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2413 | LCC = 0.8512 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.18250063873636416
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2528 | LCC = 0.8487 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.17457096775372824
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2281 | LCC = 0.8558 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.17086050334649208
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2370 | LCC = 0.8595 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.16444322915795523
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2424 | LCC = 0.8566 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.15044987440491334
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2664 | LCC = 0.8498 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.15168741316749498
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2277 | LCC = 0.8637 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.14398523993217027
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2275 | LCC = 0.8608 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.13185548008634493
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2579 | LCC = 0.8611 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.13090895861387253
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2343 | LCC = 0.8645 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.13206800961723694
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2286 | LCC = 0.8617 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.12073043552346718
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2235 | LCC = 0.8630 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.12107177250660382
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2277 | LCC = 0.8614 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.10572925343727455
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2366 | LCC = 0.8581 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.1112573283414046
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2321 | LCC = 0.8616 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.12770637057912657
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2150 | LCC = 0.8664 | SRCC = 0.8670 ]
+INFO:root:Average loss=0.10709270992531227
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2175 | LCC = 0.8650 | SRCC = 0.8664 ]
+INFO:root:Average loss=0.10094710897940856
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2249 | LCC = 0.8669 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.1002285528737001
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2159 | LCC = 0.8653 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.09020494134762348
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2201 | LCC = 0.8688 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.08908839065294999
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2162 | LCC = 0.8625 | SRCC = 0.8635 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2524 | LCC = 0.8553 | SRCC = 0.8572 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer2/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer2/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer2/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..180452eb9a2fcfbe1a17777d8c2b8591df45fe79
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01bf4c796868b8dfe1e32878df53b13d192618c7c59f2d336f180bdbacb69b7b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer2/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..19735297434a43ed3c004241ca23c6e38a6f20af
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f066fc6623f3b64c592fb0f1ec6a3ee55146993e62c7c5638bc08055cbb0ed3e
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer2/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f100ef146ab2685aae7649ca40cb0134c2fc1361
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5043551043058053
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3880 | LCC = 0.7483 | SRCC = 0.7474 ]
+INFO:root:Average loss=0.32158466695975035
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3181 | LCC = 0.7945 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.30819933594037324
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3028 | LCC = 0.8074 | SRCC = 0.8080 ]
+INFO:root:Average loss=0.26186287517731005
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2768 | LCC = 0.8235 | SRCC = 0.8237 ]
+INFO:root:Average loss=0.2515006510493083
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3445 | LCC = 0.8195 | SRCC = 0.8230 ]
+INFO:root:Average loss=0.2466420401365329
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2719 | LCC = 0.8316 | SRCC = 0.8345 ]
+INFO:root:Average loss=0.2441398560618743
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3197 | LCC = 0.8371 | SRCC = 0.8373 ]
+INFO:root:Average loss=0.22295396488446456
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2624 | LCC = 0.8389 | SRCC = 0.8387 ]
+INFO:root:Average loss=0.20400342373893812
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2698 | LCC = 0.8500 | SRCC = 0.8472 ]
+INFO:root:Average loss=0.19903526359643692
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3092 | LCC = 0.8342 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.18693790613458708
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2573 | LCC = 0.8559 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.18123306716099763
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2538 | LCC = 0.8488 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.16965399138056314
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2419 | LCC = 0.8494 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.16198541997716978
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2328 | LCC = 0.8549 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.15889191531982177
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2470 | LCC = 0.8543 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.14931924793964776
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2661 | LCC = 0.8565 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.15374425168220812
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2490 | LCC = 0.8557 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.14311337767121118
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3165 | LCC = 0.8447 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.13772323240454382
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2529 | LCC = 0.8519 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.12473351078537795
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2212 | LCC = 0.8606 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.1227820555273539
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2268 | LCC = 0.8603 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.1272132056645858
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2302 | LCC = 0.8578 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.11706842176425152
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2302 | LCC = 0.8591 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.11852550081526622
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2747 | LCC = 0.8438 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.12043083783907768
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2304 | LCC = 0.8578 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.10087116062641144
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2383 | LCC = 0.8510 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.09647660950819652
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2364 | LCC = 0.8547 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.09546238962465371
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2286 | LCC = 0.8581 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.10264475280657792
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2331 | LCC = 0.8604 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.08980958779843953
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2520 | LCC = 0.8565 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.09254464612175257
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2244 | LCC = 0.8593 | SRCC = 0.8575 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2423 | LCC = 0.8587 | SRCC = 0.8606 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer2/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer2/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer2/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..12c001ef980cd9955cb72a5a84dd69d8edb1dbc8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42050c4b48568a469d181774c22d5a8ce997ed76412a6d7a4bd18ee7d5f7d0f3
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer2/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1ca7615652345c21c1785de04d008070f9e8e046
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer2/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55e9c68a5af7f38cadeec1f6c7219fa7c867bdbd4129e78965d46eb0f27b59dd
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer2/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer2/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c82e79519ce82649ccbe9eb5c51c81679ace27fe
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5966755686662136
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4231 | LCC = 0.7345 | SRCC = 0.7418 ]
+INFO:root:Average loss=0.3413202309837708
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3781 | LCC = 0.7825 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.2821957717339198
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2912 | LCC = 0.8150 | SRCC = 0.8142 ]
+INFO:root:Average loss=0.2584948071684593
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2793 | LCC = 0.8229 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.23240535075847918
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2919 | LCC = 0.8391 | SRCC = 0.8384 ]
+INFO:root:Average loss=0.23121264825264612
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2461 | LCC = 0.8442 | SRCC = 0.8439 ]
+INFO:root:Average loss=0.21080050751184806
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2685 | LCC = 0.8487 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.20933503189530128
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2602 | LCC = 0.8498 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.20992823575551695
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2749 | LCC = 0.8530 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.18949702639992422
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2363 | LCC = 0.8579 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.1846547574760058
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2378 | LCC = 0.8586 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.1785829167526502
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2580 | LCC = 0.8595 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.15748495130966872
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2476 | LCC = 0.8567 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.16308496816036028
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2250 | LCC = 0.8569 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.14919450726264563
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2316 | LCC = 0.8610 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.14066965648761162
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2115 | LCC = 0.8668 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.13898263823909637
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2516 | LCC = 0.8626 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.13097439849605927
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2308 | LCC = 0.8609 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.12789499062376145
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2136 | LCC = 0.8635 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.11777637832057782
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2232 | LCC = 0.8600 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.11878331989431992
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2178 | LCC = 0.8682 | SRCC = 0.8691 ]
+INFO:root:Average loss=0.12059216191753364
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2112 | LCC = 0.8675 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.11308853084651324
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2283 | LCC = 0.8652 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.1185444876169547
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2269 | LCC = 0.8647 | SRCC = 0.8659 ]
+INFO:root:Average loss=0.10135682528981796
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2363 | LCC = 0.8644 | SRCC = 0.8664 ]
+INFO:root:Average loss=0.10671338816292775
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2275 | LCC = 0.8649 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.0944781159170163
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2438 | LCC = 0.8644 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.09732629511600886
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2183 | LCC = 0.8680 | SRCC = 0.8664 ]
+INFO:root:Average loss=0.08572880054513614
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2196 | LCC = 0.8661 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.0892558261179007
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2358 | LCC = 0.8646 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.08316670410717145
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2202 | LCC = 0.8639 | SRCC = 0.8635 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2549 | LCC = 0.8600 | SRCC = 0.8634 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer20/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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_large_asr_en/layer20/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer20/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d7e70b9cd17a706a7bfa696c27753d9e12e9bb13
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45c296f6d32208f9a65706484325701bf169dbff127180f0f1ccc7fa503f307c
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer20/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..09296382f8671afa5d65d1f32daebbb9ab1efb17
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43a7a86c4019d32a0981bb999ef837bc18f87bda7232e33851594591e4cf6bd8
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer20/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..292e12680d5f14a92de7c8f24449d1dc4f7bc566
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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=0.6229916142347531
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6316 | LCC = 0.6025 | SRCC = 0.6046 ]
+INFO:root:Average loss=0.4431934272631621
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4374 | LCC = 0.7077 | SRCC = 0.7078 ]
+INFO:root:Average loss=0.3708682119464263
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4263 | LCC = 0.7513 | SRCC = 0.7516 ]
+INFO:root:Average loss=0.357066358320224
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3960 | LCC = 0.7524 | SRCC = 0.7505 ]
+INFO:root:Average loss=0.33499239251399654
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3760 | LCC = 0.7697 | SRCC = 0.7680 ]
+INFO:root:Average loss=0.32679426975739306
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4630 | LCC = 0.7638 | SRCC = 0.7659 ]
+INFO:root:Average loss=0.3036182629756438
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3484 | LCC = 0.7777 | SRCC = 0.7759 ]
+INFO:root:Average loss=0.3197106306369488
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4393 | LCC = 0.7721 | SRCC = 0.7725 ]
+INFO:root:Average loss=0.2890735883743335
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3352 | LCC = 0.7878 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.2704694290191699
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3666 | LCC = 0.7838 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.2795221392924969
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3533 | LCC = 0.7770 | SRCC = 0.7729 ]
+INFO:root:Average loss=0.2553427267151001
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3613 | LCC = 0.7829 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.26934622056209123
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3367 | LCC = 0.7917 | SRCC = 0.7874 ]
+INFO:root:Average loss=0.2572085951001216
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3420 | LCC = 0.7951 | SRCC = 0.7896 ]
+INFO:root:Average loss=0.2317669904575898
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4040 | LCC = 0.7816 | SRCC = 0.7782 ]
+INFO:root:Average loss=0.23956962178150812
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3387 | LCC = 0.7892 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.2379731562657234
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3551 | LCC = 0.7882 | SRCC = 0.7865 ]
+INFO:root:Average loss=0.21905355938734153
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3456 | LCC = 0.7847 | SRCC = 0.7800 ]
+INFO:root:Average loss=0.23269043633571038
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3517 | LCC = 0.7901 | SRCC = 0.7873 ]
+INFO:root:Average loss=0.2117307866230989
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3176 | LCC = 0.7959 | SRCC = 0.7920 ]
+INFO:root:Average loss=0.19737250682635185
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3687 | LCC = 0.7857 | SRCC = 0.7840 ]
+INFO:root:Average loss=0.19814433444004792
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3392 | LCC = 0.7859 | SRCC = 0.7817 ]
+INFO:root:Average loss=0.20124542674957177
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3333 | LCC = 0.7914 | SRCC = 0.7867 ]
+INFO:root:Average loss=0.19877294833079362
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3379 | LCC = 0.7828 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.18635217090829825
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3327 | LCC = 0.7896 | SRCC = 0.7869 ]
+INFO:root:Average loss=0.19122684670564455
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3408 | LCC = 0.7936 | SRCC = 0.7905 ]
+INFO:root:Average loss=0.1812684218852948
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3413 | LCC = 0.7868 | SRCC = 0.7822 ]
+INFO:root:Average loss=0.17434646198764825
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3422 | LCC = 0.7865 | SRCC = 0.7840 ]
+INFO:root:Average loss=0.17858150954812002
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3542 | LCC = 0.7828 | SRCC = 0.7812 ]
+INFO:root:Average loss=0.1731518528973445
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3429 | LCC = 0.7864 | SRCC = 0.7843 ]
+INFO:root:Average loss=0.16330494874945053
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3273 | LCC = 0.7879 | SRCC = 0.7845 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3327 | LCC = 0.7818 | SRCC = 0.7816 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer20/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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_large_asr_en/layer20/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer20/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f3c54477f7b3c9cb7f3589cab9993605d1e9d6cd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87a491636b30b8f3cb28ea6b069207e30b564fa7aeb6e52c2715ddbc3c6b728a
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer20/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..debb815a143d333402de9f5e90e492c33fede74f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16d14df8b1f340cb97acd6e6f3e29125f3d9fc4dc5fee76cdd8915226d17fb40
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer20/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..83faaed3fd840423e0aeb00dddcc420ed4ac45cf
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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=0.6542089035113653
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6339 | LCC = 0.5366 | SRCC = 0.5336 ]
+INFO:root:Average loss=0.4502780544452178
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4283 | LCC = 0.7031 | SRCC = 0.7040 ]
+INFO:root:Average loss=0.4070443604619075
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4863 | LCC = 0.7454 | SRCC = 0.7549 ]
+INFO:root:Average loss=0.37299549732452786
+INFO:root:
+[Valid][3][UTT][ MSE = 0.5270 | LCC = 0.7518 | SRCC = 0.7596 ]
+INFO:root:Average loss=0.36118285491680485
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3868 | LCC = 0.7679 | SRCC = 0.7663 ]
+INFO:root:Average loss=0.3272971959832387
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3854 | LCC = 0.7769 | SRCC = 0.7754 ]
+INFO:root:Average loss=0.33491277618285936
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4000 | LCC = 0.7726 | SRCC = 0.7698 ]
+INFO:root:Average loss=0.3189687157670657
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3677 | LCC = 0.7769 | SRCC = 0.7738 ]
+INFO:root:Average loss=0.2943552495577397
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3507 | LCC = 0.7819 | SRCC = 0.7768 ]
+INFO:root:Average loss=0.280481297427263
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3451 | LCC = 0.7823 | SRCC = 0.7783 ]
+INFO:root:Average loss=0.2693167166450085
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3610 | LCC = 0.7812 | SRCC = 0.7788 ]
+INFO:root:Average loss=0.2654354730859781
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3722 | LCC = 0.7788 | SRCC = 0.7751 ]
+INFO:root:Average loss=0.24484441830561712
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3868 | LCC = 0.7760 | SRCC = 0.7718 ]
+INFO:root:Average loss=0.24137951586491022
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4305 | LCC = 0.7718 | SRCC = 0.7685 ]
+INFO:root:Average loss=0.2541380620155579
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4047 | LCC = 0.7745 | SRCC = 0.7745 ]
+INFO:root:Average loss=0.22993140056347236
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3419 | LCC = 0.7822 | SRCC = 0.7773 ]
+INFO:root:Average loss=0.23512060233415702
+INFO:root:
+[Valid][16][UTT][ MSE = 0.4064 | LCC = 0.7777 | SRCC = 0.7735 ]
+INFO:root:Average loss=0.22723315732601362
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3519 | LCC = 0.7781 | SRCC = 0.7732 ]
+INFO:root:Average loss=0.22203898582703027
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3660 | LCC = 0.7765 | SRCC = 0.7720 ]
+INFO:root:Average loss=0.2304110844165851
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3847 | LCC = 0.7772 | SRCC = 0.7757 ]
+INFO:root:Average loss=0.20635513388193572
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3931 | LCC = 0.7810 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.2588891524534959
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3503 | LCC = 0.7773 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.20674703747798234
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3775 | LCC = 0.7763 | SRCC = 0.7737 ]
+INFO:root:Average loss=0.2016318607597779
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3600 | LCC = 0.7779 | SRCC = 0.7737 ]
+INFO:root:Average loss=0.21666315102424377
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3477 | LCC = 0.7826 | SRCC = 0.7785 ]
+INFO:root:Average loss=0.18890720901963037
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3610 | LCC = 0.7836 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.18817025910203272
+INFO:root:
+[Valid][26][UTT][ MSE = 0.4180 | LCC = 0.7674 | SRCC = 0.7650 ]
+INFO:root:Average loss=0.1795005303544876
+INFO:root:
+[Valid][27][UTT][ MSE = 0.4063 | LCC = 0.7722 | SRCC = 0.7723 ]
+INFO:root:Average loss=0.1722664173023823
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4110 | LCC = 0.7751 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.17171561288145873
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3420 | LCC = 0.7779 | SRCC = 0.7749 ]
+INFO:root:Average loss=0.1721250235270231
+INFO:root:
+[Valid][30][UTT][ MSE = 0.4466 | LCC = 0.7784 | SRCC = 0.7789 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3235 | LCC = 0.7940 | SRCC = 0.7917 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer20/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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_large_asr_en/layer20/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer20/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..99d9535788424c67de450bdb12e59626365b1a0a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb6ec134f82d12e015ab6f41f72af6814bdbaae1ea255430325d078e7245b2d5
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer20/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fe62e61ae89eb26c6b3f971f722d5553a47b7c6b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d996fadff1dbfe9e791e8cbb15765cb55551d2b7eabb52ae72d3e3f339778964
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer20/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..95c18f54e4ef01613048730f7441fd0d504e1065
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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=0.6908769275133426
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6824 | LCC = 0.5131 | SRCC = 0.5130 ]
+INFO:root:Average loss=0.45781128452374387
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4631 | LCC = 0.7139 | SRCC = 0.7139 ]
+INFO:root:Average loss=0.40511363859360033
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4104 | LCC = 0.7592 | SRCC = 0.7584 ]
+INFO:root:Average loss=0.3456517230623808
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3696 | LCC = 0.7689 | SRCC = 0.7672 ]
+INFO:root:Average loss=0.3496958418534352
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3628 | LCC = 0.7847 | SRCC = 0.7836 ]
+INFO:root:Average loss=0.30489069051467454
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3739 | LCC = 0.7766 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.3042036044674042
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3523 | LCC = 0.7905 | SRCC = 0.7869 ]
+INFO:root:Average loss=0.2858631962384933
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3690 | LCC = 0.7661 | SRCC = 0.7641 ]
+INFO:root:Average loss=0.26964332220646053
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3629 | LCC = 0.7922 | SRCC = 0.7896 ]
+INFO:root:Average loss=0.2805905193090439
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3711 | LCC = 0.7871 | SRCC = 0.7844 ]
+INFO:root:Average loss=0.2576861633704259
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3284 | LCC = 0.7965 | SRCC = 0.7934 ]
+INFO:root:Average loss=0.2549732761123242
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3509 | LCC = 0.7975 | SRCC = 0.7956 ]
+INFO:root:Average loss=0.23582181907617128
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3650 | LCC = 0.7823 | SRCC = 0.7789 ]
+INFO:root:Average loss=0.24248983137882674
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3308 | LCC = 0.7923 | SRCC = 0.7891 ]
+INFO:root:Average loss=0.23606906410975334
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3285 | LCC = 0.7931 | SRCC = 0.7901 ]
+INFO:root:Average loss=0.22376566322950217
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3370 | LCC = 0.7947 | SRCC = 0.7913 ]
+INFO:root:Average loss=0.23098966746758193
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3326 | LCC = 0.7869 | SRCC = 0.7831 ]
+INFO:root:Average loss=0.23274932610682952
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3683 | LCC = 0.7835 | SRCC = 0.7819 ]
+INFO:root:Average loss=0.21384690568233147
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3862 | LCC = 0.7789 | SRCC = 0.7783 ]
+INFO:root:Average loss=0.20553086315974212
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3535 | LCC = 0.7889 | SRCC = 0.7884 ]
+INFO:root:Average loss=0.19836876187951136
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3333 | LCC = 0.7908 | SRCC = 0.7874 ]
+INFO:root:Average loss=0.19090498620883012
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3378 | LCC = 0.7773 | SRCC = 0.7740 ]
+INFO:root:Average loss=0.1988059708323234
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3651 | LCC = 0.7854 | SRCC = 0.7817 ]
+INFO:root:Average loss=0.18776545874201334
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3472 | LCC = 0.7789 | SRCC = 0.7773 ]
+INFO:root:Average loss=0.18013988444820428
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3362 | LCC = 0.7864 | SRCC = 0.7837 ]
+INFO:root:Average loss=0.1703842431306839
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3340 | LCC = 0.7805 | SRCC = 0.7766 ]
+INFO:root:Average loss=0.1761443252937916
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3389 | LCC = 0.7827 | SRCC = 0.7788 ]
+INFO:root:Average loss=0.1901106292811724
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3393 | LCC = 0.7782 | SRCC = 0.7742 ]
+INFO:root:Average loss=0.16217985434027818
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3318 | LCC = 0.7819 | SRCC = 0.7786 ]
+INFO:root:Average loss=0.17875289448942894
+INFO:root:
+[Valid][29][UTT][ MSE = 0.4021 | LCC = 0.7882 | SRCC = 0.7872 ]
+INFO:root:Average loss=0.1765448207465502
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3495 | LCC = 0.7770 | SRCC = 0.7763 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3707 | LCC = 0.7925 | SRCC = 0.7981 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer20/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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_large_asr_en/layer20/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer20/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e0ef4957cbca9088f54f74760d6c3582962d6c16
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe70d66670644776041f9529d05dbc35b6e2b6d718c6e262e63489fa64c899dd
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer20/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..316b53f341175ce7ac6c86d67ed5a7b2a5210d1e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed5933dd01129a58ff666fd8cdec5708f9c306aef4a0af301a59c1bb2bddc47a
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer20/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ee61719520a25d12137732ffc4f94410974b5ba5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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=0.6433453850257091
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5670 | LCC = 0.5929 | SRCC = 0.5930 ]
+INFO:root:Average loss=0.44703276722859114
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4727 | LCC = 0.7164 | SRCC = 0.7174 ]
+INFO:root:Average loss=0.3869521594964541
+INFO:root:
+[Valid][2][UTT][ MSE = 0.5233 | LCC = 0.7459 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.4033991205386626
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4352 | LCC = 0.7516 | SRCC = 0.7505 ]
+INFO:root:Average loss=0.347045059578541
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3798 | LCC = 0.7763 | SRCC = 0.7748 ]
+INFO:root:Average loss=0.33171335875223845
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3875 | LCC = 0.7675 | SRCC = 0.7659 ]
+INFO:root:Average loss=0.31523044178119075
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3768 | LCC = 0.7801 | SRCC = 0.7770 ]
+INFO:root:Average loss=0.2986804684385275
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3817 | LCC = 0.7832 | SRCC = 0.7827 ]
+INFO:root:Average loss=0.2844089749149787
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3574 | LCC = 0.7773 | SRCC = 0.7770 ]
+INFO:root:Average loss=0.2806875220476053
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3534 | LCC = 0.7876 | SRCC = 0.7859 ]
+INFO:root:Average loss=0.26458843453572345
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3595 | LCC = 0.7843 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.2736114848118562
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3619 | LCC = 0.7792 | SRCC = 0.7773 ]
+INFO:root:Average loss=0.27666906401132924
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3393 | LCC = 0.7878 | SRCC = 0.7864 ]
+INFO:root:Average loss=0.24929902033928114
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3592 | LCC = 0.7895 | SRCC = 0.7877 ]
+INFO:root:Average loss=0.24690100053946176
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3471 | LCC = 0.7870 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.2267589943531232
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3328 | LCC = 0.7855 | SRCC = 0.7830 ]
+INFO:root:Average loss=0.2261811028688382
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3673 | LCC = 0.7731 | SRCC = 0.7751 ]
+INFO:root:Average loss=0.22960782223022902
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3915 | LCC = 0.7864 | SRCC = 0.7850 ]
+INFO:root:Average loss=0.2218629987193988
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3793 | LCC = 0.7611 | SRCC = 0.7577 ]
+INFO:root:Average loss=0.22772129185688803
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3831 | LCC = 0.7801 | SRCC = 0.7778 ]
+INFO:root:Average loss=0.22244416654874116
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3715 | LCC = 0.7854 | SRCC = 0.7828 ]
+INFO:root:Average loss=0.20315144545374772
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3509 | LCC = 0.7836 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.203089938331873
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3215 | LCC = 0.7864 | SRCC = 0.7837 ]
+INFO:root:Average loss=0.1914045201280178
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3568 | LCC = 0.7844 | SRCC = 0.7816 ]
+INFO:root:Average loss=0.18469780855453932
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3357 | LCC = 0.7852 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.1823237515412844
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3348 | LCC = 0.7809 | SRCC = 0.7787 ]
+INFO:root:Average loss=0.18473496574621934
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3438 | LCC = 0.7808 | SRCC = 0.7788 ]
+INFO:root:Average loss=0.16517463135413635
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3519 | LCC = 0.7710 | SRCC = 0.7667 ]
+INFO:root:Average loss=0.17085311389886415
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3244 | LCC = 0.7894 | SRCC = 0.7893 ]
+INFO:root:Average loss=0.16510364699822205
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3606 | LCC = 0.7798 | SRCC = 0.7752 ]
+INFO:root:Average loss=0.16361612329880396
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3599 | LCC = 0.7756 | SRCC = 0.7740 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3772 | LCC = 0.7888 | SRCC = 0.7907 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer20/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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_large_asr_en/layer20/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer20/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4a231ac33e57c868021264ae15e36a41627e7374
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:929d49c0b67b5d6630ae3ca4915e811c2730e8c5d375d240a662e1cd6cd00399
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer20/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..91eaf47a4efe3466443667cd9bef382d43c90cf9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6666898a137427a48694123d0eec09a0d6208e5fc55cd3673ca2651d3622a540
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer20/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer20/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..cf07d6751ce3cee16e63b55928961e46c0bd92f1
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer20/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=0.6416271317463654
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5766 | LCC = 0.5770 | SRCC = 0.5773 ]
+INFO:root:Average loss=0.5159485332476788
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5419 | LCC = 0.6629 | SRCC = 0.6663 ]
+INFO:root:Average loss=0.3854274593102626
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4075 | LCC = 0.7478 | SRCC = 0.7470 ]
+INFO:root:Average loss=0.34977272076484484
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4601 | LCC = 0.7654 | SRCC = 0.7685 ]
+INFO:root:Average loss=0.35748951805707735
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3620 | LCC = 0.7781 | SRCC = 0.7742 ]
+INFO:root:Average loss=0.31423807545350146
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3665 | LCC = 0.7828 | SRCC = 0.7798 ]
+INFO:root:Average loss=0.30235078644294006
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3486 | LCC = 0.7894 | SRCC = 0.7852 ]
+INFO:root:Average loss=0.3045484096958087
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4009 | LCC = 0.7831 | SRCC = 0.7845 ]
+INFO:root:Average loss=0.2924057831748938
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3498 | LCC = 0.7859 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.2828804393991446
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3528 | LCC = 0.7908 | SRCC = 0.7866 ]
+INFO:root:Average loss=0.2660055214013809
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3436 | LCC = 0.7849 | SRCC = 0.7803 ]
+INFO:root:Average loss=0.26292293748030293
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3310 | LCC = 0.7886 | SRCC = 0.7847 ]
+INFO:root:Average loss=0.25146647600027233
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3433 | LCC = 0.7941 | SRCC = 0.7933 ]
+INFO:root:Average loss=0.24623468040655822
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3429 | LCC = 0.7927 | SRCC = 0.7886 ]
+INFO:root:Average loss=0.24724859515061745
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3310 | LCC = 0.7896 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.22428938211538854
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3323 | LCC = 0.7867 | SRCC = 0.7831 ]
+INFO:root:Average loss=0.21388575721245545
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3620 | LCC = 0.7832 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.2143234647810459
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3343 | LCC = 0.7933 | SRCC = 0.7890 ]
+INFO:root:Average loss=0.20692490805418062
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3359 | LCC = 0.7769 | SRCC = 0.7750 ]
+INFO:root:Average loss=0.2166112413009008
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3673 | LCC = 0.7929 | SRCC = 0.7911 ]
+INFO:root:Average loss=0.1953373350776159
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3459 | LCC = 0.7757 | SRCC = 0.7724 ]
+INFO:root:Average loss=0.1924275373801207
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3597 | LCC = 0.7861 | SRCC = 0.7831 ]
+INFO:root:Average loss=0.18850624800110474
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3534 | LCC = 0.7828 | SRCC = 0.7797 ]
+INFO:root:Average loss=0.18766360013530806
+INFO:root:
+[Valid][23][UTT][ MSE = 0.4306 | LCC = 0.7726 | SRCC = 0.7708 ]
+INFO:root:Average loss=0.18363918803441218
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3633 | LCC = 0.7773 | SRCC = 0.7748 ]
+INFO:root:Average loss=0.17677668797282073
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3456 | LCC = 0.7730 | SRCC = 0.7695 ]
+INFO:root:Average loss=0.17392522669755495
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3368 | LCC = 0.7858 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.1717687128828122
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3616 | LCC = 0.7786 | SRCC = 0.7770 ]
+INFO:root:Average loss=0.16646470855443907
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4159 | LCC = 0.7683 | SRCC = 0.7678 ]
+INFO:root:Average loss=0.16682464428819144
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3720 | LCC = 0.7737 | SRCC = 0.7706 ]
+INFO:root:Average loss=0.15743470220611647
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3587 | LCC = 0.7784 | SRCC = 0.7733 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3208 | LCC = 0.7955 | SRCC = 0.7929 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer22/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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_large_asr_en/layer22/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer22/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d8d83ee7174c6d19a05ece662e3a9ee490dafeb3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28b45baa0a45bb53799f502bb3945449077c1a8234b1a3578d69cb6589eac6c5
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer22/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..93c8cf80b64fc83917358b9e25daf1ff487f5a90
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4970bc2b8a644af1a97b530c4007047898c62c120110d59793279a35ae908d8b
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer22/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c78dfd7051e86e3b00126be703963f8312d3f0dd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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=0.6465839128463696
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6371 | LCC = 0.5033 | SRCC = 0.5034 ]
+INFO:root:Average loss=0.5048288275034
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5283 | LCC = 0.6582 | SRCC = 0.6611 ]
+INFO:root:Average loss=0.44086953424490416
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4973 | LCC = 0.7112 | SRCC = 0.7127 ]
+INFO:root:Average loss=0.3993158371020586
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4485 | LCC = 0.7205 | SRCC = 0.7231 ]
+INFO:root:Average loss=0.3616277705400418
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3982 | LCC = 0.7339 | SRCC = 0.7345 ]
+INFO:root:Average loss=0.3682943595907627
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4485 | LCC = 0.7556 | SRCC = 0.7549 ]
+INFO:root:Average loss=0.33705301678333527
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4005 | LCC = 0.7512 | SRCC = 0.7498 ]
+INFO:root:Average loss=0.3157298240141991
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3895 | LCC = 0.7629 | SRCC = 0.7622 ]
+INFO:root:Average loss=0.32175525354269224
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4732 | LCC = 0.7706 | SRCC = 0.7690 ]
+INFO:root:Average loss=0.2996225460217549
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3604 | LCC = 0.7710 | SRCC = 0.7730 ]
+INFO:root:Average loss=0.31240536731023055
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3764 | LCC = 0.7755 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.2870524817934403
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4427 | LCC = 0.7728 | SRCC = 0.7729 ]
+INFO:root:Average loss=0.30936927195542896
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3648 | LCC = 0.7731 | SRCC = 0.7703 ]
+INFO:root:Average loss=0.27827458675855243
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3578 | LCC = 0.7763 | SRCC = 0.7751 ]
+INFO:root:Average loss=0.2647558036140906
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3594 | LCC = 0.7732 | SRCC = 0.7731 ]
+INFO:root:Average loss=0.2671168538240286
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3449 | LCC = 0.7738 | SRCC = 0.7725 ]
+INFO:root:Average loss=0.24723372207238123
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3502 | LCC = 0.7778 | SRCC = 0.7768 ]
+INFO:root:Average loss=0.25076266397268343
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3593 | LCC = 0.7827 | SRCC = 0.7819 ]
+INFO:root:Average loss=0.2442904722232085
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3426 | LCC = 0.7826 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.22686027945616308
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3286 | LCC = 0.7771 | SRCC = 0.7757 ]
+INFO:root:Average loss=0.23610981152607843
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3308 | LCC = 0.7820 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.22025622265079084
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3492 | LCC = 0.7747 | SRCC = 0.7726 ]
+INFO:root:Average loss=0.24376349303966913
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3453 | LCC = 0.7725 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.2152591799505246
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3562 | LCC = 0.7741 | SRCC = 0.7719 ]
+INFO:root:Average loss=0.2022304001909036
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3592 | LCC = 0.7697 | SRCC = 0.7680 ]
+INFO:root:Average loss=0.20198181500801674
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3398 | LCC = 0.7827 | SRCC = 0.7811 ]
+INFO:root:Average loss=0.20385970337650713
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3309 | LCC = 0.7786 | SRCC = 0.7764 ]
+INFO:root:Average loss=0.19744073513608712
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3723 | LCC = 0.7705 | SRCC = 0.7694 ]
+INFO:root:Average loss=0.19240317656061587
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3906 | LCC = 0.7649 | SRCC = 0.7642 ]
+INFO:root:Average loss=0.18861712821019003
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3315 | LCC = 0.7815 | SRCC = 0.7787 ]
+INFO:root:Average loss=0.1797840643960696
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3558 | LCC = 0.7739 | SRCC = 0.7717 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3772 | LCC = 0.7799 | SRCC = 0.7837 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer22/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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_large_asr_en/layer22/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer22/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0f209cddf7b11d6757d2ea470f1139afefdba6b1
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b0f1ec074c40fe9cb1c52e4622e89f8b35593c60a39d7f7c280f420487f5fe9
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer22/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8f5411bc4e54a85cbb14f2068800e058299cebf8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fff91cef116f25ce9d174b33a5ab4c08f9f0017e9b3ac1357ef22cbfc60cf1c
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer22/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..3b18be255272b332c0dd848e37af82539fd95046
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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=0.6545122586763822
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6392 | LCC = 0.4902 | SRCC = 0.4937 ]
+INFO:root:Average loss=0.5073018849660189
+INFO:root:
+[Valid][1][UTT][ MSE = 0.6842 | LCC = 0.6229 | SRCC = 0.6518 ]
+INFO:root:Average loss=0.4508979698786369
+INFO:root:
+[Valid][2][UTT][ MSE = 0.5881 | LCC = 0.6433 | SRCC = 0.6551 ]
+INFO:root:Average loss=0.4134330399907552
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4390 | LCC = 0.7183 | SRCC = 0.7238 ]
+INFO:root:Average loss=0.3702701444809253
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4159 | LCC = 0.7422 | SRCC = 0.7447 ]
+INFO:root:Average loss=0.3613661458859077
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4027 | LCC = 0.7535 | SRCC = 0.7538 ]
+INFO:root:Average loss=0.3192377969240531
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4057 | LCC = 0.7615 | SRCC = 0.7612 ]
+INFO:root:Average loss=0.3429505282487625
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3953 | LCC = 0.7544 | SRCC = 0.7565 ]
+INFO:root:Average loss=0.318384146079039
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3688 | LCC = 0.7731 | SRCC = 0.7704 ]
+INFO:root:Average loss=0.32130552790103817
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3847 | LCC = 0.7784 | SRCC = 0.7764 ]
+INFO:root:Average loss=0.2944389793735284
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3645 | LCC = 0.7760 | SRCC = 0.7758 ]
+INFO:root:Average loss=0.2733699083328247
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3496 | LCC = 0.7777 | SRCC = 0.7774 ]
+INFO:root:Average loss=0.2858751167853673
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3777 | LCC = 0.7697 | SRCC = 0.7673 ]
+INFO:root:Average loss=0.27136043650217545
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4532 | LCC = 0.7621 | SRCC = 0.7688 ]
+INFO:root:Average loss=0.3067011082401642
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3705 | LCC = 0.7730 | SRCC = 0.7720 ]
+INFO:root:Average loss=0.2689260305502476
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3600 | LCC = 0.7848 | SRCC = 0.7816 ]
+INFO:root:Average loss=0.24474910684885123
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3889 | LCC = 0.7684 | SRCC = 0.7659 ]
+INFO:root:Average loss=0.2603062223165463
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3829 | LCC = 0.7664 | SRCC = 0.7627 ]
+INFO:root:Average loss=0.23323508504873666
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3642 | LCC = 0.7718 | SRCC = 0.7697 ]
+INFO:root:Average loss=0.24340319939148733
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3878 | LCC = 0.7707 | SRCC = 0.7737 ]
+INFO:root:Average loss=0.23664515274457443
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3284 | LCC = 0.7841 | SRCC = 0.7828 ]
+INFO:root:Average loss=0.23357953551488045
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3504 | LCC = 0.7819 | SRCC = 0.7781 ]
+INFO:root:Average loss=0.22553779815252012
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3406 | LCC = 0.7780 | SRCC = 0.7756 ]
+INFO:root:Average loss=0.20767518839774987
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3271 | LCC = 0.7864 | SRCC = 0.7844 ]
+INFO:root:Average loss=0.22089615884499672
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3449 | LCC = 0.7831 | SRCC = 0.7794 ]
+INFO:root:Average loss=0.2037420561298346
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3297 | LCC = 0.7849 | SRCC = 0.7816 ]
+INFO:root:Average loss=0.20880440450631654
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3412 | LCC = 0.7782 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.18842934253506172
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3332 | LCC = 0.7823 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.17934104723808092
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3801 | LCC = 0.7852 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.2328605483739804
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3955 | LCC = 0.7804 | SRCC = 0.7772 ]
+INFO:root:Average loss=0.19494677917697492
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3372 | LCC = 0.7741 | SRCC = 0.7740 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3471 | LCC = 0.7828 | SRCC = 0.7844 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer22/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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_large_asr_en/layer22/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer22/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5ca648c338fff5e226244e021724a63cec224ca7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dda9dfe7145761aae62515776d69b940a67b9273761bc7ddf47b7dbd1eea41f6
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer22/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..090e20f3e3f5be6d0ec5742db056a782577b35c4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f6520b1834c279bb0a9c3b6859e39e609deb4d216776a63cd2b4ec0404cd1f24
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer22/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e3e46baf18243d997b0ee52b01c3a691ed01d64f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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=0.6726182259810276
+INFO:root:
+[Valid][0][UTT][ MSE = 0.7647 | LCC = 0.4303 | SRCC = 0.4493 ]
+INFO:root:Average loss=0.5446818123261133
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5864 | LCC = 0.6008 | SRCC = 0.6044 ]
+INFO:root:Average loss=0.4508382219534654
+INFO:root:
+[Valid][2][UTT][ MSE = 0.5650 | LCC = 0.6836 | SRCC = 0.6896 ]
+INFO:root:Average loss=0.4054417623541294
+INFO:root:
+[Valid][3][UTT][ MSE = 0.5065 | LCC = 0.7205 | SRCC = 0.7290 ]
+INFO:root:Average loss=0.396134132758165
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4598 | LCC = 0.7128 | SRCC = 0.7152 ]
+INFO:root:Average loss=0.3431313505921608
+INFO:root:
+[Valid][5][UTT][ MSE = 0.5457 | LCC = 0.7379 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.3398138628556178
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4800 | LCC = 0.7466 | SRCC = 0.7523 ]
+INFO:root:Average loss=0.318364934470409
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3762 | LCC = 0.7484 | SRCC = 0.7478 ]
+INFO:root:Average loss=0.3322936756870685
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3913 | LCC = 0.7629 | SRCC = 0.7599 ]
+INFO:root:Average loss=0.30261275783563274
+INFO:root:
+[Valid][9][UTT][ MSE = 0.4629 | LCC = 0.7625 | SRCC = 0.7633 ]
+INFO:root:Average loss=0.3068606857305918
+INFO:root:
+[Valid][10][UTT][ MSE = 0.4177 | LCC = 0.7600 | SRCC = 0.7584 ]
+INFO:root:Average loss=0.2939209118485451
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3685 | LCC = 0.7739 | SRCC = 0.7710 ]
+INFO:root:Average loss=0.2910680264616624
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3584 | LCC = 0.7769 | SRCC = 0.7726 ]
+INFO:root:Average loss=0.27433275412290525
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3567 | LCC = 0.7803 | SRCC = 0.7764 ]
+INFO:root:Average loss=0.28065323944275195
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4277 | LCC = 0.7743 | SRCC = 0.7735 ]
+INFO:root:Average loss=0.24948286360655075
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3545 | LCC = 0.7814 | SRCC = 0.7782 ]
+INFO:root:Average loss=0.25638557836795467
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3477 | LCC = 0.7732 | SRCC = 0.7679 ]
+INFO:root:Average loss=0.2650247714840449
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3831 | LCC = 0.7781 | SRCC = 0.7746 ]
+INFO:root:Average loss=0.24281271680807456
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3537 | LCC = 0.7770 | SRCC = 0.7735 ]
+INFO:root:Average loss=0.24526464423308006
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3572 | LCC = 0.7712 | SRCC = 0.7679 ]
+INFO:root:Average loss=0.22803543794613618
+INFO:root:
+[Valid][20][UTT][ MSE = 0.4545 | LCC = 0.7731 | SRCC = 0.7702 ]
+INFO:root:Average loss=0.2337620451282232
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3424 | LCC = 0.7848 | SRCC = 0.7823 ]
+INFO:root:Average loss=0.23301812815360534
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3792 | LCC = 0.7765 | SRCC = 0.7755 ]
+INFO:root:Average loss=0.22225314073073557
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3868 | LCC = 0.7824 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.2129060012789873
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3739 | LCC = 0.7847 | SRCC = 0.7840 ]
+INFO:root:Average loss=0.21051938850910235
+INFO:root:
+[Valid][25][UTT][ MSE = 0.4032 | LCC = 0.7760 | SRCC = 0.7753 ]
+INFO:root:Average loss=0.2021453951795896
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3473 | LCC = 0.7835 | SRCC = 0.7803 ]
+INFO:root:Average loss=0.18700938700483397
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3393 | LCC = 0.7923 | SRCC = 0.7899 ]
+INFO:root:Average loss=0.19500003965237203
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3636 | LCC = 0.7915 | SRCC = 0.7888 ]
+INFO:root:Average loss=0.18371785355684084
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3808 | LCC = 0.7794 | SRCC = 0.7764 ]
+INFO:root:Average loss=0.19360770323337653
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3553 | LCC = 0.7814 | SRCC = 0.7781 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3343 | LCC = 0.7809 | SRCC = 0.7796 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer22/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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_large_asr_en/layer22/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer22/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..25b647033495ae194003288221fe209362ba34c6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a54f20871abb6f198cc29d28bb21d40b3c3cb19182365173d8b6a07cadf8c4c
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer22/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2e3a5d5af25460e820cbe34e9218b15f77a4bb8b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adafdf5a1eda39484564b511813e72f1c6423030ccc37b3a5c2a36d3bd851689
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer22/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f026798cbcb7d9d4aec976968bf295d7c7a15946
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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=0.6395223373785998
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6216 | LCC = 0.5143 | SRCC = 0.5117 ]
+INFO:root:Average loss=0.5041530755085822
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5478 | LCC = 0.6504 | SRCC = 0.6531 ]
+INFO:root:Average loss=0.42619985074569017
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4895 | LCC = 0.7075 | SRCC = 0.7094 ]
+INFO:root:Average loss=0.3833836416403453
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4054 | LCC = 0.7435 | SRCC = 0.7446 ]
+INFO:root:Average loss=0.39163063886838084
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4142 | LCC = 0.7460 | SRCC = 0.7454 ]
+INFO:root:Average loss=0.3603534052769343
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3890 | LCC = 0.7594 | SRCC = 0.7589 ]
+INFO:root:Average loss=0.3527109871308009
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3827 | LCC = 0.7694 | SRCC = 0.7671 ]
+INFO:root:Average loss=0.3242579299287918
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3822 | LCC = 0.7762 | SRCC = 0.7744 ]
+INFO:root:Average loss=0.3261551333543582
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3656 | LCC = 0.7734 | SRCC = 0.7697 ]
+INFO:root:Average loss=0.30315558535930437
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3608 | LCC = 0.7710 | SRCC = 0.7674 ]
+INFO:root:Average loss=0.30126762810425883
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3975 | LCC = 0.7686 | SRCC = 0.7686 ]
+INFO:root:Average loss=0.2852603061459003
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4174 | LCC = 0.7723 | SRCC = 0.7691 ]
+INFO:root:Average loss=0.313059990413678
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3775 | LCC = 0.7789 | SRCC = 0.7761 ]
+INFO:root:Average loss=0.2700687205562225
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4328 | LCC = 0.7700 | SRCC = 0.7714 ]
+INFO:root:Average loss=0.2707088918257982
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3449 | LCC = 0.7754 | SRCC = 0.7715 ]
+INFO:root:Average loss=0.2585600094917493
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3602 | LCC = 0.7794 | SRCC = 0.7760 ]
+INFO:root:Average loss=0.2562518757887376
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3546 | LCC = 0.7794 | SRCC = 0.7757 ]
+INFO:root:Average loss=0.2636430343756309
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3602 | LCC = 0.7786 | SRCC = 0.7737 ]
+INFO:root:Average loss=0.25816324276801866
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3582 | LCC = 0.7809 | SRCC = 0.7771 ]
+INFO:root:Average loss=0.24376020160241005
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3705 | LCC = 0.7825 | SRCC = 0.7790 ]
+INFO:root:Average loss=0.243467706709336
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3551 | LCC = 0.7805 | SRCC = 0.7788 ]
+INFO:root:Average loss=0.23983643910823724
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3391 | LCC = 0.7816 | SRCC = 0.7794 ]
+INFO:root:Average loss=0.23604853680500618
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3380 | LCC = 0.7835 | SRCC = 0.7800 ]
+INFO:root:Average loss=0.22639595144070113
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3506 | LCC = 0.7813 | SRCC = 0.7783 ]
+INFO:root:Average loss=0.2144407020547451
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3477 | LCC = 0.7758 | SRCC = 0.7723 ]
+INFO:root:Average loss=0.23041024230993712
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3401 | LCC = 0.7814 | SRCC = 0.7777 ]
+INFO:root:Average loss=0.21231888750424752
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3527 | LCC = 0.7749 | SRCC = 0.7711 ]
+INFO:root:Average loss=0.20067639209521124
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3425 | LCC = 0.7787 | SRCC = 0.7760 ]
+INFO:root:Average loss=0.2054739790276075
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3446 | LCC = 0.7779 | SRCC = 0.7744 ]
+INFO:root:Average loss=0.19344643789988297
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3642 | LCC = 0.7758 | SRCC = 0.7725 ]
+INFO:root:Average loss=0.1998096723587085
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3724 | LCC = 0.7792 | SRCC = 0.7752 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3605 | LCC = 0.7657 | SRCC = 0.7659 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer22/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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_large_asr_en/layer22/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer22/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..53e778ce4a661d460ce4facebfa235fd8dc03060
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d77e66bfce037a289d33f17cc366e15cee362e537b14d8776d51583bdc9cd78
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer22/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b673ab56fb88755f233ccbd8f32f823e4541923a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38ef84cec58f29d6a5a4b86c6bc14c4dbe3a509b8eb643a8377144eec102f880
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer22/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer22/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..05ad04a30e907b2d68bbf18634e5bd34827980bc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer22/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=0.6452068743797449
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6271 | LCC = 0.5096 | SRCC = 0.5098 ]
+INFO:root:Average loss=0.5145264458961976
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5236 | LCC = 0.6579 | SRCC = 0.6609 ]
+INFO:root:Average loss=0.45809286221479756
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4669 | LCC = 0.6987 | SRCC = 0.6969 ]
+INFO:root:Average loss=0.41544592456939894
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4572 | LCC = 0.7144 | SRCC = 0.7202 ]
+INFO:root:Average loss=0.37767991041525817
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4167 | LCC = 0.7326 | SRCC = 0.7318 ]
+INFO:root:Average loss=0.3577414877139605
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4075 | LCC = 0.7543 | SRCC = 0.7547 ]
+INFO:root:Average loss=0.33934019601497895
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4103 | LCC = 0.7395 | SRCC = 0.7371 ]
+INFO:root:Average loss=0.32436974843343097
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4050 | LCC = 0.7538 | SRCC = 0.7533 ]
+INFO:root:Average loss=0.30775114435416
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3941 | LCC = 0.7648 | SRCC = 0.7620 ]
+INFO:root:Average loss=0.30312073345367724
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3626 | LCC = 0.7729 | SRCC = 0.7697 ]
+INFO:root:Average loss=0.3267986642626616
+INFO:root:
+[Valid][10][UTT][ MSE = 0.4550 | LCC = 0.7706 | SRCC = 0.7692 ]
+INFO:root:Average loss=0.31390945307719403
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3571 | LCC = 0.7740 | SRCC = 0.7736 ]
+INFO:root:Average loss=0.28714533169300127
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3582 | LCC = 0.7666 | SRCC = 0.7636 ]
+INFO:root:Average loss=0.322723405674482
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3708 | LCC = 0.7731 | SRCC = 0.7726 ]
+INFO:root:Average loss=0.2739328438272843
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4105 | LCC = 0.7675 | SRCC = 0.7673 ]
+INFO:root:Average loss=0.25220842220080203
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3700 | LCC = 0.7704 | SRCC = 0.7664 ]
+INFO:root:Average loss=0.2884072069174204
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3525 | LCC = 0.7749 | SRCC = 0.7704 ]
+INFO:root:Average loss=0.2638591990256921
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3526 | LCC = 0.7757 | SRCC = 0.7722 ]
+INFO:root:Average loss=0.2511790931606904
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3726 | LCC = 0.7810 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.25428714469457286
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3576 | LCC = 0.7773 | SRCC = 0.7734 ]
+INFO:root:Average loss=0.22167799430779922
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3746 | LCC = 0.7777 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.22291515863094574
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3275 | LCC = 0.7854 | SRCC = 0.7816 ]
+INFO:root:Average loss=0.22722744101133102
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3380 | LCC = 0.7761 | SRCC = 0.7719 ]
+INFO:root:Average loss=0.2227866135728665
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3797 | LCC = 0.7793 | SRCC = 0.7769 ]
+INFO:root:Average loss=0.2160347661910913
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3338 | LCC = 0.7848 | SRCC = 0.7817 ]
+INFO:root:Average loss=0.2174350518064621
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3339 | LCC = 0.7838 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.20735572565060395
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3553 | LCC = 0.7695 | SRCC = 0.7670 ]
+INFO:root:Average loss=0.18991576612759858
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3556 | LCC = 0.7747 | SRCC = 0.7713 ]
+INFO:root:Average loss=0.19485137191338417
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3561 | LCC = 0.7717 | SRCC = 0.7685 ]
+INFO:root:Average loss=0.1918607403834661
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3320 | LCC = 0.7797 | SRCC = 0.7751 ]
+INFO:root:Average loss=0.18905905309395912
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3772 | LCC = 0.7734 | SRCC = 0.7703 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3423 | LCC = 0.7809 | SRCC = 0.7831 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer4/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer4/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer4/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a2accd2c07c2827bfa5f5afab20b0ef1f604e0b8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4137c25a97b81cde1b282ef754bcef8c67661794f34b3f7740b7d329a9f8347a
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer4/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1430819a99ecf83c74c23e70457433754b5064d0
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d881d6b0c2ae0a321ba932c7b8df4d9018366a6ee2e405868e49aefadd63d4f4
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer4/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..72c0737e2b575f184a32ea5f44489cf71d9db864
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.636781497261463
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3983 | LCC = 0.7675 | SRCC = 0.7766 ]
+INFO:root:Average loss=0.3041597226491341
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3300 | LCC = 0.8226 | SRCC = 0.8237 ]
+INFO:root:Average loss=0.2580852386278984
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2984 | LCC = 0.8365 | SRCC = 0.8359 ]
+INFO:root:Average loss=0.25057003379632264
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2904 | LCC = 0.8420 | SRCC = 0.8403 ]
+INFO:root:Average loss=0.22248924943881157
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3349 | LCC = 0.8503 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.22169464692855492
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2939 | LCC = 0.8537 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.1917128498928669
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2771 | LCC = 0.8555 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.19538176995821488
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2997 | LCC = 0.8586 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.17599794765313467
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2389 | LCC = 0.8562 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.1689317986751214
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2447 | LCC = 0.8610 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.15811163187026978
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2443 | LCC = 0.8583 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.14659458475235182
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2364 | LCC = 0.8599 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.14879292650864676
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2602 | LCC = 0.8481 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.15017955196209443
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2839 | LCC = 0.8588 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.1337945403961035
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2391 | LCC = 0.8563 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.1377496705032312
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2396 | LCC = 0.8538 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.12409372551318927
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2375 | LCC = 0.8598 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.11613194749523433
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2491 | LCC = 0.8577 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.1116243547353989
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2309 | LCC = 0.8583 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.11833249844419651
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2216 | LCC = 0.8640 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.10902192099736287
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2416 | LCC = 0.8570 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.09917388932827191
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2235 | LCC = 0.8623 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.09031969662278126
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2497 | LCC = 0.8501 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.09541012838673897
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2286 | LCC = 0.8590 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.08643527658512959
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2795 | LCC = 0.8611 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.08781587513975608
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2291 | LCC = 0.8582 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.08673023002652022
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2986 | LCC = 0.8612 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.08450862268606822
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2683 | LCC = 0.8616 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.07571362694486594
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2420 | LCC = 0.8553 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.07681375823150842
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2473 | LCC = 0.8588 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.06978087074672565
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2286 | LCC = 0.8573 | SRCC = 0.8555 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2477 | LCC = 0.8564 | SRCC = 0.8584 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer4/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer4/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer4/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8acf624fa64aff30d3772a6e6dfa55ca30a7a819
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d14d4afbbc3650f0ac3de53fdf6dda23fc738ddbb196073101ab88b65687dbee
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer4/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bce7cd53b66b1e7c14066ea0f3fff4e62034547b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8645e689286f98cb10f56acbe72f1ee8c4f3e7dc83e6688a2de3cb3d9f1bab1
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer4/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..d85ee6c5367021c291390d7a3808d96c475afd76
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5211799079791094
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3770 | LCC = 0.7807 | SRCC = 0.7863 ]
+INFO:root:Average loss=0.30526885848778945
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3084 | LCC = 0.8163 | SRCC = 0.8206 ]
+INFO:root:Average loss=0.2637748473729843
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2790 | LCC = 0.8334 | SRCC = 0.8372 ]
+INFO:root:Average loss=0.24211729031342727
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3155 | LCC = 0.8312 | SRCC = 0.8411 ]
+INFO:root:Average loss=0.2236161130743149
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2453 | LCC = 0.8494 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.20181990462617996
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2544 | LCC = 0.8584 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.20097911768616775
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2309 | LCC = 0.8593 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.18439530772276413
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2333 | LCC = 0.8635 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.17359972143402466
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2397 | LCC = 0.8672 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.16261370823933527
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2370 | LCC = 0.8591 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.1645883637934159
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2496 | LCC = 0.8673 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.14424587375460526
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2235 | LCC = 0.8632 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.13931459503678176
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2268 | LCC = 0.8609 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.13125577425727478
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2614 | LCC = 0.8680 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.1262121490943126
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2430 | LCC = 0.8655 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.12380474968216358
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2218 | LCC = 0.8641 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.12611443205521658
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3449 | LCC = 0.8677 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.12928436830257758
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2245 | LCC = 0.8668 | SRCC = 0.8652 ]
+INFO:root:Average loss=0.10644934135369766
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2228 | LCC = 0.8657 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.11149725957940786
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2432 | LCC = 0.8698 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.0990176366116756
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2335 | LCC = 0.8654 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.10604338801633088
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2194 | LCC = 0.8677 | SRCC = 0.8665 ]
+INFO:root:Average loss=0.09047352632459922
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2164 | LCC = 0.8618 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.08630562463823037
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2127 | LCC = 0.8641 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.09461555620416617
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2108 | LCC = 0.8634 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.08412921285400024
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2226 | LCC = 0.8695 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.07676853979818332
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2216 | LCC = 0.8664 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.07242902559347641
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2204 | LCC = 0.8665 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.07709039079073148
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2272 | LCC = 0.8614 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.0712589446741801
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2168 | LCC = 0.8633 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.06709826283921035
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2241 | LCC = 0.8621 | SRCC = 0.8623 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2905 | LCC = 0.8560 | SRCC = 0.8563 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer4/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer4/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer4/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..811120bb299d7c5b92c5b05e4570c305cf4b436c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08d2ccf48143044eaf58a4c4081920abbf1a2af6df4d804330e52dde64ebf4ad
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer4/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4d38cf14404393dc432a60ce23b5cde5d5206e0e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec6c422577334067dfb5e2c6041e4d4f3a178c8c23d5ea4dae65a5a55d06f61e
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer4/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..75e61bd473f672a7234a71dbfc11eb0c09815808
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5290904402350768
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3834 | LCC = 0.7808 | SRCC = 0.7853 ]
+INFO:root:Average loss=0.31868481196654147
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3264 | LCC = 0.8144 | SRCC = 0.8177 ]
+INFO:root:Average loss=0.2656069059784596
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3080 | LCC = 0.8320 | SRCC = 0.8345 ]
+INFO:root:Average loss=0.2356557045609523
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2791 | LCC = 0.8411 | SRCC = 0.8421 ]
+INFO:root:Average loss=0.2156206319729487
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2798 | LCC = 0.8494 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.20005341170307916
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2768 | LCC = 0.8466 | SRCC = 0.8487 ]
+INFO:root:Average loss=0.18813201765983534
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2616 | LCC = 0.8537 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.1847435637162282
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3003 | LCC = 0.8573 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.19249913488061
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2474 | LCC = 0.8599 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.17306911104764694
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3093 | LCC = 0.8579 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.1694744368776297
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2817 | LCC = 0.8562 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.15472735502781013
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2551 | LCC = 0.8585 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.15156691702894676
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2676 | LCC = 0.8588 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.14520762067956802
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2262 | LCC = 0.8600 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.13105058192442626
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2459 | LCC = 0.8579 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.12998413141721335
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2691 | LCC = 0.8591 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.12525418907021865
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2388 | LCC = 0.8578 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.11580211669206619
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2484 | LCC = 0.8567 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.11221618979023053
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2366 | LCC = 0.8539 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.10916013518969218
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2370 | LCC = 0.8566 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.10442758217836037
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2363 | LCC = 0.8573 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.10239835274525177
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2400 | LCC = 0.8549 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.09593776728098209
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2331 | LCC = 0.8599 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.09140429316231838
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2396 | LCC = 0.8527 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.08992339536929742
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2347 | LCC = 0.8563 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.08245390429137608
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2420 | LCC = 0.8521 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.08118152967057167
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2278 | LCC = 0.8545 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.08066742423062141
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2439 | LCC = 0.8542 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.0728699632275563
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2426 | LCC = 0.8536 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.07401530511486225
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2305 | LCC = 0.8566 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.07053369694413283
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2354 | LCC = 0.8540 | SRCC = 0.8510 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2411 | LCC = 0.8591 | SRCC = 0.8610 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer4/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer4/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer4/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4d10cd738abd86cd7ef1d4464704ea28f75c8efc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b27e78f93687b2fd3a7fafe4c07e413d8e4afff3ea6cbda7b872056850bb3be4
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer4/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..da4a8d2309e92cfaaa2df0f45a9f5aa4fb9e64a6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ac09550592fe15e2568f138916a795f29937b6f3dfb35a6df804be4dda55f80
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer4/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..5b8a29968d57e56d42a2190d0bd721d253916bba
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5727077351930814
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4081 | LCC = 0.7594 | SRCC = 0.7758 ]
+INFO:root:Average loss=0.3000529503975159
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3063 | LCC = 0.8126 | SRCC = 0.8213 ]
+INFO:root:Average loss=0.2716636382616483
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2938 | LCC = 0.8287 | SRCC = 0.8353 ]
+INFO:root:Average loss=0.24064865880287611
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2687 | LCC = 0.8417 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.22093986672086594
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2754 | LCC = 0.8459 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.20378163762581655
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2465 | LCC = 0.8513 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.18960362223860544
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2737 | LCC = 0.8494 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.18198368574182192
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2914 | LCC = 0.8607 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.17092444804998544
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2604 | LCC = 0.8566 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.16577474085184243
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2355 | LCC = 0.8618 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.15519386023665085
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2446 | LCC = 0.8580 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.1580491706919976
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2359 | LCC = 0.8639 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.13222161413003236
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2309 | LCC = 0.8587 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.13655586473834821
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2179 | LCC = 0.8640 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.12948705265537286
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2227 | LCC = 0.8650 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.12357596231576724
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2383 | LCC = 0.8646 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.11162887685574018
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2315 | LCC = 0.8609 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.11586065007708012
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2334 | LCC = 0.8545 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.10308763423027137
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2374 | LCC = 0.8646 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.09437809870220147
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2244 | LCC = 0.8632 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.09192764854583985
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2174 | LCC = 0.8621 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.08767979210003829
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2388 | LCC = 0.8646 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.0879775476283752
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2287 | LCC = 0.8629 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.08221690342403375
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2378 | LCC = 0.8582 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.07800384639547421
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2519 | LCC = 0.8602 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.07770145526872231
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2478 | LCC = 0.8604 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.07350426859771594
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2137 | LCC = 0.8635 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.07077698127772564
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2486 | LCC = 0.8652 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.06568509478790638
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2399 | LCC = 0.8633 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.06193374603604659
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2165 | LCC = 0.8647 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.06587842293083668
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2452 | LCC = 0.8652 | SRCC = 0.8630 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2878 | LCC = 0.8557 | SRCC = 0.8584 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer4/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer4/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer4/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6b0f44f034793fe5d88d257c1e68a8c45f2187cf
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11fa4e91a11280783f05fbaa47d184c2ac3d22e9e751081f4d57061b4ca02932
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer4/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..376e65083271144f1514d8926f41a7a2cc75420a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer4/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:571865381b6cfd2c4afcb3cdb2b0181d7325b3e31662f508d50b555847a5a517
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer4/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer4/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..0d7dc10e811c43200df7e1a6d3a4c396eae2c635
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5411792661135013
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3772 | LCC = 0.7869 | SRCC = 0.7928 ]
+INFO:root:Average loss=0.292506381105154
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3002 | LCC = 0.8221 | SRCC = 0.8237 ]
+INFO:root:Average loss=0.26205395276729876
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2919 | LCC = 0.8375 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.23386607414636856
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2720 | LCC = 0.8418 | SRCC = 0.8415 ]
+INFO:root:Average loss=0.21474162603800112
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2617 | LCC = 0.8438 | SRCC = 0.8460 ]
+INFO:root:Average loss=0.2046936855484278
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2682 | LCC = 0.8458 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.1904317543674738
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2421 | LCC = 0.8513 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.17438960371491236
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2389 | LCC = 0.8557 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.18474838682092154
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2542 | LCC = 0.8546 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.15982740257795042
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2413 | LCC = 0.8545 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.1652196125151255
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2528 | LCC = 0.8562 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.14915264101746756
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2911 | LCC = 0.8579 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.1477982128659884
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2451 | LCC = 0.8536 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.13134815486577842
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2463 | LCC = 0.8592 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.13166931214240882
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2241 | LCC = 0.8608 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.12644055323340955
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2764 | LCC = 0.8615 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.11153096371354201
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2400 | LCC = 0.8531 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.11592544892277473
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2462 | LCC = 0.8578 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.10382798476478992
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2577 | LCC = 0.8522 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.1093404515622518
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2386 | LCC = 0.8545 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.1048823951337582
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2323 | LCC = 0.8581 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.09569058309380825
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2501 | LCC = 0.8521 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.09838158689821377
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2421 | LCC = 0.8580 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.08766726194283901
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2451 | LCC = 0.8514 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.08760022797072545
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2323 | LCC = 0.8548 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.08215399731237155
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2293 | LCC = 0.8581 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.07556331974382584
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2246 | LCC = 0.8584 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.0729847692239743
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2251 | LCC = 0.8598 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.06998449654724354
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2503 | LCC = 0.8585 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.0806819668565041
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2278 | LCC = 0.8584 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.0705473534285258
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2372 | LCC = 0.8618 | SRCC = 0.8605 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2545 | LCC = 0.8589 | SRCC = 0.8593 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer6/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer6/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer6/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4a3f5cb9d0bcaa26deede3cdaa2e237ff25c1487
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86ee5e0383a7c0d49f9760e17b3ee28bbc7c6e96180673ab223be3f4b7c65680
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer6/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..886ff61b6be05e66c65c9261cdc6b58816682bfd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:564895484eb4a61946501f59bc5e8f9821ec844a8fc820435186a3f538a551d9
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer6/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b15e47856d6fa07ac8fab5bc0aa5f976b7329bb2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5191217431655297
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3567 | LCC = 0.8095 | SRCC = 0.8132 ]
+INFO:root:Average loss=0.2885653745287504
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3288 | LCC = 0.8316 | SRCC = 0.8333 ]
+INFO:root:Average loss=0.2704092971025369
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3061 | LCC = 0.8393 | SRCC = 0.8375 ]
+INFO:root:Average loss=0.2353519441989752
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3163 | LCC = 0.8437 | SRCC = 0.8427 ]
+INFO:root:Average loss=0.2107037112880976
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3211 | LCC = 0.8384 | SRCC = 0.8409 ]
+INFO:root:Average loss=0.19786760994257072
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2636 | LCC = 0.8489 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.1865656396899468
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2534 | LCC = 0.8525 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.1827525420066638
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2622 | LCC = 0.8514 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.1653901090224584
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2665 | LCC = 0.8508 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.1653835001664284
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2704 | LCC = 0.8514 | SRCC = 0.8464 ]
+INFO:root:Average loss=0.1571344462915873
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2997 | LCC = 0.8453 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.15418900176882744
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2427 | LCC = 0.8531 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.129369556139677
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2824 | LCC = 0.8443 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.1275286767154168
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2529 | LCC = 0.8469 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.11749750652756447
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2619 | LCC = 0.8483 | SRCC = 0.8459 ]
+INFO:root:Average loss=0.11723392256177388
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2697 | LCC = 0.8453 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.10880294833809902
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2387 | LCC = 0.8489 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.10143085674215586
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2467 | LCC = 0.8490 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.11061901002167127
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3072 | LCC = 0.8449 | SRCC = 0.8457 ]
+INFO:root:Average loss=0.09306165485236889
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2461 | LCC = 0.8471 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.08726298746963342
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2541 | LCC = 0.8462 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.08632739981970726
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2722 | LCC = 0.8443 | SRCC = 0.8451 ]
+INFO:root:Average loss=0.0807643194611256
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2710 | LCC = 0.8452 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.07973802457444179
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2612 | LCC = 0.8496 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.07122448043754467
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2478 | LCC = 0.8477 | SRCC = 0.8448 ]
+INFO:root:Average loss=0.07544274609058331
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2485 | LCC = 0.8485 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.08088702507890187
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3330 | LCC = 0.8386 | SRCC = 0.8413 ]
+INFO:root:Average loss=0.06533586219526254
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2405 | LCC = 0.8489 | SRCC = 0.8478 ]
+INFO:root:Average loss=0.06660664458878529
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2423 | LCC = 0.8498 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.06761524463311219
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2813 | LCC = 0.8411 | SRCC = 0.8407 ]
+INFO:root:Average loss=0.06128997288835354
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2639 | LCC = 0.8402 | SRCC = 0.8400 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2383 | LCC = 0.8579 | SRCC = 0.8587 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer6/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer6/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer6/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fc580ff7224d5dbf1dfcd52ea60c797c72850083
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1030200f52929f8fd24f2b9bfed334b661a4367fd4ef28ce8c2f71acc50575c3
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer6/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c22127c53a22f1a4021cc81b8b53d185799a0053
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cf8cfb61d715a3f19c4bdf5a3822f4fe953494bcba68dd51997ebd14df62817
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer6/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..2ff9a02c867c1edd1c78f0900578be10a4170067
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5091629070349228
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3437 | LCC = 0.7979 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.30357699994093335
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3814 | LCC = 0.8239 | SRCC = 0.8291 ]
+INFO:root:Average loss=0.26621937598937595
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3071 | LCC = 0.8340 | SRCC = 0.8356 ]
+INFO:root:Average loss=0.24544061987827986
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2752 | LCC = 0.8417 | SRCC = 0.8399 ]
+INFO:root:Average loss=0.22238903836562082
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2881 | LCC = 0.8479 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.2007739605047764
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2578 | LCC = 0.8501 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.1954783093279753
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2714 | LCC = 0.8495 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.17901444186766943
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2551 | LCC = 0.8505 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.1766602139060314
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2613 | LCC = 0.8504 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.15733747021892133
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2512 | LCC = 0.8514 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.15684678424627352
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2490 | LCC = 0.8512 | SRCC = 0.8487 ]
+INFO:root:Average loss=0.13909926532934874
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2530 | LCC = 0.8480 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.13216660076226944
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2723 | LCC = 0.8451 | SRCC = 0.8424 ]
+INFO:root:Average loss=0.12467315153051646
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3065 | LCC = 0.8448 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.12114304208602661
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2673 | LCC = 0.8465 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.1132668303564573
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2996 | LCC = 0.8473 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.109209927992943
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2654 | LCC = 0.8475 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.10071533822860473
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2657 | LCC = 0.8457 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.09793728284346752
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2647 | LCC = 0.8440 | SRCC = 0.8410 ]
+INFO:root:Average loss=0.09626562181764688
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2864 | LCC = 0.8439 | SRCC = 0.8424 ]
+INFO:root:Average loss=0.08956387061148118
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2483 | LCC = 0.8470 | SRCC = 0.8435 ]
+INFO:root:Average loss=0.08361273311460629
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3071 | LCC = 0.8419 | SRCC = 0.8408 ]
+INFO:root:Average loss=0.09554175425989506
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2399 | LCC = 0.8489 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.0818006831388443
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2629 | LCC = 0.8484 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.07934305525552003
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3055 | LCC = 0.8433 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.07536701280146073
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2744 | LCC = 0.8448 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.06978445385511105
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2599 | LCC = 0.8473 | SRCC = 0.8438 ]
+INFO:root:Average loss=0.06970565823408273
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3543 | LCC = 0.8384 | SRCC = 0.8413 ]
+INFO:root:Average loss=0.06908286229158059
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2381 | LCC = 0.8507 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.061694420396517485
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2517 | LCC = 0.8471 | SRCC = 0.8457 ]
+INFO:root:Average loss=0.06419033413896194
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3099 | LCC = 0.8389 | SRCC = 0.8380 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2538 | LCC = 0.8535 | SRCC = 0.8585 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer6/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer6/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer6/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..246e5f58ba6997946cbbca1a54d6069a18941963
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07ca834beec068e1d39c00e062bd6d6e35ef7b1b065f854b6e316c6c3009ac0e
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer6/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f32371f1b001e48e62b0a3459916023eb11f3b48
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:704faebc0d96733e49e5b106c8cb7508012a546ca8c09b1c3c121860295d6717
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer6/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..535a24831baf2d0bbd5349c1b6c1938be0f7948c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5435493542597845
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4067 | LCC = 0.8008 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.2858025311277463
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2805 | LCC = 0.8352 | SRCC = 0.8367 ]
+INFO:root:Average loss=0.2411745696877822
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2855 | LCC = 0.8455 | SRCC = 0.8429 ]
+INFO:root:Average loss=0.218177090661648
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2637 | LCC = 0.8464 | SRCC = 0.8446 ]
+INFO:root:Average loss=0.21996914939238474
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3043 | LCC = 0.8463 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.1916164979338646
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2516 | LCC = 0.8576 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.19093513374145216
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2432 | LCC = 0.8568 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.17316079645966873
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2428 | LCC = 0.8551 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.183893928447595
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2876 | LCC = 0.8507 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.16099189699460298
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2402 | LCC = 0.8532 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.155331317621928
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2461 | LCC = 0.8524 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.14069319277619705
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2520 | LCC = 0.8524 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.1343622968938106
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2818 | LCC = 0.8473 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.1261865796569066
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2432 | LCC = 0.8503 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.11942776579123277
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2324 | LCC = 0.8553 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.110879624214692
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2680 | LCC = 0.8470 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.11294142104303226
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2531 | LCC = 0.8496 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.10004007677810314
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2668 | LCC = 0.8488 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.1042942171677565
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2363 | LCC = 0.8525 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.0974375079266536
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2297 | LCC = 0.8545 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.08879997294682723
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2315 | LCC = 0.8513 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.08539177458255719
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2357 | LCC = 0.8508 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.08293518849099293
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2922 | LCC = 0.8471 | SRCC = 0.8478 ]
+INFO:root:Average loss=0.08149344025131984
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2292 | LCC = 0.8516 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.07241267930620755
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2301 | LCC = 0.8528 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.07494749018970208
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2358 | LCC = 0.8490 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.07015080286715275
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2820 | LCC = 0.8448 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.0714184554437032
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2385 | LCC = 0.8490 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.06819772973465614
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2393 | LCC = 0.8469 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.06250946233287835
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2496 | LCC = 0.8463 | SRCC = 0.8437 ]
+INFO:root:Average loss=0.059145834225301556
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2491 | LCC = 0.8489 | SRCC = 0.8463 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2734 | LCC = 0.8517 | SRCC = 0.8534 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer6/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer6/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer6/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..870893af0781a1a1885476093478b281ea97f933
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a1b73da5105d26866fa6cc89cf2c4c94b2bcbbcdca899c43b8e70ecb5786816
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer6/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..42f1c4d9dda49a57ce8cc7c6afae4eb924517108
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8e7c9e3d874bd2598c5e2c762cb060a7958e11e758371bc989f2e30935c9d85
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer6/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..83cf07901d8151d78cb702e402ae6a9aecbbaffb
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5089653617678545
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3450 | LCC = 0.8065 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.2804075139455306
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3102 | LCC = 0.8357 | SRCC = 0.8368 ]
+INFO:root:Average loss=0.25953011004588544
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3273 | LCC = 0.8448 | SRCC = 0.8429 ]
+INFO:root:Average loss=0.22110078865901017
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3447 | LCC = 0.8468 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.21510944496362638
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2693 | LCC = 0.8514 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.18901782291821945
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2641 | LCC = 0.8491 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.18962579163221213
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3680 | LCC = 0.8369 | SRCC = 0.8393 ]
+INFO:root:Average loss=0.18110551637334701
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3084 | LCC = 0.8481 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.16864969227940607
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2560 | LCC = 0.8540 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.16425999922630113
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2474 | LCC = 0.8531 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.1409686150459143
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2603 | LCC = 0.8552 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.15599834116605613
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2510 | LCC = 0.8564 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.14762397740895933
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2453 | LCC = 0.8508 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.13179562154870766
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2341 | LCC = 0.8548 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.11994275336082165
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2576 | LCC = 0.8501 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.11116536940710667
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2433 | LCC = 0.8541 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.11411970333220103
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2336 | LCC = 0.8525 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.10367367579004703
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2557 | LCC = 0.8519 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.10561057705527697
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2407 | LCC = 0.8521 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.10448916738805099
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2340 | LCC = 0.8550 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.09638763849551861
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2537 | LCC = 0.8542 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.09483297589497688
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2411 | LCC = 0.8527 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.07973274187399791
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2407 | LCC = 0.8519 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.0827386709742057
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2488 | LCC = 0.8529 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.08140216949276435
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2557 | LCC = 0.8517 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.07399444105342412
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2430 | LCC = 0.8505 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.06676668926882438
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2640 | LCC = 0.8437 | SRCC = 0.8398 ]
+INFO:root:Average loss=0.06938737974717067
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2365 | LCC = 0.8505 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.07049474120140076
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2563 | LCC = 0.8533 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.06948504959925628
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2429 | LCC = 0.8489 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.06568268089531323
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2547 | LCC = 0.8491 | SRCC = 0.8457 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2769 | LCC = 0.8510 | SRCC = 0.8547 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer6/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer6/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer6/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a4f7597c9252ba474502c55aae42d4e52ec5ab4d
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0bdfb017694d615049a6aef367cac8c8e3fd414b65c5b9c23c38b0d4ec916dd5
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer6/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2fc2350958e5a270892eb2023341af2c80d94269
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer6/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2812bad64c15a3612855be8dcf976eecb8351c90051e4fa5eb28e1c3e1defd7
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer6/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer6/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f0379d68dea9e952cc76a7bd089f4fab84f0043d
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5275432293613752
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3367 | LCC = 0.8072 | SRCC = 0.8118 ]
+INFO:root:Average loss=0.27928310040480053
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2846 | LCC = 0.8385 | SRCC = 0.8389 ]
+INFO:root:Average loss=0.25268142555768675
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2695 | LCC = 0.8391 | SRCC = 0.8406 ]
+INFO:root:Average loss=0.22784876938049609
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2642 | LCC = 0.8507 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.20615910433041743
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2493 | LCC = 0.8497 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.19719389959787711
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2497 | LCC = 0.8528 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.18683201915178543
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2624 | LCC = 0.8509 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.17101691013727432
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2336 | LCC = 0.8566 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.1652512015440525
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2382 | LCC = 0.8526 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.1645187534009799
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2365 | LCC = 0.8573 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.15206997965772948
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2375 | LCC = 0.8513 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.14748089149212226
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2373 | LCC = 0.8527 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.15221368884428954
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2374 | LCC = 0.8539 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.12491984512561406
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2485 | LCC = 0.8521 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.11797314729446019
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2349 | LCC = 0.8577 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.13814446435142785
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2275 | LCC = 0.8526 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.114249523394765
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2448 | LCC = 0.8501 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.10314719101939446
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2530 | LCC = 0.8483 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.0953176976300967
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2453 | LCC = 0.8503 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.09133702702820301
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2357 | LCC = 0.8469 | SRCC = 0.8462 ]
+INFO:root:Average loss=0.10559641111355561
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2361 | LCC = 0.8514 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.08211385735716575
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2502 | LCC = 0.8490 | SRCC = 0.8478 ]
+INFO:root:Average loss=0.07874680644808671
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2315 | LCC = 0.8503 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.07674830058255257
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2362 | LCC = 0.8517 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.07471877565750709
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2394 | LCC = 0.8509 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.07366108015561715
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2449 | LCC = 0.8500 | SRCC = 0.8474 ]
+INFO:root:Average loss=0.06806351855779305
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2311 | LCC = 0.8510 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.06411270892772919
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2370 | LCC = 0.8502 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.06242613780957002
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2529 | LCC = 0.8509 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.06401710737592135
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2267 | LCC = 0.8526 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.05909861968113826
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2365 | LCC = 0.8505 | SRCC = 0.8494 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2699 | LCC = 0.8459 | SRCC = 0.8513 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run0/config.gin b/results_bvcc/w2v2_large_asr_en/layer8/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer8/run0/model.pt b/results_bvcc/w2v2_large_asr_en/layer8/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f759bb636a4ad7e42320e17caf9695ba4c6a85a4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25229ca895dd1889e9c2cc1185f4a304560870624997d3b4f24f63497148a901
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run0/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer8/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3222067821ff92a1a775f2a29954af848bb418fe
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b2b2bf0299ec30f19f0dc3538646001c17776ab41d7f57fbb4cd6947c16ba8b
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run0/train.log b/results_bvcc/w2v2_large_asr_en/layer8/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..792780b037e50c43cbbe242e02c5fa1a1436800a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.6304852675933105
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4173 | LCC = 0.7737 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.3101444381933946
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3028 | LCC = 0.8232 | SRCC = 0.8238 ]
+INFO:root:Average loss=0.2692752883602411
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3746 | LCC = 0.8291 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.25009044489035237
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3228 | LCC = 0.8301 | SRCC = 0.8324 ]
+INFO:root:Average loss=0.2342588926355044
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2961 | LCC = 0.8338 | SRCC = 0.8354 ]
+INFO:root:Average loss=0.22017884178039354
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2884 | LCC = 0.8402 | SRCC = 0.8380 ]
+INFO:root:Average loss=0.198032074345228
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2656 | LCC = 0.8387 | SRCC = 0.8363 ]
+INFO:root:Average loss=0.19642159581566468
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3343 | LCC = 0.8266 | SRCC = 0.8288 ]
+INFO:root:Average loss=0.17874659836674348
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3256 | LCC = 0.8186 | SRCC = 0.8209 ]
+INFO:root:Average loss=0.159526971861338
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3262 | LCC = 0.8267 | SRCC = 0.8288 ]
+INFO:root:Average loss=0.15132203296973154
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3001 | LCC = 0.8257 | SRCC = 0.8266 ]
+INFO:root:Average loss=0.14962658066398057
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3013 | LCC = 0.8222 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.13843064927137816
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3462 | LCC = 0.8190 | SRCC = 0.8214 ]
+INFO:root:Average loss=0.12637752600205251
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3070 | LCC = 0.8234 | SRCC = 0.8224 ]
+INFO:root:Average loss=0.12813899971735784
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2992 | LCC = 0.8221 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.11103113989035289
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2876 | LCC = 0.8224 | SRCC = 0.8212 ]
+INFO:root:Average loss=0.10419490637305455
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2730 | LCC = 0.8239 | SRCC = 0.8239 ]
+INFO:root:Average loss=0.0985812234859436
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2700 | LCC = 0.8231 | SRCC = 0.8231 ]
+INFO:root:Average loss=0.09471802074366655
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2728 | LCC = 0.8212 | SRCC = 0.8200 ]
+INFO:root:Average loss=0.08544020736828828
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3010 | LCC = 0.8240 | SRCC = 0.8251 ]
+INFO:root:Average loss=0.08669998487218833
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3142 | LCC = 0.8201 | SRCC = 0.8191 ]
+INFO:root:Average loss=0.08322943193026078
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2806 | LCC = 0.8208 | SRCC = 0.8200 ]
+INFO:root:Average loss=0.07792564978202184
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2862 | LCC = 0.8149 | SRCC = 0.8137 ]
+INFO:root:Average loss=0.08095170605259064
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2733 | LCC = 0.8235 | SRCC = 0.8225 ]
+INFO:root:Average loss=0.07241971924518928
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2815 | LCC = 0.8194 | SRCC = 0.8173 ]
+INFO:root:Average loss=0.06709119257254478
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3378 | LCC = 0.8114 | SRCC = 0.8142 ]
+INFO:root:Average loss=0.06703143948927903
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3164 | LCC = 0.8200 | SRCC = 0.8188 ]
+INFO:root:Average loss=0.0638393904440678
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2823 | LCC = 0.8195 | SRCC = 0.8186 ]
+INFO:root:Average loss=0.06160924726953873
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3068 | LCC = 0.8143 | SRCC = 0.8135 ]
+INFO:root:Average loss=0.06443889644474556
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3115 | LCC = 0.8174 | SRCC = 0.8189 ]
+INFO:root:Average loss=0.05722231618486918
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3070 | LCC = 0.8185 | SRCC = 0.8185 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2997 | LCC = 0.8416 | SRCC = 0.8478 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run1/config.gin b/results_bvcc/w2v2_large_asr_en/layer8/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer8/run1/model.pt b/results_bvcc/w2v2_large_asr_en/layer8/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..af5951e569728bd192a8f4d41b717f344229508c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db350e6557782b3ae006c3da45e38da64c9c5c9dbe1a4db6ecfe57622e99ec86
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run1/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer8/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f6cab52f75950d4a74561ce65d30148ba2416814
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b2ea70ca6b4df7423c0da40dadf5470a4408a15df7f057d4245c3967b419d04
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run1/train.log b/results_bvcc/w2v2_large_asr_en/layer8/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b5d67561e4aef4c35305228ca3974bec7bf0a15b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.769368548423816
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4478 | LCC = 0.7498 | SRCC = 0.7531 ]
+INFO:root:Average loss=0.3477230339478224
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3561 | LCC = 0.8106 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.2812269328114314
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3036 | LCC = 0.8255 | SRCC = 0.8242 ]
+INFO:root:Average loss=0.24552879654444182
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3239 | LCC = 0.8309 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.23466109465330076
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3006 | LCC = 0.8333 | SRCC = 0.8329 ]
+INFO:root:Average loss=0.21913122691405126
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2899 | LCC = 0.8342 | SRCC = 0.8336 ]
+INFO:root:Average loss=0.19476691270485902
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2777 | LCC = 0.8379 | SRCC = 0.8353 ]
+INFO:root:Average loss=0.18827756226826936
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2858 | LCC = 0.8345 | SRCC = 0.8350 ]
+INFO:root:Average loss=0.17015053150363457
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3077 | LCC = 0.8321 | SRCC = 0.8324 ]
+INFO:root:Average loss=0.16006908527551553
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2723 | LCC = 0.8389 | SRCC = 0.8365 ]
+INFO:root:Average loss=0.15624022751282424
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2765 | LCC = 0.8339 | SRCC = 0.8333 ]
+INFO:root:Average loss=0.15913671809129226
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2665 | LCC = 0.8338 | SRCC = 0.8316 ]
+INFO:root:Average loss=0.13689601325835937
+INFO:root:
+[Valid][12][UTT][ MSE = 0.4595 | LCC = 0.8156 | SRCC = 0.8281 ]
+INFO:root:Average loss=0.145293340469018
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2580 | LCC = 0.8344 | SRCC = 0.8336 ]
+INFO:root:Average loss=0.11797587726360713
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2616 | LCC = 0.8324 | SRCC = 0.8320 ]
+INFO:root:Average loss=0.11355487973644184
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2780 | LCC = 0.8288 | SRCC = 0.8294 ]
+INFO:root:Average loss=0.11141859854643162
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2782 | LCC = 0.8372 | SRCC = 0.8367 ]
+INFO:root:Average loss=0.0998157137670578
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2669 | LCC = 0.8339 | SRCC = 0.8337 ]
+INFO:root:Average loss=0.09322399751116069
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2959 | LCC = 0.8303 | SRCC = 0.8299 ]
+INFO:root:Average loss=0.09732641886251095
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2936 | LCC = 0.8333 | SRCC = 0.8346 ]
+INFO:root:Average loss=0.09167037942470649
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2595 | LCC = 0.8352 | SRCC = 0.8347 ]
+INFO:root:Average loss=0.08147403459327343
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2709 | LCC = 0.8317 | SRCC = 0.8296 ]
+INFO:root:Average loss=0.08141978199665363
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2643 | LCC = 0.8311 | SRCC = 0.8293 ]
+INFO:root:Average loss=0.0818121949545084
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2705 | LCC = 0.8294 | SRCC = 0.8294 ]
+INFO:root:Average loss=0.07335440924343391
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2770 | LCC = 0.8308 | SRCC = 0.8301 ]
+INFO:root:Average loss=0.07258073187982425
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2721 | LCC = 0.8254 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.07038236471513908
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2833 | LCC = 0.8268 | SRCC = 0.8256 ]
+INFO:root:Average loss=0.06586808587114017
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2960 | LCC = 0.8234 | SRCC = 0.8247 ]
+INFO:root:Average loss=0.06269122134798612
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2853 | LCC = 0.8253 | SRCC = 0.8242 ]
+INFO:root:Average loss=0.05808411385768499
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2767 | LCC = 0.8272 | SRCC = 0.8263 ]
+INFO:root:Average loss=0.0542663235265093
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2802 | LCC = 0.8264 | SRCC = 0.8261 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2913 | LCC = 0.8363 | SRCC = 0.8412 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run2/config.gin b/results_bvcc/w2v2_large_asr_en/layer8/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer8/run2/model.pt b/results_bvcc/w2v2_large_asr_en/layer8/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..86b34a768959571347de41d1691c325ffc6efe93
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8617f73a3b2295ba642036ce75f2a6c7951d55c8b06f1260591ac3821cd347e
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run2/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer8/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cb8d3f54809dc6fff75467a7ff0116f07700209e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:703d3ba8f646af8a08b923c2505da3da16afc9325a6cc87cc7110ac5e1a049a0
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run2/train.log b/results_bvcc/w2v2_large_asr_en/layer8/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..d387ed025567bd01a8ba14d2b5bb2b054121ea21
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.5920990210695144
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3649 | LCC = 0.7757 | SRCC = 0.7778 ]
+INFO:root:Average loss=0.315941806213978
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3309 | LCC = 0.8091 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.2754261161272342
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3146 | LCC = 0.8244 | SRCC = 0.8214 ]
+INFO:root:Average loss=0.24388368485065606
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2826 | LCC = 0.8283 | SRCC = 0.8282 ]
+INFO:root:Average loss=0.23461270351440477
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2849 | LCC = 0.8308 | SRCC = 0.8300 ]
+INFO:root:Average loss=0.22730369655749735
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3172 | LCC = 0.8360 | SRCC = 0.8334 ]
+INFO:root:Average loss=0.19448089179320213
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2668 | LCC = 0.8344 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.19232425504387954
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3149 | LCC = 0.8325 | SRCC = 0.8356 ]
+INFO:root:Average loss=0.18380374155747584
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2603 | LCC = 0.8385 | SRCC = 0.8370 ]
+INFO:root:Average loss=0.15862640318198082
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2492 | LCC = 0.8418 | SRCC = 0.8400 ]
+INFO:root:Average loss=0.15488147649627465
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2783 | LCC = 0.8362 | SRCC = 0.8370 ]
+INFO:root:Average loss=0.1432800181210041
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3576 | LCC = 0.8301 | SRCC = 0.8357 ]
+INFO:root:Average loss=0.1336900950051271
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2686 | LCC = 0.8365 | SRCC = 0.8323 ]
+INFO:root:Average loss=0.13438949265923256
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2675 | LCC = 0.8375 | SRCC = 0.8357 ]
+INFO:root:Average loss=0.1270282962001287
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3244 | LCC = 0.8294 | SRCC = 0.8330 ]
+INFO:root:Average loss=0.11324216177066167
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2641 | LCC = 0.8346 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.10556095962723096
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3159 | LCC = 0.8293 | SRCC = 0.8298 ]
+INFO:root:Average loss=0.10474201798056945
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2623 | LCC = 0.8346 | SRCC = 0.8324 ]
+INFO:root:Average loss=0.09181390120050846
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2673 | LCC = 0.8364 | SRCC = 0.8363 ]
+INFO:root:Average loss=0.09121391955667581
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2641 | LCC = 0.8317 | SRCC = 0.8281 ]
+INFO:root:Average loss=0.09132993593811989
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2712 | LCC = 0.8301 | SRCC = 0.8284 ]
+INFO:root:Average loss=0.08390584774315357
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2724 | LCC = 0.8324 | SRCC = 0.8314 ]
+INFO:root:Average loss=0.07488486624489991
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2715 | LCC = 0.8280 | SRCC = 0.8255 ]
+INFO:root:Average loss=0.077513571924124
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2693 | LCC = 0.8311 | SRCC = 0.8274 ]
+INFO:root:Average loss=0.0702642400104266
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2714 | LCC = 0.8331 | SRCC = 0.8299 ]
+INFO:root:Average loss=0.06864713065517254
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2871 | LCC = 0.8298 | SRCC = 0.8298 ]
+INFO:root:Average loss=0.06657990861015442
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2775 | LCC = 0.8297 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.06300901989333141
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2723 | LCC = 0.8287 | SRCC = 0.8260 ]
+INFO:root:Average loss=0.05729062688083221
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2627 | LCC = 0.8300 | SRCC = 0.8264 ]
+INFO:root:Average loss=0.059561349547062166
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2771 | LCC = 0.8268 | SRCC = 0.8241 ]
+INFO:root:Average loss=0.055547101805225395
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2717 | LCC = 0.8299 | SRCC = 0.8272 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2575 | LCC = 0.8446 | SRCC = 0.8474 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run3/config.gin b/results_bvcc/w2v2_large_asr_en/layer8/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer8/run3/model.pt b/results_bvcc/w2v2_large_asr_en/layer8/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f5798be3d41fcee92958800da360876a1cca2195
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f22a2c4d7c0040bb775727a84013025d03f0e06a926140e2a9916ab01ef6c9da
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run3/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer8/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..61c5563c5d923c5a2cd31e1b2902c35bc187d9c5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d505dac961b7d6f8f3dd65417433c8a999dd99bc9896af80f47c15d1c69a561
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run3/train.log b/results_bvcc/w2v2_large_asr_en/layer8/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..aca83a9c9e75331c83ad5c4ab6cfda11c15a3da8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.62507202457159
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3373 | LCC = 0.7824 | SRCC = 0.7874 ]
+INFO:root:Average loss=0.3270737624321228
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3822 | LCC = 0.8185 | SRCC = 0.8211 ]
+INFO:root:Average loss=0.271847152366088
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2955 | LCC = 0.8281 | SRCC = 0.8292 ]
+INFO:root:Average loss=0.2513488504366997
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3099 | LCC = 0.8360 | SRCC = 0.8354 ]
+INFO:root:Average loss=0.22884731720655394
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2824 | LCC = 0.8413 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.21761512316954443
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2855 | LCC = 0.8381 | SRCC = 0.8380 ]
+INFO:root:Average loss=0.20151126633087793
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2792 | LCC = 0.8417 | SRCC = 0.8409 ]
+INFO:root:Average loss=0.18988199952321175
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2741 | LCC = 0.8408 | SRCC = 0.8386 ]
+INFO:root:Average loss=0.18032182456973272
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2566 | LCC = 0.8415 | SRCC = 0.8423 ]
+INFO:root:Average loss=0.16282190497104937
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2710 | LCC = 0.8351 | SRCC = 0.8329 ]
+INFO:root:Average loss=0.156348254913703
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2779 | LCC = 0.8328 | SRCC = 0.8332 ]
+INFO:root:Average loss=0.147126088062158
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3575 | LCC = 0.8296 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.14367908401749072
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2525 | LCC = 0.8385 | SRCC = 0.8373 ]
+INFO:root:Average loss=0.13543562390483344
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3128 | LCC = 0.8358 | SRCC = 0.8370 ]
+INFO:root:Average loss=0.12543009374386224
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2999 | LCC = 0.8323 | SRCC = 0.8314 ]
+INFO:root:Average loss=0.11478446003718254
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2736 | LCC = 0.8289 | SRCC = 0.8282 ]
+INFO:root:Average loss=0.1095991319952867
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2900 | LCC = 0.8276 | SRCC = 0.8285 ]
+INFO:root:Average loss=0.115093226186358
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2651 | LCC = 0.8318 | SRCC = 0.8303 ]
+INFO:root:Average loss=0.10151107532855792
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3142 | LCC = 0.8253 | SRCC = 0.8272 ]
+INFO:root:Average loss=0.09904264930922252
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3046 | LCC = 0.8281 | SRCC = 0.8282 ]
+INFO:root:Average loss=0.09665300720968308
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2675 | LCC = 0.8309 | SRCC = 0.8291 ]
+INFO:root:Average loss=0.08651875035885052
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3313 | LCC = 0.8272 | SRCC = 0.8273 ]
+INFO:root:Average loss=0.08233245161290352
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2651 | LCC = 0.8298 | SRCC = 0.8296 ]
+INFO:root:Average loss=0.07400796844218022
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2598 | LCC = 0.8337 | SRCC = 0.8334 ]
+INFO:root:Average loss=0.07835986489095749
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3028 | LCC = 0.8274 | SRCC = 0.8272 ]
+INFO:root:Average loss=0.06909838715233864
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2838 | LCC = 0.8310 | SRCC = 0.8296 ]
+INFO:root:Average loss=0.07173858826550153
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2990 | LCC = 0.8276 | SRCC = 0.8276 ]
+INFO:root:Average loss=0.07067272064682001
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2790 | LCC = 0.8283 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.06841386424807402
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3116 | LCC = 0.8285 | SRCC = 0.8282 ]
+INFO:root:Average loss=0.061001754294221215
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3243 | LCC = 0.8251 | SRCC = 0.8253 ]
+INFO:root:Average loss=0.06288361191176452
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3033 | LCC = 0.8248 | SRCC = 0.8250 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2514 | LCC = 0.8529 | SRCC = 0.8538 ]
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run4/config.gin b/results_bvcc/w2v2_large_asr_en/layer8/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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_large_asr_en/layer8/run4/model.pt b/results_bvcc/w2v2_large_asr_en/layer8/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..95abe0918a063935d65b0e1fcba0439f5beb115e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3965899a188c7550f44283e2e14e9b278d018f3877aebb9d3e503a489df2542b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run4/model_best.pt b/results_bvcc/w2v2_large_asr_en/layer8/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a2d0706b46c83274ee76844655ab8630d975a3d6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/layer8/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:885de4f6f48ba7a68a5586bcbd1d0588a18d15d2825843d1a31b0142d7d22df7
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_en/layer8/run4/train.log b/results_bvcc/w2v2_large_asr_en/layer8/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..9b935c4c019e2d3f88c2ab1d52c6e5dc5b99c331
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_en/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=0.6618998516828586
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3813 | LCC = 0.7678 | SRCC = 0.7745 ]
+INFO:root:Average loss=0.32140526213707066
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3141 | LCC = 0.8194 | SRCC = 0.8178 ]
+INFO:root:Average loss=0.2664671458112888
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2960 | LCC = 0.8264 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.24386629443138075
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2764 | LCC = 0.8313 | SRCC = 0.8293 ]
+INFO:root:Average loss=0.21986305656341407
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3138 | LCC = 0.8316 | SRCC = 0.8292 ]
+INFO:root:Average loss=0.20802140885438675
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2662 | LCC = 0.8380 | SRCC = 0.8354 ]
+INFO:root:Average loss=0.2245303722910392
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3223 | LCC = 0.8292 | SRCC = 0.8310 ]
+INFO:root:Average loss=0.1906212916932045
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2663 | LCC = 0.8327 | SRCC = 0.8314 ]
+INFO:root:Average loss=0.17657794535924226
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2703 | LCC = 0.8323 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.1814419880318336
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2643 | LCC = 0.8354 | SRCC = 0.8341 ]
+INFO:root:Average loss=0.15958523291807908
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2653 | LCC = 0.8337 | SRCC = 0.8346 ]
+INFO:root:Average loss=0.15289685454888222
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2733 | LCC = 0.8366 | SRCC = 0.8341 ]
+INFO:root:Average loss=0.14085470245052606
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2654 | LCC = 0.8349 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.13483140302392152
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2632 | LCC = 0.8318 | SRCC = 0.8299 ]
+INFO:root:Average loss=0.12815706632458246
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2709 | LCC = 0.8359 | SRCC = 0.8338 ]
+INFO:root:Average loss=0.12275465367696224
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2657 | LCC = 0.8315 | SRCC = 0.8310 ]
+INFO:root:Average loss=0.11643308515732105
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2858 | LCC = 0.8267 | SRCC = 0.8254 ]
+INFO:root:Average loss=0.10818114169897178
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2643 | LCC = 0.8313 | SRCC = 0.8291 ]
+INFO:root:Average loss=0.09976167757159625
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2759 | LCC = 0.8300 | SRCC = 0.8279 ]
+INFO:root:Average loss=0.09941325895488262
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2609 | LCC = 0.8349 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.09182443125889851
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2648 | LCC = 0.8279 | SRCC = 0.8265 ]
+INFO:root:Average loss=0.08794683007857738
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2652 | LCC = 0.8265 | SRCC = 0.8237 ]
+INFO:root:Average loss=0.08387588938841453
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2637 | LCC = 0.8325 | SRCC = 0.8296 ]
+INFO:root:Average loss=0.07581656650663951
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2690 | LCC = 0.8273 | SRCC = 0.8246 ]
+INFO:root:Average loss=0.07839438930535927
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2755 | LCC = 0.8255 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.07357982226098195
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2750 | LCC = 0.8257 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.07027270090885651
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2651 | LCC = 0.8270 | SRCC = 0.8257 ]
+INFO:root:Average loss=0.07055736667452714
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3173 | LCC = 0.8244 | SRCC = 0.8217 ]
+INFO:root:Average loss=0.06606034208566715
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2794 | LCC = 0.8229 | SRCC = 0.8195 ]
+INFO:root:Average loss=0.0651187600615697
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2612 | LCC = 0.8304 | SRCC = 0.8278 ]
+INFO:root:Average loss=0.06710948943136594
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2706 | LCC = 0.8254 | SRCC = 0.8238 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2537 | LCC = 0.8462 | SRCC = 0.8464 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer0/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer0/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a84e2d631ce32301986b8b4b602f633551d6013b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:387e6702770fe9f86d9fe5854ca3375ab69073e38a8354aee3ecd629360dba1b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..60003d24944a20dcbecf5d76b5a28b77a03c0826
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7281b72d4a0fb9ea0f545553ab5cc7f53b86f315d91f3675521ce4eeabc639b7
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer0/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..40c557f29b1b3f91fe4cbcd8d0a4d6feaec5e2f6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6153803819265121
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6401 | LCC = 0.6330 | SRCC = 0.6662 ]
+INFO:root:Average loss=0.4232792101609401
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4396 | LCC = 0.7112 | SRCC = 0.7071 ]
+INFO:root:Average loss=0.3782325499714949
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4062 | LCC = 0.7306 | SRCC = 0.7327 ]
+INFO:root:Average loss=0.3674026601589643
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3928 | LCC = 0.7361 | SRCC = 0.7342 ]
+INFO:root:Average loss=0.3479069496194522
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3755 | LCC = 0.7569 | SRCC = 0.7527 ]
+INFO:root:Average loss=0.3315584021501052
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3794 | LCC = 0.7630 | SRCC = 0.7596 ]
+INFO:root:Average loss=0.3117982189242656
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3459 | LCC = 0.7726 | SRCC = 0.7681 ]
+INFO:root:Average loss=0.30513986830527967
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3505 | LCC = 0.7765 | SRCC = 0.7711 ]
+INFO:root:Average loss=0.29888534049193066
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3392 | LCC = 0.7874 | SRCC = 0.7839 ]
+INFO:root:Average loss=0.2812736145196817
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3691 | LCC = 0.7919 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.27974004050095874
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3286 | LCC = 0.7943 | SRCC = 0.7917 ]
+INFO:root:Average loss=0.2807014573078889
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3800 | LCC = 0.7888 | SRCC = 0.7880 ]
+INFO:root:Average loss=0.28032174209753674
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3107 | LCC = 0.7977 | SRCC = 0.7939 ]
+INFO:root:Average loss=0.2575501853074783
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3809 | LCC = 0.8019 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.26048811104817265
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3499 | LCC = 0.7928 | SRCC = 0.7915 ]
+INFO:root:Average loss=0.23879220432195908
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3179 | LCC = 0.8081 | SRCC = 0.8066 ]
+INFO:root:Average loss=0.23051390825555876
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2834 | LCC = 0.8142 | SRCC = 0.8128 ]
+INFO:root:Average loss=0.23170274018477172
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3150 | LCC = 0.8071 | SRCC = 0.8074 ]
+INFO:root:Average loss=0.21125817833802638
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3039 | LCC = 0.8056 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.21651424352939314
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3076 | LCC = 0.8057 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.21577594047173476
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3195 | LCC = 0.8073 | SRCC = 0.8089 ]
+INFO:root:Average loss=0.1999552971086441
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3972 | LCC = 0.7971 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.22084282453243548
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2860 | LCC = 0.8195 | SRCC = 0.8197 ]
+INFO:root:Average loss=0.19021562143014029
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3183 | LCC = 0.8039 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.20573836402633253
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3038 | LCC = 0.8171 | SRCC = 0.8176 ]
+INFO:root:Average loss=0.18840821297504964
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3036 | LCC = 0.8159 | SRCC = 0.8146 ]
+INFO:root:Average loss=0.2014197203784417
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3202 | LCC = 0.8203 | SRCC = 0.8219 ]
+INFO:root:Average loss=0.18026252998373446
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3121 | LCC = 0.8166 | SRCC = 0.8175 ]
+INFO:root:Average loss=0.18768034044366616
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2821 | LCC = 0.8176 | SRCC = 0.8179 ]
+INFO:root:Average loss=0.18158646214466828
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2907 | LCC = 0.8191 | SRCC = 0.8207 ]
+INFO:root:Average loss=0.17247915353912574
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2737 | LCC = 0.8239 | SRCC = 0.8246 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2566 | LCC = 0.8407 | SRCC = 0.8382 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer0/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer0/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0f3e712d8761cfcf2f0de963e0e44a06a3c63f47
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:913d5b718aef2047d9370728272c958ed82fcc41da3185beaac608cb7acf28b5
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..388b8ed18cc54e868bb9af5bc7e954c2e184e715
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66578c1152cb437f41d7aea4f5b515710f6cfbd295d25c8a6ba47934a51d3215
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer0/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e9a094a968b94a3a3c779d7dc37f8bc61dbc1232
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6450774704034512
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5002 | LCC = 0.6710 | SRCC = 0.6740 ]
+INFO:root:Average loss=0.4261036969912358
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4899 | LCC = 0.6975 | SRCC = 0.6996 ]
+INFO:root:Average loss=0.38706462047038936
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4248 | LCC = 0.7266 | SRCC = 0.7156 ]
+INFO:root:Average loss=0.3680088038627918
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3885 | LCC = 0.7447 | SRCC = 0.7405 ]
+INFO:root:Average loss=0.34551261900327146
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3926 | LCC = 0.7588 | SRCC = 0.7545 ]
+INFO:root:Average loss=0.3383384375618054
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3821 | LCC = 0.7578 | SRCC = 0.7484 ]
+INFO:root:Average loss=0.324850643101411
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3613 | LCC = 0.7737 | SRCC = 0.7731 ]
+INFO:root:Average loss=0.3112261923841941
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4360 | LCC = 0.7730 | SRCC = 0.7732 ]
+INFO:root:Average loss=0.31720944933402234
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3397 | LCC = 0.7878 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.28424859257080615
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3852 | LCC = 0.7822 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.2980057927660453
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3622 | LCC = 0.7990 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.3048993144661952
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3702 | LCC = 0.7938 | SRCC = 0.7900 ]
+INFO:root:Average loss=0.2754586201447707
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3264 | LCC = 0.8014 | SRCC = 0.8006 ]
+INFO:root:Average loss=0.27401741517659944
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3481 | LCC = 0.8034 | SRCC = 0.8033 ]
+INFO:root:Average loss=0.270308456550806
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3517 | LCC = 0.8064 | SRCC = 0.8067 ]
+INFO:root:Average loss=0.2522292989186751
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3264 | LCC = 0.8053 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.25160848196500385
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3824 | LCC = 0.8057 | SRCC = 0.8071 ]
+INFO:root:Average loss=0.23654232823695892
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3313 | LCC = 0.8099 | SRCC = 0.8092 ]
+INFO:root:Average loss=0.22728628034775072
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3249 | LCC = 0.8083 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.24224627247223487
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3188 | LCC = 0.8140 | SRCC = 0.8152 ]
+INFO:root:Average loss=0.21664529255567452
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3088 | LCC = 0.8097 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.21097801931393453
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2884 | LCC = 0.8182 | SRCC = 0.8170 ]
+INFO:root:Average loss=0.2097721975774337
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3746 | LCC = 0.8049 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.2042856203057827
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3282 | LCC = 0.8144 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.2238229210369098
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2880 | LCC = 0.8130 | SRCC = 0.8131 ]
+INFO:root:Average loss=0.19801922601002914
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3012 | LCC = 0.8177 | SRCC = 0.8172 ]
+INFO:root:Average loss=0.2214040040778808
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3169 | LCC = 0.8194 | SRCC = 0.8210 ]
+INFO:root:Average loss=0.18251966484464133
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2993 | LCC = 0.8174 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.18593005960186323
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3012 | LCC = 0.8148 | SRCC = 0.8146 ]
+INFO:root:Average loss=0.17166097500385383
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3074 | LCC = 0.8165 | SRCC = 0.8136 ]
+INFO:root:Average loss=0.1892995886886731
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3432 | LCC = 0.8132 | SRCC = 0.8148 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3423 | LCC = 0.8236 | SRCC = 0.8292 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer0/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer0/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e90ebf5220047daad5d441d945eacda3ab35b2f8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50ceffd4c36059672f7633a19a435a926f44129eb03f6f14d01efc0d096ecd87
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d87f31f8bb0ac2078a501d1d039dbfec87579edc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47207aaec7f4836189dc4f242e232cd112e9f0af0bd17ce93f26dfbfacd39dd4
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer0/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..374831a8e14fe2d977c2d2214b2497bde8eb8bce
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5657562338388883
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4685 | LCC = 0.6865 | SRCC = 0.6828 ]
+INFO:root:Average loss=0.41559153222120726
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4254 | LCC = 0.7178 | SRCC = 0.7157 ]
+INFO:root:Average loss=0.3761133916485004
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3951 | LCC = 0.7305 | SRCC = 0.7241 ]
+INFO:root:Average loss=0.3468074456621439
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4250 | LCC = 0.7398 | SRCC = 0.7385 ]
+INFO:root:Average loss=0.34677179883687925
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3941 | LCC = 0.7566 | SRCC = 0.7508 ]
+INFO:root:Average loss=0.3442192039428613
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4127 | LCC = 0.7600 | SRCC = 0.7576 ]
+INFO:root:Average loss=0.308611016242932
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3850 | LCC = 0.7689 | SRCC = 0.7657 ]
+INFO:root:Average loss=0.311153199046086
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3488 | LCC = 0.7784 | SRCC = 0.7750 ]
+INFO:root:Average loss=0.3005408732554851
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3352 | LCC = 0.7888 | SRCC = 0.7855 ]
+INFO:root:Average loss=0.29458300845745283
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3252 | LCC = 0.7861 | SRCC = 0.7859 ]
+INFO:root:Average loss=0.2729555742862897
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3182 | LCC = 0.7910 | SRCC = 0.7911 ]
+INFO:root:Average loss=0.2773521920809379
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3769 | LCC = 0.7878 | SRCC = 0.7864 ]
+INFO:root:Average loss=0.27876011033852893
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3508 | LCC = 0.7937 | SRCC = 0.7907 ]
+INFO:root:Average loss=0.2625609877017828
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3278 | LCC = 0.8046 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.24638667129553282
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3118 | LCC = 0.7917 | SRCC = 0.7948 ]
+INFO:root:Average loss=0.251423465517851
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3090 | LCC = 0.8034 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.235788479160804
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3467 | LCC = 0.8051 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.24142851718725303
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3332 | LCC = 0.8080 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.22268550441815302
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3104 | LCC = 0.8087 | SRCC = 0.8106 ]
+INFO:root:Average loss=0.22597307081405932
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3102 | LCC = 0.8108 | SRCC = 0.8142 ]
+INFO:root:Average loss=0.23159974011091086
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2994 | LCC = 0.8133 | SRCC = 0.8137 ]
+INFO:root:Average loss=0.23651938235912567
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3312 | LCC = 0.8107 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.2039145087966552
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3022 | LCC = 0.8104 | SRCC = 0.8118 ]
+INFO:root:Average loss=0.20785519900994423
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3136 | LCC = 0.8047 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.2138944931137256
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3001 | LCC = 0.8156 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.20072228595232353
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2989 | LCC = 0.8157 | SRCC = 0.8146 ]
+INFO:root:Average loss=0.18797264926326582
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3145 | LCC = 0.8107 | SRCC = 0.8124 ]
+INFO:root:Average loss=0.18415662321524742
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2987 | LCC = 0.8048 | SRCC = 0.8079 ]
+INFO:root:Average loss=0.1751980813076863
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2913 | LCC = 0.8121 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.1790683510211798
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2898 | LCC = 0.8102 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.17473183763332856
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3025 | LCC = 0.8108 | SRCC = 0.8129 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2915 | LCC = 0.8319 | SRCC = 0.8294 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer0/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer0/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0b2a5aaec9d07e79a6c2a5e17772a42e24bff66a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd4aa9ba544d9ec475773f56df877a3963b6c7d391aee4f43a2b565663400b1d
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e0d199d747012e3213eaaf773138ec5e85c67814
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8979961fbb94ee5ccce03a1f5ea5cb1fea4b89f3446776293e4bf67115c554cb
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer0/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..5fa24a14e49c4970110f0c09b2fd64d4126650db
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6651877810557684
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4813 | LCC = 0.6675 | SRCC = 0.6731 ]
+INFO:root:Average loss=0.43530282913110196
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4311 | LCC = 0.7074 | SRCC = 0.7045 ]
+INFO:root:Average loss=0.4079463405486865
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4215 | LCC = 0.7250 | SRCC = 0.7271 ]
+INFO:root:Average loss=0.36429607829986477
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3944 | LCC = 0.7457 | SRCC = 0.7421 ]
+INFO:root:Average loss=0.3426483986087335
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4402 | LCC = 0.7461 | SRCC = 0.7418 ]
+INFO:root:Average loss=0.3312659122240849
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3541 | LCC = 0.7703 | SRCC = 0.7667 ]
+INFO:root:Average loss=0.3216481130474653
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3536 | LCC = 0.7748 | SRCC = 0.7739 ]
+INFO:root:Average loss=0.3214389496506789
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3711 | LCC = 0.7801 | SRCC = 0.7773 ]
+INFO:root:Average loss=0.2950542208093863
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3200 | LCC = 0.7897 | SRCC = 0.7866 ]
+INFO:root:Average loss=0.2942025535381757
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3585 | LCC = 0.7917 | SRCC = 0.7900 ]
+INFO:root:Average loss=0.2975920089162313
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3353 | LCC = 0.7867 | SRCC = 0.7868 ]
+INFO:root:Average loss=0.273456199046893
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3144 | LCC = 0.7993 | SRCC = 0.7977 ]
+INFO:root:Average loss=0.263991385889359
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3347 | LCC = 0.7998 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.24873472043336967
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3286 | LCC = 0.8004 | SRCC = 0.8007 ]
+INFO:root:Average loss=0.2477194259946163
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3280 | LCC = 0.7999 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.24804968597033086
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3269 | LCC = 0.8055 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.25174413869778317
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3089 | LCC = 0.8007 | SRCC = 0.8012 ]
+INFO:root:Average loss=0.26288133821426296
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3196 | LCC = 0.8033 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.2204845612629866
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2964 | LCC = 0.8093 | SRCC = 0.8088 ]
+INFO:root:Average loss=0.21760398187698463
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2977 | LCC = 0.8058 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.21028384661827332
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3336 | LCC = 0.8073 | SRCC = 0.8080 ]
+INFO:root:Average loss=0.20359274095449692
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2974 | LCC = 0.8115 | SRCC = 0.8104 ]
+INFO:root:Average loss=0.21926447921074355
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2957 | LCC = 0.8050 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.21156210280381715
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2877 | LCC = 0.8126 | SRCC = 0.8143 ]
+INFO:root:Average loss=0.21626374068168494
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3260 | LCC = 0.8089 | SRCC = 0.8124 ]
+INFO:root:Average loss=0.19321845605587348
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2945 | LCC = 0.8121 | SRCC = 0.8114 ]
+INFO:root:Average loss=0.1817326994660573
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2927 | LCC = 0.8168 | SRCC = 0.8175 ]
+INFO:root:Average loss=0.18405426799868926
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2946 | LCC = 0.8154 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.17709417106249395
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2810 | LCC = 0.8167 | SRCC = 0.8165 ]
+INFO:root:Average loss=0.1717715593866813
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2902 | LCC = 0.8141 | SRCC = 0.8159 ]
+INFO:root:Average loss=0.1831495240330696
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3324 | LCC = 0.8133 | SRCC = 0.8158 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3102 | LCC = 0.8229 | SRCC = 0.8195 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer0/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer0/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..01971faea1575ef0e011fe000e34e892d5ad4a66
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:730a9f1cfd06f5fd61f1f3fb1822d4c89ed108a84b7d24b44eb8ade6ba7c1c50
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer0/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e94728b07cf6803a1b66a4f550d04bfdd1bca496
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer0/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e62c91197b5bb3b989de7b88e31b3a7a5f8a0c7f7f2f1298037300302cfdc7ff
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer0/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer0/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ae16574d4aec05fab4a7cb2c0c5620f8dcbe69e9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5607142112193964
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4707 | LCC = 0.6874 | SRCC = 0.6881 ]
+INFO:root:Average loss=0.42144791514445573
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4451 | LCC = 0.7008 | SRCC = 0.6901 ]
+INFO:root:Average loss=0.38114097790840346
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4293 | LCC = 0.7358 | SRCC = 0.7299 ]
+INFO:root:Average loss=0.3571292280386656
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4312 | LCC = 0.7367 | SRCC = 0.7313 ]
+INFO:root:Average loss=0.3411806495143817
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3862 | LCC = 0.7575 | SRCC = 0.7571 ]
+INFO:root:Average loss=0.34588965658958143
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3638 | LCC = 0.7692 | SRCC = 0.7688 ]
+INFO:root:Average loss=0.3337958772213031
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3999 | LCC = 0.7728 | SRCC = 0.7691 ]
+INFO:root:Average loss=0.3234243312707314
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4050 | LCC = 0.7807 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.30625424056481093
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3823 | LCC = 0.7848 | SRCC = 0.7808 ]
+INFO:root:Average loss=0.29899209164656126
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3417 | LCC = 0.7977 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.3082099056396729
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3383 | LCC = 0.7852 | SRCC = 0.7895 ]
+INFO:root:Average loss=0.2865468254074072
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3229 | LCC = 0.7954 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.2748844377123393
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3649 | LCC = 0.7912 | SRCC = 0.7948 ]
+INFO:root:Average loss=0.25685524290952927
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3074 | LCC = 0.8061 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.2514217856984872
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3016 | LCC = 0.8062 | SRCC = 0.8069 ]
+INFO:root:Average loss=0.24578115955377236
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3762 | LCC = 0.7996 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.23930794554643142
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3051 | LCC = 0.8052 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.24138082888646004
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3264 | LCC = 0.8033 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.22697756802424407
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2916 | LCC = 0.8094 | SRCC = 0.8096 ]
+INFO:root:Average loss=0.23757219047118455
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3083 | LCC = 0.8103 | SRCC = 0.8116 ]
+INFO:root:Average loss=0.21668737534529123
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2964 | LCC = 0.8151 | SRCC = 0.8159 ]
+INFO:root:Average loss=0.2231595216271205
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3402 | LCC = 0.8060 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.2071292927631965
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3266 | LCC = 0.8124 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.228471701248334
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3089 | LCC = 0.8140 | SRCC = 0.8131 ]
+INFO:root:Average loss=0.20009363738772196
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2915 | LCC = 0.8126 | SRCC = 0.8130 ]
+INFO:root:Average loss=0.19390814340649506
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2918 | LCC = 0.8186 | SRCC = 0.8184 ]
+INFO:root:Average loss=0.20822193693274108
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3090 | LCC = 0.8194 | SRCC = 0.8190 ]
+INFO:root:Average loss=0.19226106523703307
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2942 | LCC = 0.8168 | SRCC = 0.8181 ]
+INFO:root:Average loss=0.18272897209494543
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2813 | LCC = 0.8190 | SRCC = 0.8190 ]
+INFO:root:Average loss=0.18614336838706946
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3586 | LCC = 0.8109 | SRCC = 0.8143 ]
+INFO:root:Average loss=0.18242253286716265
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3137 | LCC = 0.8120 | SRCC = 0.8129 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3084 | LCC = 0.8318 | SRCC = 0.8304 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer10/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer10/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b7593cd6e9f62ec61ea2863922cd53b8392a8cc9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f9ae3eec41fdcf465e487d3d31c2bf444dbdf81c592a9fc49eddf4d438dad13
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..929467b1ac633fff052d550e28dfbf65729a7061
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e25a354d77833659defedf5c0f028bac7a968cdcce9eac17929642728da52af
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer10/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..408e38a6e95a60d31950206fe38116ce177d4818
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6046717017889023
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4153 | LCC = 0.7776 | SRCC = 0.7819 ]
+INFO:root:Average loss=0.30716180476622706
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3166 | LCC = 0.8102 | SRCC = 0.8090 ]
+INFO:root:Average loss=0.27773914677210343
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3247 | LCC = 0.8259 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.25332910147232884
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2930 | LCC = 0.8322 | SRCC = 0.8299 ]
+INFO:root:Average loss=0.25064094135394466
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3661 | LCC = 0.8261 | SRCC = 0.8249 ]
+INFO:root:Average loss=0.2323698552373128
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3005 | LCC = 0.8320 | SRCC = 0.8299 ]
+INFO:root:Average loss=0.2113648816370047
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2673 | LCC = 0.8310 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.21382839366411552
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3310 | LCC = 0.8303 | SRCC = 0.8272 ]
+INFO:root:Average loss=0.21348972140978545
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2878 | LCC = 0.8365 | SRCC = 0.8328 ]
+INFO:root:Average loss=0.18542086380796555
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2857 | LCC = 0.8312 | SRCC = 0.8275 ]
+INFO:root:Average loss=0.1726516037224195
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2919 | LCC = 0.8345 | SRCC = 0.8307 ]
+INFO:root:Average loss=0.15772760946017045
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2688 | LCC = 0.8337 | SRCC = 0.8290 ]
+INFO:root:Average loss=0.17123856920844469
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2586 | LCC = 0.8368 | SRCC = 0.8349 ]
+INFO:root:Average loss=0.1587505470483731
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2834 | LCC = 0.8308 | SRCC = 0.8256 ]
+INFO:root:Average loss=0.17301047707979494
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2696 | LCC = 0.8351 | SRCC = 0.8327 ]
+INFO:root:Average loss=0.15832304591551805
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3476 | LCC = 0.8297 | SRCC = 0.8275 ]
+INFO:root:Average loss=0.14522287669854286
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2771 | LCC = 0.8335 | SRCC = 0.8318 ]
+INFO:root:Average loss=0.12502122297883034
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3052 | LCC = 0.8324 | SRCC = 0.8302 ]
+INFO:root:Average loss=0.12787257784452194
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2675 | LCC = 0.8334 | SRCC = 0.8340 ]
+INFO:root:Average loss=0.1097886551840183
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2922 | LCC = 0.8332 | SRCC = 0.8309 ]
+INFO:root:Average loss=0.10923143934745055
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2753 | LCC = 0.8331 | SRCC = 0.8334 ]
+INFO:root:Average loss=0.12875545607545438
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3068 | LCC = 0.8335 | SRCC = 0.8318 ]
+INFO:root:Average loss=0.10052486389684372
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2657 | LCC = 0.8339 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.09969302519964866
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2668 | LCC = 0.8345 | SRCC = 0.8335 ]
+INFO:root:Average loss=0.09224865494821316
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2602 | LCC = 0.8365 | SRCC = 0.8338 ]
+INFO:root:Average loss=0.08327445067847386
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2647 | LCC = 0.8346 | SRCC = 0.8333 ]
+INFO:root:Average loss=0.0817270096964561
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2659 | LCC = 0.8366 | SRCC = 0.8350 ]
+INFO:root:Average loss=0.0790156513834611
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2835 | LCC = 0.8363 | SRCC = 0.8341 ]
+INFO:root:Average loss=0.07113120666681191
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2637 | LCC = 0.8322 | SRCC = 0.8290 ]
+INFO:root:Average loss=0.07220061046954913
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2680 | LCC = 0.8375 | SRCC = 0.8352 ]
+INFO:root:Average loss=0.06538099852892068
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2675 | LCC = 0.8348 | SRCC = 0.8341 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2834 | LCC = 0.8343 | SRCC = 0.8331 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer10/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer10/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7bddd3e1f12a30278969fc49d3667092a76992c3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7bda14f62cb5a785ef927a12cc855db15844159e38134b3332f52c67a339113
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f570440d490e65e9b511699bfa3574d678745807
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edeccaf8d157d62c96877e9433609756bb2e08b26a5223689925d9b820bf0ddb
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer10/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..85cd7c896558106c0ccda87177ef254e2d736a90
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5884954669536688
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3432 | LCC = 0.7892 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.3182104705617978
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3347 | LCC = 0.8184 | SRCC = 0.8196 ]
+INFO:root:Average loss=0.2817520623405774
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3168 | LCC = 0.8270 | SRCC = 0.8269 ]
+INFO:root:Average loss=0.25120853422543943
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2912 | LCC = 0.8303 | SRCC = 0.8284 ]
+INFO:root:Average loss=0.23217189961518997
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2843 | LCC = 0.8360 | SRCC = 0.8346 ]
+INFO:root:Average loss=0.23431386282810798
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2857 | LCC = 0.8374 | SRCC = 0.8339 ]
+INFO:root:Average loss=0.20478718250225753
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2816 | LCC = 0.8340 | SRCC = 0.8321 ]
+INFO:root:Average loss=0.20227090020974478
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2754 | LCC = 0.8380 | SRCC = 0.8371 ]
+INFO:root:Average loss=0.18404924458800218
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3154 | LCC = 0.8305 | SRCC = 0.8308 ]
+INFO:root:Average loss=0.18087642773603782
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2803 | LCC = 0.8397 | SRCC = 0.8376 ]
+INFO:root:Average loss=0.16402200552133414
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2630 | LCC = 0.8344 | SRCC = 0.8325 ]
+INFO:root:Average loss=0.16097901962124384
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2600 | LCC = 0.8373 | SRCC = 0.8351 ]
+INFO:root:Average loss=0.16394166123026457
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2672 | LCC = 0.8352 | SRCC = 0.8327 ]
+INFO:root:Average loss=0.13821940391491622
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2636 | LCC = 0.8345 | SRCC = 0.8321 ]
+INFO:root:Average loss=0.13049951568245888
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3152 | LCC = 0.8286 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.14164443743916658
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2700 | LCC = 0.8345 | SRCC = 0.8329 ]
+INFO:root:Average loss=0.12575261648266745
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3014 | LCC = 0.8293 | SRCC = 0.8293 ]
+INFO:root:Average loss=0.12550538740096948
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2867 | LCC = 0.8305 | SRCC = 0.8287 ]
+INFO:root:Average loss=0.10825954320339057
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2640 | LCC = 0.8356 | SRCC = 0.8330 ]
+INFO:root:Average loss=0.10360764244045967
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2811 | LCC = 0.8331 | SRCC = 0.8289 ]
+INFO:root:Average loss=0.10341450137396653
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2767 | LCC = 0.8355 | SRCC = 0.8339 ]
+INFO:root:Average loss=0.09441756060681282
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2845 | LCC = 0.8330 | SRCC = 0.8325 ]
+INFO:root:Average loss=0.09943943528028634
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2644 | LCC = 0.8353 | SRCC = 0.8327 ]
+INFO:root:Average loss=0.09118016336399776
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2711 | LCC = 0.8322 | SRCC = 0.8293 ]
+INFO:root:Average loss=0.08661961483840759
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2724 | LCC = 0.8339 | SRCC = 0.8327 ]
+INFO:root:Average loss=0.07925602707725304
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2669 | LCC = 0.8377 | SRCC = 0.8369 ]
+INFO:root:Average loss=0.08350453172356655
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2583 | LCC = 0.8350 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.07401048138928719
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2633 | LCC = 0.8337 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.07155611528418003
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2790 | LCC = 0.8351 | SRCC = 0.8344 ]
+INFO:root:Average loss=0.08263139641628815
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2556 | LCC = 0.8387 | SRCC = 0.8355 ]
+INFO:root:Average loss=0.06453069055882785
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2586 | LCC = 0.8349 | SRCC = 0.8316 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2797 | LCC = 0.8464 | SRCC = 0.8478 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer10/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer10/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..07cd6c482d5e3abda19e45dc1b727c7162277375
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:610f77d5ed9186d5c758d598008eeda9cebb09cd1a6f000eb8eabb3afa4fa6b3
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..88fed89979355b4fbc29c3df4b676abeaf1f3ed9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8980268645848dcbab11669f96365bce5bc16c3f4b63fb70abf7462b52978cfd
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer10/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..83fd490bc53927614f66350ef7c263bebe20d41e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5931917421328716
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3835 | LCC = 0.7719 | SRCC = 0.7759 ]
+INFO:root:Average loss=0.324321601826411
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3319 | LCC = 0.8095 | SRCC = 0.8102 ]
+INFO:root:Average loss=0.3027867262180035
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3018 | LCC = 0.8211 | SRCC = 0.8208 ]
+INFO:root:Average loss=0.25680984766819537
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3093 | LCC = 0.8303 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.24634022475817266
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2758 | LCC = 0.8368 | SRCC = 0.8355 ]
+INFO:root:Average loss=0.2190385759832003
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2818 | LCC = 0.8324 | SRCC = 0.8316 ]
+INFO:root:Average loss=0.20712016847653267
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2860 | LCC = 0.8355 | SRCC = 0.8341 ]
+INFO:root:Average loss=0.20326993518914932
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3260 | LCC = 0.8376 | SRCC = 0.8366 ]
+INFO:root:Average loss=0.1898834373897467
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2628 | LCC = 0.8361 | SRCC = 0.8356 ]
+INFO:root:Average loss=0.17100587162451866
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2629 | LCC = 0.8396 | SRCC = 0.8397 ]
+INFO:root:Average loss=0.1725911205777755
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3837 | LCC = 0.8378 | SRCC = 0.8381 ]
+INFO:root:Average loss=0.17941920583446822
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2639 | LCC = 0.8376 | SRCC = 0.8370 ]
+INFO:root:Average loss=0.15153348780213258
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2799 | LCC = 0.8334 | SRCC = 0.8303 ]
+INFO:root:Average loss=0.1470800291460294
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2647 | LCC = 0.8368 | SRCC = 0.8367 ]
+INFO:root:Average loss=0.13664851452295596
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2573 | LCC = 0.8376 | SRCC = 0.8346 ]
+INFO:root:Average loss=0.13394603324242127
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3539 | LCC = 0.8312 | SRCC = 0.8308 ]
+INFO:root:Average loss=0.1224140668144593
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2611 | LCC = 0.8352 | SRCC = 0.8338 ]
+INFO:root:Average loss=0.13190239964005274
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2978 | LCC = 0.8364 | SRCC = 0.8365 ]
+INFO:root:Average loss=0.11100170102256995
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2834 | LCC = 0.8318 | SRCC = 0.8308 ]
+INFO:root:Average loss=0.10138241784312786
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3022 | LCC = 0.8315 | SRCC = 0.8316 ]
+INFO:root:Average loss=0.10219229647937493
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2687 | LCC = 0.8304 | SRCC = 0.8289 ]
+INFO:root:Average loss=0.0984894356284386
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2827 | LCC = 0.8370 | SRCC = 0.8353 ]
+INFO:root:Average loss=0.0922091157677082
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2685 | LCC = 0.8331 | SRCC = 0.8316 ]
+INFO:root:Average loss=0.09014845123657814
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2623 | LCC = 0.8326 | SRCC = 0.8309 ]
+INFO:root:Average loss=0.08822146846124759
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2678 | LCC = 0.8325 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.0816387758614161
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2709 | LCC = 0.8336 | SRCC = 0.8323 ]
+INFO:root:Average loss=0.07771971239111362
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2620 | LCC = 0.8352 | SRCC = 0.8335 ]
+INFO:root:Average loss=0.07175603871926284
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2688 | LCC = 0.8330 | SRCC = 0.8332 ]
+INFO:root:Average loss=0.07259797080396077
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2584 | LCC = 0.8338 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.06697327619752823
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2595 | LCC = 0.8339 | SRCC = 0.8332 ]
+INFO:root:Average loss=0.06632353642430061
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2694 | LCC = 0.8332 | SRCC = 0.8321 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2434 | LCC = 0.8502 | SRCC = 0.8507 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer10/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer10/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a71055767e25b0fbaf46d277ab7c2c5bf58c17b5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:baa6152c5e974307ddd6f863428ae6f8745210a6193146133c4dc678c9ee46ec
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e4eeb38795e9675e1bd9fa74258b4298f1eed114
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2a52df8579b406e53d919a717308954be95a880f274274b06ba6b21de83e678
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer10/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..a475c92bc3db630d19565a20f7f9aa270127b294
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5594413528839747
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4095 | LCC = 0.7881 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.3119930101510806
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3243 | LCC = 0.8222 | SRCC = 0.8217 ]
+INFO:root:Average loss=0.2815262367710089
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3248 | LCC = 0.8322 | SRCC = 0.8311 ]
+INFO:root:Average loss=0.2513461789259544
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2938 | LCC = 0.8352 | SRCC = 0.8341 ]
+INFO:root:Average loss=0.23342571254724112
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2962 | LCC = 0.8354 | SRCC = 0.8342 ]
+INFO:root:Average loss=0.22329954134348112
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3120 | LCC = 0.8350 | SRCC = 0.8335 ]
+INFO:root:Average loss=0.2078926377953627
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2780 | LCC = 0.8364 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.1873946121105781
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2735 | LCC = 0.8392 | SRCC = 0.8379 ]
+INFO:root:Average loss=0.18075285680018938
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2729 | LCC = 0.8394 | SRCC = 0.8360 ]
+INFO:root:Average loss=0.17774990449349085
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3537 | LCC = 0.8300 | SRCC = 0.8274 ]
+INFO:root:Average loss=0.16907900180190039
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3283 | LCC = 0.8312 | SRCC = 0.8297 ]
+INFO:root:Average loss=0.15824226958629412
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3596 | LCC = 0.8317 | SRCC = 0.8337 ]
+INFO:root:Average loss=0.14881741035825166
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2818 | LCC = 0.8371 | SRCC = 0.8331 ]
+INFO:root:Average loss=0.1460408387848964
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3488 | LCC = 0.8300 | SRCC = 0.8288 ]
+INFO:root:Average loss=0.12922543296829248
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2933 | LCC = 0.8328 | SRCC = 0.8315 ]
+INFO:root:Average loss=0.12441651083720036
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3028 | LCC = 0.8292 | SRCC = 0.8286 ]
+INFO:root:Average loss=0.11542973633951102
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2707 | LCC = 0.8337 | SRCC = 0.8306 ]
+INFO:root:Average loss=0.10900117160800175
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2738 | LCC = 0.8338 | SRCC = 0.8331 ]
+INFO:root:Average loss=0.11856163097306704
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2904 | LCC = 0.8364 | SRCC = 0.8347 ]
+INFO:root:Average loss=0.10856152602877372
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2898 | LCC = 0.8338 | SRCC = 0.8318 ]
+INFO:root:Average loss=0.10888848635248649
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2487 | LCC = 0.8388 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.09760746961602798
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2669 | LCC = 0.8331 | SRCC = 0.8318 ]
+INFO:root:Average loss=0.097034185169599
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2613 | LCC = 0.8360 | SRCC = 0.8344 ]
+INFO:root:Average loss=0.08610999326293285
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2634 | LCC = 0.8347 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.08116951737648402
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3690 | LCC = 0.8236 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.08852930921010482
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2902 | LCC = 0.8304 | SRCC = 0.8319 ]
+INFO:root:Average loss=0.08888748856500173
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2731 | LCC = 0.8298 | SRCC = 0.8289 ]
+INFO:root:Average loss=0.07595544064847323
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3116 | LCC = 0.8204 | SRCC = 0.8193 ]
+INFO:root:Average loss=0.07694695598612993
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4110 | LCC = 0.8154 | SRCC = 0.8186 ]
+INFO:root:Average loss=0.07310052249485101
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3198 | LCC = 0.8249 | SRCC = 0.8252 ]
+INFO:root:Average loss=0.072272886784795
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3159 | LCC = 0.8296 | SRCC = 0.8308 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2599 | LCC = 0.8442 | SRCC = 0.8448 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer10/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer10/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d35868c4b8b868ac2bad007aa5f6970fbd640d02
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9285fe6d864a5b35b394fffd5952d81d7cac0cd459ddfe7e1a561bc706a47db
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer10/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cbb0c6413face4cb5e5843892a51ba36413c02fe
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer10/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7cc22c3fbcd709c789c55086ad8950621b02523565bbb764773aa9be7a09f23c
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer10/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer10/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..a6e92e8729bdfcea58ed3a9ebf7bb491efa29854
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6585969863793789
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3984 | LCC = 0.7635 | SRCC = 0.7735 ]
+INFO:root:Average loss=0.3150277416675519
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3296 | LCC = 0.8183 | SRCC = 0.8170 ]
+INFO:root:Average loss=0.2697792947292328
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3140 | LCC = 0.8338 | SRCC = 0.8319 ]
+INFO:root:Average loss=0.24654768158992132
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2697 | LCC = 0.8336 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.2298332861600778
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3137 | LCC = 0.8377 | SRCC = 0.8347 ]
+INFO:root:Average loss=0.23006353145226455
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3849 | LCC = 0.8291 | SRCC = 0.8291 ]
+INFO:root:Average loss=0.2168125013510386
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2841 | LCC = 0.8394 | SRCC = 0.8383 ]
+INFO:root:Average loss=0.19634150656370017
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2689 | LCC = 0.8404 | SRCC = 0.8377 ]
+INFO:root:Average loss=0.1796574353789672
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2862 | LCC = 0.8375 | SRCC = 0.8329 ]
+INFO:root:Average loss=0.18154766200444636
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2772 | LCC = 0.8411 | SRCC = 0.8374 ]
+INFO:root:Average loss=0.1666212264352884
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2795 | LCC = 0.8411 | SRCC = 0.8366 ]
+INFO:root:Average loss=0.1527683609762253
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2551 | LCC = 0.8382 | SRCC = 0.8358 ]
+INFO:root:Average loss=0.1469293770690759
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2971 | LCC = 0.8346 | SRCC = 0.8301 ]
+INFO:root:Average loss=0.13259238339005372
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2780 | LCC = 0.8342 | SRCC = 0.8306 ]
+INFO:root:Average loss=0.13744249892158386
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3034 | LCC = 0.8324 | SRCC = 0.8292 ]
+INFO:root:Average loss=0.13791707186744764
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2617 | LCC = 0.8397 | SRCC = 0.8373 ]
+INFO:root:Average loss=0.12493221117899968
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2723 | LCC = 0.8368 | SRCC = 0.8334 ]
+INFO:root:Average loss=0.109848713550048
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3167 | LCC = 0.8272 | SRCC = 0.8245 ]
+INFO:root:Average loss=0.10628020481612438
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2722 | LCC = 0.8352 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.10100257377593945
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2658 | LCC = 0.8337 | SRCC = 0.8302 ]
+INFO:root:Average loss=0.09950418841953461
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3449 | LCC = 0.8243 | SRCC = 0.8228 ]
+INFO:root:Average loss=0.09276609060664971
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2705 | LCC = 0.8295 | SRCC = 0.8250 ]
+INFO:root:Average loss=0.08569110332964322
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2778 | LCC = 0.8325 | SRCC = 0.8300 ]
+INFO:root:Average loss=0.08246339914890435
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3025 | LCC = 0.8298 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.08237405431767304
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3315 | LCC = 0.8294 | SRCC = 0.8264 ]
+INFO:root:Average loss=0.07611445643198796
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2749 | LCC = 0.8309 | SRCC = 0.8287 ]
+INFO:root:Average loss=0.07343195541164814
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3355 | LCC = 0.8258 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.0677119435217136
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2841 | LCC = 0.8250 | SRCC = 0.8200 ]
+INFO:root:Average loss=0.06952373506740117
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3426 | LCC = 0.8235 | SRCC = 0.8192 ]
+INFO:root:Average loss=0.07108884099393319
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2871 | LCC = 0.8345 | SRCC = 0.8324 ]
+INFO:root:Average loss=0.0649545442026395
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3552 | LCC = 0.8240 | SRCC = 0.8215 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2581 | LCC = 0.8475 | SRCC = 0.8485 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer12/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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_large_asr_zh/layer12/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c4a25addaf7dca428b4a3c5ed734c40c5fe5a674
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88ddc005f1bb37db9107bf0488acdba5cf5e233ba537ee11fa9c901fa98d312c
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ac4bd8c2ae779df87eecc5f5eaa44c4436dabe05
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:777120cb9e7635d6ce436b26277ef09b356ae305513b78a5a84fc6c97f327c8d
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer12/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..2b41567881b134075da6eabc3481a490a86c9533
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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=0.593781458834807
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3734 | LCC = 0.7599 | SRCC = 0.7655 ]
+INFO:root:Average loss=0.33442257249202484
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3726 | LCC = 0.7995 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.29341678550610173
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3374 | LCC = 0.8113 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.26189139523567295
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3218 | LCC = 0.8287 | SRCC = 0.8272 ]
+INFO:root:Average loss=0.23733771439546195
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3046 | LCC = 0.8262 | SRCC = 0.8257 ]
+INFO:root:Average loss=0.23207828622215834
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2786 | LCC = 0.8283 | SRCC = 0.8264 ]
+INFO:root:Average loss=0.20490836103757223
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2865 | LCC = 0.8308 | SRCC = 0.8281 ]
+INFO:root:Average loss=0.197538524101942
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2821 | LCC = 0.8272 | SRCC = 0.8231 ]
+INFO:root:Average loss=0.1831949094358163
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2943 | LCC = 0.8246 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.19903829951698965
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2795 | LCC = 0.8227 | SRCC = 0.8204 ]
+INFO:root:Average loss=0.1707853557398686
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2901 | LCC = 0.8266 | SRCC = 0.8240 ]
+INFO:root:Average loss=0.1740234320362409
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2892 | LCC = 0.8299 | SRCC = 0.8286 ]
+INFO:root:Average loss=0.15343993472365233
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3217 | LCC = 0.8188 | SRCC = 0.8152 ]
+INFO:root:Average loss=0.13598756482585883
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2991 | LCC = 0.8222 | SRCC = 0.8192 ]
+INFO:root:Average loss=0.1272979783706176
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3035 | LCC = 0.8252 | SRCC = 0.8224 ]
+INFO:root:Average loss=0.11893150602013637
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2836 | LCC = 0.8244 | SRCC = 0.8214 ]
+INFO:root:Average loss=0.12479495505491893
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2727 | LCC = 0.8305 | SRCC = 0.8285 ]
+INFO:root:Average loss=0.11873011314907135
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2734 | LCC = 0.8235 | SRCC = 0.8209 ]
+INFO:root:Average loss=0.10494068647042298
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2848 | LCC = 0.8265 | SRCC = 0.8238 ]
+INFO:root:Average loss=0.10342100004737194
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2775 | LCC = 0.8312 | SRCC = 0.8287 ]
+INFO:root:Average loss=0.09774367625896747
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2686 | LCC = 0.8283 | SRCC = 0.8255 ]
+INFO:root:Average loss=0.10541871963785245
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2808 | LCC = 0.8250 | SRCC = 0.8218 ]
+INFO:root:Average loss=0.09181428165772022
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2737 | LCC = 0.8263 | SRCC = 0.8231 ]
+INFO:root:Average loss=0.07956979767634319
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2762 | LCC = 0.8227 | SRCC = 0.8193 ]
+INFO:root:Average loss=0.07916285622960482
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2919 | LCC = 0.8221 | SRCC = 0.8188 ]
+INFO:root:Average loss=0.07802203655816041
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2957 | LCC = 0.8234 | SRCC = 0.8226 ]
+INFO:root:Average loss=0.08283957394842918
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2741 | LCC = 0.8242 | SRCC = 0.8219 ]
+INFO:root:Average loss=0.08503605554310176
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2761 | LCC = 0.8236 | SRCC = 0.8223 ]
+INFO:root:Average loss=0.07319161897668472
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2843 | LCC = 0.8207 | SRCC = 0.8183 ]
+INFO:root:Average loss=0.0741046647517345
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2766 | LCC = 0.8250 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.061536975539265536
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2872 | LCC = 0.8213 | SRCC = 0.8206 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2748 | LCC = 0.8281 | SRCC = 0.8259 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer12/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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_large_asr_zh/layer12/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bae0749f4bb0d83194e9a2e72d2669d173b4057e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dc433a85271fa80deda984f31d1d86f67cb12e340c080ac5330404b243c0717
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..473489ba3ccd256be29003d869d027edbcd2c1b9
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59044345bfd6f68a9276dfa62a2a0d6a7e08cfee8d19173372db40fab4fd68d3
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer12/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..20fbccccf132a01bdb5056e06e3351481ecabfa7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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=0.6078306753665973
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4746 | LCC = 0.7645 | SRCC = 0.7779 ]
+INFO:root:Average loss=0.33928072911042434
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3498 | LCC = 0.8056 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.3091904885875873
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3738 | LCC = 0.8093 | SRCC = 0.8082 ]
+INFO:root:Average loss=0.27461593063213885
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3020 | LCC = 0.8245 | SRCC = 0.8223 ]
+INFO:root:Average loss=0.2515262260268896
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3222 | LCC = 0.8287 | SRCC = 0.8287 ]
+INFO:root:Average loss=0.24087492472086197
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3079 | LCC = 0.8258 | SRCC = 0.8220 ]
+INFO:root:Average loss=0.21345732007653284
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2811 | LCC = 0.8230 | SRCC = 0.8235 ]
+INFO:root:Average loss=0.20380681638534254
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3010 | LCC = 0.8244 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.2094504765402048
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3137 | LCC = 0.8231 | SRCC = 0.8219 ]
+INFO:root:Average loss=0.1751047320281848
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2881 | LCC = 0.8154 | SRCC = 0.8108 ]
+INFO:root:Average loss=0.17528949066614494
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2793 | LCC = 0.8257 | SRCC = 0.8224 ]
+INFO:root:Average loss=0.1711702867387197
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2974 | LCC = 0.8254 | SRCC = 0.8245 ]
+INFO:root:Average loss=0.14763514009805825
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2689 | LCC = 0.8257 | SRCC = 0.8224 ]
+INFO:root:Average loss=0.16189315943763807
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3089 | LCC = 0.8222 | SRCC = 0.8202 ]
+INFO:root:Average loss=0.14330850598903802
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3092 | LCC = 0.8186 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.14650522888853
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3050 | LCC = 0.8250 | SRCC = 0.8235 ]
+INFO:root:Average loss=0.12632365863865766
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2874 | LCC = 0.8209 | SRCC = 0.8196 ]
+INFO:root:Average loss=0.1171528295828746
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2753 | LCC = 0.8232 | SRCC = 0.8213 ]
+INFO:root:Average loss=0.1045174536605676
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2909 | LCC = 0.8190 | SRCC = 0.8147 ]
+INFO:root:Average loss=0.10569913331896831
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2794 | LCC = 0.8247 | SRCC = 0.8213 ]
+INFO:root:Average loss=0.10092585610273556
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2756 | LCC = 0.8235 | SRCC = 0.8204 ]
+INFO:root:Average loss=0.09513005814873256
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2790 | LCC = 0.8188 | SRCC = 0.8162 ]
+INFO:root:Average loss=0.08632824646356778
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2868 | LCC = 0.8160 | SRCC = 0.8135 ]
+INFO:root:Average loss=0.08375428903561372
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3085 | LCC = 0.8173 | SRCC = 0.8132 ]
+INFO:root:Average loss=0.08038457492605233
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2701 | LCC = 0.8264 | SRCC = 0.8235 ]
+INFO:root:Average loss=0.0819729179239426
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2814 | LCC = 0.8199 | SRCC = 0.8166 ]
+INFO:root:Average loss=0.07086690725424351
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2817 | LCC = 0.8171 | SRCC = 0.8129 ]
+INFO:root:Average loss=0.0696497336985209
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2867 | LCC = 0.8188 | SRCC = 0.8152 ]
+INFO:root:Average loss=0.07428869280295494
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2871 | LCC = 0.8144 | SRCC = 0.8114 ]
+INFO:root:Average loss=0.07574995315800874
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2802 | LCC = 0.8161 | SRCC = 0.8130 ]
+INFO:root:Average loss=0.06302012818363997
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2883 | LCC = 0.8184 | SRCC = 0.8151 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3819 | LCC = 0.8202 | SRCC = 0.8288 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer12/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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_large_asr_zh/layer12/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8b01bba468a145c9624d8873a1079a2a9625eb1b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ede73eba6440e318969fa92ce1841c431b5f274de9d3fb16dadc8abbf4a139d
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..527a8a057921adbf10268f360ff099ed23f5e63c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54f8c8db92387e3f9f038dc985ceaaac86bdf4482a8c7e303547331815ad16b4
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer12/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..926569f74d55b8d9e7c1e8c1a9ab6be2bec20858
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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=0.5933209508657455
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4041 | LCC = 0.7781 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.3457098702589671
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3626 | LCC = 0.7981 | SRCC = 0.8008 ]
+INFO:root:Average loss=0.2871458213298749
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3231 | LCC = 0.8152 | SRCC = 0.8151 ]
+INFO:root:Average loss=0.25893592605224025
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3402 | LCC = 0.8211 | SRCC = 0.8194 ]
+INFO:root:Average loss=0.25742118328045577
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3071 | LCC = 0.8258 | SRCC = 0.8228 ]
+INFO:root:Average loss=0.23303488202584097
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3334 | LCC = 0.8270 | SRCC = 0.8260 ]
+INFO:root:Average loss=0.23401279403613165
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3013 | LCC = 0.8283 | SRCC = 0.8278 ]
+INFO:root:Average loss=0.21119910917985132
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2875 | LCC = 0.8276 | SRCC = 0.8255 ]
+INFO:root:Average loss=0.19360822381881568
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3423 | LCC = 0.8243 | SRCC = 0.8224 ]
+INFO:root:Average loss=0.2039361646733223
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2793 | LCC = 0.8231 | SRCC = 0.8191 ]
+INFO:root:Average loss=0.18641148287898454
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2896 | LCC = 0.8257 | SRCC = 0.8220 ]
+INFO:root:Average loss=0.17090515486704996
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3204 | LCC = 0.8266 | SRCC = 0.8232 ]
+INFO:root:Average loss=0.15713604893057775
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2959 | LCC = 0.8226 | SRCC = 0.8200 ]
+INFO:root:Average loss=0.15612486673471254
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3176 | LCC = 0.8236 | SRCC = 0.8221 ]
+INFO:root:Average loss=0.15464575292590338
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2995 | LCC = 0.8211 | SRCC = 0.8179 ]
+INFO:root:Average loss=0.13224860003743416
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2861 | LCC = 0.8183 | SRCC = 0.8159 ]
+INFO:root:Average loss=0.12622674831595176
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3056 | LCC = 0.8119 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.11809795961166039
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3033 | LCC = 0.8162 | SRCC = 0.8129 ]
+INFO:root:Average loss=0.10941184651202117
+INFO:root:
+[Valid][18][UTT][ MSE = 0.4129 | LCC = 0.8115 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.12318884309094685
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3245 | LCC = 0.8142 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.11161218430751409
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2919 | LCC = 0.8165 | SRCC = 0.8120 ]
+INFO:root:Average loss=0.09631452465859744
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2882 | LCC = 0.8191 | SRCC = 0.8139 ]
+INFO:root:Average loss=0.10002572433306621
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3024 | LCC = 0.8131 | SRCC = 0.8081 ]
+INFO:root:Average loss=0.11625740462197708
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2698 | LCC = 0.8235 | SRCC = 0.8210 ]
+INFO:root:Average loss=0.086972551420331
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2879 | LCC = 0.8107 | SRCC = 0.8070 ]
+INFO:root:Average loss=0.07953985283772151
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2833 | LCC = 0.8184 | SRCC = 0.8147 ]
+INFO:root:Average loss=0.08107363069668794
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2818 | LCC = 0.8216 | SRCC = 0.8175 ]
+INFO:root:Average loss=0.07419563992283283
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2939 | LCC = 0.8168 | SRCC = 0.8125 ]
+INFO:root:Average loss=0.07319487086855449
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2951 | LCC = 0.8173 | SRCC = 0.8126 ]
+INFO:root:Average loss=0.0778065271771107
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3003 | LCC = 0.8192 | SRCC = 0.8158 ]
+INFO:root:Average loss=0.06956741275886695
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3058 | LCC = 0.8166 | SRCC = 0.8126 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3297 | LCC = 0.8301 | SRCC = 0.8379 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer12/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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_large_asr_zh/layer12/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3fe92993b544033a182b7506fc24c0724d079888
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f1223a647f50a99634c1152cddf327271a33d795477b4bda4242c50071ca963
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8e3c1a56e9e34991d87712fe3502c684916f8161
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70a515eeed42dd0f7b463967416c433116746392183880ef0643e2fd8969ec97
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer12/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c0cbc18ab60c44274301a20a5f9a791c1fdcb901
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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=0.569871092453981
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4061 | LCC = 0.7713 | SRCC = 0.7786 ]
+INFO:root:Average loss=0.35465007829360473
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3443 | LCC = 0.8022 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.2890994076927503
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3731 | LCC = 0.8127 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.26544370253880817
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3202 | LCC = 0.8232 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.239548965715445
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2865 | LCC = 0.8247 | SRCC = 0.8231 ]
+INFO:root:Average loss=0.21607589511535105
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2837 | LCC = 0.8276 | SRCC = 0.8253 ]
+INFO:root:Average loss=0.21137382281132233
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3201 | LCC = 0.8268 | SRCC = 0.8258 ]
+INFO:root:Average loss=0.19564391309634233
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2850 | LCC = 0.8303 | SRCC = 0.8284 ]
+INFO:root:Average loss=0.18471001967405662
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2805 | LCC = 0.8304 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.17245952861431318
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3318 | LCC = 0.8271 | SRCC = 0.8243 ]
+INFO:root:Average loss=0.16624163998625216
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2867 | LCC = 0.8251 | SRCC = 0.8192 ]
+INFO:root:Average loss=0.14627993641755518
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2701 | LCC = 0.8310 | SRCC = 0.8279 ]
+INFO:root:Average loss=0.13294237298079026
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2607 | LCC = 0.8321 | SRCC = 0.8285 ]
+INFO:root:Average loss=0.13406542258766982
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2878 | LCC = 0.8291 | SRCC = 0.8233 ]
+INFO:root:Average loss=0.12726635352159157
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2755 | LCC = 0.8256 | SRCC = 0.8212 ]
+INFO:root:Average loss=0.10910780527270757
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2659 | LCC = 0.8291 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.10865533810395461
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2837 | LCC = 0.8257 | SRCC = 0.8211 ]
+INFO:root:Average loss=0.1089890956496581
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2751 | LCC = 0.8194 | SRCC = 0.8147 ]
+INFO:root:Average loss=0.09874153753312734
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3017 | LCC = 0.8256 | SRCC = 0.8218 ]
+INFO:root:Average loss=0.10046967075994381
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2784 | LCC = 0.8270 | SRCC = 0.8208 ]
+INFO:root:Average loss=0.08432278538552615
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2721 | LCC = 0.8267 | SRCC = 0.8209 ]
+INFO:root:Average loss=0.09160048653108951
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2782 | LCC = 0.8257 | SRCC = 0.8219 ]
+INFO:root:Average loss=0.07835442572832108
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2639 | LCC = 0.8259 | SRCC = 0.8209 ]
+INFO:root:Average loss=0.09080215362020028
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2949 | LCC = 0.8281 | SRCC = 0.8246 ]
+INFO:root:Average loss=0.07390717011040603
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2847 | LCC = 0.8231 | SRCC = 0.8188 ]
+INFO:root:Average loss=0.06830827409449296
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2739 | LCC = 0.8196 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.06643359090846318
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2689 | LCC = 0.8254 | SRCC = 0.8208 ]
+INFO:root:Average loss=0.06650755329965016
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2936 | LCC = 0.8203 | SRCC = 0.8142 ]
+INFO:root:Average loss=0.06262619549838397
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2792 | LCC = 0.8236 | SRCC = 0.8168 ]
+INFO:root:Average loss=0.07104214978141662
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2751 | LCC = 0.8227 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.05966435186564922
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2739 | LCC = 0.8210 | SRCC = 0.8148 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2594 | LCC = 0.8350 | SRCC = 0.8337 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer12/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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_large_asr_zh/layer12/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f7bea1ae31cffa7b85bfd4df411ead0d0930b607
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1ac0fa853bd6144e999b1708d2bc11046a75cc561d08088a00fecea27105b43
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer12/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..68f7210c56d828cbdb6d9165e05f193ff789f514
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e73bdf0514da415419037ac7e483c47454581c14630a8dc2015d8c59e621d9ba
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer12/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer12/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..89220f423a52a51c5b24d7a01547fca38a954209
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer12/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=0.5925460916299087
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3698 | LCC = 0.7874 | SRCC = 0.7904 ]
+INFO:root:Average loss=0.3092254509146397
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3210 | LCC = 0.8154 | SRCC = 0.8140 ]
+INFO:root:Average loss=0.27443552093628126
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3625 | LCC = 0.8222 | SRCC = 0.8209 ]
+INFO:root:Average loss=0.28082980368381893
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2976 | LCC = 0.8308 | SRCC = 0.8282 ]
+INFO:root:Average loss=0.2536847438567724
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3766 | LCC = 0.8231 | SRCC = 0.8253 ]
+INFO:root:Average loss=0.23221366661481369
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2943 | LCC = 0.8319 | SRCC = 0.8286 ]
+INFO:root:Average loss=0.20639960085734343
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3226 | LCC = 0.8289 | SRCC = 0.8259 ]
+INFO:root:Average loss=0.1970644783324156
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2924 | LCC = 0.8274 | SRCC = 0.8249 ]
+INFO:root:Average loss=0.1790283395884893
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3406 | LCC = 0.8292 | SRCC = 0.8280 ]
+INFO:root:Average loss=0.16871051843731832
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2817 | LCC = 0.8311 | SRCC = 0.8278 ]
+INFO:root:Average loss=0.17224081720297152
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3272 | LCC = 0.8277 | SRCC = 0.8257 ]
+INFO:root:Average loss=0.15605442646222237
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2818 | LCC = 0.8296 | SRCC = 0.8272 ]
+INFO:root:Average loss=0.16145158167450857
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3074 | LCC = 0.8283 | SRCC = 0.8242 ]
+INFO:root:Average loss=0.1376730115749897
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3893 | LCC = 0.8192 | SRCC = 0.8190 ]
+INFO:root:Average loss=0.1311792851640628
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3086 | LCC = 0.8267 | SRCC = 0.8248 ]
+INFO:root:Average loss=0.1375230205937838
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2904 | LCC = 0.8330 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.1482892996416642
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3525 | LCC = 0.8269 | SRCC = 0.8264 ]
+INFO:root:Average loss=0.144559317196791
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3714 | LCC = 0.8195 | SRCC = 0.8220 ]
+INFO:root:Average loss=0.1254599156956642
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2928 | LCC = 0.8262 | SRCC = 0.8237 ]
+INFO:root:Average loss=0.10176243623479819
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3256 | LCC = 0.8228 | SRCC = 0.8234 ]
+INFO:root:Average loss=0.09771311407287915
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3295 | LCC = 0.8262 | SRCC = 0.8259 ]
+INFO:root:Average loss=0.09345725919000614
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3432 | LCC = 0.8243 | SRCC = 0.8265 ]
+INFO:root:Average loss=0.08711900390111484
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2911 | LCC = 0.8296 | SRCC = 0.8293 ]
+INFO:root:Average loss=0.08660536617613755
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2722 | LCC = 0.8306 | SRCC = 0.8301 ]
+INFO:root:Average loss=0.0829225216442958
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2922 | LCC = 0.8315 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.08165924709576827
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3376 | LCC = 0.8269 | SRCC = 0.8276 ]
+INFO:root:Average loss=0.08340184629345551
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2969 | LCC = 0.8296 | SRCC = 0.8307 ]
+INFO:root:Average loss=0.06955522833726345
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2907 | LCC = 0.8240 | SRCC = 0.8227 ]
+INFO:root:Average loss=0.06624106900432171
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2716 | LCC = 0.8290 | SRCC = 0.8281 ]
+INFO:root:Average loss=0.0648708884150554
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2889 | LCC = 0.8250 | SRCC = 0.8256 ]
+INFO:root:Average loss=0.07095540553713456
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2651 | LCC = 0.8332 | SRCC = 0.8327 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2671 | LCC = 0.8331 | SRCC = 0.8352 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer14/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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_large_asr_zh/layer14/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fe04dd8dbe5e2357d77867753c6df97edddcb84b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7dfcc32661fdc121ce64d62055d726b04abf4d2f28a3d5240b0e3a1d8da2b522
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..dfbbc405f64df7d461fa8796e564038c04184282
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:28f1a77904f6f7af84d66ca391784e845b948ba0910ad5e857151bd1e996196e
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer14/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..cfc8d4d08bc02042a56aa1880eedd290b330ea28
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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=0.5856305846037009
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5245 | LCC = 0.7370 | SRCC = 0.7661 ]
+INFO:root:Average loss=0.34069551393771785
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3410 | LCC = 0.8030 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.30694543551176023
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3474 | LCC = 0.8139 | SRCC = 0.8152 ]
+INFO:root:Average loss=0.2911164991748639
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3203 | LCC = 0.8160 | SRCC = 0.8159 ]
+INFO:root:Average loss=0.27356668542592955
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3218 | LCC = 0.8211 | SRCC = 0.8207 ]
+INFO:root:Average loss=0.2426513536618306
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3500 | LCC = 0.8203 | SRCC = 0.8192 ]
+INFO:root:Average loss=0.23375889448783335
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3125 | LCC = 0.8238 | SRCC = 0.8227 ]
+INFO:root:Average loss=0.2165706809132527
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3061 | LCC = 0.8216 | SRCC = 0.8187 ]
+INFO:root:Average loss=0.21629812721258554
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3387 | LCC = 0.8186 | SRCC = 0.8151 ]
+INFO:root:Average loss=0.19143610753310034
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3205 | LCC = 0.8213 | SRCC = 0.8192 ]
+INFO:root:Average loss=0.1876136630009382
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2895 | LCC = 0.8243 | SRCC = 0.8226 ]
+INFO:root:Average loss=0.17994730671246847
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3382 | LCC = 0.8173 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.16514565031497908
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3155 | LCC = 0.8184 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.15459752474457789
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2953 | LCC = 0.8159 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.15915734588335723
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3114 | LCC = 0.8153 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.15081660306224456
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3299 | LCC = 0.8151 | SRCC = 0.8125 ]
+INFO:root:Average loss=0.15990489453841478
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3618 | LCC = 0.8140 | SRCC = 0.8139 ]
+INFO:root:Average loss=0.13204602763438836
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3264 | LCC = 0.8133 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.12821707587975723
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2936 | LCC = 0.8160 | SRCC = 0.8126 ]
+INFO:root:Average loss=0.11174596187013847
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2947 | LCC = 0.8134 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.11426252718919362
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2952 | LCC = 0.8156 | SRCC = 0.8124 ]
+INFO:root:Average loss=0.10546606081800583
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2891 | LCC = 0.8169 | SRCC = 0.8146 ]
+INFO:root:Average loss=0.11783815987217121
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3264 | LCC = 0.8147 | SRCC = 0.8126 ]
+INFO:root:Average loss=0.09939809091007099
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3243 | LCC = 0.8113 | SRCC = 0.8070 ]
+INFO:root:Average loss=0.12598778842351374
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3035 | LCC = 0.8138 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.09041044861078262
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2895 | LCC = 0.8192 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.08559700517127147
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3081 | LCC = 0.8098 | SRCC = 0.8051 ]
+INFO:root:Average loss=0.09180701824908073
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3572 | LCC = 0.8101 | SRCC = 0.8094 ]
+INFO:root:Average loss=0.08460132739482781
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3535 | LCC = 0.8110 | SRCC = 0.8081 ]
+INFO:root:Average loss=0.07812378273751491
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3021 | LCC = 0.8132 | SRCC = 0.8096 ]
+INFO:root:Average loss=0.07754336101695514
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3221 | LCC = 0.8116 | SRCC = 0.8095 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2701 | LCC = 0.8402 | SRCC = 0.8440 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer14/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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_large_asr_zh/layer14/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b52eec626b41d521846a38b09d4e7d650af3cd8e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c33afa3966536a81e7b653345ba0e947920e9022ebaf97857130199418aea94f
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c9f08946e721d17f21afaa4d7e30a6b8958d57d6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27446991be3739f05ab6fa95b2bebc385ab1f7f4679b1d4fc433e504b768462a
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer14/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b85fbe38b9c71a5a2a8d502d99869212cd2928c8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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=0.6114558768578064
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4024 | LCC = 0.7603 | SRCC = 0.7624 ]
+INFO:root:Average loss=0.3545209805552776
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3277 | LCC = 0.8010 | SRCC = 0.8028 ]
+INFO:root:Average loss=0.3330599948381766
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3121 | LCC = 0.8052 | SRCC = 0.8072 ]
+INFO:root:Average loss=0.27552681053296113
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3131 | LCC = 0.8144 | SRCC = 0.8157 ]
+INFO:root:Average loss=0.25115230278326917
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3048 | LCC = 0.8169 | SRCC = 0.8178 ]
+INFO:root:Average loss=0.23082366299170715
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3330 | LCC = 0.8153 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.21881868289067194
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3042 | LCC = 0.8215 | SRCC = 0.8204 ]
+INFO:root:Average loss=0.21732523311407137
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2961 | LCC = 0.8193 | SRCC = 0.8182 ]
+INFO:root:Average loss=0.20274515201648077
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2817 | LCC = 0.8221 | SRCC = 0.8218 ]
+INFO:root:Average loss=0.18321098673802155
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3008 | LCC = 0.8178 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.17793953170379004
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2966 | LCC = 0.8166 | SRCC = 0.8134 ]
+INFO:root:Average loss=0.18164317386272627
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3176 | LCC = 0.8086 | SRCC = 0.8082 ]
+INFO:root:Average loss=0.17363888751237821
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2977 | LCC = 0.8193 | SRCC = 0.8179 ]
+INFO:root:Average loss=0.1501252915805731
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2918 | LCC = 0.8134 | SRCC = 0.8122 ]
+INFO:root:Average loss=0.1487183011113069
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2940 | LCC = 0.8147 | SRCC = 0.8119 ]
+INFO:root:Average loss=0.1386232492633355
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3004 | LCC = 0.8145 | SRCC = 0.8099 ]
+INFO:root:Average loss=0.1336887155014735
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2990 | LCC = 0.8104 | SRCC = 0.8079 ]
+INFO:root:Average loss=0.13148223054714692
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3009 | LCC = 0.8088 | SRCC = 0.8052 ]
+INFO:root:Average loss=0.11101685235133538
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2936 | LCC = 0.8101 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.1140589150480735
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2992 | LCC = 0.8084 | SRCC = 0.8036 ]
+INFO:root:Average loss=0.10706855939366879
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2919 | LCC = 0.8105 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.10339963760895607
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3005 | LCC = 0.8092 | SRCC = 0.8068 ]
+INFO:root:Average loss=0.10448501688929704
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2816 | LCC = 0.8182 | SRCC = 0.8163 ]
+INFO:root:Average loss=0.0908915865688752
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2981 | LCC = 0.8105 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.09168047190476687
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2996 | LCC = 0.8129 | SRCC = 0.8101 ]
+INFO:root:Average loss=0.09085412280490765
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3045 | LCC = 0.8101 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.08235531133145858
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3008 | LCC = 0.8095 | SRCC = 0.8072 ]
+INFO:root:Average loss=0.07995396102659214
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2854 | LCC = 0.8128 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.07673277061146039
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3233 | LCC = 0.8098 | SRCC = 0.8096 ]
+INFO:root:Average loss=0.08006713868906865
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2983 | LCC = 0.8074 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.08244798881694293
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3026 | LCC = 0.8033 | SRCC = 0.8001 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2743 | LCC = 0.8333 | SRCC = 0.8356 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer14/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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_large_asr_zh/layer14/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..20608c6d96ae0c20c1121e6ea053d3c687b19dcd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea313683ed6f1cf1e475019a9e0c0789d2970720d0107106bf451def1c760a7b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..74901f77b3346bd7e2c3f4baabecd5d144cbf4aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c646a9f351137c2b0f4e3ded8d7f996af4bb3eaee80db33794e8466b61b27f51
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer14/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..998093703f8a974085b1a32611b98c4d4635bda1
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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=0.5685124561572686
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4885 | LCC = 0.7398 | SRCC = 0.7580 ]
+INFO:root:Average loss=0.34813621162604064
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3367 | LCC = 0.7936 | SRCC = 0.7950 ]
+INFO:root:Average loss=0.30582126860435194
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3410 | LCC = 0.8077 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.276287830219819
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2916 | LCC = 0.8184 | SRCC = 0.8172 ]
+INFO:root:Average loss=0.25781629177240223
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3968 | LCC = 0.8120 | SRCC = 0.8180 ]
+INFO:root:Average loss=0.2922199001679054
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3850 | LCC = 0.8098 | SRCC = 0.8154 ]
+INFO:root:Average loss=0.24447709054518968
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3353 | LCC = 0.8169 | SRCC = 0.8147 ]
+INFO:root:Average loss=0.24470537786300367
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3964 | LCC = 0.8152 | SRCC = 0.8147 ]
+INFO:root:Average loss=0.21433675795411453
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3068 | LCC = 0.8220 | SRCC = 0.8207 ]
+INFO:root:Average loss=0.207186503861195
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3023 | LCC = 0.8209 | SRCC = 0.8169 ]
+INFO:root:Average loss=0.18809118064550254
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2916 | LCC = 0.8212 | SRCC = 0.8175 ]
+INFO:root:Average loss=0.18521025977455652
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3501 | LCC = 0.8082 | SRCC = 0.8069 ]
+INFO:root:Average loss=0.16592738404870033
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2940 | LCC = 0.8171 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.16094968993312272
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2847 | LCC = 0.8162 | SRCC = 0.8135 ]
+INFO:root:Average loss=0.15951652070268607
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4414 | LCC = 0.7994 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.18003677786924902
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2934 | LCC = 0.8154 | SRCC = 0.8129 ]
+INFO:root:Average loss=0.13446118319645906
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2945 | LCC = 0.8166 | SRCC = 0.8156 ]
+INFO:root:Average loss=0.12828783939282098
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2880 | LCC = 0.8142 | SRCC = 0.8105 ]
+INFO:root:Average loss=0.12479480250905721
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2968 | LCC = 0.8140 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.11975541577125207
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2851 | LCC = 0.8154 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.10956916270347741
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3041 | LCC = 0.8151 | SRCC = 0.8104 ]
+INFO:root:Average loss=0.10623372804660064
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2919 | LCC = 0.8120 | SRCC = 0.8073 ]
+INFO:root:Average loss=0.1007903641424118
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3045 | LCC = 0.8132 | SRCC = 0.8088 ]
+INFO:root:Average loss=0.10790145148833592
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3151 | LCC = 0.8137 | SRCC = 0.8096 ]
+INFO:root:Average loss=0.0900859522800415
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3366 | LCC = 0.8003 | SRCC = 0.7941 ]
+INFO:root:Average loss=0.09417940222491057
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2932 | LCC = 0.8102 | SRCC = 0.8051 ]
+INFO:root:Average loss=0.09165235914480992
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2884 | LCC = 0.8125 | SRCC = 0.8058 ]
+INFO:root:Average loss=0.08707401204185608
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2978 | LCC = 0.8121 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.08153711397869465
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2880 | LCC = 0.8156 | SRCC = 0.8111 ]
+INFO:root:Average loss=0.08456750519764729
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3080 | LCC = 0.8114 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.07837937103632169
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2900 | LCC = 0.8194 | SRCC = 0.8154 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2683 | LCC = 0.8407 | SRCC = 0.8408 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer14/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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_large_asr_zh/layer14/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3ff048ece463124c632164f27ba3bae456635ecb
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52b754ec1d94c1341e8889d2efc1b692714418163682b235b15e6a5b270d10ed
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..204af69360bd5759347bc18982d3a60870acfffc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb7bfa5a9aee5e9a6f8d2e6cc59d73975de76b774fb6e189ddaf68279c6c97f8
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer14/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..033011423a2cd09630200a2e29b43535351b1cce
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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=0.663461121229025
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4024 | LCC = 0.7468 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.3519439674340762
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3729 | LCC = 0.7963 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.30428705918483245
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3242 | LCC = 0.8057 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.2766057203213374
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3275 | LCC = 0.8140 | SRCC = 0.8115 ]
+INFO:root:Average loss=0.25431986248646027
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3209 | LCC = 0.8162 | SRCC = 0.8136 ]
+INFO:root:Average loss=0.259485838504938
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3272 | LCC = 0.8207 | SRCC = 0.8194 ]
+INFO:root:Average loss=0.22875319039210296
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2855 | LCC = 0.8212 | SRCC = 0.8185 ]
+INFO:root:Average loss=0.22528943973474014
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2901 | LCC = 0.8210 | SRCC = 0.8184 ]
+INFO:root:Average loss=0.21128987320340598
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3220 | LCC = 0.8204 | SRCC = 0.8197 ]
+INFO:root:Average loss=0.2032948491665033
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3023 | LCC = 0.8196 | SRCC = 0.8167 ]
+INFO:root:Average loss=0.18310466590218055
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2817 | LCC = 0.8221 | SRCC = 0.8202 ]
+INFO:root:Average loss=0.1712208384504685
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2820 | LCC = 0.8254 | SRCC = 0.8225 ]
+INFO:root:Average loss=0.17889408967815912
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3182 | LCC = 0.8153 | SRCC = 0.8140 ]
+INFO:root:Average loss=0.14919733628630638
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2801 | LCC = 0.8188 | SRCC = 0.8167 ]
+INFO:root:Average loss=0.1520637829716389
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2894 | LCC = 0.8139 | SRCC = 0.8114 ]
+INFO:root:Average loss=0.14403569344908762
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2938 | LCC = 0.8132 | SRCC = 0.8108 ]
+INFO:root:Average loss=0.13165531527155486
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3617 | LCC = 0.8064 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.12878435878799513
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3028 | LCC = 0.8100 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.11276467946859506
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3662 | LCC = 0.8037 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.11413414050371219
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3108 | LCC = 0.8029 | SRCC = 0.8007 ]
+INFO:root:Average loss=0.11284163087988511
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3573 | LCC = 0.8009 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.10498893041259204
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3583 | LCC = 0.8036 | SRCC = 0.8040 ]
+INFO:root:Average loss=0.10554609027428505
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3270 | LCC = 0.8103 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.10141345562461095
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2952 | LCC = 0.8102 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.09234894062273014
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3028 | LCC = 0.8053 | SRCC = 0.8002 ]
+INFO:root:Average loss=0.08851775331183886
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3504 | LCC = 0.8065 | SRCC = 0.8079 ]
+INFO:root:Average loss=0.08572460166536845
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3349 | LCC = 0.8065 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.08392574628576255
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2908 | LCC = 0.8086 | SRCC = 0.8048 ]
+INFO:root:Average loss=0.07819584618585232
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3002 | LCC = 0.8053 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.08238757387376748
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3418 | LCC = 0.8058 | SRCC = 0.8047 ]
+INFO:root:Average loss=0.09265355675075299
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3064 | LCC = 0.8056 | SRCC = 0.8021 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2722 | LCC = 0.8337 | SRCC = 0.8367 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer14/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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_large_asr_zh/layer14/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..84038f507efcc2dfd4173f98b7b59cb1f5775e9e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:119c2bd33b499aa3f99596da843ef91d0800aef6b20922a70b06458a5f0940ee
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer14/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..abbde44e829a5f3b75340fcae23dbbb6841b7ade
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21c42037c5e5590fd115167ceb08805cd73ffc8c0cf0935e5a1da625594d97c1
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer14/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer14/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..988153b7baf45f8c816cfdec5f3cde7154e0e242
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer14/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=0.6389655929345351
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4360 | LCC = 0.7422 | SRCC = 0.7463 ]
+INFO:root:Average loss=0.36016436780874544
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3837 | LCC = 0.7919 | SRCC = 0.7933 ]
+INFO:root:Average loss=0.3123921520816974
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4478 | LCC = 0.7979 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.29096480439870787
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3386 | LCC = 0.8044 | SRCC = 0.8106 ]
+INFO:root:Average loss=0.28518108622385907
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3350 | LCC = 0.8137 | SRCC = 0.8125 ]
+INFO:root:Average loss=0.24961339281155512
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3170 | LCC = 0.8188 | SRCC = 0.8179 ]
+INFO:root:Average loss=0.22904855394974732
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3055 | LCC = 0.8114 | SRCC = 0.8098 ]
+INFO:root:Average loss=0.20883450905481973
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3239 | LCC = 0.8083 | SRCC = 0.8071 ]
+INFO:root:Average loss=0.20722163459047294
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3073 | LCC = 0.8115 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.19492422063381243
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2974 | LCC = 0.8154 | SRCC = 0.8126 ]
+INFO:root:Average loss=0.18603072009789637
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2870 | LCC = 0.8197 | SRCC = 0.8174 ]
+INFO:root:Average loss=0.1751426670413751
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3084 | LCC = 0.8150 | SRCC = 0.8114 ]
+INFO:root:Average loss=0.18731801001689372
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2926 | LCC = 0.8124 | SRCC = 0.8097 ]
+INFO:root:Average loss=0.1574198784163365
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3081 | LCC = 0.8084 | SRCC = 0.8058 ]
+INFO:root:Average loss=0.15328462039813018
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2967 | LCC = 0.8191 | SRCC = 0.8156 ]
+INFO:root:Average loss=0.145640806414378
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3268 | LCC = 0.8099 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.15109135439762703
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3262 | LCC = 0.8115 | SRCC = 0.8106 ]
+INFO:root:Average loss=0.13297760104521728
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3025 | LCC = 0.8146 | SRCC = 0.8119 ]
+INFO:root:Average loss=0.12913777956213707
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2849 | LCC = 0.8185 | SRCC = 0.8155 ]
+INFO:root:Average loss=0.11382193338030423
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2931 | LCC = 0.8145 | SRCC = 0.8113 ]
+INFO:root:Average loss=0.10997446559560604
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2901 | LCC = 0.8184 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.10262294553029232
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2889 | LCC = 0.8141 | SRCC = 0.8110 ]
+INFO:root:Average loss=0.10931772108261402
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2788 | LCC = 0.8198 | SRCC = 0.8167 ]
+INFO:root:Average loss=0.09503827339563614
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3069 | LCC = 0.8172 | SRCC = 0.8148 ]
+INFO:root:Average loss=0.0925950462428423
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3071 | LCC = 0.8141 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.09703680509940171
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3828 | LCC = 0.8154 | SRCC = 0.8143 ]
+INFO:root:Average loss=0.0920566720649218
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2847 | LCC = 0.8200 | SRCC = 0.8164 ]
+INFO:root:Average loss=0.09633291312135182
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2966 | LCC = 0.8183 | SRCC = 0.8158 ]
+INFO:root:Average loss=0.0819381609176978
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2904 | LCC = 0.8148 | SRCC = 0.8107 ]
+INFO:root:Average loss=0.08359725596622014
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2853 | LCC = 0.8152 | SRCC = 0.8116 ]
+INFO:root:Average loss=0.07845558364612934
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3083 | LCC = 0.8140 | SRCC = 0.8098 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2882 | LCC = 0.8374 | SRCC = 0.8377 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer16/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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_large_asr_zh/layer16/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ebaa588faa708bf206ad6e7218f646e40abaa846
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:657b768d891df1d50ee13999a25c6027051e3d3ff656d2b39b5ae24b2f8bd0a9
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..01b5b49f4ef7b6d907b763b6d2c5bae78fd3b7a7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25ee82356194be0f02aab53f9b72a1640e9940ccdab0df19ebf60a6743d8f50d
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer16/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..cf6528b0bee59063a4437b12e0db9fdcd5f2681e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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=0.617804589561927
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4198 | LCC = 0.7379 | SRCC = 0.7411 ]
+INFO:root:Average loss=0.3832902640868456
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3842 | LCC = 0.7665 | SRCC = 0.7746 ]
+INFO:root:Average loss=0.32546988893777895
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3524 | LCC = 0.7907 | SRCC = 0.7927 ]
+INFO:root:Average loss=0.2964569763877453
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3641 | LCC = 0.7889 | SRCC = 0.7911 ]
+INFO:root:Average loss=0.26913357373231495
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3918 | LCC = 0.7936 | SRCC = 0.7955 ]
+INFO:root:Average loss=0.26239625842143327
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3358 | LCC = 0.8054 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.24196216865227774
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3471 | LCC = 0.7989 | SRCC = 0.7980 ]
+INFO:root:Average loss=0.2328253046442301
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4070 | LCC = 0.7944 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.21711012224356332
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4208 | LCC = 0.7975 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.20527806238104135
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3710 | LCC = 0.8084 | SRCC = 0.8104 ]
+INFO:root:Average loss=0.21880336784017393
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3826 | LCC = 0.8038 | SRCC = 0.8009 ]
+INFO:root:Average loss=0.19335324126176345
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3008 | LCC = 0.8107 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.17186183320024076
+INFO:root:
+[Valid][12][UTT][ MSE = 0.4026 | LCC = 0.7990 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.16985617979214743
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3747 | LCC = 0.7904 | SRCC = 0.7900 ]
+INFO:root:Average loss=0.16012472735765654
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3423 | LCC = 0.7965 | SRCC = 0.7958 ]
+INFO:root:Average loss=0.17193591995881155
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3788 | LCC = 0.7943 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.15229181467722624
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3535 | LCC = 0.7963 | SRCC = 0.7956 ]
+INFO:root:Average loss=0.13056838502868628
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3427 | LCC = 0.7958 | SRCC = 0.7940 ]
+INFO:root:Average loss=0.12885951016766903
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3458 | LCC = 0.7951 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.12146538715714063
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3178 | LCC = 0.8011 | SRCC = 0.7998 ]
+INFO:root:Average loss=0.11085419863080367
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3859 | LCC = 0.7913 | SRCC = 0.7944 ]
+INFO:root:Average loss=0.11025035343109033
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3052 | LCC = 0.7970 | SRCC = 0.7964 ]
+INFO:root:Average loss=0.10975201762257478
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3219 | LCC = 0.7936 | SRCC = 0.7912 ]
+INFO:root:Average loss=0.10082110562003575
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3182 | LCC = 0.7936 | SRCC = 0.7894 ]
+INFO:root:Average loss=0.09422457285034351
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3251 | LCC = 0.7965 | SRCC = 0.7929 ]
+INFO:root:Average loss=0.09603484399998799
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3169 | LCC = 0.7973 | SRCC = 0.7946 ]
+INFO:root:Average loss=0.09161738320611991
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3406 | LCC = 0.8001 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.0843254575649133
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3112 | LCC = 0.7965 | SRCC = 0.7936 ]
+INFO:root:Average loss=0.08538231430336451
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3321 | LCC = 0.7960 | SRCC = 0.7933 ]
+INFO:root:Average loss=0.08580643435319264
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3126 | LCC = 0.8005 | SRCC = 0.7972 ]
+INFO:root:Average loss=0.0756087307460033
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3096 | LCC = 0.8041 | SRCC = 0.8010 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3034 | LCC = 0.8193 | SRCC = 0.8206 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer16/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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_large_asr_zh/layer16/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..56024068e16c8d4260788b01503fc4de2b9730bc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c118f9cbb320059996a5831b6e909c4329b8b08a933f49a61546123e88f1270
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3608a2ec9c230488513992286143a246055e5f15
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e93f5a0334ea6df72a4295ef766e9bd97ce89434a9a60ab6b42a873338dccd81
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer16/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..1ca337ae0c18e46458192be1318bef2b9232fedb
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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=0.5999517757923175
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4916 | LCC = 0.7317 | SRCC = 0.7377 ]
+INFO:root:Average loss=0.37254038739662904
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3884 | LCC = 0.7812 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.33620757705126053
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3574 | LCC = 0.7911 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.30679143296602446
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3367 | LCC = 0.7990 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.27586541993495745
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3221 | LCC = 0.8048 | SRCC = 0.8039 ]
+INFO:root:Average loss=0.2588746729187476
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3215 | LCC = 0.8073 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.23797578651171464
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3838 | LCC = 0.8103 | SRCC = 0.8098 ]
+INFO:root:Average loss=0.22348241336070573
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3104 | LCC = 0.8114 | SRCC = 0.8075 ]
+INFO:root:Average loss=0.21066251874734193
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3108 | LCC = 0.8115 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.19586507020852503
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3106 | LCC = 0.8115 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.18464481075986838
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2976 | LCC = 0.8127 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.17674379537885004
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3097 | LCC = 0.8026 | SRCC = 0.7998 ]
+INFO:root:Average loss=0.16670650597184133
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3065 | LCC = 0.8069 | SRCC = 0.8033 ]
+INFO:root:Average loss=0.162465317795674
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3304 | LCC = 0.8066 | SRCC = 0.8020 ]
+INFO:root:Average loss=0.14429712830445704
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3395 | LCC = 0.8053 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.14724714289873075
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3193 | LCC = 0.8000 | SRCC = 0.7977 ]
+INFO:root:Average loss=0.12896038878422517
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3067 | LCC = 0.8005 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.1244250027797161
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3114 | LCC = 0.8042 | SRCC = 0.8006 ]
+INFO:root:Average loss=0.12012293514532921
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3017 | LCC = 0.8037 | SRCC = 0.8008 ]
+INFO:root:Average loss=0.11437399312853813
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3300 | LCC = 0.8042 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.10849337442181049
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2933 | LCC = 0.8110 | SRCC = 0.8083 ]
+INFO:root:Average loss=0.11781663170609719
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3025 | LCC = 0.8053 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.09839573932381776
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2991 | LCC = 0.8051 | SRCC = 0.8037 ]
+INFO:root:Average loss=0.11199819864944005
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3005 | LCC = 0.8063 | SRCC = 0.8045 ]
+INFO:root:Average loss=0.0912688955760155
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3049 | LCC = 0.8054 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.09261799818621232
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3034 | LCC = 0.8006 | SRCC = 0.7990 ]
+INFO:root:Average loss=0.0814808001025365
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3112 | LCC = 0.7986 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.0812675205942912
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3125 | LCC = 0.7973 | SRCC = 0.7932 ]
+INFO:root:Average loss=0.08486196131278308
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3100 | LCC = 0.7994 | SRCC = 0.7965 ]
+INFO:root:Average loss=0.083720055480416
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3101 | LCC = 0.8030 | SRCC = 0.8016 ]
+INFO:root:Average loss=0.07966209312853141
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3111 | LCC = 0.7970 | SRCC = 0.7953 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2800 | LCC = 0.8250 | SRCC = 0.8261 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer16/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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_large_asr_zh/layer16/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ed550b068eaf646ee7221aa71b64b2d8c4146f3e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:232d2491803178d7b44ea68185a60a206e7ed69f36df35e565a7640b4b46fa59
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8451ba7e9489f7e1f8c4cdb105d903f4ac61b4b5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9878fc7e5b45bccb8d610d7b29980832c073e4a542b88927711a8a6adf6876e2
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer16/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..aee03ad3bec6cc93518319203e82ea3960eda852
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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=0.6712896522039022
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4539 | LCC = 0.6987 | SRCC = 0.7028 ]
+INFO:root:Average loss=0.39630510982794637
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4817 | LCC = 0.7591 | SRCC = 0.7709 ]
+INFO:root:Average loss=0.3365716345799275
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3722 | LCC = 0.7868 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.3003965253249193
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3498 | LCC = 0.7894 | SRCC = 0.7886 ]
+INFO:root:Average loss=0.27240631920404923
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3280 | LCC = 0.7990 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.26967517229226917
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3544 | LCC = 0.8021 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.23091499985028535
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3203 | LCC = 0.8065 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.21692214753383246
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4644 | LCC = 0.7986 | SRCC = 0.8001 ]
+INFO:root:Average loss=0.21924362274316642
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3303 | LCC = 0.8066 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.19755806210331428
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3137 | LCC = 0.8091 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.19536881129711103
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3278 | LCC = 0.8114 | SRCC = 0.8072 ]
+INFO:root:Average loss=0.16498766056238076
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3160 | LCC = 0.8093 | SRCC = 0.8061 ]
+INFO:root:Average loss=0.16656879058633095
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3150 | LCC = 0.8046 | SRCC = 0.8010 ]
+INFO:root:Average loss=0.15717839573820433
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3724 | LCC = 0.7979 | SRCC = 0.7943 ]
+INFO:root:Average loss=0.1553824833379342
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3746 | LCC = 0.8004 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.1404804719182161
+INFO:root:
+[Valid][15][UTT][ MSE = 0.4035 | LCC = 0.8002 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.15455049610672852
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3044 | LCC = 0.8073 | SRCC = 0.8027 ]
+INFO:root:Average loss=0.12209116294980049
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3131 | LCC = 0.8015 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.12023557273623271
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3233 | LCC = 0.7964 | SRCC = 0.7919 ]
+INFO:root:Average loss=0.11715642085824257
+INFO:root:
+[Valid][19][UTT][ MSE = 0.4040 | LCC = 0.7941 | SRCC = 0.7909 ]
+INFO:root:Average loss=0.11872469366360934
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3105 | LCC = 0.8019 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.10201370668334839
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3571 | LCC = 0.7965 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.09872516812995458
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3113 | LCC = 0.8016 | SRCC = 0.7970 ]
+INFO:root:Average loss=0.09232525154948235
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3229 | LCC = 0.7961 | SRCC = 0.7908 ]
+INFO:root:Average loss=0.09973541871668437
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3307 | LCC = 0.7983 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.09301726734982087
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3196 | LCC = 0.7999 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.08213804194178337
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3219 | LCC = 0.7922 | SRCC = 0.7888 ]
+INFO:root:Average loss=0.0917213444526379
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3046 | LCC = 0.7982 | SRCC = 0.7951 ]
+INFO:root:Average loss=0.07643846274377444
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3177 | LCC = 0.7976 | SRCC = 0.7936 ]
+INFO:root:Average loss=0.07627463426727515
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3060 | LCC = 0.8028 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.08183055609846726
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3444 | LCC = 0.8021 | SRCC = 0.7989 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2744 | LCC = 0.8318 | SRCC = 0.8331 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer16/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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_large_asr_zh/layer16/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a18c8e70d3aa59347c4b306d1c7fb7f793042576
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f13cb60fa576e16d7c338cdf8e7fbec3f72274e45d3512fec712e9c30eb0365b
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..12bcdbd375563d5ff917bb9e6b93ca5ed46e27de
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c5d883dee0ccf5c5c439f74657ae784c30a7f984f22c9799a3c046ad60edb17
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer16/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f932933251426d96cf8ba9fcf40d23b4c99abc05
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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=0.6133301731867667
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4406 | LCC = 0.7151 | SRCC = 0.7191 ]
+INFO:root:Average loss=0.3947343062131833
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3723 | LCC = 0.7673 | SRCC = 0.7684 ]
+INFO:root:Average loss=0.3264686472904988
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3664 | LCC = 0.7911 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.2878621745950136
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3530 | LCC = 0.7917 | SRCC = 0.7914 ]
+INFO:root:Average loss=0.2696080284240918
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3303 | LCC = 0.8001 | SRCC = 0.7991 ]
+INFO:root:Average loss=0.27325960497061413
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3801 | LCC = 0.8017 | SRCC = 0.8028 ]
+INFO:root:Average loss=0.23502706048580316
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3410 | LCC = 0.8045 | SRCC = 0.8043 ]
+INFO:root:Average loss=0.22603305314595884
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3377 | LCC = 0.7971 | SRCC = 0.7974 ]
+INFO:root:Average loss=0.2141062290622638
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3798 | LCC = 0.8047 | SRCC = 0.8034 ]
+INFO:root:Average loss=0.2075769060697311
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3474 | LCC = 0.7924 | SRCC = 0.7891 ]
+INFO:root:Average loss=0.19724368705199316
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2972 | LCC = 0.8073 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.18850459893926597
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3045 | LCC = 0.8005 | SRCC = 0.8002 ]
+INFO:root:Average loss=0.17102261489400497
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3166 | LCC = 0.7992 | SRCC = 0.7969 ]
+INFO:root:Average loss=0.16943553042335388
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3295 | LCC = 0.7974 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.15315352170131144
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3213 | LCC = 0.8010 | SRCC = 0.7995 ]
+INFO:root:Average loss=0.1487064870695273
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3012 | LCC = 0.8053 | SRCC = 0.8031 ]
+INFO:root:Average loss=0.16366229129907411
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3616 | LCC = 0.7946 | SRCC = 0.7924 ]
+INFO:root:Average loss=0.1376529889037976
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3319 | LCC = 0.7827 | SRCC = 0.7801 ]
+INFO:root:Average loss=0.12489812696973483
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3823 | LCC = 0.7925 | SRCC = 0.7920 ]
+INFO:root:Average loss=0.13463062840776566
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3389 | LCC = 0.7882 | SRCC = 0.7857 ]
+INFO:root:Average loss=0.11293550217762971
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3911 | LCC = 0.7864 | SRCC = 0.7870 ]
+INFO:root:Average loss=0.10760122346572387
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3114 | LCC = 0.7981 | SRCC = 0.7961 ]
+INFO:root:Average loss=0.10161371318957745
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3072 | LCC = 0.7977 | SRCC = 0.7954 ]
+INFO:root:Average loss=0.10221123394484703
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3397 | LCC = 0.7935 | SRCC = 0.7926 ]
+INFO:root:Average loss=0.1036058091200315
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3754 | LCC = 0.7972 | SRCC = 0.7975 ]
+INFO:root:Average loss=0.12949891818257478
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3144 | LCC = 0.7973 | SRCC = 0.7965 ]
+INFO:root:Average loss=0.08454359938892034
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3440 | LCC = 0.7930 | SRCC = 0.7925 ]
+INFO:root:Average loss=0.0873526003307257
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3371 | LCC = 0.7949 | SRCC = 0.7941 ]
+INFO:root:Average loss=0.08430205949415
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3041 | LCC = 0.7974 | SRCC = 0.7960 ]
+INFO:root:Average loss=0.08224857775255656
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3199 | LCC = 0.7915 | SRCC = 0.7897 ]
+INFO:root:Average loss=0.07633921866997695
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3351 | LCC = 0.7864 | SRCC = 0.7836 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2893 | LCC = 0.8244 | SRCC = 0.8287 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer16/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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_large_asr_zh/layer16/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..44d1744aec66ed882372e010a3848d4d956f0d1b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0a68991c840b7baf8b0f47dd9eed48d95fb1c247db9ecea30937e2f62c2aeb7
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer16/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e434164b63f019d8d650b0d45837f610e8423f28
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd063bb24f8750f9f3e9d083ea54889634d583622056da6f0429765bd0b27b02
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer16/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer16/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..bc267e6f8c0cabaedabfd45365c0dee15c641f6a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer16/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=0.5767955741821191
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4270 | LCC = 0.7285 | SRCC = 0.7387 ]
+INFO:root:Average loss=0.38020937584149533
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3845 | LCC = 0.7718 | SRCC = 0.7767 ]
+INFO:root:Average loss=0.3307898107629556
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3490 | LCC = 0.7901 | SRCC = 0.7918 ]
+INFO:root:Average loss=0.29274248637450045
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3570 | LCC = 0.7958 | SRCC = 0.7964 ]
+INFO:root:Average loss=0.2764365984461246
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3421 | LCC = 0.7990 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.2724122905578369
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3394 | LCC = 0.8045 | SRCC = 0.8022 ]
+INFO:root:Average loss=0.23688829289032862
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3415 | LCC = 0.8011 | SRCC = 0.7986 ]
+INFO:root:Average loss=0.2289980307985575
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3242 | LCC = 0.8091 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.23124581995682839
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3337 | LCC = 0.8047 | SRCC = 0.8017 ]
+INFO:root:Average loss=0.21001119491381523
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3091 | LCC = 0.8075 | SRCC = 0.8035 ]
+INFO:root:Average loss=0.19372704176184458
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3198 | LCC = 0.8085 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.1780718872562433
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3087 | LCC = 0.8073 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.17982332857373434
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3009 | LCC = 0.8130 | SRCC = 0.8120 ]
+INFO:root:Average loss=0.17146013849056685
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3171 | LCC = 0.8046 | SRCC = 0.8034 ]
+INFO:root:Average loss=0.16012458933087495
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3718 | LCC = 0.7955 | SRCC = 0.7966 ]
+INFO:root:Average loss=0.15472842963078084
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2964 | LCC = 0.8095 | SRCC = 0.8070 ]
+INFO:root:Average loss=0.15313536979449102
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2988 | LCC = 0.8088 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.13907590880990028
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2930 | LCC = 0.8077 | SRCC = 0.8059 ]
+INFO:root:Average loss=0.1355640993286402
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3180 | LCC = 0.7973 | SRCC = 0.7961 ]
+INFO:root:Average loss=0.12286060819259056
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2961 | LCC = 0.8083 | SRCC = 0.8054 ]
+INFO:root:Average loss=0.12428751817116371
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3212 | LCC = 0.7973 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.11880805467565854
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3464 | LCC = 0.8027 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.10844255496676151
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3013 | LCC = 0.8026 | SRCC = 0.8005 ]
+INFO:root:Average loss=0.10153036335339913
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3383 | LCC = 0.7970 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.100433993129394
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3032 | LCC = 0.7980 | SRCC = 0.7957 ]
+INFO:root:Average loss=0.09386359814267892
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3035 | LCC = 0.8001 | SRCC = 0.8011 ]
+INFO:root:Average loss=0.0946582154585765
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3311 | LCC = 0.7943 | SRCC = 0.7911 ]
+INFO:root:Average loss=0.08255906699177547
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3087 | LCC = 0.7969 | SRCC = 0.7938 ]
+INFO:root:Average loss=0.08532208572022426
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3155 | LCC = 0.7973 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.08029769212962726
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3140 | LCC = 0.7995 | SRCC = 0.7967 ]
+INFO:root:Average loss=0.07536508864126144
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3317 | LCC = 0.7934 | SRCC = 0.7919 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2768 | LCC = 0.8260 | SRCC = 0.8254 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer18/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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_large_asr_zh/layer18/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6a9330e4adb16080e438056cdbaa575a93055991
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f106e5dacd3b10d646c1f0389c9062816b0c4417ab807800a0fae9891213f54
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2ec02c8a64f303e0d576f438a9c5ed1d6235c14e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:435473a717f42b09c21e1f4377752a36ecb4142d39dedcb5c740178ce03c0f8f
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer18/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..403373f9d468ba77e4543fb1f3934de1b540936f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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=0.6048095715351594
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5268 | LCC = 0.7085 | SRCC = 0.7267 ]
+INFO:root:Average loss=0.38417714796005153
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3880 | LCC = 0.7724 | SRCC = 0.7746 ]
+INFO:root:Average loss=0.3589881799924068
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3872 | LCC = 0.7811 | SRCC = 0.7845 ]
+INFO:root:Average loss=0.32873297616457325
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4083 | LCC = 0.7894 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.3061356768012047
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3412 | LCC = 0.7903 | SRCC = 0.7882 ]
+INFO:root:Average loss=0.27566553289309526
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4067 | LCC = 0.7772 | SRCC = 0.7809 ]
+INFO:root:Average loss=0.2630172855196855
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3726 | LCC = 0.7892 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.30549786793879974
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3334 | LCC = 0.7984 | SRCC = 0.7984 ]
+INFO:root:Average loss=0.23387060218896621
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3449 | LCC = 0.7883 | SRCC = 0.7880 ]
+INFO:root:Average loss=0.21670354176790285
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3348 | LCC = 0.7943 | SRCC = 0.7930 ]
+INFO:root:Average loss=0.201355048479178
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3367 | LCC = 0.7836 | SRCC = 0.7855 ]
+INFO:root:Average loss=0.24008895036501762
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3426 | LCC = 0.7880 | SRCC = 0.7875 ]
+INFO:root:Average loss=0.20787447003217843
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3553 | LCC = 0.7821 | SRCC = 0.7809 ]
+INFO:root:Average loss=0.1969221253425647
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3383 | LCC = 0.7872 | SRCC = 0.7852 ]
+INFO:root:Average loss=0.19040902140430915
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3675 | LCC = 0.7770 | SRCC = 0.7787 ]
+INFO:root:Average loss=0.16251453728629991
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3160 | LCC = 0.7905 | SRCC = 0.7900 ]
+INFO:root:Average loss=0.15018922816484404
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3747 | LCC = 0.7769 | SRCC = 0.7825 ]
+INFO:root:Average loss=0.14922086904064202
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3406 | LCC = 0.7880 | SRCC = 0.7885 ]
+INFO:root:Average loss=0.13902705315596017
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3385 | LCC = 0.7806 | SRCC = 0.7807 ]
+INFO:root:Average loss=0.14468975183673394
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3398 | LCC = 0.7917 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.12312141442910218
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3390 | LCC = 0.7806 | SRCC = 0.7790 ]
+INFO:root:Average loss=0.12822190987376067
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3306 | LCC = 0.7847 | SRCC = 0.7836 ]
+INFO:root:Average loss=0.12146608493266961
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3227 | LCC = 0.7866 | SRCC = 0.7860 ]
+INFO:root:Average loss=0.12444523827005656
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3461 | LCC = 0.7833 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.10664599909423253
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3228 | LCC = 0.7856 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.10354644776536868
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3943 | LCC = 0.7697 | SRCC = 0.7711 ]
+INFO:root:Average loss=0.10306118624523664
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3291 | LCC = 0.7804 | SRCC = 0.7794 ]
+INFO:root:Average loss=0.11547293026859944
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3581 | LCC = 0.7835 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.09728895911039451
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3180 | LCC = 0.7891 | SRCC = 0.7880 ]
+INFO:root:Average loss=0.08416045385484512
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3396 | LCC = 0.7805 | SRCC = 0.7791 ]
+INFO:root:Average loss=0.0949158327988325
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3369 | LCC = 0.7799 | SRCC = 0.7780 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3005 | LCC = 0.8226 | SRCC = 0.8242 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer18/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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_large_asr_zh/layer18/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..73c6cbc1efea1e49e46f06cf99cd4f707b04ca5a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2da4d9af80cb25b79810f73adbb1bd2ec5f84c4cc18ae66fc430a90bcddc6d34
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e4d039373271ac3fa1feb441009b11e75a56c456
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:422a7f1c7506e1dbeb3977d83facf5b8f39ef2d205d99c15132420ce584ee180
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer18/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e8b10c0e339351ba0649fa1363743927d35776b7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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=0.6794399214096558
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4819 | LCC = 0.6921 | SRCC = 0.7077 ]
+INFO:root:Average loss=0.4108802829033289
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3670 | LCC = 0.7727 | SRCC = 0.7747 ]
+INFO:root:Average loss=0.34309792537719774
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3716 | LCC = 0.7760 | SRCC = 0.7743 ]
+INFO:root:Average loss=0.34598140284801143
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3492 | LCC = 0.7845 | SRCC = 0.7834 ]
+INFO:root:Average loss=0.2927417760858169
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3371 | LCC = 0.7894 | SRCC = 0.7884 ]
+INFO:root:Average loss=0.27826601572525805
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3566 | LCC = 0.7925 | SRCC = 0.7940 ]
+INFO:root:Average loss=0.2658075103775049
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3611 | LCC = 0.8028 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.2611243346562752
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3510 | LCC = 0.7937 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.2398724662951934
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3840 | LCC = 0.7901 | SRCC = 0.7910 ]
+INFO:root:Average loss=0.22630749681057075
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3535 | LCC = 0.7905 | SRCC = 0.7889 ]
+INFO:root:Average loss=0.2219956925091071
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3398 | LCC = 0.7930 | SRCC = 0.7922 ]
+INFO:root:Average loss=0.1972171975634037
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3944 | LCC = 0.7883 | SRCC = 0.7875 ]
+INFO:root:Average loss=0.20580201185284516
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3582 | LCC = 0.7883 | SRCC = 0.7882 ]
+INFO:root:Average loss=0.17552859909259355
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3252 | LCC = 0.7899 | SRCC = 0.7889 ]
+INFO:root:Average loss=0.16693091831910303
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3511 | LCC = 0.7873 | SRCC = 0.7863 ]
+INFO:root:Average loss=0.15405845966858742
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3237 | LCC = 0.7886 | SRCC = 0.7877 ]
+INFO:root:Average loss=0.15309990474428886
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3535 | LCC = 0.7848 | SRCC = 0.7847 ]
+INFO:root:Average loss=0.1451931330255973
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3335 | LCC = 0.7852 | SRCC = 0.7834 ]
+INFO:root:Average loss=0.13299646601080894
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3211 | LCC = 0.7908 | SRCC = 0.7893 ]
+INFO:root:Average loss=0.1307629980146885
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3701 | LCC = 0.7802 | SRCC = 0.7779 ]
+INFO:root:Average loss=0.12026849923989712
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3757 | LCC = 0.7793 | SRCC = 0.7784 ]
+INFO:root:Average loss=0.12076909629962383
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3493 | LCC = 0.7821 | SRCC = 0.7800 ]
+INFO:root:Average loss=0.11172437247557518
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3517 | LCC = 0.7828 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.10021271389455368
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3916 | LCC = 0.7813 | SRCC = 0.7821 ]
+INFO:root:Average loss=0.11269735879240891
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3803 | LCC = 0.7734 | SRCC = 0.7744 ]
+INFO:root:Average loss=0.10615711936201805
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3785 | LCC = 0.7803 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.10072148639040115
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3394 | LCC = 0.7807 | SRCC = 0.7796 ]
+INFO:root:Average loss=0.08804661193146156
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3363 | LCC = 0.7786 | SRCC = 0.7781 ]
+INFO:root:Average loss=0.09718333495159943
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3898 | LCC = 0.7822 | SRCC = 0.7836 ]
+INFO:root:Average loss=0.08739403979136394
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3883 | LCC = 0.7755 | SRCC = 0.7758 ]
+INFO:root:Average loss=0.09493732313888195
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3330 | LCC = 0.7808 | SRCC = 0.7800 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3768 | LCC = 0.8135 | SRCC = 0.8184 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer18/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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_large_asr_zh/layer18/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ed521e1e2b3ed311e9ca188b05bfddc8d047d264
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93b37fe3b21e7bd7d0c9f85c72a8d49953fbd174153f4396e20ade657541e6a7
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..bcd4d4f93c07af7157424b585b1bbc5782a868ab
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f63684498c0ce00f17b5ff17892e93bb63ee6f826c040f2a204cae4525fbd608
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer18/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..22215191399c5d6941dce17ea7a154106a226854
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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=0.7088329814947568
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4776 | LCC = 0.7024 | SRCC = 0.7104 ]
+INFO:root:Average loss=0.411259609919328
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4478 | LCC = 0.7678 | SRCC = 0.7724 ]
+INFO:root:Average loss=0.3442196204112126
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3621 | LCC = 0.7831 | SRCC = 0.7813 ]
+INFO:root:Average loss=0.30573488485354644
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3618 | LCC = 0.7894 | SRCC = 0.7870 ]
+INFO:root:Average loss=0.275793531193183
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3620 | LCC = 0.8004 | SRCC = 0.7983 ]
+INFO:root:Average loss=0.2649319202472002
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3565 | LCC = 0.8043 | SRCC = 0.8035 ]
+INFO:root:Average loss=0.2447364506049034
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3510 | LCC = 0.7966 | SRCC = 0.7963 ]
+INFO:root:Average loss=0.23429816025189865
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3404 | LCC = 0.8005 | SRCC = 0.7992 ]
+INFO:root:Average loss=0.21495886261646563
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3124 | LCC = 0.8038 | SRCC = 0.8023 ]
+INFO:root:Average loss=0.22531813249374047
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3392 | LCC = 0.7991 | SRCC = 0.7981 ]
+INFO:root:Average loss=0.2029418357862876
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3232 | LCC = 0.8010 | SRCC = 0.7997 ]
+INFO:root:Average loss=0.17873231961559027
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3431 | LCC = 0.7949 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.16936125988379502
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3274 | LCC = 0.7911 | SRCC = 0.7898 ]
+INFO:root:Average loss=0.16092888437784636
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3333 | LCC = 0.7899 | SRCC = 0.7870 ]
+INFO:root:Average loss=0.15700197353577003
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3169 | LCC = 0.7979 | SRCC = 0.7989 ]
+INFO:root:Average loss=0.14581671118354186
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3495 | LCC = 0.7845 | SRCC = 0.7846 ]
+INFO:root:Average loss=0.1397852058020922
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3453 | LCC = 0.7836 | SRCC = 0.7830 ]
+INFO:root:Average loss=0.12901236728215829
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3235 | LCC = 0.7898 | SRCC = 0.7892 ]
+INFO:root:Average loss=0.12951037235175952
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3470 | LCC = 0.7823 | SRCC = 0.7798 ]
+INFO:root:Average loss=0.12870305843460253
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3755 | LCC = 0.7815 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.1140361535243499
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3356 | LCC = 0.7854 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.1364881430680935
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3387 | LCC = 0.7785 | SRCC = 0.7777 ]
+INFO:root:Average loss=0.10637553227253449
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3297 | LCC = 0.7862 | SRCC = 0.7845 ]
+INFO:root:Average loss=0.1018494680428352
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3622 | LCC = 0.7722 | SRCC = 0.7719 ]
+INFO:root:Average loss=0.0967354361827557
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3408 | LCC = 0.7783 | SRCC = 0.7775 ]
+INFO:root:Average loss=0.09333522995121968
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3609 | LCC = 0.7738 | SRCC = 0.7721 ]
+INFO:root:Average loss=0.09094826461604008
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3443 | LCC = 0.7771 | SRCC = 0.7747 ]
+INFO:root:Average loss=0.08315262059943798
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3627 | LCC = 0.7706 | SRCC = 0.7686 ]
+INFO:root:Average loss=0.08257628437609245
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3367 | LCC = 0.7760 | SRCC = 0.7740 ]
+INFO:root:Average loss=0.07360127109747666
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3465 | LCC = 0.7749 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.08850868533437069
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3380 | LCC = 0.7789 | SRCC = 0.7769 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3058 | LCC = 0.8280 | SRCC = 0.8296 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer18/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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_large_asr_zh/layer18/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..eac7d5f999a8bfe1d612bdf6a1ce6b925ce532b0
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14f2619dfdcfa5fd5cceb1b95098fc0894151a9db05443057af03e1fa63ddcee
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..32b081a4bdf277c9cd9afe9111acc976ffbba43c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0711ada9218a08c0dd1de216adb091fbfd06943f91b54444e55751f73027b7a
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer18/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..07ab2f714382aae2f2c1b192ff5ff1b342dc55e7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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=0.6511736676478997
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4505 | LCC = 0.7058 | SRCC = 0.7134 ]
+INFO:root:Average loss=0.3974240384040735
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4071 | LCC = 0.7688 | SRCC = 0.7708 ]
+INFO:root:Average loss=0.34283036127304417
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3700 | LCC = 0.7786 | SRCC = 0.7784 ]
+INFO:root:Average loss=0.31350990957938707
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3493 | LCC = 0.7915 | SRCC = 0.7906 ]
+INFO:root:Average loss=0.2744106627427615
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3520 | LCC = 0.7883 | SRCC = 0.7878 ]
+INFO:root:Average loss=0.26965529662676346
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3650 | LCC = 0.7853 | SRCC = 0.7843 ]
+INFO:root:Average loss=0.25312094390392303
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3533 | LCC = 0.7865 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.22926663339902192
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3440 | LCC = 0.7812 | SRCC = 0.7792 ]
+INFO:root:Average loss=0.2347284849637594
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3628 | LCC = 0.7831 | SRCC = 0.7825 ]
+INFO:root:Average loss=0.21773495037968343
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3381 | LCC = 0.7810 | SRCC = 0.7791 ]
+INFO:root:Average loss=0.20148373758181548
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3581 | LCC = 0.7750 | SRCC = 0.7721 ]
+INFO:root:Average loss=0.20381291105579108
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3794 | LCC = 0.7754 | SRCC = 0.7763 ]
+INFO:root:Average loss=0.1914100442559291
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3484 | LCC = 0.7827 | SRCC = 0.7807 ]
+INFO:root:Average loss=0.1936397960361762
+INFO:root:
+[Valid][13][UTT][ MSE = 0.5131 | LCC = 0.7775 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.23486301216941613
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3311 | LCC = 0.7883 | SRCC = 0.7865 ]
+INFO:root:Average loss=0.15242975166974923
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3352 | LCC = 0.7849 | SRCC = 0.7826 ]
+INFO:root:Average loss=0.15213058506831145
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3322 | LCC = 0.7802 | SRCC = 0.7785 ]
+INFO:root:Average loss=0.15613674620787302
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3552 | LCC = 0.7810 | SRCC = 0.7797 ]
+INFO:root:Average loss=0.16196670984992614
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3313 | LCC = 0.7831 | SRCC = 0.7813 ]
+INFO:root:Average loss=0.13588166724030787
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3437 | LCC = 0.7819 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.12462152483371589
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3454 | LCC = 0.7752 | SRCC = 0.7743 ]
+INFO:root:Average loss=0.11608865178930454
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3464 | LCC = 0.7746 | SRCC = 0.7721 ]
+INFO:root:Average loss=0.15427978871724543
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3392 | LCC = 0.7840 | SRCC = 0.7839 ]
+INFO:root:Average loss=0.11411105793638107
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3566 | LCC = 0.7744 | SRCC = 0.7719 ]
+INFO:root:Average loss=0.10815377953724983
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3504 | LCC = 0.7717 | SRCC = 0.7708 ]
+INFO:root:Average loss=0.11463041956990193
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3497 | LCC = 0.7764 | SRCC = 0.7748 ]
+INFO:root:Average loss=0.11967766638367604
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3797 | LCC = 0.7702 | SRCC = 0.7691 ]
+INFO:root:Average loss=0.09879504502392732
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3736 | LCC = 0.7628 | SRCC = 0.7606 ]
+INFO:root:Average loss=0.09336482991392796
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3387 | LCC = 0.7767 | SRCC = 0.7744 ]
+INFO:root:Average loss=0.09100249992349209
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3623 | LCC = 0.7759 | SRCC = 0.7742 ]
+INFO:root:Average loss=0.09572055763923205
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3485 | LCC = 0.7748 | SRCC = 0.7729 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3270 | LCC = 0.8142 | SRCC = 0.8157 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer18/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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_large_asr_zh/layer18/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..73102e65ec22779c23e82b63daebd3150994cc33
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec39386fc8ca83167ac300259864372634efb6feb4c4472263b3d24201339802
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer18/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..da5a291977a57559a4d63b169937150c9bbb6b77
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27dd201926bfd42ebba3d03e23041d9ae2af7cd8e33ab1be36e3df15400e5e15
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer18/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer18/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..07dd35cda455f32ace68bc3844f9701a9fbc5c58
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer18/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=0.6375979773509197
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4764 | LCC = 0.7126 | SRCC = 0.7257 ]
+INFO:root:Average loss=0.40801817446182936
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4176 | LCC = 0.7676 | SRCC = 0.7767 ]
+INFO:root:Average loss=0.35329034332281506
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4347 | LCC = 0.7805 | SRCC = 0.7861 ]
+INFO:root:Average loss=0.32646018228469753
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3751 | LCC = 0.7919 | SRCC = 0.7897 ]
+INFO:root:Average loss=0.3223370882945183
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3850 | LCC = 0.7929 | SRCC = 0.7947 ]
+INFO:root:Average loss=0.27787793198457134
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3458 | LCC = 0.7931 | SRCC = 0.7932 ]
+INFO:root:Average loss=0.26084507294954395
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3487 | LCC = 0.7999 | SRCC = 0.7984 ]
+INFO:root:Average loss=0.2636784118337509
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3942 | LCC = 0.8007 | SRCC = 0.7999 ]
+INFO:root:Average loss=0.24447268362228686
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3543 | LCC = 0.7986 | SRCC = 0.7965 ]
+INFO:root:Average loss=0.24607433894505867
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3318 | LCC = 0.8015 | SRCC = 0.7993 ]
+INFO:root:Average loss=0.21287147691234565
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3646 | LCC = 0.7952 | SRCC = 0.7940 ]
+INFO:root:Average loss=0.20981448296553049
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3372 | LCC = 0.7953 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.2135737188733541
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3267 | LCC = 0.7948 | SRCC = 0.7937 ]
+INFO:root:Average loss=0.19202730842889884
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3336 | LCC = 0.7953 | SRCC = 0.7931 ]
+INFO:root:Average loss=0.17521271396141785
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3470 | LCC = 0.7873 | SRCC = 0.7832 ]
+INFO:root:Average loss=0.1580854438436337
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3261 | LCC = 0.7886 | SRCC = 0.7846 ]
+INFO:root:Average loss=0.1483658828223363
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3232 | LCC = 0.7915 | SRCC = 0.7899 ]
+INFO:root:Average loss=0.18064516314711326
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3346 | LCC = 0.7904 | SRCC = 0.7888 ]
+INFO:root:Average loss=0.14232454764155242
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3214 | LCC = 0.7916 | SRCC = 0.7905 ]
+INFO:root:Average loss=0.14035675827509317
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3317 | LCC = 0.7853 | SRCC = 0.7825 ]
+INFO:root:Average loss=0.13014732014674407
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3378 | LCC = 0.7845 | SRCC = 0.7814 ]
+INFO:root:Average loss=0.13305159142384163
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3237 | LCC = 0.7871 | SRCC = 0.7839 ]
+INFO:root:Average loss=0.12629582713811827
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3215 | LCC = 0.7862 | SRCC = 0.7837 ]
+INFO:root:Average loss=0.11616324318143037
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3243 | LCC = 0.7874 | SRCC = 0.7848 ]
+INFO:root:Average loss=0.1151853138819719
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3323 | LCC = 0.7850 | SRCC = 0.7828 ]
+INFO:root:Average loss=0.10194966044181432
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3532 | LCC = 0.7852 | SRCC = 0.7824 ]
+INFO:root:Average loss=0.11036440090109141
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3408 | LCC = 0.7796 | SRCC = 0.7780 ]
+INFO:root:Average loss=0.10675919781892727
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3272 | LCC = 0.7898 | SRCC = 0.7858 ]
+INFO:root:Average loss=0.10038706220877476
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3269 | LCC = 0.7855 | SRCC = 0.7846 ]
+INFO:root:Average loss=0.08733069963562183
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3219 | LCC = 0.7858 | SRCC = 0.7842 ]
+INFO:root:Average loss=0.09205064349449597
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3298 | LCC = 0.7901 | SRCC = 0.7869 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3335 | LCC = 0.8150 | SRCC = 0.8178 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer2/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer2/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d2d6fc11f0b0acaa65803aec50aea15ea63451bd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5c7ce0dc862af08e717d0f2f2a229ebb2da5349752f2a5f25e33b5ff87bfd89
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..87d700d7cff941a30a5b34bcb7ffa8bcce12c518
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ba62e865b8847e4cfda80d5ca6aef8242142d4c8b541e8418e3339bc5d18646
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer2/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..fe6d7f2f0d509efcfa8c572c98f2b64171357fcd
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5083602458620683
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3976 | LCC = 0.7302 | SRCC = 0.7362 ]
+INFO:root:Average loss=0.35402559259763133
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3990 | LCC = 0.7777 | SRCC = 0.7805 ]
+INFO:root:Average loss=0.3067695332261232
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3256 | LCC = 0.8031 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.2770666642448841
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2988 | LCC = 0.8152 | SRCC = 0.8173 ]
+INFO:root:Average loss=0.2533111872199254
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2904 | LCC = 0.8290 | SRCC = 0.8283 ]
+INFO:root:Average loss=0.25765139189286107
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2969 | LCC = 0.8283 | SRCC = 0.8336 ]
+INFO:root:Average loss=0.23569648139751875
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3264 | LCC = 0.8359 | SRCC = 0.8414 ]
+INFO:root:Average loss=0.25441447091408265
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2497 | LCC = 0.8474 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.21636617737702835
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2512 | LCC = 0.8439 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.21629000588869438
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2411 | LCC = 0.8550 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.19065759722621012
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2790 | LCC = 0.8577 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.1900341933927475
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2578 | LCC = 0.8524 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.18174921138546404
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2412 | LCC = 0.8636 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.1771188706923754
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2269 | LCC = 0.8634 | SRCC = 0.8647 ]
+INFO:root:Average loss=0.18427710559887764
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2267 | LCC = 0.8604 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.16292956547859389
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2513 | LCC = 0.8491 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.1653949335599557
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2284 | LCC = 0.8639 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.15068205446004868
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2212 | LCC = 0.8719 | SRCC = 0.8714 ]
+INFO:root:Average loss=0.13930451984588915
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2256 | LCC = 0.8611 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.14620928475872064
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2169 | LCC = 0.8653 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.17036526268109298
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2301 | LCC = 0.8660 | SRCC = 0.8685 ]
+INFO:root:Average loss=0.14292859037717184
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2305 | LCC = 0.8636 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.1258738681387443
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2130 | LCC = 0.8622 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.12786418495652002
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2253 | LCC = 0.8672 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.12839812852251223
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2430 | LCC = 0.8724 | SRCC = 0.8718 ]
+INFO:root:Average loss=0.11225139473875363
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2094 | LCC = 0.8709 | SRCC = 0.8708 ]
+INFO:root:Average loss=0.11467277354154831
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2468 | LCC = 0.8714 | SRCC = 0.8705 ]
+INFO:root:Average loss=0.1286105989263608
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2150 | LCC = 0.8696 | SRCC = 0.8711 ]
+INFO:root:Average loss=0.10358170458139518
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2221 | LCC = 0.8708 | SRCC = 0.8714 ]
+INFO:root:Average loss=0.10560793076188137
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2209 | LCC = 0.8660 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.10911964004238446
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2201 | LCC = 0.8667 | SRCC = 0.8665 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2996 | LCC = 0.8577 | SRCC = 0.8594 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer2/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer2/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0a8dfa8154492d102df62ebaac4bf3ab15201b0b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52dc8bea41ec54ff100edb1c05e021ce7b2e1ff1316c6c7951127edf7febfc59
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c0bd1fc4ed3e1be5af4a41dd050899e1cd9561ed
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b85df917c003a9a79712e8fbc51e5e8f5ebe79258c9c04c44040d1a769477b2
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer2/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e4cd80af92a4b004ebdd1faf681d824d26190004
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5069185036879319
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4021 | LCC = 0.7339 | SRCC = 0.7307 ]
+INFO:root:Average loss=0.34805116076499987
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3548 | LCC = 0.7755 | SRCC = 0.7723 ]
+INFO:root:Average loss=0.30278632331352967
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3480 | LCC = 0.8028 | SRCC = 0.8025 ]
+INFO:root:Average loss=0.2999036197478955
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3198 | LCC = 0.8085 | SRCC = 0.8050 ]
+INFO:root:Average loss=0.2648198759326568
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2953 | LCC = 0.8250 | SRCC = 0.8189 ]
+INFO:root:Average loss=0.2699900779586572
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2858 | LCC = 0.8346 | SRCC = 0.8323 ]
+INFO:root:Average loss=0.2717033435518925
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2766 | LCC = 0.8327 | SRCC = 0.8328 ]
+INFO:root:Average loss=0.2462893600265185
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2813 | LCC = 0.8408 | SRCC = 0.8364 ]
+INFO:root:Average loss=0.2191823531801884
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2428 | LCC = 0.8456 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.20908284550293899
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2693 | LCC = 0.8476 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.21583111947163558
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2573 | LCC = 0.8489 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.19556081705750564
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2511 | LCC = 0.8558 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.19103736745623442
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2443 | LCC = 0.8618 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.18358467805844086
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2413 | LCC = 0.8609 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.1784610583040959
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2426 | LCC = 0.8548 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.17016060144091263
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2630 | LCC = 0.8526 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.16412888295375383
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2486 | LCC = 0.8619 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.15165715378064376
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2359 | LCC = 0.8557 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.14378994130171263
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2498 | LCC = 0.8586 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.15493129174678755
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2312 | LCC = 0.8667 | SRCC = 0.8665 ]
+INFO:root:Average loss=0.1412997902967991
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2161 | LCC = 0.8680 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.1443354727174991
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2267 | LCC = 0.8704 | SRCC = 0.8706 ]
+INFO:root:Average loss=0.13059568863648635
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2257 | LCC = 0.8669 | SRCC = 0.8697 ]
+INFO:root:Average loss=0.12818883302120063
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2204 | LCC = 0.8707 | SRCC = 0.8701 ]
+INFO:root:Average loss=0.12435411356198482
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2218 | LCC = 0.8704 | SRCC = 0.8697 ]
+INFO:root:Average loss=0.11948952613732754
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2138 | LCC = 0.8638 | SRCC = 0.8666 ]
+INFO:root:Average loss=0.12403913558675693
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2083 | LCC = 0.8676 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.11586600274611743
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2423 | LCC = 0.8595 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.11296626581595494
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2206 | LCC = 0.8686 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.10393781405992997
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2253 | LCC = 0.8674 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.10559598996471135
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2148 | LCC = 0.8673 | SRCC = 0.8676 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2342 | LCC = 0.8665 | SRCC = 0.8647 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer2/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer2/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..299942d5b6d4daf1be020a4fad0b1e161be4a3f4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:443a174340f981e6f6160c05b60be810a701fe718eee4a8efaaeedc44389f639
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..01339b3bd4cad1143ee276583ba5a3344639c952
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55adccb3e47ede0f641b95591037632b77f2ce5481b88e8059ca7b48afb4d7cd
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer2/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..2e2e0e961c01e454d84e1108f6a32f8fdd6bda90
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5071774411659974
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4010 | LCC = 0.7370 | SRCC = 0.7436 ]
+INFO:root:Average loss=0.3341799735640868
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3345 | LCC = 0.7871 | SRCC = 0.7874 ]
+INFO:root:Average loss=0.2859563512297777
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3132 | LCC = 0.8081 | SRCC = 0.8077 ]
+INFO:root:Average loss=0.2784053996587411
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3151 | LCC = 0.8135 | SRCC = 0.8164 ]
+INFO:root:Average loss=0.2518202901268617
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2809 | LCC = 0.8273 | SRCC = 0.8250 ]
+INFO:root:Average loss=0.2730556638577046
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3801 | LCC = 0.8176 | SRCC = 0.8298 ]
+INFO:root:Average loss=0.23840354211055315
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2800 | LCC = 0.8359 | SRCC = 0.8374 ]
+INFO:root:Average loss=0.22062070935200423
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2648 | LCC = 0.8422 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.2391769425609173
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3277 | LCC = 0.8299 | SRCC = 0.8385 ]
+INFO:root:Average loss=0.20694726896591675
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2574 | LCC = 0.8468 | SRCC = 0.8460 ]
+INFO:root:Average loss=0.2051760914425055
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3067 | LCC = 0.8422 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.2003936727459614
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2579 | LCC = 0.8442 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.1885785967684709
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3252 | LCC = 0.8426 | SRCC = 0.8464 ]
+INFO:root:Average loss=0.18013938382650033
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3040 | LCC = 0.8496 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.18321645823426735
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2319 | LCC = 0.8574 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.17756526372753656
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3202 | LCC = 0.8507 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.16212370552313635
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2314 | LCC = 0.8632 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.15835311053655085
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2265 | LCC = 0.8608 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.15917920149289644
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2349 | LCC = 0.8609 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.16181919398980263
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3057 | LCC = 0.8483 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.147420633966342
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2570 | LCC = 0.8680 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.16530314383980554
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2227 | LCC = 0.8673 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.13597037767370543
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2294 | LCC = 0.8573 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.13608333926934463
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2191 | LCC = 0.8671 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.12184918394837624
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2231 | LCC = 0.8601 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.11397998388379048
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2213 | LCC = 0.8653 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.11923427574145488
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2288 | LCC = 0.8664 | SRCC = 0.8660 ]
+INFO:root:Average loss=0.11915780985966706
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2264 | LCC = 0.8689 | SRCC = 0.8682 ]
+INFO:root:Average loss=0.12889650473609948
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2909 | LCC = 0.8567 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.11436087217850563
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2319 | LCC = 0.8632 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.10374408057675912
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2848 | LCC = 0.8560 | SRCC = 0.8594 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2345 | LCC = 0.8668 | SRCC = 0.8657 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer2/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer2/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..db94484ebb3a4203066f8a676ae872e6ada5fe5e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f74b17828c89a23d6a6076fd103875c7ead7985183f799297b2b17543911d53
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cb23ecc3cbb03260e5c214f8aac9c65bf520ed6f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:975d8a8b6c43adcf1c494b9d715216c9ccfba06cb7dd118d4d306e3172bd83d7
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer2/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..70cdd4c59ed2ae8782ddd7998478d7d23b93f28f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5612030132458761
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4096 | LCC = 0.7252 | SRCC = 0.7345 ]
+INFO:root:Average loss=0.3334534697425671
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3677 | LCC = 0.7820 | SRCC = 0.7777 ]
+INFO:root:Average loss=0.3021611326780075
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3298 | LCC = 0.8068 | SRCC = 0.8044 ]
+INFO:root:Average loss=0.27409811337024736
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3493 | LCC = 0.8183 | SRCC = 0.8181 ]
+INFO:root:Average loss=0.2571243212009088
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2817 | LCC = 0.8271 | SRCC = 0.8285 ]
+INFO:root:Average loss=0.2686785597067613
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2827 | LCC = 0.8297 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.23854077511873
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3011 | LCC = 0.8373 | SRCC = 0.8346 ]
+INFO:root:Average loss=0.2742406958952928
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3098 | LCC = 0.8384 | SRCC = 0.8389 ]
+INFO:root:Average loss=0.2131561734355413
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2860 | LCC = 0.8452 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.21517092486222586
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2770 | LCC = 0.8536 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.1986798857076046
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2391 | LCC = 0.8541 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.18907710164785385
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2331 | LCC = 0.8544 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.1883221100538205
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2492 | LCC = 0.8602 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.1881886634689111
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2320 | LCC = 0.8558 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.17590207864458746
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2433 | LCC = 0.8568 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.16475846150364631
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3238 | LCC = 0.8484 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.16788766628656632
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2546 | LCC = 0.8550 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.1644525965436911
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2336 | LCC = 0.8630 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.1502154913659279
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2403 | LCC = 0.8637 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.14027584267732424
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2383 | LCC = 0.8639 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.13637595824324167
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2197 | LCC = 0.8643 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.13283158026826689
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2204 | LCC = 0.8624 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.13291151850269392
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2421 | LCC = 0.8621 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.13231234796918356
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2158 | LCC = 0.8643 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.11730577529240878
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2305 | LCC = 0.8636 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.12321401358797
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2467 | LCC = 0.8604 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.11272625672893646
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2288 | LCC = 0.8627 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.10558885460098584
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2093 | LCC = 0.8666 | SRCC = 0.8669 ]
+INFO:root:Average loss=0.10266241708245033
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2525 | LCC = 0.8592 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.10469394076902133
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2241 | LCC = 0.8629 | SRCC = 0.8647 ]
+INFO:root:Average loss=0.11024648619767947
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2351 | LCC = 0.8612 | SRCC = 0.8646 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2282 | LCC = 0.8624 | SRCC = 0.8617 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer2/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer2/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5b8b28c9e778dfd0de7249fe0492ae3878a2de2b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cfc11481ea546242a11aead67cf7db1a43b6d95848a094ec0fc87a5c90b946f
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer2/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..81fad7e18197894f4141f72ede9035f3f59af3e3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer2/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96202c470070ba9f8eff5605d31cac31dd37b9343ee932e61e1777f641e84578
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer2/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer2/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..2c050f384b3f21c09c9cf9763e33e1885ab6574e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.7051406491261262
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4581 | LCC = 0.6828 | SRCC = 0.6923 ]
+INFO:root:Average loss=0.3733328903714816
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3850 | LCC = 0.7675 | SRCC = 0.7694 ]
+INFO:root:Average loss=0.3159919866384604
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3589 | LCC = 0.8028 | SRCC = 0.7990 ]
+INFO:root:Average loss=0.28271813174852956
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3133 | LCC = 0.8220 | SRCC = 0.8184 ]
+INFO:root:Average loss=0.25751918458785766
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2806 | LCC = 0.8264 | SRCC = 0.8260 ]
+INFO:root:Average loss=0.24010417113701502
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2993 | LCC = 0.8314 | SRCC = 0.8301 ]
+INFO:root:Average loss=0.22395582821888801
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2593 | LCC = 0.8365 | SRCC = 0.8422 ]
+INFO:root:Average loss=0.20776851437030694
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3089 | LCC = 0.8488 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.22118838742757455
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2519 | LCC = 0.8491 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.19510377160249612
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2733 | LCC = 0.8574 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.21327871600022683
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2346 | LCC = 0.8552 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.18851707168878654
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2476 | LCC = 0.8492 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.18161020972407782
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2650 | LCC = 0.8618 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.17137994063206208
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2365 | LCC = 0.8586 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.16426294669508934
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2240 | LCC = 0.8597 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.16193443737350977
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2196 | LCC = 0.8618 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.14940429450227663
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2390 | LCC = 0.8689 | SRCC = 0.8673 ]
+INFO:root:Average loss=0.14336266674292394
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2221 | LCC = 0.8663 | SRCC = 0.8680 ]
+INFO:root:Average loss=0.13886740096868613
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2328 | LCC = 0.8646 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.1362924123994815
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2434 | LCC = 0.8637 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.12660251089777702
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2147 | LCC = 0.8663 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.12353054109292153
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2356 | LCC = 0.8655 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.13907029202733284
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2355 | LCC = 0.8635 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.12441356929066853
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2316 | LCC = 0.8593 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.12325577925031002
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2601 | LCC = 0.8685 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.12481236381408496
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2454 | LCC = 0.8694 | SRCC = 0.8680 ]
+INFO:root:Average loss=0.1066495629075246
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2222 | LCC = 0.8704 | SRCC = 0.8698 ]
+INFO:root:Average loss=0.1101575361039394
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2282 | LCC = 0.8656 | SRCC = 0.8655 ]
+INFO:root:Average loss=0.10051059312163255
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2223 | LCC = 0.8673 | SRCC = 0.8678 ]
+INFO:root:Average loss=0.09367028461434902
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2212 | LCC = 0.8686 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.0977355143867242
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2183 | LCC = 0.8667 | SRCC = 0.8662 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2641 | LCC = 0.8634 | SRCC = 0.8632 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer20/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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_large_asr_zh/layer20/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d9a0518a4d1171f10ac2944e2b0bdce5dbc85930
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aeb09d224a19ebd685a1bb8e3f33ed345b756847dc2650b2b23f066d0a226c07
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..30ca1af5f49eaf8727d86ab62bb85cbfeea15740
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7e2bcc92e1b19b42d83665657b4fb46b1af070cb3c0e34e74a71aebc70f9c56
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer20/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f1cf800a34641b769abf79d1ffbb52cecd2949a3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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=0.6632398011592718
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5148 | LCC = 0.6622 | SRCC = 0.6778 ]
+INFO:root:Average loss=0.46740298966566723
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4189 | LCC = 0.7296 | SRCC = 0.7334 ]
+INFO:root:Average loss=0.3888704900940259
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4046 | LCC = 0.7541 | SRCC = 0.7549 ]
+INFO:root:Average loss=0.34326152216929656
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3762 | LCC = 0.7650 | SRCC = 0.7644 ]
+INFO:root:Average loss=0.34191366991935634
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4342 | LCC = 0.7712 | SRCC = 0.7740 ]
+INFO:root:Average loss=0.3060128534069428
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3697 | LCC = 0.7666 | SRCC = 0.7649 ]
+INFO:root:Average loss=0.3106897244086632
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4688 | LCC = 0.7612 | SRCC = 0.7674 ]
+INFO:root:Average loss=0.2869296589723
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3696 | LCC = 0.7696 | SRCC = 0.7697 ]
+INFO:root:Average loss=0.26794787763785094
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3940 | LCC = 0.7755 | SRCC = 0.7724 ]
+INFO:root:Average loss=0.28162557096817553
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3438 | LCC = 0.7750 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.24867606315857324
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3368 | LCC = 0.7847 | SRCC = 0.7812 ]
+INFO:root:Average loss=0.2218360545543524
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3710 | LCC = 0.7803 | SRCC = 0.7811 ]
+INFO:root:Average loss=0.2112891511657299
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3404 | LCC = 0.7850 | SRCC = 0.7822 ]
+INFO:root:Average loss=0.20182187549578837
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4210 | LCC = 0.7784 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.19820093057858637
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3368 | LCC = 0.7884 | SRCC = 0.7854 ]
+INFO:root:Average loss=0.17923618719363824
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3997 | LCC = 0.7705 | SRCC = 0.7732 ]
+INFO:root:Average loss=0.18837513726873276
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3543 | LCC = 0.7732 | SRCC = 0.7699 ]
+INFO:root:Average loss=0.16979063254518387
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3375 | LCC = 0.7780 | SRCC = 0.7750 ]
+INFO:root:Average loss=0.16048130689141077
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3339 | LCC = 0.7757 | SRCC = 0.7743 ]
+INFO:root:Average loss=0.16153281659652025
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3403 | LCC = 0.7792 | SRCC = 0.7760 ]
+INFO:root:Average loss=0.145131155084341
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3441 | LCC = 0.7757 | SRCC = 0.7739 ]
+INFO:root:Average loss=0.1620444330649498
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3264 | LCC = 0.7813 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.1336769325993
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3288 | LCC = 0.7784 | SRCC = 0.7747 ]
+INFO:root:Average loss=0.11818990101798987
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3491 | LCC = 0.7667 | SRCC = 0.7648 ]
+INFO:root:Average loss=0.1199498237707676
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3473 | LCC = 0.7728 | SRCC = 0.7710 ]
+INFO:root:Average loss=0.12139831760372871
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3570 | LCC = 0.7579 | SRCC = 0.7538 ]
+INFO:root:Average loss=0.11545010751638657
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3294 | LCC = 0.7795 | SRCC = 0.7775 ]
+INFO:root:Average loss=0.11023678611486386
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3568 | LCC = 0.7740 | SRCC = 0.7722 ]
+INFO:root:Average loss=0.11291766501007935
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3693 | LCC = 0.7739 | SRCC = 0.7699 ]
+INFO:root:Average loss=0.10607639833902702
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3671 | LCC = 0.7675 | SRCC = 0.7651 ]
+INFO:root:Average loss=0.09268914970258872
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3488 | LCC = 0.7694 | SRCC = 0.7646 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3099 | LCC = 0.8016 | SRCC = 0.7990 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer20/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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_large_asr_zh/layer20/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e25b5a12a8804346b236c06a67ea15a9057090f8
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4cfc49891a08db584d3bcb557fb5b7b0e770d588c1f33ae5e5c5de12f4aed05e
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a3a3bd4d885e43d215e68475935ddc9e1cc949d6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81bbe06132c49cc38daa2ce3789269107787c71b24436f4b5fbaec4f6d1b3bf8
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer20/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..7f9d097e17f503dc44b5cf479a12dd91b92e4f8c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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=0.600587282043237
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4695 | LCC = 0.7119 | SRCC = 0.7148 ]
+INFO:root:Average loss=0.42657664953134
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4107 | LCC = 0.7411 | SRCC = 0.7432 ]
+INFO:root:Average loss=0.3883203550791129
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4370 | LCC = 0.7566 | SRCC = 0.7578 ]
+INFO:root:Average loss=0.3546573011538921
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3959 | LCC = 0.7649 | SRCC = 0.7641 ]
+INFO:root:Average loss=0.3318097551281636
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3641 | LCC = 0.7718 | SRCC = 0.7711 ]
+INFO:root:Average loss=0.32642864092038226
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3640 | LCC = 0.7768 | SRCC = 0.7763 ]
+INFO:root:Average loss=0.3022905055146951
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4067 | LCC = 0.7670 | SRCC = 0.7680 ]
+INFO:root:Average loss=0.30359247154914415
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3634 | LCC = 0.7723 | SRCC = 0.7713 ]
+INFO:root:Average loss=0.26777647817746186
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3441 | LCC = 0.7826 | SRCC = 0.7823 ]
+INFO:root:Average loss=0.26324794785334515
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3640 | LCC = 0.7786 | SRCC = 0.7779 ]
+INFO:root:Average loss=0.23404085445098388
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3519 | LCC = 0.7731 | SRCC = 0.7711 ]
+INFO:root:Average loss=0.23051960823627618
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3985 | LCC = 0.7600 | SRCC = 0.7582 ]
+INFO:root:Average loss=0.21764407880031145
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3385 | LCC = 0.7773 | SRCC = 0.7733 ]
+INFO:root:Average loss=0.2049061377079059
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3466 | LCC = 0.7799 | SRCC = 0.7772 ]
+INFO:root:Average loss=0.22113671411688512
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3759 | LCC = 0.7653 | SRCC = 0.7611 ]
+INFO:root:Average loss=0.21934194347033134
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3481 | LCC = 0.7772 | SRCC = 0.7762 ]
+INFO:root:Average loss=0.20336272300053865
+INFO:root:
+[Valid][16][UTT][ MSE = 0.4147 | LCC = 0.7628 | SRCC = 0.7620 ]
+INFO:root:Average loss=0.20981275213834566
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3814 | LCC = 0.7713 | SRCC = 0.7692 ]
+INFO:root:Average loss=0.18110525216429663
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3594 | LCC = 0.7718 | SRCC = 0.7692 ]
+INFO:root:Average loss=0.16537658526347235
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3495 | LCC = 0.7712 | SRCC = 0.7690 ]
+INFO:root:Average loss=0.15763057730136773
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3543 | LCC = 0.7701 | SRCC = 0.7683 ]
+INFO:root:Average loss=0.14636910773622683
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4262 | LCC = 0.7630 | SRCC = 0.7645 ]
+INFO:root:Average loss=0.14556372375824514
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3329 | LCC = 0.7796 | SRCC = 0.7774 ]
+INFO:root:Average loss=0.140464825125841
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3727 | LCC = 0.7562 | SRCC = 0.7523 ]
+INFO:root:Average loss=0.12897175044203416
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3825 | LCC = 0.7658 | SRCC = 0.7638 ]
+INFO:root:Average loss=0.14301232869426408
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3603 | LCC = 0.7588 | SRCC = 0.7573 ]
+INFO:root:Average loss=0.11780552298594744
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3574 | LCC = 0.7642 | SRCC = 0.7616 ]
+INFO:root:Average loss=0.12287284472049811
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3549 | LCC = 0.7664 | SRCC = 0.7632 ]
+INFO:root:Average loss=0.10741588778984852
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3635 | LCC = 0.7575 | SRCC = 0.7524 ]
+INFO:root:Average loss=0.10190426835264915
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3787 | LCC = 0.7499 | SRCC = 0.7452 ]
+INFO:root:Average loss=0.1046123087215118
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3578 | LCC = 0.7585 | SRCC = 0.7560 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3372 | LCC = 0.7956 | SRCC = 0.7956 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer20/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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_large_asr_zh/layer20/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f0ff2162dadbb0dae6ffeacb5faaa2ba91e38a69
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb8c77dee58fc1eee7eb7e0b2d7447e98568511dc1c1d6e1ae7fb8a870a0e16f
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9949ea56701d7d493c8f3b7379abc263723f8323
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fb8088a120f1e67a293e1d8a8ee3269c01a3c844b45cb572be9c0b30b841011
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer20/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c640beeeecf78e21737f3c571ffd40f73422259f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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=0.6263935031035007
+INFO:root:
+[Valid][0][UTT][ MSE = 0.6022 | LCC = 0.6423 | SRCC = 0.6880 ]
+INFO:root:Average loss=0.46869384822172994
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4400 | LCC = 0.7316 | SRCC = 0.7406 ]
+INFO:root:Average loss=0.3957036974338385
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3992 | LCC = 0.7599 | SRCC = 0.7613 ]
+INFO:root:Average loss=0.3540324463676184
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4114 | LCC = 0.7666 | SRCC = 0.7673 ]
+INFO:root:Average loss=0.3394456391151135
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3869 | LCC = 0.7681 | SRCC = 0.7697 ]
+INFO:root:Average loss=0.30166948854159087
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3682 | LCC = 0.7721 | SRCC = 0.7701 ]
+INFO:root:Average loss=0.2895612097703494
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3544 | LCC = 0.7806 | SRCC = 0.7824 ]
+INFO:root:Average loss=0.28486486199574595
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3420 | LCC = 0.7848 | SRCC = 0.7838 ]
+INFO:root:Average loss=0.2688640061861429
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3675 | LCC = 0.7772 | SRCC = 0.7753 ]
+INFO:root:Average loss=0.24947591737294808
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3374 | LCC = 0.7798 | SRCC = 0.7765 ]
+INFO:root:Average loss=0.22880168373768145
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3471 | LCC = 0.7749 | SRCC = 0.7727 ]
+INFO:root:Average loss=0.22442287989915946
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4136 | LCC = 0.7743 | SRCC = 0.7726 ]
+INFO:root:Average loss=0.21515542765458426
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3569 | LCC = 0.7715 | SRCC = 0.7706 ]
+INFO:root:Average loss=0.2028130290982051
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3498 | LCC = 0.7792 | SRCC = 0.7784 ]
+INFO:root:Average loss=0.1831431729862323
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3582 | LCC = 0.7673 | SRCC = 0.7628 ]
+INFO:root:Average loss=0.1900701162715753
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3759 | LCC = 0.7682 | SRCC = 0.7652 ]
+INFO:root:Average loss=0.17161312393653086
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3574 | LCC = 0.7673 | SRCC = 0.7630 ]
+INFO:root:Average loss=0.17162430200439233
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3564 | LCC = 0.7744 | SRCC = 0.7730 ]
+INFO:root:Average loss=0.16859554231931
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3446 | LCC = 0.7692 | SRCC = 0.7657 ]
+INFO:root:Average loss=0.15726999671031267
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3761 | LCC = 0.7692 | SRCC = 0.7634 ]
+INFO:root:Average loss=0.1567945867203749
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3454 | LCC = 0.7693 | SRCC = 0.7661 ]
+INFO:root:Average loss=0.14630049304702344
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4096 | LCC = 0.7737 | SRCC = 0.7742 ]
+INFO:root:Average loss=0.13494106726004526
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3524 | LCC = 0.7723 | SRCC = 0.7682 ]
+INFO:root:Average loss=0.1573687408788082
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3515 | LCC = 0.7729 | SRCC = 0.7692 ]
+INFO:root:Average loss=0.13327597817167258
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3640 | LCC = 0.7601 | SRCC = 0.7561 ]
+INFO:root:Average loss=0.1147385258705188
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3476 | LCC = 0.7631 | SRCC = 0.7586 ]
+INFO:root:Average loss=0.11744687238182777
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3641 | LCC = 0.7558 | SRCC = 0.7535 ]
+INFO:root:Average loss=0.10385043489245269
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3500 | LCC = 0.7652 | SRCC = 0.7638 ]
+INFO:root:Average loss=0.11374727579263541
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3678 | LCC = 0.7609 | SRCC = 0.7588 ]
+INFO:root:Average loss=0.09752895148136677
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3774 | LCC = 0.7667 | SRCC = 0.7649 ]
+INFO:root:Average loss=0.09710762019340809
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3603 | LCC = 0.7605 | SRCC = 0.7583 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3180 | LCC = 0.8017 | SRCC = 0.8022 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer20/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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_large_asr_zh/layer20/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4735a95754134a69750aeb590bb09fc7cc888b40
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e18754c8673da392b63533ac3ba90d6f68d630017dd47d4e3339a2f18781d6b5
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1dc42213b6f56e2895e136c983b48ce63ec405a6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef3bb0f1110be001470ce8f135ee36b6996213a824e47926055c49a0ac342492
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer20/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..3e907b8193e77106b8a463d1270a375909d9ffe0
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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=0.5853057159827306
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4845 | LCC = 0.6951 | SRCC = 0.6990 ]
+INFO:root:Average loss=0.4109705342696263
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4149 | LCC = 0.7395 | SRCC = 0.7402 ]
+INFO:root:Average loss=0.3777213046948115
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3899 | LCC = 0.7600 | SRCC = 0.7577 ]
+INFO:root:Average loss=0.34652732465511715
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4013 | LCC = 0.7623 | SRCC = 0.7618 ]
+INFO:root:Average loss=0.3288575924741916
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3745 | LCC = 0.7702 | SRCC = 0.7676 ]
+INFO:root:Average loss=0.3326081360379855
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4293 | LCC = 0.7661 | SRCC = 0.7689 ]
+INFO:root:Average loss=0.3187400719676262
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3674 | LCC = 0.7768 | SRCC = 0.7751 ]
+INFO:root:Average loss=0.28858411140166795
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3621 | LCC = 0.7712 | SRCC = 0.7681 ]
+INFO:root:Average loss=0.2633737474679947
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3643 | LCC = 0.7705 | SRCC = 0.7658 ]
+INFO:root:Average loss=0.3013844256981825
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3663 | LCC = 0.7714 | SRCC = 0.7679 ]
+INFO:root:Average loss=0.26054390978354675
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3905 | LCC = 0.7618 | SRCC = 0.7583 ]
+INFO:root:Average loss=0.23961107490154412
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4212 | LCC = 0.7621 | SRCC = 0.7611 ]
+INFO:root:Average loss=0.22209566430403635
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3398 | LCC = 0.7765 | SRCC = 0.7718 ]
+INFO:root:Average loss=0.21241968898819044
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3482 | LCC = 0.7719 | SRCC = 0.7680 ]
+INFO:root:Average loss=0.21746775183158043
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3558 | LCC = 0.7661 | SRCC = 0.7624 ]
+INFO:root:Average loss=0.20439457922027662
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3560 | LCC = 0.7737 | SRCC = 0.7726 ]
+INFO:root:Average loss=0.18056595860383448
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3530 | LCC = 0.7616 | SRCC = 0.7589 ]
+INFO:root:Average loss=0.1729416875885083
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3948 | LCC = 0.7510 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.16958974559719747
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3753 | LCC = 0.7639 | SRCC = 0.7616 ]
+INFO:root:Average loss=0.15975119154422712
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3580 | LCC = 0.7584 | SRCC = 0.7546 ]
+INFO:root:Average loss=0.14910823937792045
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3583 | LCC = 0.7598 | SRCC = 0.7567 ]
+INFO:root:Average loss=0.1431641678015391
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3607 | LCC = 0.7590 | SRCC = 0.7550 ]
+INFO:root:Average loss=0.14003033573046708
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3571 | LCC = 0.7588 | SRCC = 0.7556 ]
+INFO:root:Average loss=0.13590105431966293
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3666 | LCC = 0.7591 | SRCC = 0.7572 ]
+INFO:root:Average loss=0.12672371178483352
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3577 | LCC = 0.7583 | SRCC = 0.7553 ]
+INFO:root:Average loss=0.12689522109352624
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3512 | LCC = 0.7622 | SRCC = 0.7584 ]
+INFO:root:Average loss=0.11532302783467831
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3514 | LCC = 0.7646 | SRCC = 0.7628 ]
+INFO:root:Average loss=0.10687631454605323
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3614 | LCC = 0.7540 | SRCC = 0.7515 ]
+INFO:root:Average loss=0.11246077897839057
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3597 | LCC = 0.7589 | SRCC = 0.7571 ]
+INFO:root:Average loss=0.0980640625437865
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3651 | LCC = 0.7514 | SRCC = 0.7473 ]
+INFO:root:Average loss=0.09402982455988725
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3684 | LCC = 0.7527 | SRCC = 0.7493 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3668 | LCC = 0.7964 | SRCC = 0.7984 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer20/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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_large_asr_zh/layer20/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c08da1ebb0d8ace339253454ddb4759d9dfc482f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9290f3d448fa0d4e5d7274cd37376726e8616db01c79b2386db272339957783
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer20/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..98bb28c92082f458ddb9e41fa81ba2ee06719f3a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93bec005dc67f7302d29a058a7eac2d771ffcd9b8c605c1065b52a995d183738
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer20/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer20/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..51c6e01123619191213bf264e672fc6af2ca24e7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer20/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=0.7191585695896393
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5797 | LCC = 0.6406 | SRCC = 0.6460 ]
+INFO:root:Average loss=0.47084368497897416
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5314 | LCC = 0.7339 | SRCC = 0.7445 ]
+INFO:root:Average loss=0.4114570208849051
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4380 | LCC = 0.7561 | SRCC = 0.7578 ]
+INFO:root:Average loss=0.3585513020173097
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4445 | LCC = 0.7660 | SRCC = 0.7661 ]
+INFO:root:Average loss=0.3338749691461905
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3945 | LCC = 0.7607 | SRCC = 0.7612 ]
+INFO:root:Average loss=0.31052596943500715
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3830 | LCC = 0.7692 | SRCC = 0.7675 ]
+INFO:root:Average loss=0.2989879718575722
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3673 | LCC = 0.7683 | SRCC = 0.7723 ]
+INFO:root:Average loss=0.28445901855444294
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3632 | LCC = 0.7747 | SRCC = 0.7752 ]
+INFO:root:Average loss=0.26981407977067506
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4091 | LCC = 0.7762 | SRCC = 0.7767 ]
+INFO:root:Average loss=0.25752416272193956
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3468 | LCC = 0.7756 | SRCC = 0.7739 ]
+INFO:root:Average loss=0.22583653405308723
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3329 | LCC = 0.7816 | SRCC = 0.7809 ]
+INFO:root:Average loss=0.23004111208212683
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3433 | LCC = 0.7798 | SRCC = 0.7789 ]
+INFO:root:Average loss=0.20959724285281622
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3717 | LCC = 0.7812 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.20570531668953407
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3994 | LCC = 0.7742 | SRCC = 0.7729 ]
+INFO:root:Average loss=0.18211723329165044
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3909 | LCC = 0.7742 | SRCC = 0.7738 ]
+INFO:root:Average loss=0.17353330619442156
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3523 | LCC = 0.7703 | SRCC = 0.7674 ]
+INFO:root:Average loss=0.16743323531670448
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3857 | LCC = 0.7578 | SRCC = 0.7545 ]
+INFO:root:Average loss=0.17009326395316002
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3469 | LCC = 0.7723 | SRCC = 0.7691 ]
+INFO:root:Average loss=0.16017044593508428
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3578 | LCC = 0.7736 | SRCC = 0.7712 ]
+INFO:root:Average loss=0.14405476664885497
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3650 | LCC = 0.7690 | SRCC = 0.7656 ]
+INFO:root:Average loss=0.1344329500809694
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3577 | LCC = 0.7638 | SRCC = 0.7587 ]
+INFO:root:Average loss=0.14290905505036697
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3540 | LCC = 0.7718 | SRCC = 0.7672 ]
+INFO:root:Average loss=0.12719267654495361
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3888 | LCC = 0.7578 | SRCC = 0.7514 ]
+INFO:root:Average loss=0.1164169453848631
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3929 | LCC = 0.7595 | SRCC = 0.7548 ]
+INFO:root:Average loss=0.13836739231378603
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3527 | LCC = 0.7650 | SRCC = 0.7614 ]
+INFO:root:Average loss=0.10641760264451687
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3831 | LCC = 0.7647 | SRCC = 0.7611 ]
+INFO:root:Average loss=0.10376305868610358
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3943 | LCC = 0.7640 | SRCC = 0.7604 ]
+INFO:root:Average loss=0.10701486344138782
+INFO:root:
+[Valid][27][UTT][ MSE = 0.4119 | LCC = 0.7622 | SRCC = 0.7589 ]
+INFO:root:Average loss=0.09808271053509834
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3479 | LCC = 0.7653 | SRCC = 0.7611 ]
+INFO:root:Average loss=0.09055642162760098
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3828 | LCC = 0.7574 | SRCC = 0.7512 ]
+INFO:root:Average loss=0.09504021059435147
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3592 | LCC = 0.7614 | SRCC = 0.7563 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3190 | LCC = 0.8028 | SRCC = 0.8009 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer22/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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_large_asr_zh/layer22/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b631fcb097febabfe197a7be60b72c4ce3683c22
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1be3a83a52afa7160fbc9bc6b9da729d4e89b09ffff9f12a0edc942501107519
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f9765930a7e05e84fd8175cba1991000785519e4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:169314b47dcfe98312c48895eae765ff57a00eef144d98ba5a7a11ac65ee13c9
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer22/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..531e8182365d908b89758cce82e71c2e354f598e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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=0.6650302043327918
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5484 | LCC = 0.6487 | SRCC = 0.6545 ]
+INFO:root:Average loss=0.4745697069626588
+INFO:root:
+[Valid][1][UTT][ MSE = 0.5493 | LCC = 0.6731 | SRCC = 0.6985 ]
+INFO:root:Average loss=0.4246594390043846
+INFO:root:
+[Valid][2][UTT][ MSE = 0.5066 | LCC = 0.7326 | SRCC = 0.7325 ]
+INFO:root:Average loss=0.3795925837296706
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4356 | LCC = 0.7347 | SRCC = 0.7371 ]
+INFO:root:Average loss=0.3429738340469507
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4100 | LCC = 0.7374 | SRCC = 0.7351 ]
+INFO:root:Average loss=0.33716459285754424
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4061 | LCC = 0.7502 | SRCC = 0.7506 ]
+INFO:root:Average loss=0.30319274427034915
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4959 | LCC = 0.7388 | SRCC = 0.7499 ]
+INFO:root:Average loss=0.30050679850272644
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3864 | LCC = 0.7501 | SRCC = 0.7535 ]
+INFO:root:Average loss=0.27378338002241576
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4135 | LCC = 0.7486 | SRCC = 0.7466 ]
+INFO:root:Average loss=0.2564665491764362
+INFO:root:
+[Valid][9][UTT][ MSE = 0.4027 | LCC = 0.7539 | SRCC = 0.7527 ]
+INFO:root:Average loss=0.23163543278590226
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3853 | LCC = 0.7474 | SRCC = 0.7457 ]
+INFO:root:Average loss=0.23767256488402685
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4928 | LCC = 0.7349 | SRCC = 0.7410 ]
+INFO:root:Average loss=0.21981779199380141
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3889 | LCC = 0.7501 | SRCC = 0.7475 ]
+INFO:root:Average loss=0.19247518651760542
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3941 | LCC = 0.7383 | SRCC = 0.7343 ]
+INFO:root:Average loss=0.17808562641342482
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3854 | LCC = 0.7526 | SRCC = 0.7510 ]
+INFO:root:Average loss=0.17107547408877274
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3664 | LCC = 0.7554 | SRCC = 0.7517 ]
+INFO:root:Average loss=0.1671833427479634
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3850 | LCC = 0.7486 | SRCC = 0.7444 ]
+INFO:root:Average loss=0.17690056161238596
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3976 | LCC = 0.7466 | SRCC = 0.7443 ]
+INFO:root:Average loss=0.14002073661257058
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3789 | LCC = 0.7449 | SRCC = 0.7421 ]
+INFO:root:Average loss=0.13500085979318008
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3945 | LCC = 0.7401 | SRCC = 0.7378 ]
+INFO:root:Average loss=0.13661994794622445
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3928 | LCC = 0.7294 | SRCC = 0.7259 ]
+INFO:root:Average loss=0.1302244379543341
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4182 | LCC = 0.7321 | SRCC = 0.7300 ]
+INFO:root:Average loss=0.12706461677757594
+INFO:root:
+[Valid][22][UTT][ MSE = 0.4354 | LCC = 0.7368 | SRCC = 0.7351 ]
+INFO:root:Average loss=0.10957530074012585
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3802 | LCC = 0.7420 | SRCC = 0.7396 ]
+INFO:root:Average loss=0.11067616223142697
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3915 | LCC = 0.7315 | SRCC = 0.7282 ]
+INFO:root:Average loss=0.09428083887084937
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3998 | LCC = 0.7370 | SRCC = 0.7345 ]
+INFO:root:Average loss=0.09728563490968484
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3991 | LCC = 0.7310 | SRCC = 0.7270 ]
+INFO:root:Average loss=0.08652987116231368
+INFO:root:
+[Valid][27][UTT][ MSE = 0.4122 | LCC = 0.7156 | SRCC = 0.7120 ]
+INFO:root:Average loss=0.08548385721559708
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4189 | LCC = 0.7238 | SRCC = 0.7204 ]
+INFO:root:Average loss=0.07938963241684131
+INFO:root:
+[Valid][29][UTT][ MSE = 0.4036 | LCC = 0.7259 | SRCC = 0.7234 ]
+INFO:root:Average loss=0.07590151401475453
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3955 | LCC = 0.7323 | SRCC = 0.7279 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3461 | LCC = 0.7866 | SRCC = 0.7882 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer22/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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_large_asr_zh/layer22/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2a3cda5baafb10f74a97580db4f5d560abf91957
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42291cac7cb9711606e14a833a523e741bcba0b32af1767e7cb5aa21cb7dc67d
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..be4efc74af8d6a43882d63fe0f7131f5deb9608f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11d44b53995d8804be3d503b83cbfa981112c0a363ec6d367a249c6f45fb7bbd
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer22/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b081fc361c57c1a501729c9a7e62830028f0d1a7
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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=0.6299353188429123
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5747 | LCC = 0.6476 | SRCC = 0.6731 ]
+INFO:root:Average loss=0.44161230134658325
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4533 | LCC = 0.7291 | SRCC = 0.7290 ]
+INFO:root:Average loss=0.395966391532849
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4411 | LCC = 0.7464 | SRCC = 0.7477 ]
+INFO:root:Average loss=0.3550349974479431
+INFO:root:
+[Valid][3][UTT][ MSE = 0.5113 | LCC = 0.7524 | SRCC = 0.7584 ]
+INFO:root:Average loss=0.3419808890575018
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3941 | LCC = 0.7624 | SRCC = 0.7604 ]
+INFO:root:Average loss=0.329758882140502
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4531 | LCC = 0.7517 | SRCC = 0.7535 ]
+INFO:root:Average loss=0.31165724266798067
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3798 | LCC = 0.7534 | SRCC = 0.7536 ]
+INFO:root:Average loss=0.3111856134656148
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4379 | LCC = 0.7588 | SRCC = 0.7613 ]
+INFO:root:Average loss=0.285092361844503
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4038 | LCC = 0.7618 | SRCC = 0.7619 ]
+INFO:root:Average loss=0.25474361024605924
+INFO:root:
+[Valid][9][UTT][ MSE = 0.4085 | LCC = 0.7647 | SRCC = 0.7634 ]
+INFO:root:Average loss=0.23642526624294427
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3939 | LCC = 0.7630 | SRCC = 0.7605 ]
+INFO:root:Average loss=0.2345930020778607
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4168 | LCC = 0.7589 | SRCC = 0.7592 ]
+INFO:root:Average loss=0.21110722231559265
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3734 | LCC = 0.7682 | SRCC = 0.7680 ]
+INFO:root:Average loss=0.20730953300610566
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3791 | LCC = 0.7671 | SRCC = 0.7668 ]
+INFO:root:Average loss=0.19656080714403054
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4054 | LCC = 0.7623 | SRCC = 0.7619 ]
+INFO:root:Average loss=0.17987061674014115
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3995 | LCC = 0.7585 | SRCC = 0.7579 ]
+INFO:root:Average loss=0.1751115005940963
+INFO:root:
+[Valid][16][UTT][ MSE = 0.4077 | LCC = 0.7555 | SRCC = 0.7557 ]
+INFO:root:Average loss=0.1609952837610856
+INFO:root:
+[Valid][17][UTT][ MSE = 0.4153 | LCC = 0.7537 | SRCC = 0.7563 ]
+INFO:root:Average loss=0.13941536471247673
+INFO:root:
+[Valid][18][UTT][ MSE = 0.4347 | LCC = 0.7537 | SRCC = 0.7537 ]
+INFO:root:Average loss=0.14922825056008804
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3734 | LCC = 0.7628 | SRCC = 0.7643 ]
+INFO:root:Average loss=0.148209652170921
+INFO:root:
+[Valid][20][UTT][ MSE = 0.4582 | LCC = 0.7609 | SRCC = 0.7624 ]
+INFO:root:Average loss=0.11961610510181157
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4871 | LCC = 0.7490 | SRCC = 0.7518 ]
+INFO:root:Average loss=0.1201657561155466
+INFO:root:
+[Valid][22][UTT][ MSE = 0.4097 | LCC = 0.7601 | SRCC = 0.7604 ]
+INFO:root:Average loss=0.11042547875489944
+INFO:root:
+[Valid][23][UTT][ MSE = 0.4504 | LCC = 0.7602 | SRCC = 0.7609 ]
+INFO:root:Average loss=0.11560445794692406
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3727 | LCC = 0.7560 | SRCC = 0.7541 ]
+INFO:root:Average loss=0.10645339809931241
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3734 | LCC = 0.7534 | SRCC = 0.7546 ]
+INFO:root:Average loss=0.11639674552358113
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3862 | LCC = 0.7632 | SRCC = 0.7627 ]
+INFO:root:Average loss=0.09633695018979219
+INFO:root:
+[Valid][27][UTT][ MSE = 0.4907 | LCC = 0.7468 | SRCC = 0.7482 ]
+INFO:root:Average loss=0.09266737466439223
+INFO:root:
+[Valid][28][UTT][ MSE = 0.4160 | LCC = 0.7532 | SRCC = 0.7532 ]
+INFO:root:Average loss=0.09458559173613022
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3755 | LCC = 0.7635 | SRCC = 0.7632 ]
+INFO:root:Average loss=0.0914552172120565
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3854 | LCC = 0.7627 | SRCC = 0.7625 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3237 | LCC = 0.7953 | SRCC = 0.7978 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer22/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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_large_asr_zh/layer22/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2b7d2f368cb0bb19da0980974f8f2b058ae26b5d
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c997d1871f8eb198aa8169cfb0b94e6e9c202d31e02e1c6d6ea00ae2ac273ea8
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ffd02008e3289a1085baa4924a1abbdbd9afbc62
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df7b8ac7a80e8e6b9dc0b7e720dbaa18e049050b544c47d2af3d961b92ca3a6d
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer22/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e84afffa3d956311413a6d1bf1aaddfc4fe17555
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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=0.6708231033423008
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5293 | LCC = 0.6548 | SRCC = 0.6612 ]
+INFO:root:Average loss=0.451181350228114
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4757 | LCC = 0.7189 | SRCC = 0.7230 ]
+INFO:root:Average loss=0.4085739008509196
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4718 | LCC = 0.7304 | SRCC = 0.7372 ]
+INFO:root:Average loss=0.3700524488320717
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4566 | LCC = 0.7431 | SRCC = 0.7495 ]
+INFO:root:Average loss=0.3415147421451715
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4128 | LCC = 0.7591 | SRCC = 0.7567 ]
+INFO:root:Average loss=0.33773642701980394
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3836 | LCC = 0.7545 | SRCC = 0.7521 ]
+INFO:root:Average loss=0.336365184149681
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4917 | LCC = 0.7455 | SRCC = 0.7504 ]
+INFO:root:Average loss=0.29701420741203505
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3953 | LCC = 0.7599 | SRCC = 0.7594 ]
+INFO:root:Average loss=0.2852860058729465
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3734 | LCC = 0.7578 | SRCC = 0.7561 ]
+INFO:root:Average loss=0.25660150670088255
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3933 | LCC = 0.7630 | SRCC = 0.7632 ]
+INFO:root:Average loss=0.24679934882964844
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3724 | LCC = 0.7573 | SRCC = 0.7559 ]
+INFO:root:Average loss=0.24384789436291426
+INFO:root:
+[Valid][11][UTT][ MSE = 0.4074 | LCC = 0.7597 | SRCC = 0.7613 ]
+INFO:root:Average loss=0.22759564813130942
+INFO:root:
+[Valid][12][UTT][ MSE = 0.4246 | LCC = 0.7561 | SRCC = 0.7581 ]
+INFO:root:Average loss=0.23561849903601867
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3841 | LCC = 0.7647 | SRCC = 0.7639 ]
+INFO:root:Average loss=0.20105452568103105
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3643 | LCC = 0.7649 | SRCC = 0.7621 ]
+INFO:root:Average loss=0.1810010962952406
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3773 | LCC = 0.7620 | SRCC = 0.7610 ]
+INFO:root:Average loss=0.17447959994658446
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3533 | LCC = 0.7656 | SRCC = 0.7655 ]
+INFO:root:Average loss=0.16487266180606988
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3883 | LCC = 0.7533 | SRCC = 0.7551 ]
+INFO:root:Average loss=0.16320982871529383
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3825 | LCC = 0.7545 | SRCC = 0.7536 ]
+INFO:root:Average loss=0.15216042005862945
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3698 | LCC = 0.7601 | SRCC = 0.7596 ]
+INFO:root:Average loss=0.14729247137140006
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3784 | LCC = 0.7541 | SRCC = 0.7541 ]
+INFO:root:Average loss=0.13724363079437843
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4171 | LCC = 0.7464 | SRCC = 0.7454 ]
+INFO:root:Average loss=0.11826919162502655
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3667 | LCC = 0.7505 | SRCC = 0.7499 ]
+INFO:root:Average loss=0.11903292112625562
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3751 | LCC = 0.7464 | SRCC = 0.7456 ]
+INFO:root:Average loss=0.12333307854640178
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3767 | LCC = 0.7507 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.10705172447248912
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3735 | LCC = 0.7475 | SRCC = 0.7439 ]
+INFO:root:Average loss=0.11507734780510266
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3957 | LCC = 0.7509 | SRCC = 0.7481 ]
+INFO:root:Average loss=0.0971514601738025
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3862 | LCC = 0.7498 | SRCC = 0.7472 ]
+INFO:root:Average loss=0.09346274248300454
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3710 | LCC = 0.7485 | SRCC = 0.7468 ]
+INFO:root:Average loss=0.0969108182650346
+INFO:root:
+[Valid][29][UTT][ MSE = 0.4182 | LCC = 0.7458 | SRCC = 0.7449 ]
+INFO:root:Average loss=0.09096066295527495
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3779 | LCC = 0.7492 | SRCC = 0.7512 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3334 | LCC = 0.7855 | SRCC = 0.7877 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer22/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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_large_asr_zh/layer22/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..25d0503ac8c0c00edf69e3c41c15091fd363aade
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9dcd677968b05ed111dea9d91d908fe7ea69424b7171cfe41103345d27afcfef
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..347acf751f72e63388637dee85e1179d6260f7ce
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed55ed53c90c0f07d4af9c069cf9c8cc0cb9803ab204d27cde2203ea5ab9567d
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer22/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..3c8d9abce4ba4e6421b000bc0ab2ad112171fbc6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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=0.6902234603961309
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5318 | LCC = 0.6469 | SRCC = 0.6551 ]
+INFO:root:Average loss=0.46131656223382705
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4353 | LCC = 0.7270 | SRCC = 0.7296 ]
+INFO:root:Average loss=0.3955636811562074
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4705 | LCC = 0.7279 | SRCC = 0.7350 ]
+INFO:root:Average loss=0.382514636294964
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4164 | LCC = 0.7391 | SRCC = 0.7425 ]
+INFO:root:Average loss=0.3519209732229893
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4954 | LCC = 0.7516 | SRCC = 0.7551 ]
+INFO:root:Average loss=0.3179828555156023
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4490 | LCC = 0.7493 | SRCC = 0.7481 ]
+INFO:root:Average loss=0.3107380328270105
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4062 | LCC = 0.7448 | SRCC = 0.7459 ]
+INFO:root:Average loss=0.2747305675576895
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4021 | LCC = 0.7586 | SRCC = 0.7596 ]
+INFO:root:Average loss=0.25895742460703236
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4096 | LCC = 0.7561 | SRCC = 0.7518 ]
+INFO:root:Average loss=0.24455769207233038
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3878 | LCC = 0.7434 | SRCC = 0.7429 ]
+INFO:root:Average loss=0.22499395467531988
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3828 | LCC = 0.7514 | SRCC = 0.7493 ]
+INFO:root:Average loss=0.2220307124348787
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3922 | LCC = 0.7408 | SRCC = 0.7433 ]
+INFO:root:Average loss=0.2047243397205304
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3687 | LCC = 0.7594 | SRCC = 0.7569 ]
+INFO:root:Average loss=0.19341749525987184
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4075 | LCC = 0.7442 | SRCC = 0.7420 ]
+INFO:root:Average loss=0.1918508214637255
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4184 | LCC = 0.7560 | SRCC = 0.7560 ]
+INFO:root:Average loss=0.16911634965202746
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3682 | LCC = 0.7497 | SRCC = 0.7487 ]
+INFO:root:Average loss=0.15942148577708465
+INFO:root:
+[Valid][16][UTT][ MSE = 0.4080 | LCC = 0.7450 | SRCC = 0.7416 ]
+INFO:root:Average loss=0.15731602515547705
+INFO:root:
+[Valid][17][UTT][ MSE = 0.4206 | LCC = 0.7454 | SRCC = 0.7435 ]
+INFO:root:Average loss=0.13525494598807433
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3764 | LCC = 0.7473 | SRCC = 0.7447 ]
+INFO:root:Average loss=0.12736641453244749
+INFO:root:
+[Valid][19][UTT][ MSE = 0.4122 | LCC = 0.7513 | SRCC = 0.7500 ]
+INFO:root:Average loss=0.11953187294495411
+INFO:root:
+[Valid][20][UTT][ MSE = 0.4231 | LCC = 0.7462 | SRCC = 0.7435 ]
+INFO:root:Average loss=0.12539275812032896
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3693 | LCC = 0.7459 | SRCC = 0.7461 ]
+INFO:root:Average loss=0.11051746610647593
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3681 | LCC = 0.7473 | SRCC = 0.7466 ]
+INFO:root:Average loss=0.11278767492144536
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3797 | LCC = 0.7448 | SRCC = 0.7445 ]
+INFO:root:Average loss=0.11185538119230515
+INFO:root:
+[Valid][24][UTT][ MSE = 0.4219 | LCC = 0.7376 | SRCC = 0.7401 ]
+INFO:root:Average loss=0.09939980836441883
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3879 | LCC = 0.7473 | SRCC = 0.7497 ]
+INFO:root:Average loss=0.10044797347524227
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3719 | LCC = 0.7482 | SRCC = 0.7480 ]
+INFO:root:Average loss=0.09433902312929814
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3837 | LCC = 0.7430 | SRCC = 0.7413 ]
+INFO:root:Average loss=0.08819971214502285
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3763 | LCC = 0.7436 | SRCC = 0.7429 ]
+INFO:root:Average loss=0.09303078222542237
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3775 | LCC = 0.7472 | SRCC = 0.7466 ]
+INFO:root:Average loss=0.07500620840642697
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3803 | LCC = 0.7386 | SRCC = 0.7371 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3420 | LCC = 0.7803 | SRCC = 0.7840 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer22/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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_large_asr_zh/layer22/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8f5791d4b62dc08a21955545fe6ef020e762ba19
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de6a7ad079f73f58d352a10585e5696f03c0e7b0f19b532255ea3b961c263c7a
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer22/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ca711b52ff15de2569d1119ca32fdd614d3a08dc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85c0ad7360ecda95063a6b64e3faa0394739f5f0e60c86d0c1c7ed68a090bb52
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer22/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer22/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b4b13997583eba54f2ac87c99ac8f515690213c3
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer22/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=0.6786296272125
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5668 | LCC = 0.6509 | SRCC = 0.6539 ]
+INFO:root:Average loss=0.4555126000673343
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4329 | LCC = 0.7135 | SRCC = 0.7143 ]
+INFO:root:Average loss=0.39360282359979093
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4052 | LCC = 0.7343 | SRCC = 0.7340 ]
+INFO:root:Average loss=0.38305416531287706
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3951 | LCC = 0.7420 | SRCC = 0.7416 ]
+INFO:root:Average loss=0.35117992816063076
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4209 | LCC = 0.7432 | SRCC = 0.7471 ]
+INFO:root:Average loss=0.3297766922758176
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4196 | LCC = 0.7523 | SRCC = 0.7520 ]
+INFO:root:Average loss=0.3163017300076974
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3948 | LCC = 0.7558 | SRCC = 0.7565 ]
+INFO:root:Average loss=0.29205029629744017
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3932 | LCC = 0.7541 | SRCC = 0.7546 ]
+INFO:root:Average loss=0.27730640253195393
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3970 | LCC = 0.7498 | SRCC = 0.7497 ]
+INFO:root:Average loss=0.251902995010217
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3617 | LCC = 0.7586 | SRCC = 0.7558 ]
+INFO:root:Average loss=0.24220418280515915
+INFO:root:
+[Valid][10][UTT][ MSE = 0.4093 | LCC = 0.7501 | SRCC = 0.7505 ]
+INFO:root:Average loss=0.2297061847952696
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3708 | LCC = 0.7606 | SRCC = 0.7611 ]
+INFO:root:Average loss=0.24170072777913168
+INFO:root:
+[Valid][12][UTT][ MSE = 0.4811 | LCC = 0.7482 | SRCC = 0.7521 ]
+INFO:root:Average loss=0.2057895444524594
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3839 | LCC = 0.7562 | SRCC = 0.7549 ]
+INFO:root:Average loss=0.1880005101362864
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4366 | LCC = 0.7437 | SRCC = 0.7398 ]
+INFO:root:Average loss=0.21074029851036194
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3766 | LCC = 0.7581 | SRCC = 0.7555 ]
+INFO:root:Average loss=0.18371777599438643
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3990 | LCC = 0.7449 | SRCC = 0.7432 ]
+INFO:root:Average loss=0.16305071449814698
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3527 | LCC = 0.7655 | SRCC = 0.7635 ]
+INFO:root:Average loss=0.15172974879925066
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3860 | LCC = 0.7439 | SRCC = 0.7399 ]
+INFO:root:Average loss=0.14270437671205935
+INFO:root:
+[Valid][19][UTT][ MSE = 0.4184 | LCC = 0.7320 | SRCC = 0.7308 ]
+INFO:root:Average loss=0.1443466273828959
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3908 | LCC = 0.7432 | SRCC = 0.7399 ]
+INFO:root:Average loss=0.12789062534769377
+INFO:root:
+[Valid][21][UTT][ MSE = 0.4186 | LCC = 0.7495 | SRCC = 0.7511 ]
+INFO:root:Average loss=0.12789206808576217
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3698 | LCC = 0.7505 | SRCC = 0.7493 ]
+INFO:root:Average loss=0.11549884061782788
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3838 | LCC = 0.7487 | SRCC = 0.7492 ]
+INFO:root:Average loss=0.11132048834592868
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3615 | LCC = 0.7540 | SRCC = 0.7530 ]
+INFO:root:Average loss=0.10566488362084596
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3654 | LCC = 0.7503 | SRCC = 0.7478 ]
+INFO:root:Average loss=0.10165370141084377
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3983 | LCC = 0.7383 | SRCC = 0.7348 ]
+INFO:root:Average loss=0.0994757942091196
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3743 | LCC = 0.7535 | SRCC = 0.7544 ]
+INFO:root:Average loss=0.08697933713212991
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3655 | LCC = 0.7495 | SRCC = 0.7470 ]
+INFO:root:Average loss=0.08863076873314686
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3725 | LCC = 0.7483 | SRCC = 0.7443 ]
+INFO:root:Average loss=0.08377211966002598
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3784 | LCC = 0.7468 | SRCC = 0.7463 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3322 | LCC = 0.7865 | SRCC = 0.7887 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer4/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer4/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ce242a1aaf8d020415a93379e7e47e3255dca59c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32444bd4983dcde81e06c595f155aeaebc4f70e1da9482c007f338b0c0772030
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..21c5db5cf4b17c5404ee67d9eea53e95bf19b044
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee4900558efa2b0e3b9c70c1fac39f5c95b375d0c003f8bb2939f366813c6814
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer4/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c9319e8bdc7ec2a9aed2daf7c0ab99162bfbaede
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.4927096212139496
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3416 | LCC = 0.7769 | SRCC = 0.7822 ]
+INFO:root:Average loss=0.2935680659153523
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3007 | LCC = 0.8220 | SRCC = 0.8223 ]
+INFO:root:Average loss=0.26378782399189776
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2923 | LCC = 0.8267 | SRCC = 0.8305 ]
+INFO:root:Average loss=0.25331890621246433
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2866 | LCC = 0.8439 | SRCC = 0.8407 ]
+INFO:root:Average loss=0.23189956102615747
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2832 | LCC = 0.8430 | SRCC = 0.8423 ]
+INFO:root:Average loss=0.21497830939598572
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2684 | LCC = 0.8527 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.21358217012423736
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2706 | LCC = 0.8575 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.1928108151142414
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2581 | LCC = 0.8614 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.18526669333760554
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2381 | LCC = 0.8564 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.17506635867250273
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2394 | LCC = 0.8646 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.17183597987660995
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2437 | LCC = 0.8621 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.15867974036014998
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2242 | LCC = 0.8648 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.1707592565470781
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2348 | LCC = 0.8656 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.1506320127309897
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2553 | LCC = 0.8658 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.1529702079983858
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2435 | LCC = 0.8646 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.13358657014293548
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2188 | LCC = 0.8700 | SRCC = 0.8663 ]
+INFO:root:Average loss=0.13606941766845873
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2278 | LCC = 0.8657 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.12661057013349655
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2192 | LCC = 0.8658 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.12644653604962888
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2310 | LCC = 0.8696 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.1159812056292326
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2217 | LCC = 0.8701 | SRCC = 0.8680 ]
+INFO:root:Average loss=0.1144947703832235
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2263 | LCC = 0.8668 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.10503664383521447
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2586 | LCC = 0.8613 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.12140201414242768
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2361 | LCC = 0.8659 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.10817911733801548
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2306 | LCC = 0.8666 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.0959852505952884
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2246 | LCC = 0.8660 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.09312110604384007
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2308 | LCC = 0.8652 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.09270697416594395
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2317 | LCC = 0.8627 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.09654772415375099
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2229 | LCC = 0.8651 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.08220153664931273
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2306 | LCC = 0.8668 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.0849593786092905
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2130 | LCC = 0.8666 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.07613822889442627
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2244 | LCC = 0.8642 | SRCC = 0.8615 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2300 | LCC = 0.8706 | SRCC = 0.8700 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer4/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer4/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1a29982ba7f8da1b16dc84eaf2cd837583c44677
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9fc077821e5c0e65e6c07bb18c72355677f914f180028d353a9c7b20a0754e6
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cf112a01fd794441eaee674ac5425d2e7b2c814c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f35d9c183c0355be0ac0db7d4c65664c2b6c99ecc79a2ddc27f9c1352235a6b
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer4/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..9a00575fc2569d2a701e7492f8fa152508b11501
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.4973654513939833
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3848 | LCC = 0.7591 | SRCC = 0.7732 ]
+INFO:root:Average loss=0.29882863526925063
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3084 | LCC = 0.8213 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.26372782026345915
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3079 | LCC = 0.8365 | SRCC = 0.8351 ]
+INFO:root:Average loss=0.23942636450131735
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2679 | LCC = 0.8454 | SRCC = 0.8430 ]
+INFO:root:Average loss=0.2261255505757454
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2796 | LCC = 0.8529 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.21614355269150856
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2536 | LCC = 0.8555 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.20711707190061227
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2393 | LCC = 0.8608 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.18601855167593712
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2342 | LCC = 0.8627 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.18342419111957917
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2412 | LCC = 0.8644 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.16523108535852188
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2423 | LCC = 0.8646 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.1621386517698948
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2742 | LCC = 0.8623 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.16208322174274004
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2388 | LCC = 0.8713 | SRCC = 0.8702 ]
+INFO:root:Average loss=0.1452120612256038
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2248 | LCC = 0.8708 | SRCC = 0.8678 ]
+INFO:root:Average loss=0.13834419407141516
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2286 | LCC = 0.8719 | SRCC = 0.8695 ]
+INFO:root:Average loss=0.13282864349774826
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2323 | LCC = 0.8685 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.12823199920165232
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2411 | LCC = 0.8686 | SRCC = 0.8684 ]
+INFO:root:Average loss=0.12524861049575683
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2165 | LCC = 0.8706 | SRCC = 0.8684 ]
+INFO:root:Average loss=0.11135942441148636
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2284 | LCC = 0.8682 | SRCC = 0.8663 ]
+INFO:root:Average loss=0.11437538313942078
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2254 | LCC = 0.8682 | SRCC = 0.8670 ]
+INFO:root:Average loss=0.11365938329925904
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2224 | LCC = 0.8707 | SRCC = 0.8695 ]
+INFO:root:Average loss=0.10722904079235517
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2137 | LCC = 0.8701 | SRCC = 0.8678 ]
+INFO:root:Average loss=0.09867839792217964
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2614 | LCC = 0.8670 | SRCC = 0.8667 ]
+INFO:root:Average loss=0.09864260729115742
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2248 | LCC = 0.8688 | SRCC = 0.8651 ]
+INFO:root:Average loss=0.09034211073930447
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2256 | LCC = 0.8685 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.0921002741998587
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2231 | LCC = 0.8702 | SRCC = 0.8672 ]
+INFO:root:Average loss=0.08326955478733931
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2236 | LCC = 0.8670 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.0781061882391954
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2301 | LCC = 0.8676 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.08339923390975365
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2421 | LCC = 0.8679 | SRCC = 0.8652 ]
+INFO:root:Average loss=0.07651789314471759
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2158 | LCC = 0.8697 | SRCC = 0.8680 ]
+INFO:root:Average loss=0.07276817755057262
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2159 | LCC = 0.8683 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.07370402472905624
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2127 | LCC = 0.8696 | SRCC = 0.8694 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2300 | LCC = 0.8683 | SRCC = 0.8684 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer4/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer4/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..01215e5fd847aa9a921a70f27c54dc83da044051
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1835f3a682159c849dbe24bdfa36e75883d296ab9bb43f7e1d54816aaf90fe17
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8af2ddbd98b2d76b3fa28a5d2b6075d80358fbba
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:286c9a1c0c94bfc4b34cb2b8365ed44b3ad2e382ccf03de6ebc80db17485cd40
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer4/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..5819c8113e1be8ade3d6c7346ecf461a9f3279c4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5490286056047831
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3664 | LCC = 0.7662 | SRCC = 0.7795 ]
+INFO:root:Average loss=0.31338958346690887
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3235 | LCC = 0.8119 | SRCC = 0.8190 ]
+INFO:root:Average loss=0.26043597436868227
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2938 | LCC = 0.8314 | SRCC = 0.8330 ]
+INFO:root:Average loss=0.23698720106711754
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2925 | LCC = 0.8399 | SRCC = 0.8400 ]
+INFO:root:Average loss=0.2277935626797187
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2988 | LCC = 0.8437 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.2145732901035211
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2731 | LCC = 0.8508 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.20044701718367064
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2992 | LCC = 0.8467 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.18782105487890732
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2465 | LCC = 0.8571 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.18095665377302048
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2467 | LCC = 0.8544 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.18213730993179175
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2666 | LCC = 0.8543 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.16661655807342285
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2489 | LCC = 0.8552 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.1487781069217584
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2202 | LCC = 0.8633 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.15161679006921938
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2218 | LCC = 0.8597 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.15844353202443856
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2318 | LCC = 0.8616 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.14008759802732712
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2433 | LCC = 0.8606 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.13482943186775231
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2401 | LCC = 0.8611 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.12395696905560982
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2233 | LCC = 0.8659 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.12080446688028482
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2263 | LCC = 0.8620 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.11268445410025425
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2258 | LCC = 0.8661 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.10623035097542481
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2209 | LCC = 0.8638 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.10145806521177292
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2110 | LCC = 0.8661 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.10330022288820682
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2218 | LCC = 0.8646 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.1046461513122687
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2121 | LCC = 0.8680 | SRCC = 0.8661 ]
+INFO:root:Average loss=0.10365624840442951
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2244 | LCC = 0.8617 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.10525494479598144
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2269 | LCC = 0.8634 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.10308632751305898
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2454 | LCC = 0.8585 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.09607535113508885
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2390 | LCC = 0.8624 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.07681613176679
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2170 | LCC = 0.8638 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.08236386225773738
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2169 | LCC = 0.8622 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.07299563369880883
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2251 | LCC = 0.8641 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.07558740245608184
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2256 | LCC = 0.8609 | SRCC = 0.8582 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2196 | LCC = 0.8736 | SRCC = 0.8758 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer4/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer4/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b71476ec47b0265b3d5b0eb8a6a5be1105324416
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57c060904852ce95e3735922213f1c77a502921c8c53168dc3febc3501f7fdf8
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8d66844f062da487daf9a103dc22a0aa65844d7c
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c1e2009f9b71f26382760fbc37189b22988ff24b205f56b8436b8b54b1dd2e8
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer4/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b4e31dae380be572a0d4c90a19502d4fa2302056
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.4718177589850548
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3901 | LCC = 0.7853 | SRCC = 0.7887 ]
+INFO:root:Average loss=0.29702738252205724
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4029 | LCC = 0.8126 | SRCC = 0.8135 ]
+INFO:root:Average loss=0.27171572431539875
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3560 | LCC = 0.8287 | SRCC = 0.8296 ]
+INFO:root:Average loss=0.24811838185175872
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3176 | LCC = 0.8417 | SRCC = 0.8397 ]
+INFO:root:Average loss=0.227980898358883
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2708 | LCC = 0.8424 | SRCC = 0.8426 ]
+INFO:root:Average loss=0.21052432977236235
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2692 | LCC = 0.8469 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.19274834638986832
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2670 | LCC = 0.8478 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.20020054605526802
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2872 | LCC = 0.8373 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.19222261279057234
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2926 | LCC = 0.8569 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.17171880727012953
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2896 | LCC = 0.8553 | SRCC = 0.8532 ]
+INFO:root:Average loss=0.16771008800237608
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2615 | LCC = 0.8584 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.16056376523696458
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2817 | LCC = 0.8579 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.15988394904595155
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2709 | LCC = 0.8558 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.14895882591223106
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2367 | LCC = 0.8628 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.1453320922759863
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2538 | LCC = 0.8599 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.14269909845330778
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2314 | LCC = 0.8625 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.13452829764439508
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2238 | LCC = 0.8633 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.12637319167455038
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2384 | LCC = 0.8612 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.12279244225758773
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2306 | LCC = 0.8634 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.11327045334455295
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2220 | LCC = 0.8617 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.1128666775348859
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2423 | LCC = 0.8646 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.10902755411389546
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2337 | LCC = 0.8611 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.09605857897072266
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2495 | LCC = 0.8602 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.09672724455595016
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2299 | LCC = 0.8640 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.09222385287284851
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2253 | LCC = 0.8640 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.08612744744198444
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2369 | LCC = 0.8582 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.08686334620683621
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2292 | LCC = 0.8631 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.08156557056384209
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2315 | LCC = 0.8594 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.08257770008192612
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2375 | LCC = 0.8645 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.08022115780757023
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2335 | LCC = 0.8612 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.08011057113225643
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2255 | LCC = 0.8626 | SRCC = 0.8634 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2666 | LCC = 0.8666 | SRCC = 0.8693 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer4/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer4/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a000d5e47bf5501f402e8bae626b33dc1aa5840d
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30fd21d5f375aecb78517d453c10dd1b49901bb78ee85c2f888545fad044c381
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer4/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e1b364f23f76f51a59472b531e1d0880ffb66dcc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer4/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e79e82cf8e62ee9ad0e7ddb9871e3f9b1f701a82db0beef01b93d5f92cb181d6
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer4/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer4/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b52aea06ac665043c2c725dbb87d625c159b0b77
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.651791723874899
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4399 | LCC = 0.7387 | SRCC = 0.7558 ]
+INFO:root:Average loss=0.30757331599791843
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3125 | LCC = 0.8082 | SRCC = 0.8104 ]
+INFO:root:Average loss=0.26386806961053455
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2866 | LCC = 0.8306 | SRCC = 0.8325 ]
+INFO:root:Average loss=0.25324883885108507
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2698 | LCC = 0.8390 | SRCC = 0.8382 ]
+INFO:root:Average loss=0.24950652760572922
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2781 | LCC = 0.8430 | SRCC = 0.8424 ]
+INFO:root:Average loss=0.22421841208751386
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2806 | LCC = 0.8507 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.21595904164207289
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2696 | LCC = 0.8479 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.21136413839383003
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2796 | LCC = 0.8499 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.18567812452331567
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2608 | LCC = 0.8581 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.1712361342058732
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2433 | LCC = 0.8596 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.167875539798003
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2436 | LCC = 0.8651 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.15938017536432314
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2473 | LCC = 0.8633 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.15017328162988028
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2361 | LCC = 0.8673 | SRCC = 0.8667 ]
+INFO:root:Average loss=0.14993203049286818
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2639 | LCC = 0.8675 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.138557529124694
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2272 | LCC = 0.8644 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.137259462227424
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2233 | LCC = 0.8698 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.12544126560290655
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2206 | LCC = 0.8690 | SRCC = 0.8678 ]
+INFO:root:Average loss=0.12193599524788368
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2276 | LCC = 0.8680 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.11177744429845077
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2326 | LCC = 0.8698 | SRCC = 0.8685 ]
+INFO:root:Average loss=0.10778957662674096
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2358 | LCC = 0.8650 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.1035642174956126
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2222 | LCC = 0.8692 | SRCC = 0.8675 ]
+INFO:root:Average loss=0.09645586766493626
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2361 | LCC = 0.8640 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.09847583822332896
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2284 | LCC = 0.8648 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.0885091957469017
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2110 | LCC = 0.8686 | SRCC = 0.8667 ]
+INFO:root:Average loss=0.08918119742511174
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2127 | LCC = 0.8718 | SRCC = 0.8694 ]
+INFO:root:Average loss=0.08677102744770356
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2103 | LCC = 0.8690 | SRCC = 0.8679 ]
+INFO:root:Average loss=0.07919434181008583
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2384 | LCC = 0.8702 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.08450653318029183
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2263 | LCC = 0.8686 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.07873261084732337
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2155 | LCC = 0.8708 | SRCC = 0.8694 ]
+INFO:root:Average loss=0.07838947583849613
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2169 | LCC = 0.8691 | SRCC = 0.8674 ]
+INFO:root:Average loss=0.07662648874788712
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2282 | LCC = 0.8660 | SRCC = 0.8644 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2311 | LCC = 0.8634 | SRCC = 0.8641 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer6/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer6/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f2e5b1b94ea4390587c9a81f707b967e11f7c7ec
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9787e5d71953ffef3b77307ced320a918471b97f41a75bd119183423e9cad4ae
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e7b07705ff8fa82abf53549ca71d62b11ab97343
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f44da18ee0cd945e06c390c9dc44e5aa14347d332c9ee2f36e98db1ec06b4884
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer6/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..5833fa3ebed1a2572db5bf59bc889cd7c9d9d41e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.49319432503901994
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3436 | LCC = 0.7997 | SRCC = 0.8062 ]
+INFO:root:Average loss=0.27467142332058686
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2941 | LCC = 0.8393 | SRCC = 0.8411 ]
+INFO:root:Average loss=0.2635112605415858
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2905 | LCC = 0.8472 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.23730496890269792
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2718 | LCC = 0.8562 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.21706791031054962
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2519 | LCC = 0.8590 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.20735569498859918
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2562 | LCC = 0.8591 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.19664444621556845
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2594 | LCC = 0.8568 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.18193010947643182
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2469 | LCC = 0.8608 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.17107011979589096
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2426 | LCC = 0.8639 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.16087762658030558
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2656 | LCC = 0.8614 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.1818190946792945
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2475 | LCC = 0.8690 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.15292056038593635
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2419 | LCC = 0.8612 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.13834510877346382
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2342 | LCC = 0.8617 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.1356915015822802
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2307 | LCC = 0.8641 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.13522388910253844
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2633 | LCC = 0.8617 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.1227852792885059
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2364 | LCC = 0.8642 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.11174232235703713
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2568 | LCC = 0.8591 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.11200726318817872
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2730 | LCC = 0.8593 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.10302360374958087
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2820 | LCC = 0.8573 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.09856922045732155
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2314 | LCC = 0.8632 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.09484370726232345
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2496 | LCC = 0.8637 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.1000973760890655
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2800 | LCC = 0.8628 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.09169781633103505
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2464 | LCC = 0.8641 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.08259352014806026
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2442 | LCC = 0.8597 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.08715979597316338
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2594 | LCC = 0.8638 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.07932824168640834
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2343 | LCC = 0.8610 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.07336872008939584
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2799 | LCC = 0.8568 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.07635018602013588
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2297 | LCC = 0.8596 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.06990425603894088
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2312 | LCC = 0.8613 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.06926144067293559
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2372 | LCC = 0.8627 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.06404425638417403
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2322 | LCC = 0.8592 | SRCC = 0.8565 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2521 | LCC = 0.8699 | SRCC = 0.8696 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer6/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer6/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ccb3f1b74b2a0bdcd3c56fefe2f325a1955aa4da
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac90b65674097928b2b56cf29f70b230755266ce7af25c37cb5d159619117756
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..38307c4a6ff148fa4f2089a9bb1bf461d26486ee
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35f2f6e04cd9a2deb078999df8b24948d94ba31ee01817082923418ef64f6bb8
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer6/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ac3635442a489feaf871a9f64e8f8e9644c639d4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5055393716081594
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3234 | LCC = 0.8102 | SRCC = 0.8145 ]
+INFO:root:Average loss=0.28752159002499705
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2818 | LCC = 0.8393 | SRCC = 0.8398 ]
+INFO:root:Average loss=0.24382897065236017
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3166 | LCC = 0.8396 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.22617639696750885
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2813 | LCC = 0.8544 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.21963214406218284
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2655 | LCC = 0.8513 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.19988727168394968
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3324 | LCC = 0.8458 | SRCC = 0.8453 ]
+INFO:root:Average loss=0.19273112351313615
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2462 | LCC = 0.8585 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.1857919584100063
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2442 | LCC = 0.8625 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.16500998488985574
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2440 | LCC = 0.8629 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.17415925631156334
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2486 | LCC = 0.8626 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.15038836231598487
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2793 | LCC = 0.8569 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.15020842353502908
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2323 | LCC = 0.8638 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.13265016923348108
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2627 | LCC = 0.8577 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.12804275216200414
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2305 | LCC = 0.8631 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.1356116809332982
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2318 | LCC = 0.8629 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.11416845195568524
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2300 | LCC = 0.8614 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.11336425854227482
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2373 | LCC = 0.8592 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.1244609918540869
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2394 | LCC = 0.8591 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.09963471814990044
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2390 | LCC = 0.8574 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.0953349436704929
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2299 | LCC = 0.8585 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.09191969304512708
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2325 | LCC = 0.8583 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.08585436217104778
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2294 | LCC = 0.8559 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.09459765121722832
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2719 | LCC = 0.8568 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.0834753937446154
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2380 | LCC = 0.8591 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.07610214238938613
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2506 | LCC = 0.8552 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.07285681147223864
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2351 | LCC = 0.8539 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.0723539418899096
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3034 | LCC = 0.8493 | SRCC = 0.8451 ]
+INFO:root:Average loss=0.07165297994819972
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2568 | LCC = 0.8526 | SRCC = 0.8476 ]
+INFO:root:Average loss=0.07199508720674576
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2582 | LCC = 0.8516 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.06707858299024594
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2685 | LCC = 0.8510 | SRCC = 0.8469 ]
+INFO:root:Average loss=0.0715217149983614
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2278 | LCC = 0.8554 | SRCC = 0.8501 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2559 | LCC = 0.8623 | SRCC = 0.8651 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer6/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer6/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9a252a4cbf1b499e4270ed95e70cb2884613b2d5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5599374352ff80611589f0552ca01d6078e2033bb7c48187d4cef23c9c08861a
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..13cc9499f209dd01057542cf17dfd0bf970ea3c2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32b8c030bb26d1ca3309d9a7ddbce91a19b55b4a362baf56a9bee667fae294a5
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer6/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..371e97d3c981cd141412252fed1da701bcca9d09
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5019452533660791
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3356 | LCC = 0.8091 | SRCC = 0.8114 ]
+INFO:root:Average loss=0.28501268380727524
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3710 | LCC = 0.8325 | SRCC = 0.8313 ]
+INFO:root:Average loss=0.24668588164525154
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3047 | LCC = 0.8492 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.2283014273032164
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2784 | LCC = 0.8537 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.20575481003675705
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3301 | LCC = 0.8510 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.20428773741691542
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4476 | LCC = 0.8350 | SRCC = 0.8375 ]
+INFO:root:Average loss=0.19204532698943064
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2853 | LCC = 0.8527 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.18370382994031295
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2639 | LCC = 0.8621 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.17215839095222643
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2873 | LCC = 0.8525 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.1697538103430699
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2978 | LCC = 0.8610 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.15539873677950639
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2665 | LCC = 0.8578 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.14902810169718203
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3225 | LCC = 0.8512 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.13721888818037817
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3483 | LCC = 0.8563 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.13177507638167113
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2856 | LCC = 0.8586 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.12953307756628746
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2439 | LCC = 0.8634 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.12057070921246822
+INFO:root:
+[Valid][15][UTT][ MSE = 0.4172 | LCC = 0.8458 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.1282075639718618
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2589 | LCC = 0.8589 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.1098620508534786
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2239 | LCC = 0.8661 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.11379756405949593
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2421 | LCC = 0.8632 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.10131743172995555
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2753 | LCC = 0.8600 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.09888072727391353
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2763 | LCC = 0.8618 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.09431812081199425
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2393 | LCC = 0.8655 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.08766807400836395
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2257 | LCC = 0.8659 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.08805000848877124
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2558 | LCC = 0.8625 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.08140808788056557
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2523 | LCC = 0.8546 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.076956193941908
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2785 | LCC = 0.8596 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.07510767862773858
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3061 | LCC = 0.8606 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.07296722969756676
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2379 | LCC = 0.8580 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.06822248419316915
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3123 | LCC = 0.8549 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.06308909290685104
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3045 | LCC = 0.8502 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.06338760113486877
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2471 | LCC = 0.8547 | SRCC = 0.8499 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2350 | LCC = 0.8563 | SRCC = 0.8609 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer6/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer6/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c375cf6a42d8cc4555efde1da2616b2f205bdff2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b983076f524154c0f34be998339741aa537bc244df4f5be2871ac1cce5e49741
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..21654bc4a879e007904f86760c6b9a71ad339aaf
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd24093d074566a11976d8a10b22ff399f93f61055436d3fdf0a34bdd475e8de
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer6/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e953e87ca50bbc1ac970a870bd9830b7edd265c6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.761470126800048
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5302 | LCC = 0.7695 | SRCC = 0.7708 ]
+INFO:root:Average loss=0.33660822877517116
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3056 | LCC = 0.8238 | SRCC = 0.8266 ]
+INFO:root:Average loss=0.2582632026229149
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2993 | LCC = 0.8476 | SRCC = 0.8451 ]
+INFO:root:Average loss=0.23848616885833251
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2870 | LCC = 0.8478 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.21320183938130355
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2486 | LCC = 0.8561 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.19493550826341677
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3502 | LCC = 0.8515 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.19056100083085206
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2576 | LCC = 0.8586 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.17759927027882674
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2522 | LCC = 0.8568 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.1695050334510131
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2499 | LCC = 0.8520 | SRCC = 0.8458 ]
+INFO:root:Average loss=0.15596866617218041
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2456 | LCC = 0.8602 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.15606844711762208
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2664 | LCC = 0.8671 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.14276186988139764
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2421 | LCC = 0.8639 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.1322892229908552
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2517 | LCC = 0.8650 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.13371015960971513
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2400 | LCC = 0.8652 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.12165213892093071
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2412 | LCC = 0.8638 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.11826121310393016
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2620 | LCC = 0.8630 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.11299182216708477
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2564 | LCC = 0.8634 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.10159331378646386
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2248 | LCC = 0.8638 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.09424230690376881
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2851 | LCC = 0.8568 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.09647298943346892
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2447 | LCC = 0.8639 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.09668825237223735
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2741 | LCC = 0.8598 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.08719229215803818
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2459 | LCC = 0.8592 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.0833438810152121
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2629 | LCC = 0.8599 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.08159560829592057
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2838 | LCC = 0.8578 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.07767639524088456
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2176 | LCC = 0.8630 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.07938503545637314
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2431 | LCC = 0.8605 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.06863245009802856
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2746 | LCC = 0.8585 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.07047030067023559
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2625 | LCC = 0.8606 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.0635722073702476
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2479 | LCC = 0.8624 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.06365344682947183
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2683 | LCC = 0.8598 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.05909067105788451
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2357 | LCC = 0.8613 | SRCC = 0.8571 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2326 | LCC = 0.8701 | SRCC = 0.8684 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer6/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer6/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..15e7884fc62997795e71b3d85161ca2aa8470cbc
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:393cc0c50d9c29abcb1d3e8f1e556f2bc8af0202aab98d177b8d9e31f6bc8a07
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer6/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0151363f592743b3300bd6e41ed0f4953edefd1b
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer6/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d84557fe8411307d109541e46d29cede6ef92080bb738c211e7add222462e763
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer6/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer6/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..d6d96c2bcf507b0ef9f71d45c63f7e47d79acb49
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5243661124736835
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3377 | LCC = 0.7939 | SRCC = 0.8032 ]
+INFO:root:Average loss=0.28825451605595076
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3001 | LCC = 0.8239 | SRCC = 0.8270 ]
+INFO:root:Average loss=0.24706623473992714
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2854 | LCC = 0.8421 | SRCC = 0.8408 ]
+INFO:root:Average loss=0.24403023987244338
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3217 | LCC = 0.8489 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.2261888163212018
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3312 | LCC = 0.8488 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.20833001553248137
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2583 | LCC = 0.8555 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.19457923945708153
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2405 | LCC = 0.8594 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.17574977693267357
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2544 | LCC = 0.8591 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.16404308636601156
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2329 | LCC = 0.8586 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.15383539310632607
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2400 | LCC = 0.8650 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.15554215472478133
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2487 | LCC = 0.8648 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.14547239606961226
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2311 | LCC = 0.8638 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.13327226529900843
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2207 | LCC = 0.8642 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.13531611429957244
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2217 | LCC = 0.8655 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.12233222677157475
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2198 | LCC = 0.8669 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.1103487189572591
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2273 | LCC = 0.8643 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.11546283301252586
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2251 | LCC = 0.8672 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.10523425453366378
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2134 | LCC = 0.8675 | SRCC = 0.8660 ]
+INFO:root:Average loss=0.0985158428740807
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2321 | LCC = 0.8650 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.09962424526038842
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2389 | LCC = 0.8614 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.08905119854861344
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2243 | LCC = 0.8640 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.08525659105716607
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2601 | LCC = 0.8614 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.08895859805246194
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2285 | LCC = 0.8643 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.08234580539357968
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2321 | LCC = 0.8618 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.08236360487838586
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2239 | LCC = 0.8674 | SRCC = 0.8654 ]
+INFO:root:Average loss=0.07216755338968375
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2338 | LCC = 0.8649 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.07325092970560758
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2191 | LCC = 0.8631 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.07322734985978176
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2347 | LCC = 0.8631 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.07190227990922256
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2340 | LCC = 0.8632 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.06761062981035465
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2198 | LCC = 0.8662 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.06503913494256827
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2282 | LCC = 0.8643 | SRCC = 0.8628 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2578 | LCC = 0.8598 | SRCC = 0.8638 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run0/config.gin b/results_bvcc/w2v2_large_asr_zh/layer8/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer8/run0/model.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4b029e00fec044d439ccd5798e32697f81ab80d4
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d81af2fd170a53b6815de28da12f6f0f85d53ab74643690a6c35938628bf6da
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run0/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..33e4750426d943441f515825d4eb5f4b635c2bb2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5d18eefd69ee1f3bd48057a0e4f0583be1331e16bc6822f88e76333f23cb0c5
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run0/train.log b/results_bvcc/w2v2_large_asr_zh/layer8/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f42881912d6d584abc13d1c0561b83c88fe70d40
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5390540149349433
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3583 | LCC = 0.8106 | SRCC = 0.8125 ]
+INFO:root:Average loss=0.2888979355876262
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3243 | LCC = 0.8300 | SRCC = 0.8286 ]
+INFO:root:Average loss=0.2569037354909457
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2800 | LCC = 0.8388 | SRCC = 0.8388 ]
+INFO:root:Average loss=0.25953454123093533
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2955 | LCC = 0.8441 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.24391213938211784
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2681 | LCC = 0.8481 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.21410788748508844
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2767 | LCC = 0.8487 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.20770963816306529
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2789 | LCC = 0.8492 | SRCC = 0.8474 ]
+INFO:root:Average loss=0.197322595004852
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2548 | LCC = 0.8498 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.17832923155182448
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2455 | LCC = 0.8582 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.17201272006600332
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2467 | LCC = 0.8529 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.16739992873790938
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2416 | LCC = 0.8569 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.16184356598517832
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2517 | LCC = 0.8558 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.1466890828540692
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2492 | LCC = 0.8563 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.13499481221422172
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2408 | LCC = 0.8574 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.13172442408708426
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2370 | LCC = 0.8528 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.12016776242317298
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2486 | LCC = 0.8557 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.12338629527351795
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2421 | LCC = 0.8544 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.11299338487860484
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2405 | LCC = 0.8523 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.10609249627360931
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2398 | LCC = 0.8511 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.0982115259155249
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2515 | LCC = 0.8478 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.09896219302064334
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2560 | LCC = 0.8465 | SRCC = 0.8435 ]
+INFO:root:Average loss=0.09493110433985026
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2448 | LCC = 0.8493 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.08904905244708061
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2582 | LCC = 0.8480 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.09270019623904656
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2551 | LCC = 0.8481 | SRCC = 0.8457 ]
+INFO:root:Average loss=0.07899114881188442
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2450 | LCC = 0.8486 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.07539194994247876
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2467 | LCC = 0.8483 | SRCC = 0.8453 ]
+INFO:root:Average loss=0.07875140928305112
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2507 | LCC = 0.8480 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.07625656598844589
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2882 | LCC = 0.8461 | SRCC = 0.8439 ]
+INFO:root:Average loss=0.07095634058499947
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2499 | LCC = 0.8521 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.07051818750989743
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2443 | LCC = 0.8496 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.06719740738089268
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2419 | LCC = 0.8508 | SRCC = 0.8485 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2504 | LCC = 0.8593 | SRCC = 0.8612 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run1/config.gin b/results_bvcc/w2v2_large_asr_zh/layer8/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer8/run1/model.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9772a7081afbfaa67bf86b525a2dc635256d2b9e
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5e0fb4f2cb34786868aab4b2112efec446c016e5b5d89e85710655d59e98927
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run1/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7c15a1153054828c5956d76faf99fd358235d87f
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:763a579eb54fd312f000238e965f29db1c750c518eb51fcb7b81e0b3cb8ff3d5
+size 3404998
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run1/train.log b/results_bvcc/w2v2_large_asr_zh/layer8/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..98e99c2f9a8d33b5cc63561d3802a87f88fe9ea2
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5821645340094199
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3741 | LCC = 0.7873 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.308383363752793
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3478 | LCC = 0.8202 | SRCC = 0.8222 ]
+INFO:root:Average loss=0.2704595113411928
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2871 | LCC = 0.8392 | SRCC = 0.8377 ]
+INFO:root:Average loss=0.25656021462801176
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3084 | LCC = 0.8450 | SRCC = 0.8436 ]
+INFO:root:Average loss=0.22792000399950224
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2908 | LCC = 0.8478 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.2121824087240757
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2515 | LCC = 0.8497 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.2113118332165938
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2550 | LCC = 0.8531 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.1964752169755789
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2781 | LCC = 0.8453 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.18110414546651718
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2713 | LCC = 0.8455 | SRCC = 0.8410 ]
+INFO:root:Average loss=0.1733131720087467
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2558 | LCC = 0.8457 | SRCC = 0.8410 ]
+INFO:root:Average loss=0.16656515518060097
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2647 | LCC = 0.8491 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.1607415352303248
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2486 | LCC = 0.8501 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.14996527154476214
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2601 | LCC = 0.8438 | SRCC = 0.8402 ]
+INFO:root:Average loss=0.15788143739486352
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2603 | LCC = 0.8474 | SRCC = 0.8417 ]
+INFO:root:Average loss=0.13821087672542304
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2494 | LCC = 0.8515 | SRCC = 0.8474 ]
+INFO:root:Average loss=0.13113244384145126
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2646 | LCC = 0.8453 | SRCC = 0.8412 ]
+INFO:root:Average loss=0.13427004132133263
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2787 | LCC = 0.8484 | SRCC = 0.8463 ]
+INFO:root:Average loss=0.11785982253077702
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3228 | LCC = 0.8349 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.12079087678247537
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2452 | LCC = 0.8505 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.11736865198382965
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2524 | LCC = 0.8487 | SRCC = 0.8452 ]
+INFO:root:Average loss=0.11251816220390491
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2514 | LCC = 0.8453 | SRCC = 0.8414 ]
+INFO:root:Average loss=0.10438678556909928
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2641 | LCC = 0.8509 | SRCC = 0.8473 ]
+INFO:root:Average loss=0.11387381951014201
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2432 | LCC = 0.8489 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.09908581983584624
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2657 | LCC = 0.8440 | SRCC = 0.8416 ]
+INFO:root:Average loss=0.09560277847907482
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2496 | LCC = 0.8446 | SRCC = 0.8399 ]
+INFO:root:Average loss=0.10925959910337742
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3238 | LCC = 0.8498 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.08950952865565434
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2472 | LCC = 0.8456 | SRCC = 0.8423 ]
+INFO:root:Average loss=0.0812704151448531
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2472 | LCC = 0.8465 | SRCC = 0.8410 ]
+INFO:root:Average loss=0.07914436159607692
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2681 | LCC = 0.8404 | SRCC = 0.8377 ]
+INFO:root:Average loss=0.10416386295587589
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2462 | LCC = 0.8498 | SRCC = 0.8459 ]
+INFO:root:Average loss=0.07298699069099548
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2630 | LCC = 0.8463 | SRCC = 0.8427 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2776 | LCC = 0.8544 | SRCC = 0.8603 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run2/config.gin b/results_bvcc/w2v2_large_asr_zh/layer8/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer8/run2/model.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..e3b6989f752b8c1cd56ce831e17813c7c70f6733
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8418dcf04add1d06f537dac7e2e70f930dee83c973c96c4587e31386a2024c6d
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run2/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..729562800e40c4b7f21e157844d2a3e126c187e6
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c59949f6100177fce9d29c803c34a68810bcafab89cf29396a210ea887e1e929
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run2/train.log b/results_bvcc/w2v2_large_asr_zh/layer8/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..8c54df8ce2528bfe288167a9b80e13cfb2b89752
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.515836106852079
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3545 | LCC = 0.8018 | SRCC = 0.8041 ]
+INFO:root:Average loss=0.2865717248656811
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2907 | LCC = 0.8269 | SRCC = 0.8276 ]
+INFO:root:Average loss=0.26852219666425997
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2837 | LCC = 0.8380 | SRCC = 0.8351 ]
+INFO:root:Average loss=0.23654651164244384
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2495 | LCC = 0.8419 | SRCC = 0.8400 ]
+INFO:root:Average loss=0.23672136053060874
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3190 | LCC = 0.8464 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.213877251992623
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2912 | LCC = 0.8391 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.20159975190957388
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2763 | LCC = 0.8554 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.19117344724826324
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2672 | LCC = 0.8527 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.17144134602485558
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2518 | LCC = 0.8475 | SRCC = 0.8449 ]
+INFO:root:Average loss=0.18049428048424232
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3067 | LCC = 0.8425 | SRCC = 0.8418 ]
+INFO:root:Average loss=0.17281607519357634
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3037 | LCC = 0.8395 | SRCC = 0.8376 ]
+INFO:root:Average loss=0.1584569279773113
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2461 | LCC = 0.8529 | SRCC = 0.8500 ]
+INFO:root:Average loss=0.14570103824520722
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2597 | LCC = 0.8457 | SRCC = 0.8416 ]
+INFO:root:Average loss=0.13356884502065489
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2471 | LCC = 0.8531 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.13883645440905523
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2640 | LCC = 0.8510 | SRCC = 0.8471 ]
+INFO:root:Average loss=0.12861156492279127
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2784 | LCC = 0.8443 | SRCC = 0.8403 ]
+INFO:root:Average loss=0.12786910807092985
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2488 | LCC = 0.8532 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.11385954228731301
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2503 | LCC = 0.8481 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.11465225607538834
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2393 | LCC = 0.8511 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.12510227230496895
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2472 | LCC = 0.8511 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.11266142110794018
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2542 | LCC = 0.8489 | SRCC = 0.8448 ]
+INFO:root:Average loss=0.09777564297501858
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2519 | LCC = 0.8495 | SRCC = 0.8455 ]
+INFO:root:Average loss=0.09853741646004029
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2603 | LCC = 0.8471 | SRCC = 0.8441 ]
+INFO:root:Average loss=0.0843460169644692
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2500 | LCC = 0.8471 | SRCC = 0.8420 ]
+INFO:root:Average loss=0.09718821770869769
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2466 | LCC = 0.8480 | SRCC = 0.8431 ]
+INFO:root:Average loss=0.07997034265635869
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2704 | LCC = 0.8468 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.08000807502330878
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2559 | LCC = 0.8426 | SRCC = 0.8365 ]
+INFO:root:Average loss=0.07248784214831315
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2705 | LCC = 0.8410 | SRCC = 0.8362 ]
+INFO:root:Average loss=0.07402072278543925
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2506 | LCC = 0.8503 | SRCC = 0.8450 ]
+INFO:root:Average loss=0.08323130899896988
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2530 | LCC = 0.8476 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.07613030023490772
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2505 | LCC = 0.8440 | SRCC = 0.8399 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2928 | LCC = 0.8523 | SRCC = 0.8537 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run3/config.gin b/results_bvcc/w2v2_large_asr_zh/layer8/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer8/run3/model.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8f3159f22c15acd274d2ce2105bbb1aea336ab8a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91f20c6d57f79d984f9fa7e10e27afaedd698a1bf5cd5841c6b8a49e174c0d04
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run3/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c28f57863d79b6fdf71347b6769acd8f163fe0b5
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37a5481d83062702846315be278cd72b1d17a7dcfa5d026d3696e7af6cad26d1
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run3/train.log b/results_bvcc/w2v2_large_asr_zh/layer8/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..feeebbca72832f96709304dcdd01f014dd0d5fed
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.6028855507954572
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3562 | LCC = 0.7926 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.2946079149842262
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3245 | LCC = 0.8323 | SRCC = 0.8310 ]
+INFO:root:Average loss=0.2598664403343812
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2945 | LCC = 0.8404 | SRCC = 0.8404 ]
+INFO:root:Average loss=0.23241421532554504
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3201 | LCC = 0.8381 | SRCC = 0.8403 ]
+INFO:root:Average loss=0.23899052196588272
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3611 | LCC = 0.8443 | SRCC = 0.8472 ]
+INFO:root:Average loss=0.214586007862519
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2594 | LCC = 0.8486 | SRCC = 0.8447 ]
+INFO:root:Average loss=0.21055263529221216
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3055 | LCC = 0.8460 | SRCC = 0.8438 ]
+INFO:root:Average loss=0.195206019645318
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2600 | LCC = 0.8479 | SRCC = 0.8447 ]
+INFO:root:Average loss=0.19012217280956414
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2980 | LCC = 0.8484 | SRCC = 0.8448 ]
+INFO:root:Average loss=0.17066540072361627
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2753 | LCC = 0.8467 | SRCC = 0.8429 ]
+INFO:root:Average loss=0.18197163022481477
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2687 | LCC = 0.8494 | SRCC = 0.8450 ]
+INFO:root:Average loss=0.1478666412906769
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2601 | LCC = 0.8513 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.14941895953737772
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2489 | LCC = 0.8533 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.13767725334335595
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2900 | LCC = 0.8448 | SRCC = 0.8416 ]
+INFO:root:Average loss=0.13754436383262658
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3177 | LCC = 0.8446 | SRCC = 0.8434 ]
+INFO:root:Average loss=0.1271607321806443
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2490 | LCC = 0.8486 | SRCC = 0.8441 ]
+INFO:root:Average loss=0.1191246227767223
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2593 | LCC = 0.8441 | SRCC = 0.8385 ]
+INFO:root:Average loss=0.10927421742906937
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2695 | LCC = 0.8449 | SRCC = 0.8427 ]
+INFO:root:Average loss=0.11902685014483257
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2464 | LCC = 0.8456 | SRCC = 0.8418 ]
+INFO:root:Average loss=0.09744060244888832
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2730 | LCC = 0.8420 | SRCC = 0.8390 ]
+INFO:root:Average loss=0.10273259190412667
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2772 | LCC = 0.8404 | SRCC = 0.8338 ]
+INFO:root:Average loss=0.08968925829499196
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2596 | LCC = 0.8411 | SRCC = 0.8360 ]
+INFO:root:Average loss=0.09129423886919633
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2654 | LCC = 0.8435 | SRCC = 0.8395 ]
+INFO:root:Average loss=0.09624250194965264
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2559 | LCC = 0.8451 | SRCC = 0.8402 ]
+INFO:root:Average loss=0.08406234876467632
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2565 | LCC = 0.8414 | SRCC = 0.8375 ]
+INFO:root:Average loss=0.07784033261048488
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2833 | LCC = 0.8394 | SRCC = 0.8342 ]
+INFO:root:Average loss=0.09090761711391118
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2617 | LCC = 0.8445 | SRCC = 0.8418 ]
+INFO:root:Average loss=0.07627092855863082
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2519 | LCC = 0.8453 | SRCC = 0.8407 ]
+INFO:root:Average loss=0.0747095642086023
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2494 | LCC = 0.8474 | SRCC = 0.8433 ]
+INFO:root:Average loss=0.0784442809720834
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2760 | LCC = 0.8403 | SRCC = 0.8372 ]
+INFO:root:Average loss=0.06716251516571412
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2855 | LCC = 0.8409 | SRCC = 0.8375 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2585 | LCC = 0.8502 | SRCC = 0.8527 ]
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run4/config.gin b/results_bvcc/w2v2_large_asr_zh/layer8/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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_large_asr_zh/layer8/run4/model.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..98eefcb7e27864d4c320783ec25249234293464a
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2fdabbcfe201732fd9c1b023881aee5ae4615a2dd084915214bfd15c210bf92
+size 3403339
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run4/model_best.pt b/results_bvcc/w2v2_large_asr_zh/layer8/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..eed75a777eeedc47c7998908bc4c7125e594ce06
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/layer8/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:181f362438e13c766fd8ea203549081763647668bd5cb5eb36b386c18265fca4
+size 3405062
diff --git a/results_bvcc/w2v2_large_asr_zh/layer8/run4/train.log b/results_bvcc/w2v2_large_asr_zh/layer8/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..061f5125e448a19d728975236e784fcc164df3c1
--- /dev/null
+++ b/results_bvcc/w2v2_large_asr_zh/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=0.5737339000289257
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3674 | LCC = 0.8022 | SRCC = 0.8050 ]
+INFO:root:Average loss=0.3090967681163397
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3492 | LCC = 0.8235 | SRCC = 0.8276 ]
+INFO:root:Average loss=0.27174573811965114
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3199 | LCC = 0.8376 | SRCC = 0.8356 ]
+INFO:root:Average loss=0.24749373166989058
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2928 | LCC = 0.8404 | SRCC = 0.8373 ]
+INFO:root:Average loss=0.2313535463733551
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2958 | LCC = 0.8417 | SRCC = 0.8399 ]
+INFO:root:Average loss=0.21522232574912217
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2690 | LCC = 0.8473 | SRCC = 0.8435 ]
+INFO:root:Average loss=0.21244685399608734
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2591 | LCC = 0.8515 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.1997464390901419
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2568 | LCC = 0.8498 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.19060121170985392
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2689 | LCC = 0.8542 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.17335190872351328
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3061 | LCC = 0.8496 | SRCC = 0.8464 ]
+INFO:root:Average loss=0.17066822707270965
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3328 | LCC = 0.8442 | SRCC = 0.8430 ]
+INFO:root:Average loss=0.1532439518815432
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2913 | LCC = 0.8417 | SRCC = 0.8345 ]
+INFO:root:Average loss=0.16185226874091688
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2428 | LCC = 0.8550 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.14233214073838332
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2955 | LCC = 0.8495 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.14803360010950994
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2502 | LCC = 0.8504 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.12700983060476106
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3088 | LCC = 0.8465 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.12901545344636992
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3835 | LCC = 0.8390 | SRCC = 0.8375 ]
+INFO:root:Average loss=0.1279229797804967
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2439 | LCC = 0.8535 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.1209788020604696
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3060 | LCC = 0.8528 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.11011154644000225
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2396 | LCC = 0.8542 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.11482790532784584
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3062 | LCC = 0.8511 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.09609596254542851
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2618 | LCC = 0.8468 | SRCC = 0.8424 ]
+INFO:root:Average loss=0.09398430394820678
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2613 | LCC = 0.8492 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.09477801472903827
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2379 | LCC = 0.8538 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.08344364935197891
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2488 | LCC = 0.8500 | SRCC = 0.8466 ]
+INFO:root:Average loss=0.08223314373156963
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2483 | LCC = 0.8554 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.09222152313360801
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2423 | LCC = 0.8546 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.08017368060656083
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2726 | LCC = 0.8526 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.07251809733227277
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2563 | LCC = 0.8459 | SRCC = 0.8395 ]
+INFO:root:Average loss=0.07088689954999165
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2979 | LCC = 0.8466 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.0816816465021708
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2974 | LCC = 0.8472 | SRCC = 0.8435 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2501 | LCC = 0.8474 | SRCC = 0.8485 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer0/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer0/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..db3e1ab45c6144e5a957e9a208dcc97772ef2530
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4c8457aeff9bb22bc01d043f29837a49f2b31658454142f996cee9cf2b5baaa
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6d99289f3b56b6a412c02fce1029addf00873482
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b6add3e71a1bda1910b6a315a878c5c862107df14c8c2806a4f2f378589156f
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer0/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..542039032c5792a97897ca2dc41cf28028faff2d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6251133126326096
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4692 | LCC = 0.6976 | SRCC = 0.7030 ]
+INFO:root:Average loss=0.42585486517502713
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4600 | LCC = 0.7274 | SRCC = 0.7185 ]
+INFO:root:Average loss=0.38827314055882967
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4619 | LCC = 0.7239 | SRCC = 0.7182 ]
+INFO:root:Average loss=0.3725201774101991
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3908 | LCC = 0.7491 | SRCC = 0.7425 ]
+INFO:root:Average loss=0.3615777353063608
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4059 | LCC = 0.7507 | SRCC = 0.7465 ]
+INFO:root:Average loss=0.3672500451405843
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3967 | LCC = 0.7642 | SRCC = 0.7565 ]
+INFO:root:Average loss=0.3485498109307045
+INFO:root:
+[Valid][6][UTT][ MSE = 0.4161 | LCC = 0.7601 | SRCC = 0.7606 ]
+INFO:root:Average loss=0.32498349994421005
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3877 | LCC = 0.7655 | SRCC = 0.7614 ]
+INFO:root:Average loss=0.30702712463262755
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3623 | LCC = 0.7785 | SRCC = 0.7746 ]
+INFO:root:Average loss=0.3383145540570601
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3635 | LCC = 0.7707 | SRCC = 0.7688 ]
+INFO:root:Average loss=0.3051603593123265
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3500 | LCC = 0.7813 | SRCC = 0.7752 ]
+INFO:root:Average loss=0.30404745577237546
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3536 | LCC = 0.7863 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.28370224913725484
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3523 | LCC = 0.7820 | SRCC = 0.7780 ]
+INFO:root:Average loss=0.29721082288485307
+INFO:root:
+[Valid][13][UTT][ MSE = 0.4264 | LCC = 0.7825 | SRCC = 0.7831 ]
+INFO:root:Average loss=0.28791000159123004
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4134 | LCC = 0.7771 | SRCC = 0.7710 ]
+INFO:root:Average loss=0.27314491952076936
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3439 | LCC = 0.7856 | SRCC = 0.7821 ]
+INFO:root:Average loss=0.26088775312289214
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3207 | LCC = 0.7954 | SRCC = 0.7951 ]
+INFO:root:Average loss=0.25688703969503063
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3073 | LCC = 0.7947 | SRCC = 0.7939 ]
+INFO:root:Average loss=0.2526438383337779
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3148 | LCC = 0.7968 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.24960350283445457
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3237 | LCC = 0.7946 | SRCC = 0.7940 ]
+INFO:root:Average loss=0.24955711857630655
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3409 | LCC = 0.7991 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.2544173732017859
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3078 | LCC = 0.7981 | SRCC = 0.7954 ]
+INFO:root:Average loss=0.2366629325044461
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3710 | LCC = 0.8017 | SRCC = 0.8026 ]
+INFO:root:Average loss=0.2301323862794118
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3576 | LCC = 0.7999 | SRCC = 0.7965 ]
+INFO:root:Average loss=0.22272698160929558
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3242 | LCC = 0.8080 | SRCC = 0.8052 ]
+INFO:root:Average loss=0.2262789398813859
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3262 | LCC = 0.8081 | SRCC = 0.8068 ]
+INFO:root:Average loss=0.221257731031913
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3631 | LCC = 0.8002 | SRCC = 0.8021 ]
+INFO:root:Average loss=0.2105725843172807
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3429 | LCC = 0.8015 | SRCC = 0.7994 ]
+INFO:root:Average loss=0.2295877251487512
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3338 | LCC = 0.8097 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.1986099049831048
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3041 | LCC = 0.8121 | SRCC = 0.8110 ]
+INFO:root:Average loss=0.20230114039702293
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2987 | LCC = 0.8123 | SRCC = 0.8103 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2831 | LCC = 0.8349 | SRCC = 0.8359 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer0/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer0/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6489128abbb66cb4f287b2bf543c0ea5b6b10a2c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b98e2981fd88c2bbccd99f3600658113b98250dd19599cd5d393d91059e1ebe1
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..76c217bdca71bacca39a3b7a32bd96e8afe4ab3c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:192bb8cdf861a1bba2f0674b81801ee636aff9214aeb3c1dda3736228a856cd2
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer0/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..28a8a7552f93b3002830637d49a27ae08a2c73eb
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6421047659256519
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4719 | LCC = 0.6904 | SRCC = 0.6908 ]
+INFO:root:Average loss=0.41013930145746624
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4388 | LCC = 0.7190 | SRCC = 0.7119 ]
+INFO:root:Average loss=0.4018075303771557
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3930 | LCC = 0.7294 | SRCC = 0.7297 ]
+INFO:root:Average loss=0.35960478870532453
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3794 | LCC = 0.7494 | SRCC = 0.7440 ]
+INFO:root:Average loss=0.3644473789594112
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3931 | LCC = 0.7548 | SRCC = 0.7507 ]
+INFO:root:Average loss=0.3570439941607989
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3875 | LCC = 0.7632 | SRCC = 0.7576 ]
+INFO:root:Average loss=0.33671483130027086
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3693 | LCC = 0.7668 | SRCC = 0.7603 ]
+INFO:root:Average loss=0.3265153788603269
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3597 | LCC = 0.7695 | SRCC = 0.7642 ]
+INFO:root:Average loss=0.31214276395547086
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3783 | LCC = 0.7727 | SRCC = 0.7663 ]
+INFO:root:Average loss=0.30584143274105513
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3940 | LCC = 0.7798 | SRCC = 0.7755 ]
+INFO:root:Average loss=0.3116814885766078
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3296 | LCC = 0.7813 | SRCC = 0.7774 ]
+INFO:root:Average loss=0.29946465694751495
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3750 | LCC = 0.7853 | SRCC = 0.7845 ]
+INFO:root:Average loss=0.3036601503308003
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3819 | LCC = 0.7882 | SRCC = 0.7881 ]
+INFO:root:Average loss=0.3082962420124274
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3298 | LCC = 0.7878 | SRCC = 0.7859 ]
+INFO:root:Average loss=0.2826255684097608
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3241 | LCC = 0.7899 | SRCC = 0.7849 ]
+INFO:root:Average loss=0.26369154300445163
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3308 | LCC = 0.7915 | SRCC = 0.7876 ]
+INFO:root:Average loss=0.2663578312748518
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3769 | LCC = 0.7912 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.2712905152868002
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3354 | LCC = 0.7962 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.28724452222769076
+INFO:root:
+[Valid][18][UTT][ MSE = 0.4115 | LCC = 0.7847 | SRCC = 0.7884 ]
+INFO:root:Average loss=0.28145511008990115
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3595 | LCC = 0.7945 | SRCC = 0.7943 ]
+INFO:root:Average loss=0.24281845814906633
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3159 | LCC = 0.7993 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.23209569354852042
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3546 | LCC = 0.7983 | SRCC = 0.8002 ]
+INFO:root:Average loss=0.23264503135130957
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3128 | LCC = 0.8060 | SRCC = 0.8029 ]
+INFO:root:Average loss=0.2332152227560679
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3136 | LCC = 0.8077 | SRCC = 0.8069 ]
+INFO:root:Average loss=0.22181645799905825
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3134 | LCC = 0.8023 | SRCC = 0.8024 ]
+INFO:root:Average loss=0.22245705557557252
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2925 | LCC = 0.8091 | SRCC = 0.8066 ]
+INFO:root:Average loss=0.2137207425175569
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3046 | LCC = 0.8026 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.1981793483480429
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3003 | LCC = 0.8101 | SRCC = 0.8103 ]
+INFO:root:Average loss=0.2038805422683557
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2920 | LCC = 0.8144 | SRCC = 0.8130 ]
+INFO:root:Average loss=0.20219833938739237
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3141 | LCC = 0.7990 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.20545620413926932
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2988 | LCC = 0.8082 | SRCC = 0.8050 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2640 | LCC = 0.8345 | SRCC = 0.8324 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer0/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer0/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..90e259679f264825c15c89d1bf7c500b77fd9527
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06c8624c91a72e323dd01ee2799f5e8c0cef015a03444bf11327b49bb66ad290
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2bbc34ef49183a5b6c4db284456cf329edbe3f67
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0d10bc4eefb7987bcbf63d5907f5c2ecbf20e9c2ca830b88dba1e4d54acb99c
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer0/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..50cb7d7cbadb423ed28a075e23175028ca55ec2b
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6533797016510596
+INFO:root:
+[Valid][0][UTT][ MSE = 0.5285 | LCC = 0.6900 | SRCC = 0.6885 ]
+INFO:root:Average loss=0.42998539484464204
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4302 | LCC = 0.7160 | SRCC = 0.7152 ]
+INFO:root:Average loss=0.4080772124803983
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4222 | LCC = 0.7263 | SRCC = 0.7211 ]
+INFO:root:Average loss=0.380254885325065
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4245 | LCC = 0.7414 | SRCC = 0.7361 ]
+INFO:root:Average loss=0.35259735851715773
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3979 | LCC = 0.7569 | SRCC = 0.7531 ]
+INFO:root:Average loss=0.3319364885489146
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4375 | LCC = 0.7638 | SRCC = 0.7554 ]
+INFO:root:Average loss=0.3221005875712786
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3986 | LCC = 0.7782 | SRCC = 0.7742 ]
+INFO:root:Average loss=0.3405996939310661
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3604 | LCC = 0.7746 | SRCC = 0.7703 ]
+INFO:root:Average loss=0.3177570125613457
+INFO:root:
+[Valid][8][UTT][ MSE = 0.4251 | LCC = 0.7741 | SRCC = 0.7696 ]
+INFO:root:Average loss=0.30600648449781614
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3656 | LCC = 0.7763 | SRCC = 0.7694 ]
+INFO:root:Average loss=0.3321141077157779
+INFO:root:
+[Valid][10][UTT][ MSE = 0.4059 | LCC = 0.7800 | SRCC = 0.7762 ]
+INFO:root:Average loss=0.29502717902263004
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3284 | LCC = 0.7802 | SRCC = 0.7802 ]
+INFO:root:Average loss=0.2807892904831813
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3636 | LCC = 0.7931 | SRCC = 0.7942 ]
+INFO:root:Average loss=0.28576508985880095
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3273 | LCC = 0.7934 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.2865075149979347
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3578 | LCC = 0.7900 | SRCC = 0.7861 ]
+INFO:root:Average loss=0.27903416466254455
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3349 | LCC = 0.8001 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.26262708734243345
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3252 | LCC = 0.8020 | SRCC = 0.7962 ]
+INFO:root:Average loss=0.2633181625069716
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3205 | LCC = 0.8068 | SRCC = 0.8066 ]
+INFO:root:Average loss=0.25813736136143023
+INFO:root:
+[Valid][18][UTT][ MSE = 0.4375 | LCC = 0.7965 | SRCC = 0.7950 ]
+INFO:root:Average loss=0.2638626060424707
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3061 | LCC = 0.8003 | SRCC = 0.8007 ]
+INFO:root:Average loss=0.23916902660559386
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3155 | LCC = 0.8058 | SRCC = 0.8067 ]
+INFO:root:Average loss=0.25079347078616804
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3606 | LCC = 0.8090 | SRCC = 0.8056 ]
+INFO:root:Average loss=0.22747960361914757
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3137 | LCC = 0.8106 | SRCC = 0.8084 ]
+INFO:root:Average loss=0.22817132965876505
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3173 | LCC = 0.8119 | SRCC = 0.8074 ]
+INFO:root:Average loss=0.22042992214361826
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2992 | LCC = 0.8171 | SRCC = 0.8143 ]
+INFO:root:Average loss=0.2607633160092892
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3304 | LCC = 0.8117 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.2207375543239789
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2987 | LCC = 0.8160 | SRCC = 0.8133 ]
+INFO:root:Average loss=0.20507909070986968
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3879 | LCC = 0.8051 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.2035784901907811
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3116 | LCC = 0.8097 | SRCC = 0.8081 ]
+INFO:root:Average loss=0.21262204045286545
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3418 | LCC = 0.8151 | SRCC = 0.8184 ]
+INFO:root:Average loss=0.20951864008720106
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2904 | LCC = 0.8199 | SRCC = 0.8191 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2781 | LCC = 0.8287 | SRCC = 0.8260 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer0/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer0/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4c9e0d9435308b7708d35ba1e3e68ad5a97bf3ee
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c21be92d7a95e7dabb365d9f543f0fd9df87aa785dd55a291262ff08c9f3df17
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3c062c9d3cb2c1fea247d91444487d36a312579d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3673a79c84b4370ff4763bb2ffcf40d423b602f0dd8624f5ac5afa948c7ad067
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer0/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..e8cc28fd90bbbe2b3c0f913c36b642c748b71331
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.685496585491376
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4820 | LCC = 0.6831 | SRCC = 0.6869 ]
+INFO:root:Average loss=0.4273303850339009
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4453 | LCC = 0.7120 | SRCC = 0.7062 ]
+INFO:root:Average loss=0.40624534835418064
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4075 | LCC = 0.7265 | SRCC = 0.7256 ]
+INFO:root:Average loss=0.3765263964350407
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4038 | LCC = 0.7361 | SRCC = 0.7335 ]
+INFO:root:Average loss=0.35444220976951796
+INFO:root:
+[Valid][4][UTT][ MSE = 0.4051 | LCC = 0.7518 | SRCC = 0.7443 ]
+INFO:root:Average loss=0.35024357319642335
+INFO:root:
+[Valid][5][UTT][ MSE = 0.4157 | LCC = 0.7577 | SRCC = 0.7551 ]
+INFO:root:Average loss=0.3365925842752823
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3709 | LCC = 0.7572 | SRCC = 0.7567 ]
+INFO:root:Average loss=0.3237278641034395
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3737 | LCC = 0.7705 | SRCC = 0.7647 ]
+INFO:root:Average loss=0.306555358454203
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3582 | LCC = 0.7704 | SRCC = 0.7664 ]
+INFO:root:Average loss=0.299364509108739
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3431 | LCC = 0.7761 | SRCC = 0.7710 ]
+INFO:root:Average loss=0.31766079251582807
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3562 | LCC = 0.7786 | SRCC = 0.7768 ]
+INFO:root:Average loss=0.29120394320060045
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3418 | LCC = 0.7863 | SRCC = 0.7818 ]
+INFO:root:Average loss=0.27853957315286
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3194 | LCC = 0.7862 | SRCC = 0.7820 ]
+INFO:root:Average loss=0.29143356302609813
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3958 | LCC = 0.7740 | SRCC = 0.7704 ]
+INFO:root:Average loss=0.28352353397088176
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4087 | LCC = 0.7788 | SRCC = 0.7798 ]
+INFO:root:Average loss=0.26411342811890137
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3516 | LCC = 0.7915 | SRCC = 0.7895 ]
+INFO:root:Average loss=0.2728886952002843
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3194 | LCC = 0.7933 | SRCC = 0.7916 ]
+INFO:root:Average loss=0.2522257102223543
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3433 | LCC = 0.8008 | SRCC = 0.7976 ]
+INFO:root:Average loss=0.2454447960242247
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3169 | LCC = 0.7968 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.24128585652663156
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3170 | LCC = 0.7892 | SRCC = 0.7862 ]
+INFO:root:Average loss=0.2634295702744753
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3047 | LCC = 0.8001 | SRCC = 0.7970 ]
+INFO:root:Average loss=0.2440574858815242
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3295 | LCC = 0.7949 | SRCC = 0.7908 ]
+INFO:root:Average loss=0.23554500670005113
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3489 | LCC = 0.8047 | SRCC = 0.8046 ]
+INFO:root:Average loss=0.24840768579489145
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3125 | LCC = 0.8012 | SRCC = 0.8019 ]
+INFO:root:Average loss=0.21986533644107673
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3122 | LCC = 0.8019 | SRCC = 0.7973 ]
+INFO:root:Average loss=0.22942414440405676
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3316 | LCC = 0.8095 | SRCC = 0.8076 ]
+INFO:root:Average loss=0.21187829551024315
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3037 | LCC = 0.8100 | SRCC = 0.8088 ]
+INFO:root:Average loss=0.2206225527020601
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2958 | LCC = 0.8085 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.195725564009104
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2963 | LCC = 0.8113 | SRCC = 0.8099 ]
+INFO:root:Average loss=0.20883784137475184
+INFO:root:
+[Valid][29][UTT][ MSE = 0.3228 | LCC = 0.8110 | SRCC = 0.8100 ]
+INFO:root:Average loss=0.2019476198997253
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3779 | LCC = 0.8073 | SRCC = 0.8085 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2731 | LCC = 0.8379 | SRCC = 0.8347 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer0/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer0/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1679550a95a2c368c879ed7265721520a6ce0090
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46509956b831f0ccc18a6bff2f38012f029ddaebfe9f0f24a854b8f9ece3d921
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer0/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f16085659b29e522f2e605252d008cab19d89d0b
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer0/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83c38ba659fd0f56dc79fdd30053d1c801afc179088ad61fdf1eac7bf963a272
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer0/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer0/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..430e5846c9a3842966fb0057a13392f82c39c7de
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6269668493515406
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4907 | LCC = 0.6881 | SRCC = 0.6885 ]
+INFO:root:Average loss=0.4315877224390323
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4434 | LCC = 0.7118 | SRCC = 0.7087 ]
+INFO:root:Average loss=0.39771880973608065
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4300 | LCC = 0.7234 | SRCC = 0.7264 ]
+INFO:root:Average loss=0.3980563741463881
+INFO:root:
+[Valid][3][UTT][ MSE = 0.4357 | LCC = 0.7377 | SRCC = 0.7337 ]
+INFO:root:Average loss=0.346786804879323
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3899 | LCC = 0.7493 | SRCC = 0.7429 ]
+INFO:root:Average loss=0.3411435446678064
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3749 | LCC = 0.7501 | SRCC = 0.7435 ]
+INFO:root:Average loss=0.32882853654714733
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3611 | LCC = 0.7648 | SRCC = 0.7593 ]
+INFO:root:Average loss=0.3251516750225654
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3701 | LCC = 0.7596 | SRCC = 0.7622 ]
+INFO:root:Average loss=0.32645231351638454
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3627 | LCC = 0.7595 | SRCC = 0.7506 ]
+INFO:root:Average loss=0.3220098266998927
+INFO:root:
+[Valid][9][UTT][ MSE = 0.3573 | LCC = 0.7686 | SRCC = 0.7659 ]
+INFO:root:Average loss=0.2978205151664905
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3613 | LCC = 0.7770 | SRCC = 0.7678 ]
+INFO:root:Average loss=0.29106219991659504
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3347 | LCC = 0.7829 | SRCC = 0.7753 ]
+INFO:root:Average loss=0.28500003730639434
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3537 | LCC = 0.7876 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.2868137665283986
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3471 | LCC = 0.7814 | SRCC = 0.7771 ]
+INFO:root:Average loss=0.2815840855622903
+INFO:root:
+[Valid][14][UTT][ MSE = 0.4633 | LCC = 0.7681 | SRCC = 0.7657 ]
+INFO:root:Average loss=0.27716456028895503
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3287 | LCC = 0.7913 | SRCC = 0.7883 ]
+INFO:root:Average loss=0.2635840557706662
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3156 | LCC = 0.7926 | SRCC = 0.7891 ]
+INFO:root:Average loss=0.2642155912442085
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3701 | LCC = 0.7833 | SRCC = 0.7790 ]
+INFO:root:Average loss=0.2631396758250701
+INFO:root:
+[Valid][18][UTT][ MSE = 0.3583 | LCC = 0.7826 | SRCC = 0.7771 ]
+INFO:root:Average loss=0.2511937811206549
+INFO:root:
+[Valid][19][UTT][ MSE = 0.3316 | LCC = 0.7872 | SRCC = 0.7856 ]
+INFO:root:Average loss=0.2880649925806584
+INFO:root:
+[Valid][20][UTT][ MSE = 0.3234 | LCC = 0.7893 | SRCC = 0.7876 ]
+INFO:root:Average loss=0.25893506751610684
+INFO:root:
+[Valid][21][UTT][ MSE = 0.3235 | LCC = 0.7936 | SRCC = 0.7935 ]
+INFO:root:Average loss=0.2261346138249605
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3175 | LCC = 0.7962 | SRCC = 0.7920 ]
+INFO:root:Average loss=0.2333993048239977
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3231 | LCC = 0.7991 | SRCC = 0.7951 ]
+INFO:root:Average loss=0.22010634018060488
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3502 | LCC = 0.7937 | SRCC = 0.7921 ]
+INFO:root:Average loss=0.2253642657246345
+INFO:root:
+[Valid][25][UTT][ MSE = 0.3061 | LCC = 0.8020 | SRCC = 0.7981 ]
+INFO:root:Average loss=0.23142349977905935
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3219 | LCC = 0.7953 | SRCC = 0.7953 ]
+INFO:root:Average loss=0.20765936699433205
+INFO:root:
+[Valid][27][UTT][ MSE = 0.3241 | LCC = 0.8031 | SRCC = 0.8015 ]
+INFO:root:Average loss=0.20353240156785035
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2935 | LCC = 0.8043 | SRCC = 0.8030 ]
+INFO:root:Average loss=0.20294240422737905
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2975 | LCC = 0.8060 | SRCC = 0.8028 ]
+INFO:root:Average loss=0.2302976388197679
+INFO:root:
+[Valid][30][UTT][ MSE = 0.3052 | LCC = 0.8061 | SRCC = 0.8025 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.3005 | LCC = 0.8274 | SRCC = 0.8271 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer10/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer10/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0f1a12e248d534a2247f3215243f183931d94829
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9bae05c0ab00b75f4deba5fb5d765ecfeb672137c4153e96bff1f03c04c73b30
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..73ee7b621d749839051984eecdae0e52232fdc01
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e03150056a01cf4eaf5982ecd9bcd5973576993994af79b1b22fa56230280b4
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer10/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..7a74921221bba231d057565a90407c676e30b7b7
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6087003210798289
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3612 | LCC = 0.7974 | SRCC = 0.8039 ]
+INFO:root:Average loss=0.30045762256934094
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3026 | LCC = 0.8393 | SRCC = 0.8403 ]
+INFO:root:Average loss=0.26430882055025834
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3293 | LCC = 0.8448 | SRCC = 0.8456 ]
+INFO:root:Average loss=0.23864195400323623
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2468 | LCC = 0.8553 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.21530131097787467
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2487 | LCC = 0.8598 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.2143582224081724
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2606 | LCC = 0.8606 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.202375003351615
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2505 | LCC = 0.8632 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.18683583661913872
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2394 | LCC = 0.8648 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.18269292609049723
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2549 | LCC = 0.8628 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.16896921501327783
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2454 | LCC = 0.8658 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.1883069933989109
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2777 | LCC = 0.8596 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.17051062303093764
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2219 | LCC = 0.8668 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.16055597536838973
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2387 | LCC = 0.8673 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.14131035598424765
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2388 | LCC = 0.8649 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.13768170659358686
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2459 | LCC = 0.8635 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.13259852362366822
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2569 | LCC = 0.8623 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.12454598559400974
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2274 | LCC = 0.8634 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.12809184461067885
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2273 | LCC = 0.8651 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.10798022160545373
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2259 | LCC = 0.8639 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.11710841495257157
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2358 | LCC = 0.8641 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.10620607359287067
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2452 | LCC = 0.8590 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.11046675095955531
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2399 | LCC = 0.8614 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.09979210115778141
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2284 | LCC = 0.8624 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.09161472296676575
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2335 | LCC = 0.8602 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.08733070837572598
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2195 | LCC = 0.8601 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.08106572161882351
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2406 | LCC = 0.8608 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.08075310920293514
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2277 | LCC = 0.8606 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.08081225998317584
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2506 | LCC = 0.8576 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.07918853193330459
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2722 | LCC = 0.8602 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.07464820977586967
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2238 | LCC = 0.8621 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.0730158884364825
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2275 | LCC = 0.8580 | SRCC = 0.8562 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2358 | LCC = 0.8642 | SRCC = 0.8647 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer10/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer10/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d5d00bee5d19b1cf5932e3852e7c7eaa35425c3b
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:229ab99325972167cf25a3c2b8ed61535d88685a875d6c728f4fcb8131128a46
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..734d552f19b20f036969adc6986067d3021fcdee
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b78a7d5ad5dba26a81a8f792a7c01f66ea3345dbcf42d5becef48b479506bee
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer10/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..1a7ee33b21b2881b2519f3d08cc8e82f33facc1c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.5177691884529896
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3392 | LCC = 0.8164 | SRCC = 0.8190 ]
+INFO:root:Average loss=0.3038797974586487
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3069 | LCC = 0.8390 | SRCC = 0.8373 ]
+INFO:root:Average loss=0.2706706961378073
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2683 | LCC = 0.8481 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.26055509329606324
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2727 | LCC = 0.8506 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.21807880069200808
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2941 | LCC = 0.8574 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.2074204013706782
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2547 | LCC = 0.8610 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.2111218159015362
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2704 | LCC = 0.8605 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.19447844456403685
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2580 | LCC = 0.8615 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.18151671850146392
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2320 | LCC = 0.8617 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.17392347953640497
+INFO:root:
+[Valid][9][UTT][ MSE = 0.4327 | LCC = 0.8405 | SRCC = 0.8483 ]
+INFO:root:Average loss=0.18709646853116843
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2819 | LCC = 0.8593 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.15929897903249815
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2206 | LCC = 0.8626 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.15440436863364318
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2315 | LCC = 0.8672 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.1447365949742305
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2259 | LCC = 0.8649 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.1376891385477323
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2336 | LCC = 0.8643 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.1366729157475325
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2273 | LCC = 0.8666 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.11982852249191357
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2436 | LCC = 0.8595 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.12278676052124073
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2213 | LCC = 0.8640 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.11325088219764905
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2478 | LCC = 0.8619 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.11166794187365434
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2254 | LCC = 0.8623 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.10730319794936058
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2645 | LCC = 0.8598 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.10291515906842855
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2944 | LCC = 0.8590 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.1139884067651553
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2185 | LCC = 0.8636 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.09685007592615409
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2254 | LCC = 0.8606 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.08788243991633256
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2274 | LCC = 0.8599 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.08319355967717293
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2232 | LCC = 0.8612 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.08703215386814032
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2292 | LCC = 0.8619 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.07752011988598567
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2226 | LCC = 0.8592 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.07347340265718791
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2311 | LCC = 0.8595 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.07107782960893252
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2353 | LCC = 0.8595 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.07445369140268901
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2320 | LCC = 0.8602 | SRCC = 0.8592 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2375 | LCC = 0.8656 | SRCC = 0.8704 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer10/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer10/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..daab3d003b1752e403ed3b667a47b28ecf5ab813
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ab516f94375544049be49fff86216cef7272c4e1be8a0b29c63d7ed52edffd4
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..fadaa43933920d4cbb18d85929f49bd70701fa53
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4edb788e6e591b9fbb83bdc4258d20b7f703379b2fa6922ee2074d43bff6cb11
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer10/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..a0688f511c6beb5f7667a0bb2cc3fd6e2e42f966
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.538689170892422
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3525 | LCC = 0.8041 | SRCC = 0.8086 ]
+INFO:root:Average loss=0.3051311341233743
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3489 | LCC = 0.8347 | SRCC = 0.8366 ]
+INFO:root:Average loss=0.26011367103992367
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2693 | LCC = 0.8489 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.24588327319958272
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2505 | LCC = 0.8573 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.2248967985312144
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2507 | LCC = 0.8573 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.19973939485274828
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2376 | LCC = 0.8626 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.19752497608080888
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2287 | LCC = 0.8648 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.19078645052818152
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2619 | LCC = 0.8621 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.1855845237389589
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2335 | LCC = 0.8621 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.1953767542846692
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2267 | LCC = 0.8651 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.17205945698496622
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2404 | LCC = 0.8650 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.16698153909200278
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2386 | LCC = 0.8659 | SRCC = 0.8648 ]
+INFO:root:Average loss=0.16458932797496134
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2381 | LCC = 0.8640 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.14916733423104653
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2346 | LCC = 0.8643 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.14921947654623252
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2984 | LCC = 0.8621 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.14445078506683692
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2708 | LCC = 0.8634 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.12925017749269804
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2266 | LCC = 0.8628 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.12408966886309478
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2261 | LCC = 0.8648 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.11867862061048165
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2158 | LCC = 0.8672 | SRCC = 0.8651 ]
+INFO:root:Average loss=0.11106271994037506
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2395 | LCC = 0.8636 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.10965577073586293
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2378 | LCC = 0.8597 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.10531614357844377
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2464 | LCC = 0.8626 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.09833339415490627
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2257 | LCC = 0.8619 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.10052524941662948
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3019 | LCC = 0.8555 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.09129322635439727
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2456 | LCC = 0.8635 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.08559188614480007
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2297 | LCC = 0.8614 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.08353474411444786
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2760 | LCC = 0.8588 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.08713466401856679
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2839 | LCC = 0.8544 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.07715122893643685
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2289 | LCC = 0.8582 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.07364914351357864
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2294 | LCC = 0.8613 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.07828734041406558
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2235 | LCC = 0.8599 | SRCC = 0.8567 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2217 | LCC = 0.8656 | SRCC = 0.8666 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer10/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer10/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..db6e8c018572b43e990aa102f607d523ff4fce71
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bf8fddce51a1a96d862d539fa1ba2a60d55f482a82fe984c8d881f11a03a81c
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..68beaa82c930c8e19142efac6a467530d361494c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9a74c1679aeb2ef2f9508e6ab2e1d053dcd350b95836e72526f88784c176ba8
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer10/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..da7d00638bcc43795d9ae8b40f2d7e9cf20a9da2
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6435123280836985
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3917 | LCC = 0.7860 | SRCC = 0.7898 ]
+INFO:root:Average loss=0.3103034475292915
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3045 | LCC = 0.8301 | SRCC = 0.8339 ]
+INFO:root:Average loss=0.2631254698603581
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2679 | LCC = 0.8450 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.23954157798718184
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2675 | LCC = 0.8542 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.2110922519977276
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2566 | LCC = 0.8587 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.20669457736687782
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2621 | LCC = 0.8599 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.1999117449308053
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3189 | LCC = 0.8609 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.20302856207276002
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2805 | LCC = 0.8601 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.17857924380745643
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3595 | LCC = 0.8582 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.16849845533187574
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2454 | LCC = 0.8629 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.16174749619303605
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2524 | LCC = 0.8628 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.16459400483813041
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2434 | LCC = 0.8657 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.14577002947529158
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2602 | LCC = 0.8630 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.15094528442774063
+INFO:root:
+[Valid][13][UTT][ MSE = 0.3051 | LCC = 0.8646 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.1465405417749515
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2336 | LCC = 0.8644 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.12908968377189758
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2547 | LCC = 0.8615 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.12291118512168908
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2190 | LCC = 0.8650 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.11378679376764175
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2255 | LCC = 0.8622 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.11032975999972759
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2161 | LCC = 0.8654 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.11351344591149917
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2991 | LCC = 0.8617 | SRCC = 0.8621 ]
+INFO:root:Average loss=0.09957222267985344
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2724 | LCC = 0.8590 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.11326485136762643
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2448 | LCC = 0.8625 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.08993005700027332
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2362 | LCC = 0.8577 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.09304508528648278
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2299 | LCC = 0.8613 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.08715366681989951
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2364 | LCC = 0.8595 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.07969591374962758
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2228 | LCC = 0.8605 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.07776981878739136
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2228 | LCC = 0.8618 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.0707411350061496
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2403 | LCC = 0.8581 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.07504627137229992
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2280 | LCC = 0.8604 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.06763677447079083
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2346 | LCC = 0.8593 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.06371279132480805
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2244 | LCC = 0.8562 | SRCC = 0.8548 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2242 | LCC = 0.8690 | SRCC = 0.8711 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer10/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer10/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6044931523b269bd14018ff27e75c108dc43fd83
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5568fc23091483141ba5aac2c2141d9165ac6d6fb9a5ce4b9265ca4670f9de3
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer10/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9f9b5f4d1b5897bb5acdf72fffa591050c63b9be
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer10/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36f718e1028f70bd597d8dcb2f0aeb4b423ab5ef7dc828624ad06fa9b47ad6a7
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer10/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer10/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b695bc5c3877bafbe83715acd58b0c9fb0285052
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.5297759475234227
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3238 | LCC = 0.8040 | SRCC = 0.8123 ]
+INFO:root:Average loss=0.2979575107112909
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3077 | LCC = 0.8335 | SRCC = 0.8354 ]
+INFO:root:Average loss=0.2559567854190484
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2957 | LCC = 0.8436 | SRCC = 0.8425 ]
+INFO:root:Average loss=0.23543176093162635
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2759 | LCC = 0.8536 | SRCC = 0.8531 ]
+INFO:root:Average loss=0.23215899196190712
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2735 | LCC = 0.8527 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.21974438925584158
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2579 | LCC = 0.8573 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.20949960691042435
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2613 | LCC = 0.8592 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.19370769317715597
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2723 | LCC = 0.8594 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.1813895324101815
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2558 | LCC = 0.8576 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.17558160280952087
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2344 | LCC = 0.8588 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.17421745776365966
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2856 | LCC = 0.8519 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.1741144175712879
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2357 | LCC = 0.8573 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.1646432648293483
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2621 | LCC = 0.8551 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.1433298417773002
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2475 | LCC = 0.8594 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.14616038182224983
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2184 | LCC = 0.8596 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.13068531625545943
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2713 | LCC = 0.8562 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.1363635758558909
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2348 | LCC = 0.8615 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.12194129413901231
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2586 | LCC = 0.8590 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.1297381536509746
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2474 | LCC = 0.8551 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.11593807507783939
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2358 | LCC = 0.8592 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.1156122322456959
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2437 | LCC = 0.8550 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.11102591378566547
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2462 | LCC = 0.8547 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.10148917205440693
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2475 | LCC = 0.8578 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.09813321353151248
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2567 | LCC = 0.8582 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.09190538625877637
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2332 | LCC = 0.8586 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.08512028311498654
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2367 | LCC = 0.8523 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.08457158099955474
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2816 | LCC = 0.8490 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.08239038216953094
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2313 | LCC = 0.8546 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.07644519214637768
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2289 | LCC = 0.8571 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.08933828551417743
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2328 | LCC = 0.8543 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.06884586686889331
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2550 | LCC = 0.8540 | SRCC = 0.8546 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2350 | LCC = 0.8703 | SRCC = 0.8710 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer12/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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_large_xlsr53/layer12/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f9f2ec62f698164db8dbbcf38ae63b7d1ba708c2
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1ce385a504530d58567f196059ddc8455da10e340c1cb58d6d2036978aeb2a0
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..cb3dc86bf0c6591814c3c858b4deda70aa51ecfc
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c35e68ac154aeab3eca2714fc0006ce01c13f05532b36f02649e2c4819fe2538
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer12/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..bb0d7c612b77e3d7d4d2858bec25c2a37eb0ce8f
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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=0.5229130576436336
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3385 | LCC = 0.8108 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.3049738061351654
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2771 | LCC = 0.8339 | SRCC = 0.8349 ]
+INFO:root:Average loss=0.2734187153669504
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2995 | LCC = 0.8462 | SRCC = 0.8461 ]
+INFO:root:Average loss=0.24464780168655592
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2819 | LCC = 0.8515 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.2306946968803039
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2586 | LCC = 0.8569 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.22100988049537706
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3033 | LCC = 0.8529 | SRCC = 0.8479 ]
+INFO:root:Average loss=0.2068831918713374
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2748 | LCC = 0.8586 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.189126175087996
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2623 | LCC = 0.8537 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.1871159344147413
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2744 | LCC = 0.8610 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.17984647466204104
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2730 | LCC = 0.8582 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.1701448427943083
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2669 | LCC = 0.8576 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.15647208127073753
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2569 | LCC = 0.8584 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.15343748644376412
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2416 | LCC = 0.8595 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.1381178924288505
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2649 | LCC = 0.8543 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.14945160062649313
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2593 | LCC = 0.8519 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.14208885664359117
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2758 | LCC = 0.8525 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.12546119504632094
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2754 | LCC = 0.8491 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.11718843753139178
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2508 | LCC = 0.8522 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.12357078989346822
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2622 | LCC = 0.8539 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.10993560078816536
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2534 | LCC = 0.8556 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.10392669139382167
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2698 | LCC = 0.8512 | SRCC = 0.8487 ]
+INFO:root:Average loss=0.11394888644990249
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2900 | LCC = 0.8550 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.10237611142488626
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2568 | LCC = 0.8511 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.08736337587619439
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2513 | LCC = 0.8518 | SRCC = 0.8487 ]
+INFO:root:Average loss=0.08731702619638199
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2421 | LCC = 0.8530 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.08192463042453313
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2615 | LCC = 0.8555 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.0924319176433178
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2525 | LCC = 0.8552 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.08029843088334952
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2352 | LCC = 0.8535 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.07636897175166851
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2354 | LCC = 0.8549 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.07023321440777718
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2510 | LCC = 0.8524 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.06738632740691686
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2421 | LCC = 0.8545 | SRCC = 0.8527 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2742 | LCC = 0.8632 | SRCC = 0.8660 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer12/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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_large_xlsr53/layer12/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2516747edc3ae156185446ad74679886e4826bfd
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b5ca264e77ccf13a6bea969e9173d84cdb3f8178da593b66ec04f4ae5603439
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..885033a07ef9db6e144c65a637753bf18dcb3a0f
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00f55adcbf856f4769473143a10cc6ce5509d454a2e18d892c0ba64a38b9c624
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer12/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..fbea17cbc9803714241c2024cdee4a4467702461
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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=0.5415312540836823
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3529 | LCC = 0.8034 | SRCC = 0.8060 ]
+INFO:root:Average loss=0.3077265509428122
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3090 | LCC = 0.8321 | SRCC = 0.8330 ]
+INFO:root:Average loss=0.27685530407306475
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2773 | LCC = 0.8405 | SRCC = 0.8397 ]
+INFO:root:Average loss=0.23760040429158089
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2465 | LCC = 0.8490 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.2252227301016832
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2740 | LCC = 0.8485 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.21623763690392175
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3365 | LCC = 0.8440 | SRCC = 0.8472 ]
+INFO:root:Average loss=0.21610047715978745
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2769 | LCC = 0.8531 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.20660209617553613
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2473 | LCC = 0.8555 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.1819861861757743
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2483 | LCC = 0.8579 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.1833928072681794
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2530 | LCC = 0.8544 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.1700728555711416
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2503 | LCC = 0.8575 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.162475273395196
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2414 | LCC = 0.8576 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.15906749092615569
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2434 | LCC = 0.8598 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.13660780904002678
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2422 | LCC = 0.8526 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.13791070544184783
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2441 | LCC = 0.8537 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.13387139810201448
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2379 | LCC = 0.8569 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.13290187640067858
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2513 | LCC = 0.8481 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.1161755709311901
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2377 | LCC = 0.8527 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.13057350071194845
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2378 | LCC = 0.8572 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.1148239258581247
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2366 | LCC = 0.8552 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.1063500210069693
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2469 | LCC = 0.8521 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.10772804209055045
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2427 | LCC = 0.8529 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.09499864294551887
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2707 | LCC = 0.8481 | SRCC = 0.8460 ]
+INFO:root:Average loss=0.0945633509888863
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2472 | LCC = 0.8523 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.0821206549134774
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2521 | LCC = 0.8471 | SRCC = 0.8445 ]
+INFO:root:Average loss=0.08611181235084167
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2795 | LCC = 0.8550 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.08950534878441921
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2525 | LCC = 0.8507 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.095865134579631
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2492 | LCC = 0.8506 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.09366114800557113
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2751 | LCC = 0.8578 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.07821511587080283
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2428 | LCC = 0.8524 | SRCC = 0.8508 ]
+INFO:root:Average loss=0.06598333957103583
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2490 | LCC = 0.8486 | SRCC = 0.8485 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2624 | LCC = 0.8636 | SRCC = 0.8680 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer12/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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_large_xlsr53/layer12/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..95f313caf977a2646e81c3d9044072417ebe528c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e150cfcbca8cc1d440657841c0e711de7aefc591255143e0bf2590b7fe288d00
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..754f851b25a9aa2bdd3b68e9dbd7387cb090f4e4
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:970b6b7caaccd126e9b69244e8f6f12ad7f6339725ae62ad38944fc66ff49cb7
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer12/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..6802f99d23f9887f1bac50610cf9cd4198bbe65e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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=0.5083920585039334
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3529 | LCC = 0.8034 | SRCC = 0.8116 ]
+INFO:root:Average loss=0.3017737959057857
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2667 | LCC = 0.8401 | SRCC = 0.8413 ]
+INFO:root:Average loss=0.29146131108968687
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2738 | LCC = 0.8500 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.25746616300864095
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3012 | LCC = 0.8546 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.23187206914791694
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3498 | LCC = 0.8473 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.21723926812410355
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2548 | LCC = 0.8572 | SRCC = 0.8543 ]
+INFO:root:Average loss=0.20451157196209982
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2589 | LCC = 0.8598 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.19330659327216637
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2638 | LCC = 0.8582 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.18109489881839508
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2744 | LCC = 0.8593 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.17821004308569124
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2479 | LCC = 0.8603 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.16268849325103638
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2357 | LCC = 0.8591 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.16069675084107962
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2517 | LCC = 0.8593 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.1546083617095764
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2463 | LCC = 0.8582 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.15534748509526253
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2745 | LCC = 0.8587 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.15243857592726365
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2671 | LCC = 0.8601 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.14588268129871443
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2249 | LCC = 0.8607 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.14212337413277382
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2443 | LCC = 0.8582 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.12423176891528644
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2412 | LCC = 0.8571 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.11337881611707883
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2311 | LCC = 0.8598 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.10974197786970016
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2474 | LCC = 0.8596 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.10561656775191808
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2400 | LCC = 0.8598 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.09720053409154598
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2629 | LCC = 0.8536 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.10931689745913713
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2314 | LCC = 0.8596 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.10128250550956298
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2389 | LCC = 0.8584 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.09018137353735092
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2373 | LCC = 0.8552 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.08637191880589877
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2264 | LCC = 0.8577 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.08112903531545247
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2305 | LCC = 0.8577 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.08217781720062096
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2413 | LCC = 0.8567 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.10583357522502923
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3170 | LCC = 0.8558 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.08111953295958348
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2484 | LCC = 0.8567 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.0674248354939314
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2316 | LCC = 0.8567 | SRCC = 0.8568 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2376 | LCC = 0.8612 | SRCC = 0.8636 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer12/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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_large_xlsr53/layer12/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..911f4215e945727b27f395bb30786036830f19e2
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3feee78855e167ac49daf1eefaf9b3b9cd9692cdd70a4c86271a6839703957e3
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0283cf7f2eab1ed9da3f98a49acffe9b3fc92109
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69440452b4bfc1b180cca972fda89e8dcb1b8b216ba32faf45b5ff8cf90a9610
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer12/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..031d464679cdcf6b1f5745f4effb1f2be6c12b0a
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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=0.5426031736991345
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3532 | LCC = 0.8070 | SRCC = 0.8127 ]
+INFO:root:Average loss=0.29503473276511216
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3167 | LCC = 0.8344 | SRCC = 0.8351 ]
+INFO:root:Average loss=0.2746521857304451
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2756 | LCC = 0.8456 | SRCC = 0.8444 ]
+INFO:root:Average loss=0.24432956752104637
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2676 | LCC = 0.8518 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.22289872417847315
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2781 | LCC = 0.8554 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.22040411314139
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3370 | LCC = 0.8555 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.21049836593178603
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2628 | LCC = 0.8575 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.20456149963996348
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2440 | LCC = 0.8597 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.19072614323634368
+INFO:root:
+[Valid][8][UTT][ MSE = 0.3426 | LCC = 0.8523 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.18575481965373725
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2640 | LCC = 0.8562 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.16801489679477152
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2453 | LCC = 0.8584 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.16255213091006646
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2487 | LCC = 0.8575 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.1519991223437664
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2347 | LCC = 0.8597 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.1498062190337059
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2529 | LCC = 0.8563 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.13554332634577385
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2261 | LCC = 0.8615 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.13632952622496164
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2567 | LCC = 0.8499 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.12649503846963248
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2620 | LCC = 0.8580 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.13238115312579352
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3892 | LCC = 0.8564 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.12279191422156799
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2299 | LCC = 0.8582 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.10519617108198312
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2465 | LCC = 0.8523 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.11164696562366608
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2845 | LCC = 0.8496 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.10740389794302292
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2688 | LCC = 0.8573 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.09691917743438329
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2821 | LCC = 0.8523 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.09423696073010945
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2542 | LCC = 0.8533 | SRCC = 0.8484 ]
+INFO:root:Average loss=0.09013823653833988
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2521 | LCC = 0.8523 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.08669641332175487
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2454 | LCC = 0.8563 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.08380321308206289
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2395 | LCC = 0.8540 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.08093878770103821
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2676 | LCC = 0.8549 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.07208390576908222
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2308 | LCC = 0.8528 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.07509228553718482
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2678 | LCC = 0.8507 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.06471182062075688
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2329 | LCC = 0.8527 | SRCC = 0.8492 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2439 | LCC = 0.8609 | SRCC = 0.8637 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer12/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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_large_xlsr53/layer12/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..418b2b27115dc922a64e51d94a40ba8514149804
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:920454ecaf937fe3632d7792f18d3a50e237dd18d41c0dc6c21df1be5482bedf
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer12/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f2c2691b7094bbb843a9587cc9a9ac3d7b481463
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdae6b37a43d8b5dbfde56672fa3334652926438444d768060bfb82e04a8802c
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer12/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer12/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..6071f9be56ea2424d2a7d8f7c7b07a6ef28b5875
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer12/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=0.5319788694763795
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3330 | LCC = 0.8097 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.30101522975243056
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2966 | LCC = 0.8378 | SRCC = 0.8384 ]
+INFO:root:Average loss=0.2611725899653557
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2834 | LCC = 0.8432 | SRCC = 0.8394 ]
+INFO:root:Average loss=0.24721199178542846
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3469 | LCC = 0.8521 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.24994137338720834
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2450 | LCC = 0.8545 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.21274942236068922
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2826 | LCC = 0.8523 | SRCC = 0.8501 ]
+INFO:root:Average loss=0.2016927164334517
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3886 | LCC = 0.8455 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.19779254018496245
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2643 | LCC = 0.8572 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.1819632274982257
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2367 | LCC = 0.8587 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.17687589694292116
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2413 | LCC = 0.8556 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.1921975126442237
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2377 | LCC = 0.8582 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.16572594432494578
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2465 | LCC = 0.8560 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.16239066441089678
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2608 | LCC = 0.8563 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.14513309099353278
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2982 | LCC = 0.8421 | SRCC = 0.8494 ]
+INFO:root:Average loss=0.1369285188997403
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2437 | LCC = 0.8528 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.1338784556167248
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2574 | LCC = 0.8530 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.1374043556742179
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2485 | LCC = 0.8527 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.11739992388547996
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2392 | LCC = 0.8538 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.11265210329722135
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2353 | LCC = 0.8544 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.11214267089962959
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2992 | LCC = 0.8496 | SRCC = 0.8487 ]
+INFO:root:Average loss=0.10670997412540974
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2386 | LCC = 0.8517 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.09959840000821994
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2428 | LCC = 0.8519 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.09924865910449089
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2324 | LCC = 0.8512 | SRCC = 0.8491 ]
+INFO:root:Average loss=0.10269785280793141
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3408 | LCC = 0.8457 | SRCC = 0.8468 ]
+INFO:root:Average loss=0.0932943303233538
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2523 | LCC = 0.8537 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.08528446783431065
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2327 | LCC = 0.8538 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.07683258580091672
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2776 | LCC = 0.8511 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.08846053280509435
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2408 | LCC = 0.8528 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.07158010161648957
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2510 | LCC = 0.8528 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.07027591912983319
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2313 | LCC = 0.8526 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.06783072344767742
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2611 | LCC = 0.8437 | SRCC = 0.8408 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2353 | LCC = 0.8650 | SRCC = 0.8680 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer14/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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_large_xlsr53/layer14/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f438aed995f2263601d3d160233309beed264d4e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a48eb5f1fc3bbb8802b901136b8e03cd0d85de99a5d3287460cf40315b17c61
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2eb88eedaea2f1e3f47979d999cc6ccaaa7d0700
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d1b794c479a7a5c8a372e1ca633ccf5ab1d6b4b95268825d475848d36e1fda4
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer14/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f0f4b15d98188f6151ee8271ae35c6d5eec398ed
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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=0.6377177758094592
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4197 | LCC = 0.7760 | SRCC = 0.7902 ]
+INFO:root:Average loss=0.30346459054794067
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3146 | LCC = 0.8368 | SRCC = 0.8394 ]
+INFO:root:Average loss=0.26094600902153897
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3769 | LCC = 0.8403 | SRCC = 0.8447 ]
+INFO:root:Average loss=0.231740038555402
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3221 | LCC = 0.8474 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.22008869548638663
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2868 | LCC = 0.8539 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.21269434929276124
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2453 | LCC = 0.8590 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.20495427208833206
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3010 | LCC = 0.8541 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.1980606491367022
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2714 | LCC = 0.8599 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.1793183110272273
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2312 | LCC = 0.8627 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.1784592255567893
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2256 | LCC = 0.8623 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.17145175907092217
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3224 | LCC = 0.8513 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.15872295067096368
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2690 | LCC = 0.8566 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.16907801020603913
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2474 | LCC = 0.8602 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.15631271984714729
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2490 | LCC = 0.8597 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.14779499173164368
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2400 | LCC = 0.8584 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.13995959600194907
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2274 | LCC = 0.8624 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.13326992562566048
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2599 | LCC = 0.8608 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.1214907496021344
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2499 | LCC = 0.8574 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.1249652147675172
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2700 | LCC = 0.8487 | SRCC = 0.8467 ]
+INFO:root:Average loss=0.11559975519776344
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2338 | LCC = 0.8590 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.11076885605087647
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2559 | LCC = 0.8553 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.11962494053519689
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2286 | LCC = 0.8580 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.1028830082848286
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3254 | LCC = 0.8479 | SRCC = 0.8474 ]
+INFO:root:Average loss=0.10304131377965976
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2275 | LCC = 0.8555 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.09006107932864091
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2379 | LCC = 0.8527 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.0922596597423156
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2303 | LCC = 0.8557 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.08730082935056625
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2342 | LCC = 0.8530 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.08444626634128582
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2527 | LCC = 0.8517 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.07583755068480968
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2669 | LCC = 0.8477 | SRCC = 0.8459 ]
+INFO:root:Average loss=0.0782900559118925
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2893 | LCC = 0.8475 | SRCC = 0.8448 ]
+INFO:root:Average loss=0.07234098193928218
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2442 | LCC = 0.8496 | SRCC = 0.8475 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2280 | LCC = 0.8647 | SRCC = 0.8684 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer14/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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_large_xlsr53/layer14/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4375fada3b5fb169bb1bf4c03280d9bbfedf49ce
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ade647842c56e5fcb0ec1458cac0f556eb2369798e59930c73b42ad7e938628
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..4ea5f240ddbd221ea1117ba0e5ca5677cbfde94d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3a3a56547d19ed96fe0ed892286d25c6941e2549c4ecef33a73024eb8cfcdd2
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer14/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..3497ebc204117d64d92377ad0c355c513516d3fb
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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=0.5554496302054479
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3780 | LCC = 0.8034 | SRCC = 0.8112 ]
+INFO:root:Average loss=0.32430396592005706
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3574 | LCC = 0.8281 | SRCC = 0.8307 ]
+INFO:root:Average loss=0.2773826164312852
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4328 | LCC = 0.8313 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.26556658095274216
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2888 | LCC = 0.8475 | SRCC = 0.8458 ]
+INFO:root:Average loss=0.23400942943035027
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2480 | LCC = 0.8544 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.21895222251231855
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2604 | LCC = 0.8545 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.22236716689971778
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2718 | LCC = 0.8553 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.2102835856569119
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2840 | LCC = 0.8538 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.1849993642133016
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2325 | LCC = 0.8555 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.18955166475513044
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2312 | LCC = 0.8602 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.18822792583169082
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2800 | LCC = 0.8600 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.18812511708491889
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3779 | LCC = 0.8399 | SRCC = 0.8409 ]
+INFO:root:Average loss=0.16712110336774436
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3357 | LCC = 0.8402 | SRCC = 0.8428 ]
+INFO:root:Average loss=0.1595154212644467
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2905 | LCC = 0.8503 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.15580938345728776
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2589 | LCC = 0.8575 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.14415670654330498
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2565 | LCC = 0.8579 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.13382426057106409
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2451 | LCC = 0.8556 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.1266811468089238
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2271 | LCC = 0.8550 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.13478597501913706
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2439 | LCC = 0.8530 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.12329835683489457
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2520 | LCC = 0.8534 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.1198885831504296
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2364 | LCC = 0.8560 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.10578467324376106
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2509 | LCC = 0.8479 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.10239235875316155
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2507 | LCC = 0.8505 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.10399790509388997
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2534 | LCC = 0.8509 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.09645997904814206
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2687 | LCC = 0.8449 | SRCC = 0.8437 ]
+INFO:root:Average loss=0.09687125792679115
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2391 | LCC = 0.8477 | SRCC = 0.8470 ]
+INFO:root:Average loss=0.10246050873628029
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2298 | LCC = 0.8528 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.08621127335115886
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2556 | LCC = 0.8503 | SRCC = 0.8472 ]
+INFO:root:Average loss=0.08966433968490516
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2397 | LCC = 0.8511 | SRCC = 0.8472 ]
+INFO:root:Average loss=0.07752279579066314
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2880 | LCC = 0.8416 | SRCC = 0.8386 ]
+INFO:root:Average loss=0.07722145481369434
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2638 | LCC = 0.8432 | SRCC = 0.8388 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2344 | LCC = 0.8648 | SRCC = 0.8692 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer14/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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_large_xlsr53/layer14/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c139f72a4d3d8b2df23a2b35097e7a246fa71984
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:758ca306c9700b26a5553746e8ab81c99fae405c88125ddb676d4dd75e16784a
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1b39b28415031ef5f6040a426a010991cab4c0d0
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b56bbc5c2120e934f69a2dfaa737d60b24376e2c611c8c0af104b28a273b49f
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer14/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..bd07fa1a0abd3450965cadfc3dccebcc4d7735ac
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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=0.5510767434652035
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3442 | LCC = 0.8029 | SRCC = 0.8064 ]
+INFO:root:Average loss=0.2985539537591812
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3515 | LCC = 0.8408 | SRCC = 0.8407 ]
+INFO:root:Average loss=0.26845811956968063
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2623 | LCC = 0.8478 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.23501436355022284
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2561 | LCC = 0.8499 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.22920069117576647
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2755 | LCC = 0.8566 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.21401728957127303
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2390 | LCC = 0.8540 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.20325946922485644
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2484 | LCC = 0.8586 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.19901255604166251
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2530 | LCC = 0.8549 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.1880508986994242
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2254 | LCC = 0.8574 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.17574994227825066
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2312 | LCC = 0.8587 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.19118150228109115
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2532 | LCC = 0.8622 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.1635187162229648
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2436 | LCC = 0.8583 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.15018718756544283
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2380 | LCC = 0.8601 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.14953246884621108
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2451 | LCC = 0.8568 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.1451362127868029
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2429 | LCC = 0.8575 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.13707023257246384
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2388 | LCC = 0.8566 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.13310499517963484
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2310 | LCC = 0.8577 | SRCC = 0.8544 ]
+INFO:root:Average loss=0.12950529186771467
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2300 | LCC = 0.8573 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.12426639033051637
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2356 | LCC = 0.8563 | SRCC = 0.8523 ]
+INFO:root:Average loss=0.12183011199037234
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2409 | LCC = 0.8536 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.11125013748040566
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2599 | LCC = 0.8523 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.11558080459825504
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2484 | LCC = 0.8531 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.10967826642669164
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2314 | LCC = 0.8581 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.09924226235120724
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2334 | LCC = 0.8541 | SRCC = 0.8498 ]
+INFO:root:Average loss=0.10675105404777405
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2398 | LCC = 0.8515 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.09119460908457255
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2732 | LCC = 0.8472 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.08723153761373116
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2355 | LCC = 0.8529 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.08631423101402246
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2322 | LCC = 0.8526 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.08516873108843963
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2542 | LCC = 0.8447 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.07893028191457956
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2550 | LCC = 0.8481 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.07406979961655079
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2384 | LCC = 0.8545 | SRCC = 0.8531 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2758 | LCC = 0.8594 | SRCC = 0.8659 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer14/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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_large_xlsr53/layer14/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..74d2669eedec4e85270e5b218991d91902cb7a60
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2075cc7e9d77d4c315d41238873b250d6427c134c9451f74cc8031607d8042c1
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..c558b13a3628e7e7de5e503cca029e0115ee8386
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da9ba7aeb080040d02f8e4831bebc64a762f0b1f9d6a9ffc74071d6e468a6315
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer14/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..47c8d69d6b133487627e0d9d28330e9b323fe25d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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=0.5007903031431712
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4416 | LCC = 0.8098 | SRCC = 0.8182 ]
+INFO:root:Average loss=0.29422732710074156
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3023 | LCC = 0.8439 | SRCC = 0.8433 ]
+INFO:root:Average loss=0.2676052225705905
+INFO:root:
+[Valid][2][UTT][ MSE = 0.4074 | LCC = 0.8432 | SRCC = 0.8478 ]
+INFO:root:Average loss=0.2346277374487657
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2568 | LCC = 0.8585 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.2220818195014428
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2521 | LCC = 0.8561 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.23435873939440802
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3014 | LCC = 0.8580 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.19955169619658056
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2686 | LCC = 0.8590 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.1915367802557273
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2590 | LCC = 0.8610 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.17832379205486712
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2773 | LCC = 0.8613 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.1758087799908259
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2576 | LCC = 0.8595 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.1676784351659127
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2553 | LCC = 0.8586 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.16533527007469764
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2323 | LCC = 0.8651 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.16059207180753732
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2391 | LCC = 0.8629 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.15345832648185584
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2451 | LCC = 0.8630 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.14285220540104768
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2185 | LCC = 0.8657 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.14353159194191298
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2567 | LCC = 0.8635 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.1287395864343032
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2831 | LCC = 0.8550 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.12012758831947278
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2610 | LCC = 0.8585 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.12590513206445253
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2475 | LCC = 0.8587 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.11678577491488212
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2549 | LCC = 0.8582 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.11388057814194606
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2636 | LCC = 0.8583 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.10192470959363839
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2413 | LCC = 0.8631 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.10072796572095309
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2375 | LCC = 0.8546 | SRCC = 0.8516 ]
+INFO:root:Average loss=0.09056994925515774
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2316 | LCC = 0.8571 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.09334234367960538
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2552 | LCC = 0.8555 | SRCC = 0.8514 ]
+INFO:root:Average loss=0.09517521721621354
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2434 | LCC = 0.8542 | SRCC = 0.8504 ]
+INFO:root:Average loss=0.08554224106363761
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2491 | LCC = 0.8563 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.08512963364139581
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2714 | LCC = 0.8583 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.08358807565692143
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2432 | LCC = 0.8528 | SRCC = 0.8488 ]
+INFO:root:Average loss=0.07785516337324412
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2577 | LCC = 0.8544 | SRCC = 0.8509 ]
+INFO:root:Average loss=0.07070467821680583
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2349 | LCC = 0.8553 | SRCC = 0.8510 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2212 | LCC = 0.8628 | SRCC = 0.8660 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer14/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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_large_xlsr53/layer14/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a1f928b01f152cba2d9723adca6aaa8939cda085
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95f9b2944154b1b98c167aae67e6080d51366823a748527799a6eab21c406ba6
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer14/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..89741e405e692d220b88ec7ef138397e588ae71e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4965c0db5bcd013a08d44ad6b189e3d6256991ba2741c0e363b38aacc8d0d5c9
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer14/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer14/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..fead7ca589b21cb8a4f4c95588c5eb3b64a14754
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer14/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=0.4942127405068813
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3201 | LCC = 0.8119 | SRCC = 0.8138 ]
+INFO:root:Average loss=0.2960547086520073
+INFO:root:
+[Valid][1][UTT][ MSE = 0.4159 | LCC = 0.8223 | SRCC = 0.8319 ]
+INFO:root:Average loss=0.2729759069207387
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2978 | LCC = 0.8447 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.24741354833046594
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2809 | LCC = 0.8558 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.22419307734339666
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2794 | LCC = 0.8580 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.23912876443221018
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2369 | LCC = 0.8591 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.22454379594478852
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2518 | LCC = 0.8594 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.20234141747156778
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2545 | LCC = 0.8601 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.18685102271728027
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2679 | LCC = 0.8613 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.1788761644409253
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2366 | LCC = 0.8641 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.1705268701681724
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2565 | LCC = 0.8653 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.16498512029647827
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2301 | LCC = 0.8668 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.16140372382524687
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2221 | LCC = 0.8629 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.15667039069991845
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2529 | LCC = 0.8620 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.14934798798117882
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2480 | LCC = 0.8648 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.140453241001337
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2265 | LCC = 0.8632 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.14080578509049538
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2148 | LCC = 0.8640 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.12885944029459587
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2426 | LCC = 0.8611 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.1263223113730932
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2387 | LCC = 0.8625 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.11280184095868698
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2426 | LCC = 0.8539 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.10999988162746796
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2346 | LCC = 0.8607 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.11666844059259464
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2541 | LCC = 0.8529 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.10759853705381736
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2297 | LCC = 0.8607 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.10516392153043014
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2227 | LCC = 0.8615 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.09643846549666844
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2377 | LCC = 0.8592 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.0970881258447965
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2253 | LCC = 0.8639 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.0924050987053376
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2297 | LCC = 0.8611 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.09157167561352253
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2226 | LCC = 0.8609 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.08752626777650455
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2424 | LCC = 0.8563 | SRCC = 0.8536 ]
+INFO:root:Average loss=0.08024174075287122
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2292 | LCC = 0.8562 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.08211523986970767
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2356 | LCC = 0.8582 | SRCC = 0.8550 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2289 | LCC = 0.8685 | SRCC = 0.8712 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer16/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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_large_xlsr53/layer16/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..19e558e9164c4947ca49ea8f6d6e6045913f214f
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b464e7736ab7ade8b9977f5487c0610f0ae44b1007eb5ec875726ae70c33c72
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..862818f6acee5b78d9bc38cad2437fb7cd1c98bd
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e275e6f33f7c537c9e5506c9892f7a6203c676acc87e73f063b1d45caf3a61a
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer16/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..7c9c0eaabaf151857c369f8568e365768a8bea7f
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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=0.549296165123964
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4474 | LCC = 0.7965 | SRCC = 0.8065 ]
+INFO:root:Average loss=0.2814397320915491
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2852 | LCC = 0.8401 | SRCC = 0.8404 ]
+INFO:root:Average loss=0.25467026157256883
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3350 | LCC = 0.8446 | SRCC = 0.8455 ]
+INFO:root:Average loss=0.24599542526098397
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2767 | LCC = 0.8503 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.22644929339488348
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2903 | LCC = 0.8571 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.21250931460123795
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2493 | LCC = 0.8577 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.2293084262846372
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2780 | LCC = 0.8598 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.21578812933503053
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2407 | LCC = 0.8621 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.18257067763270476
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2734 | LCC = 0.8590 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.18744600869906256
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2768 | LCC = 0.8623 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.16535278705832285
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2766 | LCC = 0.8611 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.16407409090644273
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3084 | LCC = 0.8620 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.15296474968393645
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2372 | LCC = 0.8671 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.14874040927642432
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2328 | LCC = 0.8682 | SRCC = 0.8647 ]
+INFO:root:Average loss=0.1412718759324306
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2221 | LCC = 0.8635 | SRCC = 0.8609 ]
+INFO:root:Average loss=0.12885510644469506
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2668 | LCC = 0.8597 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.12576797060095346
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2591 | LCC = 0.8592 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.12338019401217118
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2493 | LCC = 0.8641 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.1167804805132059
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2449 | LCC = 0.8597 | SRCC = 0.8560 ]
+INFO:root:Average loss=0.11158327892040595
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2342 | LCC = 0.8634 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.10314218413371307
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2483 | LCC = 0.8601 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.10206232344110806
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2344 | LCC = 0.8640 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.0925306701220763
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2327 | LCC = 0.8615 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.09083973434873116
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2317 | LCC = 0.8616 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.0892332628942453
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2705 | LCC = 0.8563 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.08910695277154446
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2408 | LCC = 0.8572 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.0778169711239827
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2336 | LCC = 0.8613 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.08548496868938972
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2353 | LCC = 0.8571 | SRCC = 0.8539 ]
+INFO:root:Average loss=0.07468801746383691
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2389 | LCC = 0.8590 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.07358211441299854
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2351 | LCC = 0.8600 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.07069605412200475
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2489 | LCC = 0.8586 | SRCC = 0.8554 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2306 | LCC = 0.8709 | SRCC = 0.8759 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer16/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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_large_xlsr53/layer16/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..23d155301b6ec554efd07f41e7b643e2204bf6b6
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09318cde77afdea028e6e8ff792a24cd5193ba70ccb0eb6b3b41359fdbb88c48
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1433c33d658ac0ff73a0eaa14425b23a56c667f1
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e25fea56cc2838be97cb9e638e3bd342337cb374ef7f137188b98810bcb2791
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer16/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..5a4f65001bfdeb1b7cf7f59c582d2a77e3521363
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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=0.5268183690615189
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3360 | LCC = 0.8110 | SRCC = 0.8168 ]
+INFO:root:Average loss=0.27939223039608735
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3177 | LCC = 0.8425 | SRCC = 0.8427 ]
+INFO:root:Average loss=0.27267833340626496
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2903 | LCC = 0.8505 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.2279875093163588
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3681 | LCC = 0.8398 | SRCC = 0.8432 ]
+INFO:root:Average loss=0.20655014996345228
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2399 | LCC = 0.8573 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.20283748381412947
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2646 | LCC = 0.8611 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.19051378038831246
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2319 | LCC = 0.8638 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.17919908912900168
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2534 | LCC = 0.8613 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.16129395384819079
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2429 | LCC = 0.8637 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.16294702333517563
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2549 | LCC = 0.8597 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.15832113885344604
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2621 | LCC = 0.8593 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.1517282647964282
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2320 | LCC = 0.8637 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.13509684848861817
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2694 | LCC = 0.8589 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.14351300465372893
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2533 | LCC = 0.8572 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.13828678133013922
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2453 | LCC = 0.8557 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.12329283122641918
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2636 | LCC = 0.8518 | SRCC = 0.8451 ]
+INFO:root:Average loss=0.12324314497602291
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2891 | LCC = 0.8534 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.11741260611093961
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2370 | LCC = 0.8547 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.12022850156212464
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2681 | LCC = 0.8538 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.11499397609478389
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2403 | LCC = 0.8562 | SRCC = 0.8505 ]
+INFO:root:Average loss=0.10130923547041722
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2394 | LCC = 0.8600 | SRCC = 0.8569 ]
+INFO:root:Average loss=0.09575368358920781
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2413 | LCC = 0.8566 | SRCC = 0.8527 ]
+INFO:root:Average loss=0.09229040728547634
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2451 | LCC = 0.8533 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.0885161250256575
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2407 | LCC = 0.8554 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.08992506945744538
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2645 | LCC = 0.8571 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.07810718561403263
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2598 | LCC = 0.8504 | SRCC = 0.8453 ]
+INFO:root:Average loss=0.07709860453047813
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2345 | LCC = 0.8558 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.0759624086606961
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2362 | LCC = 0.8548 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.07303297911316921
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2321 | LCC = 0.8570 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.06880355478288272
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2562 | LCC = 0.8560 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.06913847629076396
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2426 | LCC = 0.8500 | SRCC = 0.8455 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2237 | LCC = 0.8691 | SRCC = 0.8717 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer16/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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_large_xlsr53/layer16/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9c9b192632d4a1663354a8b90d70fea73ddc8e5e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ced87793e24f949f2632e70ae6000e31df46a0f9a6bb12b443b7b0409fed2dea
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..926953404b64a11499ef93d5cb9bf49181768637
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a80ab04de737150aebce389ecd3686f0e442cb6d08465ed330d4929d658792d9
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer16/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..c00ac5d0599d9bb769ee867ab3286b6a636120e5
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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=0.5057094895686859
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4039 | LCC = 0.8026 | SRCC = 0.8149 ]
+INFO:root:Average loss=0.27349784072393024
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3348 | LCC = 0.8476 | SRCC = 0.8481 ]
+INFO:root:Average loss=0.23881260305643082
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3589 | LCC = 0.8469 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.2272036916170365
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2626 | LCC = 0.8585 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.21215608314825937
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2656 | LCC = 0.8626 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.21371001320389602
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2416 | LCC = 0.8604 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.2032833980062069
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2741 | LCC = 0.8649 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.18974210140414727
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2846 | LCC = 0.8636 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.18220622570086747
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2738 | LCC = 0.8620 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.17813834939629603
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2540 | LCC = 0.8688 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.16077468066643447
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2652 | LCC = 0.8683 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.15492601711780596
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2246 | LCC = 0.8656 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.14702522477660424
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2792 | LCC = 0.8648 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.14312984803930306
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2391 | LCC = 0.8662 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.13566699480781189
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2404 | LCC = 0.8653 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.13765621404999343
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2439 | LCC = 0.8653 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.13335117449363074
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2869 | LCC = 0.8577 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.12745704110234213
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2514 | LCC = 0.8664 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.13414326338813856
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2732 | LCC = 0.8650 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.11413344822059839
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2355 | LCC = 0.8662 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.12086986033962323
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2489 | LCC = 0.8655 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.10014239139854908
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2347 | LCC = 0.8619 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.09716901097160119
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2187 | LCC = 0.8649 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.0995783510689552
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2295 | LCC = 0.8668 | SRCC = 0.8621 ]
+INFO:root:Average loss=0.09126611073047687
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2212 | LCC = 0.8619 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.10096494246942875
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2327 | LCC = 0.8646 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.08491003842881092
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3081 | LCC = 0.8610 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.08567550477500145
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2687 | LCC = 0.8535 | SRCC = 0.8492 ]
+INFO:root:Average loss=0.07830696958952989
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2501 | LCC = 0.8609 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.07690010301004617
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2225 | LCC = 0.8644 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.07512496053599395
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2265 | LCC = 0.8646 | SRCC = 0.8609 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2293 | LCC = 0.8735 | SRCC = 0.8743 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer16/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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_large_xlsr53/layer16/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f073ddf3404ad76fd5df00b65d9d6d5c5f5a94b3
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1944fa15a0f112deeea3509a1abf9e7ac1b44f6076b5ea6d0e8e16fe6bc618e5
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5672321e113ea005be67bb248a38592236dd0da2
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:534f5ec191e07fb9946864915accbcc06cf277f7a7403fa314046bfcb4d0c5dd
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer16/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..d3340a33673b31a625496025a9e76e9a4004f69c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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=0.4747970151977661
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3172 | LCC = 0.8199 | SRCC = 0.8216 ]
+INFO:root:Average loss=0.2805781087432152
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3256 | LCC = 0.8405 | SRCC = 0.8401 ]
+INFO:root:Average loss=0.24233578279232368
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2727 | LCC = 0.8528 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.254819576556866
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2618 | LCC = 0.8564 | SRCC = 0.8545 ]
+INFO:root:Average loss=0.2259552545654468
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3239 | LCC = 0.8439 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.21080531657506257
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2871 | LCC = 0.8499 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.20858058276084754
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2625 | LCC = 0.8585 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.18857590242838249
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2529 | LCC = 0.8617 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.1738054714141748
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2427 | LCC = 0.8618 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.16891788758146456
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2631 | LCC = 0.8629 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.16804016983279815
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2354 | LCC = 0.8635 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.15093148795840067
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2532 | LCC = 0.8593 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.15673816949129105
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2462 | LCC = 0.8648 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.1511241852855071
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2324 | LCC = 0.8646 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.13158575703318304
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2600 | LCC = 0.8542 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.13742568515814269
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2583 | LCC = 0.8563 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.11781342174762334
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2316 | LCC = 0.8615 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.1242143601561204
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2425 | LCC = 0.8628 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.11436045150726269
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2345 | LCC = 0.8551 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.12181995675349847
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2692 | LCC = 0.8555 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.10040534545595829
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2320 | LCC = 0.8594 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.09252928655881149
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2309 | LCC = 0.8576 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.09633958635803981
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2339 | LCC = 0.8596 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.09994914793433288
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2396 | LCC = 0.8564 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.09033834867370434
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2279 | LCC = 0.8564 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.08537545948265454
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2323 | LCC = 0.8551 | SRCC = 0.8506 ]
+INFO:root:Average loss=0.07902559929360183
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2479 | LCC = 0.8557 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.07542022919425598
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2334 | LCC = 0.8559 | SRCC = 0.8530 ]
+INFO:root:Average loss=0.07310572457619202
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2411 | LCC = 0.8576 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.0745118807714719
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2285 | LCC = 0.8588 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.07188550000771499
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2753 | LCC = 0.8511 | SRCC = 0.8511 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2462 | LCC = 0.8671 | SRCC = 0.8684 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer16/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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_large_xlsr53/layer16/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..00b1b1ff4926b6d073a1317aa8126e68aa6fd2da
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc62254595539338fe52c631d41a23cf5bbd1bca8639a9364ad919ba7b1aa1b8
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer16/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2fbd5c030696fd9ae8a93cc3ce9ba2648e801d13
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:322f91d2382b09cb94e0c267579aa9db80e7054d5c86e5d5f94261146d0853f5
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer16/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer16/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..fa05721e2f7437da103815c101d7f9979175954b
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer16/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=0.5660264079387372
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3521 | LCC = 0.8011 | SRCC = 0.8108 ]
+INFO:root:Average loss=0.26522525648276013
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2505 | LCC = 0.8447 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.2629090054677083
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2691 | LCC = 0.8507 | SRCC = 0.8515 ]
+INFO:root:Average loss=0.2261966151686815
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2856 | LCC = 0.8577 | SRCC = 0.8599 ]
+INFO:root:Average loss=0.21788338113289613
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2285 | LCC = 0.8600 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.19757069781040534
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2432 | LCC = 0.8618 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.19838895706030038
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2432 | LCC = 0.8616 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.21027925696510535
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3118 | LCC = 0.8553 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.17938848584890366
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2289 | LCC = 0.8639 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.17368002560658333
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2305 | LCC = 0.8641 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.17002115809382537
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2287 | LCC = 0.8680 | SRCC = 0.8653 ]
+INFO:root:Average loss=0.16292516008401528
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2327 | LCC = 0.8632 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.15635761771446618
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2353 | LCC = 0.8656 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.14995289851839727
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2265 | LCC = 0.8652 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.14557673973150742
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2400 | LCC = 0.8646 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.13592664964306048
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2264 | LCC = 0.8633 | SRCC = 0.8597 ]
+INFO:root:Average loss=0.12313351665551846
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2213 | LCC = 0.8650 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.11984357648552993
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2382 | LCC = 0.8556 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.11654616787265508
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2535 | LCC = 0.8622 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.11070188650718102
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2466 | LCC = 0.8584 | SRCC = 0.8580 ]
+INFO:root:Average loss=0.10641018110208023
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2422 | LCC = 0.8562 | SRCC = 0.8534 ]
+INFO:root:Average loss=0.09969593250216582
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2327 | LCC = 0.8549 | SRCC = 0.8529 ]
+INFO:root:Average loss=0.09599123560847381
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2178 | LCC = 0.8608 | SRCC = 0.8591 ]
+INFO:root:Average loss=0.09793216085586792
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2255 | LCC = 0.8598 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.08688966609919682
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2182 | LCC = 0.8601 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.09783052118160786
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2462 | LCC = 0.8561 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.07746113024843045
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2481 | LCC = 0.8546 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.07581212271291476
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2575 | LCC = 0.8560 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.07926514806846778
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2566 | LCC = 0.8505 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.07539076410616055
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2337 | LCC = 0.8555 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.07129306179017593
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2316 | LCC = 0.8552 | SRCC = 0.8528 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2168 | LCC = 0.8741 | SRCC = 0.8752 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer18/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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_large_xlsr53/layer18/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..41b0e0ae263441a5e111ee8b0c3769458c35c07a
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d9b2f5d62c8f60adcf08cab27ae3d39aa7f60b3f168f50fc49f3b9229b47087
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6c03ac75f9c7bbfc7cf7b4c141b51c15558be9ed
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b0db6843cb0befb4cdb06a18ad451745460e11822dd851e5aade79ec47f5ee4
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer18/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..9a56b095d5c6dae46a0f2f9d30dd599dd711375e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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=0.4971846525485699
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3408 | LCC = 0.8099 | SRCC = 0.8178 ]
+INFO:root:Average loss=0.28522754422365093
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3380 | LCC = 0.8413 | SRCC = 0.8385 ]
+INFO:root:Average loss=0.30190649189246005
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3251 | LCC = 0.8420 | SRCC = 0.8442 ]
+INFO:root:Average loss=0.23421919918977296
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2867 | LCC = 0.8565 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.21529662112394968
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3049 | LCC = 0.8502 | SRCC = 0.8495 ]
+INFO:root:Average loss=0.21180473420864496
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3529 | LCC = 0.8535 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.20125764379134545
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2790 | LCC = 0.8601 | SRCC = 0.8578 ]
+INFO:root:Average loss=0.18570858459824172
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2688 | LCC = 0.8557 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.17995665356134757
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2567 | LCC = 0.8628 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.17594374878666338
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2706 | LCC = 0.8598 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.17853074033672994
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2362 | LCC = 0.8646 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.1675518876275955
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2593 | LCC = 0.8594 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.18974485678168443
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2724 | LCC = 0.8664 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.15363696265297058
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2282 | LCC = 0.8648 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.15000660450030595
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2435 | LCC = 0.8642 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.14066420247157416
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2665 | LCC = 0.8645 | SRCC = 0.8637 ]
+INFO:root:Average loss=0.13662056195048186
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2543 | LCC = 0.8650 | SRCC = 0.8623 ]
+INFO:root:Average loss=0.1285053872718261
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2800 | LCC = 0.8562 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.11993546077074149
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2325 | LCC = 0.8618 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.12303215637803078
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2355 | LCC = 0.8619 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.1157129002878299
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2385 | LCC = 0.8657 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.12222903154981442
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2360 | LCC = 0.8612 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.11487506654782173
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2443 | LCC = 0.8628 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.10968999130030473
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2927 | LCC = 0.8558 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.09346701357609187
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2678 | LCC = 0.8565 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.09169094025706634
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2290 | LCC = 0.8596 | SRCC = 0.8564 ]
+INFO:root:Average loss=0.09329492656084207
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2435 | LCC = 0.8609 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.08985214756849484
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2465 | LCC = 0.8542 | SRCC = 0.8525 ]
+INFO:root:Average loss=0.08647012605498998
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2389 | LCC = 0.8606 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.08165009510822785
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2688 | LCC = 0.8626 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.09522509407729675
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2358 | LCC = 0.8538 | SRCC = 0.8519 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2656 | LCC = 0.8689 | SRCC = 0.8737 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer18/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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_large_xlsr53/layer18/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..84d14ceeaccdf459476d9fd2874f1db56fa3d078
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f8a58cf916c2f11ffd45c5902d8e80ab53e9af3e0639573ac788ef8ac952b50
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..9ee54f7a2e158da281932a4ff2a93c4639005aad
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:397cdc746602b4f99b13f7a570d50c0a29f116b0574afac8ceae06fa5a205b3a
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer18/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f853e865b89e80370967b3b020ff0865d323b3da
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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=0.5686741152252907
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3349 | LCC = 0.8154 | SRCC = 0.8198 ]
+INFO:root:Average loss=0.2739389248383351
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2647 | LCC = 0.8488 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.2645046475988168
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2688 | LCC = 0.8523 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.24343398270698693
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2425 | LCC = 0.8565 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.22021976113319397
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2628 | LCC = 0.8574 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.2128464019833467
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2333 | LCC = 0.8617 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.1893839010825524
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2401 | LCC = 0.8619 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.2007237872443138
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2547 | LCC = 0.8658 | SRCC = 0.8614 ]
+INFO:root:Average loss=0.18224674243575487
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2596 | LCC = 0.8680 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.18003081176907587
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2422 | LCC = 0.8623 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.17467280763846177
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2456 | LCC = 0.8628 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.17274723201990128
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2784 | LCC = 0.8551 | SRCC = 0.8511 ]
+INFO:root:Average loss=0.1595378276438285
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2193 | LCC = 0.8654 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.14708203851030424
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2265 | LCC = 0.8670 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.15652922368966615
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2429 | LCC = 0.8673 | SRCC = 0.8647 ]
+INFO:root:Average loss=0.1412223297625016
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2311 | LCC = 0.8655 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.14041341124818876
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2355 | LCC = 0.8654 | SRCC = 0.8631 ]
+INFO:root:Average loss=0.1322904299849119
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2155 | LCC = 0.8664 | SRCC = 0.8640 ]
+INFO:root:Average loss=0.12581961659284738
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2475 | LCC = 0.8630 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.12416928185102268
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2336 | LCC = 0.8631 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.11377716112213257
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2274 | LCC = 0.8654 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.11254502364840263
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2372 | LCC = 0.8662 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.11821447265071747
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2541 | LCC = 0.8626 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.10500003321048541
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2345 | LCC = 0.8612 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.10186112404633792
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2263 | LCC = 0.8650 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.10305000502520646
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2280 | LCC = 0.8635 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.09602528609908544
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2299 | LCC = 0.8602 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.1082378757210114
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2346 | LCC = 0.8613 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.08831003771569484
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2341 | LCC = 0.8579 | SRCC = 0.8535 ]
+INFO:root:Average loss=0.0898404247008073
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2395 | LCC = 0.8603 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.08401936469360805
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2320 | LCC = 0.8621 | SRCC = 0.8591 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2627 | LCC = 0.8706 | SRCC = 0.8707 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer18/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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_large_xlsr53/layer18/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a74a5555b7c451b23e648959f5a097179d093dfb
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5423f37119b3f9d96b42ee4fe01fad649fdece4efe7a0109ffc5e3decfc0cae
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..a002f3cc410e10892ee4616bbbceffe5610b446d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:002be8ede5f135ce98946d75df01fbd71bbe3bbf8f550c9e1c0cfb7dd80ca3d2
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer18/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..f3cf2b8635f7223fb14229b43ade6e7300eb2059
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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=0.44875079469803053
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3622 | LCC = 0.8139 | SRCC = 0.8193 ]
+INFO:root:Average loss=0.27708634638633484
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2969 | LCC = 0.8370 | SRCC = 0.8339 ]
+INFO:root:Average loss=0.27427230641627925
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3089 | LCC = 0.8468 | SRCC = 0.8440 ]
+INFO:root:Average loss=0.2407580730624688
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2927 | LCC = 0.8532 | SRCC = 0.8499 ]
+INFO:root:Average loss=0.21774287273486456
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2512 | LCC = 0.8592 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.20799775956532893
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2266 | LCC = 0.8610 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.211577815696215
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2763 | LCC = 0.8593 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.18842628254340246
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2620 | LCC = 0.8628 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.1821849342340078
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2249 | LCC = 0.8669 | SRCC = 0.8635 ]
+INFO:root:Average loss=0.17135804194288376
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2673 | LCC = 0.8649 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.16451404769069108
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2364 | LCC = 0.8604 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.15550127912026185
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2381 | LCC = 0.8618 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.15170929514062712
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2635 | LCC = 0.8595 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.14859600709034845
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2323 | LCC = 0.8660 | SRCC = 0.8638 ]
+INFO:root:Average loss=0.1476218350804769
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2503 | LCC = 0.8547 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.1442226121822993
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2446 | LCC = 0.8601 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.13015274254557413
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2267 | LCC = 0.8669 | SRCC = 0.8639 ]
+INFO:root:Average loss=0.1402759624597354
+INFO:root:
+[Valid][17][UTT][ MSE = 0.3053 | LCC = 0.8572 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.12397800787137105
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2429 | LCC = 0.8642 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.12693906785585943
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2937 | LCC = 0.8548 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.1189798996425592
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2749 | LCC = 0.8576 | SRCC = 0.8550 ]
+INFO:root:Average loss=0.10997354974731421
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2389 | LCC = 0.8628 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.11532460793088643
+INFO:root:
+[Valid][22][UTT][ MSE = 0.3094 | LCC = 0.8555 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.0999842088860579
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2179 | LCC = 0.8657 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.09585396391459
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2695 | LCC = 0.8528 | SRCC = 0.8512 ]
+INFO:root:Average loss=0.1022173794798362
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2296 | LCC = 0.8630 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.08848893847794105
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3016 | LCC = 0.8564 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.09240396631260712
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2444 | LCC = 0.8575 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.09281975881029399
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2321 | LCC = 0.8616 | SRCC = 0.8583 ]
+INFO:root:Average loss=0.0810229687545544
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2927 | LCC = 0.8489 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.07936903667182495
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2524 | LCC = 0.8558 | SRCC = 0.8521 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2222 | LCC = 0.8716 | SRCC = 0.8750 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer18/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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_large_xlsr53/layer18/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..6dde1358de4228aaedb5c9f80076c3f5c9c451b0
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f942e3fc2799a2290ca75c5853220201a82d095bdc4e041cdb32260ac10e3e48
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..167b5fd8c2da06d602a57ab1d12a8c03aa882878
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c46722b60d1f92c7f524325a75b98b1797f7b535e8bf89fdb2a7b2a1e4017cd
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer18/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..73867d1bbc0dc3f52d4bc012d96cdcf308321d3c
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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=0.5342514255107977
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3128 | LCC = 0.8186 | SRCC = 0.8216 ]
+INFO:root:Average loss=0.28380395223697025
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3198 | LCC = 0.8371 | SRCC = 0.8370 ]
+INFO:root:Average loss=0.25883785826273453
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2795 | LCC = 0.8510 | SRCC = 0.8475 ]
+INFO:root:Average loss=0.23649571262873137
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2551 | LCC = 0.8559 | SRCC = 0.8540 ]
+INFO:root:Average loss=0.2251943208468266
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2796 | LCC = 0.8591 | SRCC = 0.8552 ]
+INFO:root:Average loss=0.20893713048635384
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2631 | LCC = 0.8634 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.21406995304501975
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2649 | LCC = 0.8631 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.19283120744885543
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2828 | LCC = 0.8624 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.18423355523592386
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2572 | LCC = 0.8657 | SRCC = 0.8613 ]
+INFO:root:Average loss=0.17083848430178103
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2223 | LCC = 0.8661 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.17321157780213234
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2322 | LCC = 0.8660 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.17338865431837547
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2617 | LCC = 0.8653 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.16031705120053047
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2594 | LCC = 0.8642 | SRCC = 0.8603 ]
+INFO:root:Average loss=0.15209124285059097
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2545 | LCC = 0.8661 | SRCC = 0.8632 ]
+INFO:root:Average loss=0.14465851737902716
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2549 | LCC = 0.8663 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.14407624609959432
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2434 | LCC = 0.8661 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.14969639795330855
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3706 | LCC = 0.8562 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.13597669194524103
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2250 | LCC = 0.8678 | SRCC = 0.8646 ]
+INFO:root:Average loss=0.1287708304440364
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2388 | LCC = 0.8679 | SRCC = 0.8642 ]
+INFO:root:Average loss=0.13103407793320143
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2486 | LCC = 0.8649 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.12278724079712844
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2370 | LCC = 0.8618 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.11855120020799148
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2398 | LCC = 0.8624 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.12086461599056537
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2428 | LCC = 0.8648 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.10363842699772273
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2315 | LCC = 0.8601 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.10550379944153321
+INFO:root:
+[Valid][24][UTT][ MSE = 0.3018 | LCC = 0.8590 | SRCC = 0.8533 ]
+INFO:root:Average loss=0.10759677776159385
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2546 | LCC = 0.8621 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.09804145552409001
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2928 | LCC = 0.8620 | SRCC = 0.8573 ]
+INFO:root:Average loss=0.09553377239558941
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2884 | LCC = 0.8607 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.08889905286905093
+INFO:root:
+[Valid][28][UTT][ MSE = 0.3006 | LCC = 0.8597 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.08368818533535187
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2418 | LCC = 0.8582 | SRCC = 0.8526 ]
+INFO:root:Average loss=0.08447391224595216
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2347 | LCC = 0.8594 | SRCC = 0.8551 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2296 | LCC = 0.8749 | SRCC = 0.8754 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer18/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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_large_xlsr53/layer18/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..8e68b3f2d4e2aadd2fe0500e7bd5b86791b02cf4
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf0d32a55e3dec4ad18361ac564b903320f30074c02dfa8f43fa567cb8fa419b
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer18/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..f97335b0ae623c2eaaf29fd5cde13c7ef6c885f8
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93940d6a580e26bbe03fc0d8ecf6b2f0de20160ce8498b96aff195ac840476ed
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer18/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer18/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..ea407ca281dbce0223045e32a55a21811e9d96ea
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer18/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=0.44078543132696396
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3075 | LCC = 0.8340 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.2858843919940484
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2658 | LCC = 0.8401 | SRCC = 0.8435 ]
+INFO:root:Average loss=0.2548126091177647
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2780 | LCC = 0.8576 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.24341513101871198
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2922 | LCC = 0.8582 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.2257133160646145
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3506 | LCC = 0.8484 | SRCC = 0.8551 ]
+INFO:root:Average loss=0.23106022217334846
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2605 | LCC = 0.8613 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.20142971351742744
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2809 | LCC = 0.8588 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.18869605717750695
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2417 | LCC = 0.8645 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.18977311922189516
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2349 | LCC = 0.8663 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.17192986875008315
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2189 | LCC = 0.8637 | SRCC = 0.8634 ]
+INFO:root:Average loss=0.16870022202149415
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2584 | LCC = 0.8619 | SRCC = 0.8620 ]
+INFO:root:Average loss=0.16097086142653075
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2375 | LCC = 0.8675 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.16547420811958802
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2281 | LCC = 0.8666 | SRCC = 0.8630 ]
+INFO:root:Average loss=0.15273824372352698
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2238 | LCC = 0.8655 | SRCC = 0.8643 ]
+INFO:root:Average loss=0.14998569396825936
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2505 | LCC = 0.8632 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.14653104610550097
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2268 | LCC = 0.8629 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.14863089653543937
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2843 | LCC = 0.8608 | SRCC = 0.8612 ]
+INFO:root:Average loss=0.13167025082004377
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2218 | LCC = 0.8624 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.13128639423312286
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2440 | LCC = 0.8630 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.11932740771235564
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2455 | LCC = 0.8619 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.1194965569063639
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2294 | LCC = 0.8614 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.11155245911616546
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2968 | LCC = 0.8558 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.10903995264417086
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2308 | LCC = 0.8628 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.11845288760004899
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2691 | LCC = 0.8579 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.10165799953616582
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2209 | LCC = 0.8657 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.10226528618771297
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2235 | LCC = 0.8621 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.10295278196915603
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2535 | LCC = 0.8630 | SRCC = 0.8629 ]
+INFO:root:Average loss=0.09583482705056667
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2282 | LCC = 0.8596 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.08928130824978535
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2315 | LCC = 0.8598 | SRCC = 0.8588 ]
+INFO:root:Average loss=0.10246748539308707
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2288 | LCC = 0.8576 | SRCC = 0.8553 ]
+INFO:root:Average loss=0.08863186903106861
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2202 | LCC = 0.8609 | SRCC = 0.8594 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2285 | LCC = 0.8738 | SRCC = 0.8738 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer2/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer2/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..31ae5285b23c82d55fe95df2a5361f81e063e6e1
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4532147fd4f0d1b47a867fae30dcb055163ce878d965af4a47378843bf3a61eb
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3bd2dff26ded96eb212878c062ff681f47921b71
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3d64b951c987fb748f09c140c43fbf621d18bf3e6d83fda50866a7272023b6f
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer2/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..b2d52344ce0a03d7ae41ca15224a10851a75eabd
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.5929750784849509
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4078 | LCC = 0.7370 | SRCC = 0.7415 ]
+INFO:root:Average loss=0.3456364971322891
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3563 | LCC = 0.7783 | SRCC = 0.7819 ]
+INFO:root:Average loss=0.29959009014643156
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3348 | LCC = 0.8027 | SRCC = 0.8038 ]
+INFO:root:Average loss=0.27279968082140654
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2961 | LCC = 0.8166 | SRCC = 0.8171 ]
+INFO:root:Average loss=0.25154274167158663
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2782 | LCC = 0.8276 | SRCC = 0.8259 ]
+INFO:root:Average loss=0.2370137205490699
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2895 | LCC = 0.8397 | SRCC = 0.8360 ]
+INFO:root:Average loss=0.2368054055632689
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3877 | LCC = 0.8271 | SRCC = 0.8322 ]
+INFO:root:Average loss=0.2210994290235715
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2853 | LCC = 0.8453 | SRCC = 0.8455 ]
+INFO:root:Average loss=0.23162634708942512
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2705 | LCC = 0.8515 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.2116609710531357
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2529 | LCC = 0.8478 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.2223581503599118
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2867 | LCC = 0.8516 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.20329457273085913
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3046 | LCC = 0.8447 | SRCC = 0.8517 ]
+INFO:root:Average loss=0.1872141471084876
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2483 | LCC = 0.8588 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.18039099556895402
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2574 | LCC = 0.8537 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.17127923533702508
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2763 | LCC = 0.8574 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.17222903793056807
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2298 | LCC = 0.8658 | SRCC = 0.8650 ]
+INFO:root:Average loss=0.18056693739998034
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2461 | LCC = 0.8597 | SRCC = 0.8594 ]
+INFO:root:Average loss=0.17454533546398848
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2676 | LCC = 0.8564 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.1525247492469274
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2421 | LCC = 0.8644 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.15073555001081565
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2227 | LCC = 0.8671 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.1417191325662992
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2397 | LCC = 0.8652 | SRCC = 0.8641 ]
+INFO:root:Average loss=0.1405433602631092
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2253 | LCC = 0.8663 | SRCC = 0.8656 ]
+INFO:root:Average loss=0.16024759612404382
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2596 | LCC = 0.8612 | SRCC = 0.8610 ]
+INFO:root:Average loss=0.1303031227718561
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3170 | LCC = 0.8467 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.12754793407825324
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2418 | LCC = 0.8630 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.12300908269408421
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2640 | LCC = 0.8575 | SRCC = 0.8616 ]
+INFO:root:Average loss=0.1267167180776596
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2320 | LCC = 0.8617 | SRCC = 0.8621 ]
+INFO:root:Average loss=0.11295439579930061
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2371 | LCC = 0.8619 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.12014639740570998
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2396 | LCC = 0.8612 | SRCC = 0.8645 ]
+INFO:root:Average loss=0.10684989803494552
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2523 | LCC = 0.8637 | SRCC = 0.8636 ]
+INFO:root:Average loss=0.10924137823092632
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2239 | LCC = 0.8644 | SRCC = 0.8646 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2188 | LCC = 0.8664 | SRCC = 0.8661 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer2/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer2/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..05dcbf1f1726a088dacc620021338067bc10b9e5
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cb8db16a47372ca68fd8b46b309a044bf326bde1e454f34cb039d0c5067459b
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..59ebe6e561f8e4f650e64d49ca6298dc752f74a9
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8753b5cf8789dff49a58211ec90b51b5ea008e6bf01764e40ba0031007b30873
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer2/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..7946c3c452eff8be7951700fc64c9ec665ddb74e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.577119808166455
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3954 | LCC = 0.7427 | SRCC = 0.7448 ]
+INFO:root:Average loss=0.34476271711098844
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3569 | LCC = 0.7864 | SRCC = 0.7829 ]
+INFO:root:Average loss=0.30708818290478146
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3367 | LCC = 0.7987 | SRCC = 0.8018 ]
+INFO:root:Average loss=0.27942181416811085
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2926 | LCC = 0.8171 | SRCC = 0.8130 ]
+INFO:root:Average loss=0.31265338452962727
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2896 | LCC = 0.8223 | SRCC = 0.8229 ]
+INFO:root:Average loss=0.2538654687695014
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2813 | LCC = 0.8313 | SRCC = 0.8309 ]
+INFO:root:Average loss=0.2645117004330342
+INFO:root:
+[Valid][6][UTT][ MSE = 0.3173 | LCC = 0.8295 | SRCC = 0.8314 ]
+INFO:root:Average loss=0.24624581711414534
+INFO:root:
+[Valid][7][UTT][ MSE = 0.4030 | LCC = 0.8179 | SRCC = 0.8263 ]
+INFO:root:Average loss=0.24331172679861388
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2620 | LCC = 0.8441 | SRCC = 0.8408 ]
+INFO:root:Average loss=0.20799508881874573
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2556 | LCC = 0.8531 | SRCC = 0.8503 ]
+INFO:root:Average loss=0.20758341243251777
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2331 | LCC = 0.8537 | SRCC = 0.8520 ]
+INFO:root:Average loss=0.20853939461402404
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2324 | LCC = 0.8550 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.19738949300386968
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2700 | LCC = 0.8511 | SRCC = 0.8524 ]
+INFO:root:Average loss=0.18709134960021728
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2441 | LCC = 0.8543 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.1863673019867677
+INFO:root:
+[Valid][14][UTT][ MSE = 0.3028 | LCC = 0.8575 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.18238311900924414
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2306 | LCC = 0.8648 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.16393615353183869
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2466 | LCC = 0.8585 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.1658866268893083
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2454 | LCC = 0.8594 | SRCC = 0.8602 ]
+INFO:root:Average loss=0.15910716058734137
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2342 | LCC = 0.8604 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.15228757864007583
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2140 | LCC = 0.8632 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.16661855607078627
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2239 | LCC = 0.8635 | SRCC = 0.8628 ]
+INFO:root:Average loss=0.14351297036195412
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2453 | LCC = 0.8615 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.13454610176193407
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2591 | LCC = 0.8589 | SRCC = 0.8589 ]
+INFO:root:Average loss=0.1333655615647634
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2497 | LCC = 0.8581 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.12696116217053854
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2162 | LCC = 0.8669 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.1250633686207808
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2256 | LCC = 0.8656 | SRCC = 0.8644 ]
+INFO:root:Average loss=0.12008439128597577
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2153 | LCC = 0.8690 | SRCC = 0.8662 ]
+INFO:root:Average loss=0.12730576317661849
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2231 | LCC = 0.8675 | SRCC = 0.8659 ]
+INFO:root:Average loss=0.12417003283133873
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2378 | LCC = 0.8661 | SRCC = 0.8658 ]
+INFO:root:Average loss=0.1235647605588803
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2387 | LCC = 0.8615 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.10471592193994766
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2183 | LCC = 0.8687 | SRCC = 0.8667 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2427 | LCC = 0.8663 | SRCC = 0.8654 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer2/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer2/run2/model.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run2/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..db333631ba742163e2f9aa0cfeee1feb9cfd93d4
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run2/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:569f79ba853d4e1a4786c86ebc4faf6b45de27fd30542e9d57aea74183970aaa
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run2/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run2/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2cdf227bb7a97685e5a3affbc520a802b181f1df
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run2/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe8dd160a3e40c193a517e03992234286ef7db76f1def8159f9cd8c091ba1d1b
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run2/train.log b/results_bvcc/w2v2_large_xlsr53/layer2/run2/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..3ddec27a683c0ee611d7881e008563d18f3788f6
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.563461205898187
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3877 | LCC = 0.7516 | SRCC = 0.7575 ]
+INFO:root:Average loss=0.3353806487642802
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3485 | LCC = 0.7970 | SRCC = 0.7948 ]
+INFO:root:Average loss=0.3098092008477602
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3550 | LCC = 0.8019 | SRCC = 0.8022 ]
+INFO:root:Average loss=0.30285963186850917
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3319 | LCC = 0.8136 | SRCC = 0.8109 ]
+INFO:root:Average loss=0.263142276268739
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3058 | LCC = 0.8226 | SRCC = 0.8193 ]
+INFO:root:Average loss=0.25002123625614703
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3162 | LCC = 0.8198 | SRCC = 0.8162 ]
+INFO:root:Average loss=0.23413045207659403
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2789 | LCC = 0.8358 | SRCC = 0.8331 ]
+INFO:root:Average loss=0.22040749933475104
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2850 | LCC = 0.8453 | SRCC = 0.8418 ]
+INFO:root:Average loss=0.22002424815526375
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2635 | LCC = 0.8384 | SRCC = 0.8388 ]
+INFO:root:Average loss=0.2521914791984436
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2732 | LCC = 0.8445 | SRCC = 0.8426 ]
+INFO:root:Average loss=0.19741697810017145
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2672 | LCC = 0.8532 | SRCC = 0.8502 ]
+INFO:root:Average loss=0.20803830600701845
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2701 | LCC = 0.8469 | SRCC = 0.8486 ]
+INFO:root:Average loss=0.20357371045228762
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3167 | LCC = 0.8576 | SRCC = 0.8568 ]
+INFO:root:Average loss=0.1942050131276632
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2549 | LCC = 0.8521 | SRCC = 0.8518 ]
+INFO:root:Average loss=0.17379580285304633
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2690 | LCC = 0.8561 | SRCC = 0.8555 ]
+INFO:root:Average loss=0.16605104095278642
+INFO:root:
+[Valid][15][UTT][ MSE = 0.3465 | LCC = 0.8431 | SRCC = 0.8513 ]
+INFO:root:Average loss=0.1680787809384175
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2526 | LCC = 0.8560 | SRCC = 0.8562 ]
+INFO:root:Average loss=0.15412277852495512
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2369 | LCC = 0.8604 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.15607098241647085
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2275 | LCC = 0.8618 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.1548545095974054
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2241 | LCC = 0.8621 | SRCC = 0.8590 ]
+INFO:root:Average loss=0.13600976640979448
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2331 | LCC = 0.8590 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.14369639200277817
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2286 | LCC = 0.8571 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.1403532262222889
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2278 | LCC = 0.8630 | SRCC = 0.8608 ]
+INFO:root:Average loss=0.1327283234359362
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2296 | LCC = 0.8620 | SRCC = 0.8596 ]
+INFO:root:Average loss=0.12207843602085724
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2267 | LCC = 0.8585 | SRCC = 0.8581 ]
+INFO:root:Average loss=0.1174970782147004
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2293 | LCC = 0.8552 | SRCC = 0.8566 ]
+INFO:root:Average loss=0.11612954649787682
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2299 | LCC = 0.8585 | SRCC = 0.8593 ]
+INFO:root:Average loss=0.12072048393579629
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2436 | LCC = 0.8633 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.11825937777757645
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2484 | LCC = 0.8607 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.11471253041273509
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2244 | LCC = 0.8635 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.10326044328319721
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2164 | LCC = 0.8616 | SRCC = 0.8629 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2226 | LCC = 0.8685 | SRCC = 0.8663 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run3/config.gin b/results_bvcc/w2v2_large_xlsr53/layer2/run3/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer2/run3/model.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run3/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..7fac1a80ba6d0ecbd3517a4eff60c5a3a0abe806
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run3/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b2ae5ee7dfad34bd0463ef9a5287b3b9363e8c796b3bc21426d05d7fc17bd1a
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run3/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run3/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..2c9446affb51135fd9c0f1ef78ab2a8fc214d225
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run3/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef6da9991e76a27f087e85624ed5d8df2e2bcae4c50ca3e999a132e84c41a2c5
+size 3405062
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run3/train.log b/results_bvcc/w2v2_large_xlsr53/layer2/run3/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..59e9d6330add8a7f31c56b19f50a62a8b5987f05
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.6563875656097363
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4322 | LCC = 0.7314 | SRCC = 0.7368 ]
+INFO:root:Average loss=0.3626988608485613
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3623 | LCC = 0.7727 | SRCC = 0.7799 ]
+INFO:root:Average loss=0.3146055592940404
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3557 | LCC = 0.8035 | SRCC = 0.8057 ]
+INFO:root:Average loss=0.2979992035871897
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3114 | LCC = 0.8192 | SRCC = 0.8184 ]
+INFO:root:Average loss=0.25859257808098424
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3639 | LCC = 0.8204 | SRCC = 0.8235 ]
+INFO:root:Average loss=0.2786229814474399
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3015 | LCC = 0.8289 | SRCC = 0.8312 ]
+INFO:root:Average loss=0.23665199352380556
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2765 | LCC = 0.8467 | SRCC = 0.8430 ]
+INFO:root:Average loss=0.24825419581089264
+INFO:root:
+[Valid][7][UTT][ MSE = 0.3183 | LCC = 0.8417 | SRCC = 0.8420 ]
+INFO:root:Average loss=0.21832896004884672
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2737 | LCC = 0.8478 | SRCC = 0.8446 ]
+INFO:root:Average loss=0.2223016077127212
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2662 | LCC = 0.8511 | SRCC = 0.8497 ]
+INFO:root:Average loss=0.20443501533606115
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3161 | LCC = 0.8461 | SRCC = 0.8490 ]
+INFO:root:Average loss=0.1935145401228697
+INFO:root:
+[Valid][11][UTT][ MSE = 0.3344 | LCC = 0.8470 | SRCC = 0.8519 ]
+INFO:root:Average loss=0.1872095822905883
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2519 | LCC = 0.8567 | SRCC = 0.8567 ]
+INFO:root:Average loss=0.1865205307228443
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2633 | LCC = 0.8593 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.17886203049849242
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2831 | LCC = 0.8597 | SRCC = 0.8607 ]
+INFO:root:Average loss=0.18613210358680823
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2402 | LCC = 0.8618 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.18536449978366876
+INFO:root:
+[Valid][16][UTT][ MSE = 0.3230 | LCC = 0.8577 | SRCC = 0.8619 ]
+INFO:root:Average loss=0.16457857239322785
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2248 | LCC = 0.8619 | SRCC = 0.8617 ]
+INFO:root:Average loss=0.16332166613294527
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2708 | LCC = 0.8560 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.15693221928981635
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2384 | LCC = 0.8680 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.14241065849096346
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2640 | LCC = 0.8620 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.14809257231461695
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2347 | LCC = 0.8653 | SRCC = 0.8657 ]
+INFO:root:Average loss=0.14758247824815604
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2184 | LCC = 0.8683 | SRCC = 0.8685 ]
+INFO:root:Average loss=0.15096520584745285
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2855 | LCC = 0.8631 | SRCC = 0.8633 ]
+INFO:root:Average loss=0.13649573320379624
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2903 | LCC = 0.8620 | SRCC = 0.8622 ]
+INFO:root:Average loss=0.12859808099575532
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2263 | LCC = 0.8680 | SRCC = 0.8671 ]
+INFO:root:Average loss=0.12911979309641397
+INFO:root:
+[Valid][26][UTT][ MSE = 0.3023 | LCC = 0.8618 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.12048750122388203
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2493 | LCC = 0.8660 | SRCC = 0.8668 ]
+INFO:root:Average loss=0.1265702624924672
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2255 | LCC = 0.8659 | SRCC = 0.8667 ]
+INFO:root:Average loss=0.11162384398854695
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2199 | LCC = 0.8688 | SRCC = 0.8691 ]
+INFO:root:Average loss=0.10768346679516327
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2494 | LCC = 0.8677 | SRCC = 0.8696 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2264 | LCC = 0.8615 | SRCC = 0.8640 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run4/config.gin b/results_bvcc/w2v2_large_xlsr53/layer2/run4/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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_large_xlsr53/layer2/run4/model.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run4/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..71f9091456c581a8a37bfde592deb218f8c29f9e
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run4/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fd04040f4f5d1b20a604a7d0d359eb6b7d3b82495ef9b3d74078a4f36179302
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run4/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer2/run4/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5cae5cb9a276825cf94f65c61afd2727522acb2d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer2/run4/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c53d333fb9aaf50c26aaad8a0d387e81992a5ba3bfe9ad4429f80b08316e3897
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer2/run4/train.log b/results_bvcc/w2v2_large_xlsr53/layer2/run4/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..272be6c0c4efd0b07597a930c6d49ccfb6b1a10d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/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=0.5403771140636542
+INFO:root:
+[Valid][0][UTT][ MSE = 0.4458 | LCC = 0.7328 | SRCC = 0.7370 ]
+INFO:root:Average loss=0.3670667994480867
+INFO:root:
+[Valid][1][UTT][ MSE = 0.3436 | LCC = 0.7803 | SRCC = 0.7778 ]
+INFO:root:Average loss=0.30482433908260786
+INFO:root:
+[Valid][2][UTT][ MSE = 0.3501 | LCC = 0.8001 | SRCC = 0.7979 ]
+INFO:root:Average loss=0.2911053687716142
+INFO:root:
+[Valid][3][UTT][ MSE = 0.3167 | LCC = 0.8112 | SRCC = 0.8091 ]
+INFO:root:Average loss=0.25995553246675396
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2803 | LCC = 0.8190 | SRCC = 0.8171 ]
+INFO:root:Average loss=0.26645475167494553
+INFO:root:
+[Valid][5][UTT][ MSE = 0.3318 | LCC = 0.8268 | SRCC = 0.8245 ]
+INFO:root:Average loss=0.2575546163014876
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2898 | LCC = 0.8342 | SRCC = 0.8304 ]
+INFO:root:Average loss=0.22878147738102156
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2919 | LCC = 0.8185 | SRCC = 0.8311 ]
+INFO:root:Average loss=0.21569628287584353
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2589 | LCC = 0.8407 | SRCC = 0.8398 ]
+INFO:root:Average loss=0.21625440300275117
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2759 | LCC = 0.8408 | SRCC = 0.8429 ]
+INFO:root:Average loss=0.21570277787171876
+INFO:root:
+[Valid][10][UTT][ MSE = 0.3045 | LCC = 0.8282 | SRCC = 0.8394 ]
+INFO:root:Average loss=0.20092009829405028
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2666 | LCC = 0.8387 | SRCC = 0.8465 ]
+INFO:root:Average loss=0.19545073444262528
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2332 | LCC = 0.8514 | SRCC = 0.8528 ]
+INFO:root:Average loss=0.18475950566621926
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2572 | LCC = 0.8533 | SRCC = 0.8541 ]
+INFO:root:Average loss=0.178208499478224
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2445 | LCC = 0.8572 | SRCC = 0.8563 ]
+INFO:root:Average loss=0.19069231521242705
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2439 | LCC = 0.8531 | SRCC = 0.8557 ]
+INFO:root:Average loss=0.16175546458898446
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2346 | LCC = 0.8600 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.1553522674128031
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2262 | LCC = 0.8599 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.15752447167268166
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2362 | LCC = 0.8601 | SRCC = 0.8582 ]
+INFO:root:Average loss=0.16854566746415237
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2478 | LCC = 0.8584 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.14759149087163118
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2360 | LCC = 0.8549 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.13201203397833383
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2324 | LCC = 0.8590 | SRCC = 0.8618 ]
+INFO:root:Average loss=0.1315164377865119
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2402 | LCC = 0.8551 | SRCC = 0.8558 ]
+INFO:root:Average loss=0.12691326821461701
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2305 | LCC = 0.8556 | SRCC = 0.8579 ]
+INFO:root:Average loss=0.12245874985670432
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2301 | LCC = 0.8605 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.12064063042784348
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2254 | LCC = 0.8603 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.1219965976018172
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2339 | LCC = 0.8588 | SRCC = 0.8584 ]
+INFO:root:Average loss=0.1108860911276096
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2451 | LCC = 0.8610 | SRCC = 0.8606 ]
+INFO:root:Average loss=0.10282768777165657
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2286 | LCC = 0.8611 | SRCC = 0.8625 ]
+INFO:root:Average loss=0.10348600540787746
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2311 | LCC = 0.8576 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.10839441552376136
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2161 | LCC = 0.8617 | SRCC = 0.8644 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2122 | LCC = 0.8702 | SRCC = 0.8690 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run0/config.gin b/results_bvcc/w2v2_large_xlsr53/layer20/run0/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/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_large_xlsr53/layer20/run0/model.pt b/results_bvcc/w2v2_large_xlsr53/layer20/run0/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d2c25b83922cd41674cd3be2b8d1ad50d0279ec9
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/run0/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1da9e0a9c5c61660d908a9bcaa719e6ac2d86e1f12ee5139223326e9513095d
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run0/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer20/run0/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..5a10bd325041da14f9038c1fa2a8630ee1042a06
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/run0/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc3f0a13bbb93c8eb22bea0e70193248020f6d1bce96bade7fde3eee8a2d73a0
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run0/train.log b/results_bvcc/w2v2_large_xlsr53/layer20/run0/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..565ec285af9224a54869904e68a0fbdac5116e50
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/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=0.5310186048348745
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3646 | LCC = 0.8102 | SRCC = 0.8087 ]
+INFO:root:Average loss=0.3283000038220332
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2828 | LCC = 0.8320 | SRCC = 0.8314 ]
+INFO:root:Average loss=0.26916085489285296
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2831 | LCC = 0.8420 | SRCC = 0.8408 ]
+INFO:root:Average loss=0.2617176041389123
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2999 | LCC = 0.8477 | SRCC = 0.8454 ]
+INFO:root:Average loss=0.24296084944254312
+INFO:root:
+[Valid][4][UTT][ MSE = 0.3781 | LCC = 0.8483 | SRCC = 0.8477 ]
+INFO:root:Average loss=0.2559342365234326
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2539 | LCC = 0.8496 | SRCC = 0.8496 ]
+INFO:root:Average loss=0.23808298909511322
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2828 | LCC = 0.8497 | SRCC = 0.8482 ]
+INFO:root:Average loss=0.23274561151480064
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2946 | LCC = 0.8470 | SRCC = 0.8455 ]
+INFO:root:Average loss=0.21520998424444443
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2716 | LCC = 0.8532 | SRCC = 0.8510 ]
+INFO:root:Average loss=0.22193758476238984
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2417 | LCC = 0.8518 | SRCC = 0.8485 ]
+INFO:root:Average loss=0.20291762130382734
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2380 | LCC = 0.8535 | SRCC = 0.8522 ]
+INFO:root:Average loss=0.21668207702728418
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2376 | LCC = 0.8572 | SRCC = 0.8571 ]
+INFO:root:Average loss=0.22970228241040155
+INFO:root:
+[Valid][12][UTT][ MSE = 0.3055 | LCC = 0.8591 | SRCC = 0.8575 ]
+INFO:root:Average loss=0.18984982657891053
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2566 | LCC = 0.8595 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.1920203386973112
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2419 | LCC = 0.8561 | SRCC = 0.8521 ]
+INFO:root:Average loss=0.1965953402985365
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2389 | LCC = 0.8621 | SRCC = 0.8598 ]
+INFO:root:Average loss=0.18659448556793043
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2376 | LCC = 0.8603 | SRCC = 0.8554 ]
+INFO:root:Average loss=0.1780585770805677
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2493 | LCC = 0.8529 | SRCC = 0.8507 ]
+INFO:root:Average loss=0.17087783683569002
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2322 | LCC = 0.8572 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.17179253850227746
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2245 | LCC = 0.8635 | SRCC = 0.8601 ]
+INFO:root:Average loss=0.1662751182149618
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2256 | LCC = 0.8593 | SRCC = 0.8561 ]
+INFO:root:Average loss=0.17762407307059336
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2798 | LCC = 0.8632 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.1690527101357778
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2367 | LCC = 0.8639 | SRCC = 0.8605 ]
+INFO:root:Average loss=0.15920619351359513
+INFO:root:
+[Valid][23][UTT][ MSE = 0.3108 | LCC = 0.8608 | SRCC = 0.8595 ]
+INFO:root:Average loss=0.15111040199796358
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2316 | LCC = 0.8555 | SRCC = 0.8548 ]
+INFO:root:Average loss=0.155826329993896
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2314 | LCC = 0.8562 | SRCC = 0.8542 ]
+INFO:root:Average loss=0.15303680998010513
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2315 | LCC = 0.8580 | SRCC = 0.8537 ]
+INFO:root:Average loss=0.14303776191977355
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2333 | LCC = 0.8614 | SRCC = 0.8565 ]
+INFO:root:Average loss=0.13736250748236975
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2471 | LCC = 0.8548 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.13766746137004632
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2266 | LCC = 0.8565 | SRCC = 0.8559 ]
+INFO:root:Average loss=0.131136689048547
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2448 | LCC = 0.8552 | SRCC = 0.8517 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2419 | LCC = 0.8697 | SRCC = 0.8733 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run1/config.gin b/results_bvcc/w2v2_large_xlsr53/layer20/run1/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/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_large_xlsr53/layer20/run1/model.pt b/results_bvcc/w2v2_large_xlsr53/layer20/run1/model.pt
new file mode 100644
index 0000000000000000000000000000000000000000..841f4d6a6ec4794fc9e2b02085286877962a253b
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/run1/model.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72dc3d3d27d94c270452fd5e0813baffb9e030780bde9600de46de6e23c28ace
+size 3403339
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run1/model_best.pt b/results_bvcc/w2v2_large_xlsr53/layer20/run1/model_best.pt
new file mode 100644
index 0000000000000000000000000000000000000000..870760ae93048241ec5cb1987ffc178aa25a0af0
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/run1/model_best.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c531e9648fccbbfa820059f13deda9c37902101b38780bdf0d67ffb3dd3dddb
+size 3404998
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run1/train.log b/results_bvcc/w2v2_large_xlsr53/layer20/run1/train.log
new file mode 100644
index 0000000000000000000000000000000000000000..a7928ef3ebddb552cd2e827615cc8a2ab164417d
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/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=0.6168944251078826
+INFO:root:
+[Valid][0][UTT][ MSE = 0.3525 | LCC = 0.7863 | SRCC = 0.7968 ]
+INFO:root:Average loss=0.30537938269285053
+INFO:root:
+[Valid][1][UTT][ MSE = 0.2850 | LCC = 0.8290 | SRCC = 0.8343 ]
+INFO:root:Average loss=0.2899744439965639
+INFO:root:
+[Valid][2][UTT][ MSE = 0.2604 | LCC = 0.8386 | SRCC = 0.8413 ]
+INFO:root:Average loss=0.2945835787134293
+INFO:root:
+[Valid][3][UTT][ MSE = 0.2970 | LCC = 0.8438 | SRCC = 0.8443 ]
+INFO:root:Average loss=0.2387998872078382
+INFO:root:
+[Valid][4][UTT][ MSE = 0.2742 | LCC = 0.8460 | SRCC = 0.8489 ]
+INFO:root:Average loss=0.24168413571822336
+INFO:root:
+[Valid][5][UTT][ MSE = 0.2885 | LCC = 0.8505 | SRCC = 0.8493 ]
+INFO:root:Average loss=0.23345306620765954
+INFO:root:
+[Valid][6][UTT][ MSE = 0.2392 | LCC = 0.8549 | SRCC = 0.8538 ]
+INFO:root:Average loss=0.22555589981568167
+INFO:root:
+[Valid][7][UTT][ MSE = 0.2421 | LCC = 0.8570 | SRCC = 0.8547 ]
+INFO:root:Average loss=0.22779112423841769
+INFO:root:
+[Valid][8][UTT][ MSE = 0.2601 | LCC = 0.8549 | SRCC = 0.8546 ]
+INFO:root:Average loss=0.20734750345731393
+INFO:root:
+[Valid][9][UTT][ MSE = 0.2590 | LCC = 0.8512 | SRCC = 0.8480 ]
+INFO:root:Average loss=0.20908308831545022
+INFO:root:
+[Valid][10][UTT][ MSE = 0.2619 | LCC = 0.8560 | SRCC = 0.8570 ]
+INFO:root:Average loss=0.1982511681241867
+INFO:root:
+[Valid][11][UTT][ MSE = 0.2356 | LCC = 0.8601 | SRCC = 0.8572 ]
+INFO:root:Average loss=0.19380512652106774
+INFO:root:
+[Valid][12][UTT][ MSE = 0.2534 | LCC = 0.8604 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.19008538232017785
+INFO:root:
+[Valid][13][UTT][ MSE = 0.2221 | LCC = 0.8624 | SRCC = 0.8604 ]
+INFO:root:Average loss=0.19825636194302484
+INFO:root:
+[Valid][14][UTT][ MSE = 0.2719 | LCC = 0.8584 | SRCC = 0.8585 ]
+INFO:root:Average loss=0.19132152209297204
+INFO:root:
+[Valid][15][UTT][ MSE = 0.2528 | LCC = 0.8586 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.1818822694894595
+INFO:root:
+[Valid][16][UTT][ MSE = 0.2426 | LCC = 0.8658 | SRCC = 0.8627 ]
+INFO:root:Average loss=0.167657654827986
+INFO:root:
+[Valid][17][UTT][ MSE = 0.2487 | LCC = 0.8606 | SRCC = 0.8576 ]
+INFO:root:Average loss=0.17315203782457572
+INFO:root:
+[Valid][18][UTT][ MSE = 0.2280 | LCC = 0.8611 | SRCC = 0.8587 ]
+INFO:root:Average loss=0.16842666554909486
+INFO:root:
+[Valid][19][UTT][ MSE = 0.2348 | LCC = 0.8619 | SRCC = 0.8600 ]
+INFO:root:Average loss=0.1568243329723676
+INFO:root:
+[Valid][20][UTT][ MSE = 0.2446 | LCC = 0.8608 | SRCC = 0.8586 ]
+INFO:root:Average loss=0.16591870584166968
+INFO:root:
+[Valid][21][UTT][ MSE = 0.2359 | LCC = 0.8596 | SRCC = 0.8549 ]
+INFO:root:Average loss=0.15941721783616605
+INFO:root:
+[Valid][22][UTT][ MSE = 0.2333 | LCC = 0.8628 | SRCC = 0.8611 ]
+INFO:root:Average loss=0.14707703305742678
+INFO:root:
+[Valid][23][UTT][ MSE = 0.2396 | LCC = 0.8556 | SRCC = 0.8556 ]
+INFO:root:Average loss=0.14017440073001078
+INFO:root:
+[Valid][24][UTT][ MSE = 0.2356 | LCC = 0.8644 | SRCC = 0.8615 ]
+INFO:root:Average loss=0.14041211713965124
+INFO:root:
+[Valid][25][UTT][ MSE = 0.2631 | LCC = 0.8561 | SRCC = 0.8592 ]
+INFO:root:Average loss=0.14984417362855032
+INFO:root:
+[Valid][26][UTT][ MSE = 0.2221 | LCC = 0.8656 | SRCC = 0.8624 ]
+INFO:root:Average loss=0.14128602429842338
+INFO:root:
+[Valid][27][UTT][ MSE = 0.2381 | LCC = 0.8652 | SRCC = 0.8626 ]
+INFO:root:Average loss=0.12980026360123587
+INFO:root:
+[Valid][28][UTT][ MSE = 0.2171 | LCC = 0.8613 | SRCC = 0.8574 ]
+INFO:root:Average loss=0.13083719070523214
+INFO:root:
+[Valid][29][UTT][ MSE = 0.2489 | LCC = 0.8594 | SRCC = 0.8577 ]
+INFO:root:Average loss=0.12593997938510698
+INFO:root:
+[Valid][30][UTT][ MSE = 0.2284 | LCC = 0.8610 | SRCC = 0.8574 ]
+INFO:root:
+[Test][31][UTT][ MSE = 0.2454 | LCC = 0.8768 | SRCC = 0.8787 ]
diff --git a/results_bvcc/w2v2_large_xlsr53/layer20/run2/config.gin b/results_bvcc/w2v2_large_xlsr53/layer20/run2/config.gin
new file mode 100644
index 0000000000000000000000000000000000000000..d36a6964ab640c490fd7442e2d46a6fbe5fc84aa
--- /dev/null
+++ b/results_bvcc/w2v2_large_xlsr53/layer20/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