Upload StyleTTS2 checkpoint epoch_2nd_00004.pth with all inference components

README.md

@@ -23,9 +23,9 @@ This model is a fine-tuned version of StyleTTS2, containing all necessary compon
 - **Completed Epochs:** 4
 - **Total Iterations:** 1695
 - **Batch Size:** 2
-- **Max Length:** 120
+- **Max Length:** 650
 - **Learning Rate:** 0.0001
-- **Final Validation Loss:** 0.405064
+- **Final Validation Loss:** 0.378118
 
 ## Model Components
 The repository includes all necessary components for inference:

bert.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e49e69e5264f2a0e7cbb6a1b499a5aa1aa46eef0eccc3f7a012c29e9886beeb
+oid sha256:a24f11aa84caa53c97528d60e23c8fde361ac70509b1856708a877708487ed9a
 size 25178740

bert_encoder.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5047745d7308724dae2c71c11df03953ea5bfa2a750722fb63df041b215957d6
+oid sha256:bdcf25066d8db01ce4a3d1bffda2f18b8ec79362bbd79fca31cb6c67a4b47b74
 size 1576502

checkpoint.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2cd8fda54995a54c0eb58f04c7b3289a57ffe133e5818031cd7234dc1ca4c096
+oid sha256:6ba745754d758823a16f82d32961942c42be722ed4c28ffb92c89713828bfd49
 size 2040092478

config.json

@@ -74,7 +74,7 @@
   "training_config": {
     "epochs": 5,
     "batch_size": 2,
-    "max_len": 120,
+    "max_len": 650,
     "optimizer": {
       "bert_lr": 1e-05,
       "ft_lr": 0.0001,
@@ -114,30 +114,30 @@
   "model_state": {
     "epoch": 4,
     "iterations": 1695,
-    "val_loss": 0.40506434440612793
+    "val_loss": 0.37811824679374695
   },
   "training_metrics": {
     "train_loss": [],
     "val_loss": [
-      55.0,
-      45.0,
-      36.0,
-      23.0,
-      8.0
+      52.0,
+      33.0,
+      32.0,
+      43.0,
+      5.0
     ],
     "dur_loss": [
-      0.42,
-      0.416,
-      0.412,
-      0.428,
-      0.405
+      0.409,
+      0.398,
+      0.405,
+      0.396,
+      0.378
     ],
     "F0_loss": [
-      4.102,
-      4.104,
-      4.31,
-      4.422,
-      4.151
+      4.094,
+      4.151,
+      3.98,
+      4.119,
+      4.047
     ],
     "epochs": [
       1,

decoder.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8658d831905e0d0dc0183b3f166f5309ab51dddcf9df9a59f57cbe64fbfe67dc
+oid sha256:0accbfa7a1bfd40fb9a832657a5687c15fc33160c663ec45aea16c117b3071c6
 size 217409318

mpd.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6979b3d82a4b3e253328971e3346c47e694434a2230b33787dcc613e6288a20b
+oid sha256:de1788e93d598211a7646d0a627dd319913a0cd23e7e55f3b2d3be2ed8e6d806
 size 164447824

msd.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b2a72433fa97ca5fc2d05c6417ef758bdc3c6325b0536dbee6c30190196555f
+oid sha256:29ffbcddd26d12835e129d9a846b9a0be9d3b7b612b8d8369d06e9c36a3122ed
 size 1139020

predictor.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5143aa7d6e62813a0a9f2377aea102b0c9f2b6e88a1e2789a2b7f4182307c734
+oid sha256:b55ec3244d0f809ee069f154453ced44fa723b90619f847d3d4ecc69c0f8d3bd
 size 64813639

predictor_encoder.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a09a8c00fa28ad9881c0673bcb58fa0a38f9ad4f7b3a1549e00555a68532605
+oid sha256:b80bf87930cd2a5f39307d200075f57f02a5e02d268af9dedc8d2ef877ce15f5
 size 55547155

style_encoder.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26d6da3ba01021f9141be3055d9494a7c137c4e5382ae62128cd651d7cd4f312
+oid sha256:475c81db13655d9b84cc060f6bd1f721e04c2fb4a521617237564375d94ca38f
 size 55546871

text_aligner.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d287dde07f96fa817e7ba582f70bcc5907be35c687a4cead46bcd06cf8695b23
+oid sha256:4ec7663d9ec3a84a6e6453f9f06db9c53d5ef8814ea145260989960dbdc55983
 size 31531315

text_encoder.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0852ffa6c581525112390042558b97d994c4ba5fa48a7c07fb5d78913c413b4
+oid sha256:c884b97b322548c1a5ebc15e40de97a2fb049a38ea0d55344843c2ad6348239f
 size 22432460