From 6286b0ffc966ac7b774cfdffffe165ffa5352c7c Mon Sep 17 00:00:00 2001
From: kemuriririn <10inspiral@gmail.com>
Date: Wed, 2 Apr 2025 17:40:41 +0800
Subject: [PATCH] =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=97=B6=E5=8A=A0=E8=BD=BDbp?=
 =?UTF-8?q?e=20model=E4=BD=BF=E7=94=A8=E7=9B=B8=E5=AF=B9=E4=BA=8E=E6=A8=A1?=
 =?UTF-8?q?=E5=9E=8B=E6=A0=B9=E7=9B=AE=E5=BD=95=E7=9A=84=E8=B7=AF=E5=BE=84?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 README.md               |   2 -
 checkpoints/config.yaml | 112 ++++++++++++++++++++++++++++++++++++++++
 indextts/infer.py       |   2 +-
 3 files changed, 113 insertions(+), 3 deletions(-)
 create mode 100644 checkpoints/config.yaml

diff --git a/README.md b/README.md
index ad7737c..2d5eea3 100644
--- a/README.md
+++ b/README.md
@@ -105,11 +105,9 @@ apt-get install ffmpeg
 ```
 3. Download models:
 ```bash
-mkdir checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/bigvgan_discriminator.pth -P checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/bigvgan_generator.pth -P checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/bpe.model -P checkpoints
-wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/config.yaml -P checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/dvae.pth -P checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/gpt.pth -P checkpoints
 wget https://huggingface.co/IndexTeam/Index-TTS/resolve/main/unigram_12000.vocab -P checkpoints
diff --git a/checkpoints/config.yaml b/checkpoints/config.yaml
new file mode 100644
index 0000000..e24336d
--- /dev/null
+++ b/checkpoints/config.yaml
@@ -0,0 +1,112 @@
+dataset:
+    bpe_model: bpe.model
+    sample_rate: 24000
+    squeeze: false
+    mel:
+        sample_rate: 24000
+        n_fft: 1024
+        hop_length: 256
+        win_length: 1024
+        n_mels: 100
+        mel_fmin: 0
+        normalize: false
+
+gpt:
+    model_dim: 1024
+    max_mel_tokens: 605
+    max_text_tokens: 402
+    heads: 16
+    use_mel_codes_as_input: true
+    mel_length_compression: 1024
+    layers: 20
+    number_text_tokens: 12000
+    number_mel_codes: 8194
+    start_mel_token: 8192
+    stop_mel_token: 8193
+    start_text_token: 0
+    stop_text_token: 1
+    train_solo_embeddings: false
+    condition_type: "conformer_perceiver"
+    condition_module:
+        output_size: 512
+        linear_units: 2048
+        attention_heads: 8
+        num_blocks: 6
+        input_layer: "conv2d2"
+        perceiver_mult: 2
+
+vqvae:
+    channels: 100
+    num_tokens: 8192
+    hidden_dim: 512
+    num_resnet_blocks: 3
+    codebook_dim: 512
+    num_layers: 2
+    positional_dims: 1
+    kernel_size: 3
+    smooth_l1_loss: true
+    use_transposed_convs: false
+
+bigvgan:
+    adam_b1: 0.8
+    adam_b2: 0.99
+    lr_decay: 0.999998
+    seed: 1234
+
+    resblock: "1"
+    upsample_rates: [4,4,4,4,2,2]
+    upsample_kernel_sizes: [8,8,4,4,4,4]
+    upsample_initial_channel: 1536
+    resblock_kernel_sizes: [3,7,11]
+    resblock_dilation_sizes: [[1,3,5], [1,3,5], [1,3,5]]
+    feat_upsample: false
+    speaker_embedding_dim: 512
+    cond_d_vector_in_each_upsampling_layer: true
+
+    gpt_dim: 1024
+
+    activation: "snakebeta"
+    snake_logscale: true
+
+    use_cqtd_instead_of_mrd: true
+    cqtd_filters: 128
+    cqtd_max_filters: 1024
+    cqtd_filters_scale: 1
+    cqtd_dilations: [1, 2, 4]
+    cqtd_hop_lengths: [512, 256, 256]
+    cqtd_n_octaves: [9, 9, 9]
+    cqtd_bins_per_octaves: [24, 36, 48]
+
+    resolutions: [[1024, 120, 600], [2048, 240, 1200], [512, 50, 240]]
+    mpd_reshapes: [2, 3, 5, 7, 11]
+    use_spectral_norm: false
+    discriminator_channel_mult: 1
+
+    use_multiscale_melloss: true
+    lambda_melloss: 15
+
+    clip_grad_norm: 1000
+
+    segment_size: 16384
+    num_mels: 100
+    num_freq: 1025
+    n_fft: 1024
+    hop_size: 256
+    win_size: 1024
+
+    sampling_rate: 24000
+
+    fmin: 0
+    fmax: null
+    fmax_for_loss: null
+    mel_type: "pytorch"
+
+    num_workers: 2
+    dist_config:
+        dist_backend: "nccl"
+        dist_url: "tcp://localhost:54321"
+        world_size: 1
+
+dvae_checkpoint: dvae.pth
+gpt_checkpoint: gpt.pth
+bigvgan_checkpoint: bigvgan_generator.pth
\ No newline at end of file
diff --git a/indextts/infer.py b/indextts/infer.py
index 6cb432d..f1f419e 100644
--- a/indextts/infer.py
+++ b/indextts/infer.py
@@ -74,7 +74,7 @@ class IndexTTS:
         auto_conditioning = cond_mel
 
         tokenizer = spm.SentencePieceProcessor()
-        tokenizer.load(self.cfg.dataset['bpe_model'])
+        tokenizer.load(os.path.join(self.model_dir,self.cfg.dataset['bpe_model']))
 
         punctuation = ["!", "?", ".", ";", "！", "？", "。", "；"]
         pattern = r"(?<=[{0}])\s*".format("".join(punctuation))