add model

model_index.json

@@ -0,0 +1,21 @@
+{
+  "_class_name": "GradTTS",
+  "_diffusers_version": "0.0.3",
+  "_module": "pipeline_grad_tts",
+  "noise_scheduler": [
+    "diffusers",
+    "GradTTSScheduler"
+  ],
+  "text_encoder": [
+    "pipeline_grad_tts",
+    "TextEncoder"
+  ],
+  "tokenizer": [
+    "grad_tts_utils",
+    "GradTTSTokenizer"
+  ],
+  "unet": [
+    "diffusers",
+    "UNetGradTTSModel"
+  ]
+}

noise_scheduler/scheduler_config.json

@@ -0,0 +1,7 @@
+{
+  "_class_name": "GradTTSScheduler",
+  "_diffusers_version": "0.0.3",
+  "beta_end": 20.0,
+  "beta_start": 0.05,
+  "timesteps": 50
+}

text_encoder/config.json

@@ -0,0 +1,17 @@
+{
+  "_class_name": "TextEncoder",
+  "_diffusers_version": "0.0.3",
+  "filter_channels": 768,
+  "filter_channels_dp": 256,
+  "kernel_size": 3,
+  "n_channels": 192,
+  "n_feats": 80,
+  "n_heads": 2,
+  "n_layers": 6,
+  "n_spks": 1,
+  "n_vocab": 149,
+  "name_or_path": "grad-tts/text_encoder",
+  "p_dropout": 0.1,
+  "spk_emb_dim": 64,
+  "window_size": 4
+}

text_encoder/diffusion_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:00f47ce1fe4f1f2eee3b3f3716d0382bbfc8c56708887f721a44ff4944cb88f3
+size 28845619

tokenizer/special_tokens_map.json

@@ -0,0 +1 @@
+{}

tokenizer/tokenizer_config.json

@@ -0,0 +1,5 @@
+{
+  "name_or_path": "grad-tts/tokenizer",
+  "special_tokens_map_file": "./grad-tts-tokenizer/special_tokens_map.json",
+  "tokenizer_class": "GradTTSTokenizer"
+}

unet/config.json

@@ -0,0 +1,16 @@
+{
+  "_class_name": "UNetGradTTSModel",
+  "_diffusers_version": "0.0.3",
+  "dim": 64,
+  "dim_mults": [
+    1,
+    2,
+    4
+  ],
+  "groups": 8,
+  "n_feats": 80,
+  "n_spks": 247,
+  "name_or_path": "grad-tts/unet",
+  "pe_scale": 1000,
+  "spk_emb_dim": 64
+}

unet/diffusion_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df39c3b01dbc76477f61a5403663a00df92e9c26cbbd6974fce3e8801a5084e4
+size 30816799