enable fp16 for all inference scripts by default

Plachtaa · Nov 28, 2024 · ec0b27f · ec0b27f
1 parent c83ae7a
commit ec0b27f
Show file tree

Hide file tree

Showing 7 changed files with 124 additions and 87 deletions.
diff --git a/README.md b/README.md
@@ -45,6 +45,7 @@ python inference.py --source <source-wav>
 --semi-tone-shift 0 # pitch shift in semitones for singing voice conversion
 --checkpoint <path-to-checkpoint>
 --config <path-to-config>
+ --fp16 True
 ```
 where:
 - `source` is the path to the speech file to convert to reference voice
@@ -58,11 +59,11 @@ where:
 - `semi-tone-shift` is the pitch shift in semitones for singing voice conversion, default is 0  
 - `checkpoint` is the path to the model checkpoint if you have trained or fine-tuned your own model, leave to blank to auto-download default model from huggingface.(`seed-uvit-whisper-small-wavenet` if `f0-condition` is `False` else `seed-uvit-whisper-base`)
 - `config` is the path to the model config if you have trained or fine-tuned your own model, leave to blank to auto-download default config from huggingface  
-
+- `fp16` is the flag to use float16 inference, default is True
 
 Voice Conversion Web UI:
 ```bash
-python app_vc.py --checkpoint <path-to-checkpoint> --config <path-to-config>
+python app_vc.py --checkpoint <path-to-checkpoint> --config <path-to-config> --fp16 True
 ```
 - `checkpoint` is the path to the model checkpoint if you have trained or fine-tuned your own model, leave to blank to auto-download default model from huggingface. (`seed-uvit-whisper-small-wavenet`)
 - `config` is the path to the model config if you have trained or fine-tuned your own model, leave to blank to auto-download default config from huggingface  
@@ -71,7 +72,7 @@ Then open the browser and go to `http://localhost:7860/` to use the web interfac
 
 Singing Voice Conversion Web UI:
 ```bash
-python app_svc.py --checkpoint <path-to-checkpoint> --config <path-to-config>
+python app_svc.py --checkpoint <path-to-checkpoint> --config <path-to-config> --fp16 True
 ```
 - `checkpoint` is the path to the model checkpoint if you have trained or fine-tuned your own model, leave to blank to auto-download default model from huggingface. (`seed-uvit-whisper-base`)
 - `config` is the path to the model config if you have trained or fine-tuned your own model, leave to blank to auto-download default config from huggingface  

diff --git a/app.py b/app.py
@@ -269,13 +269,14 @@ def voice_conversion(source, target, diffusion_steps, length_adjust, inference_c
         chunk_cond = cond[:, processed_frames:processed_frames + max_source_window]
         is_last_chunk = processed_frames + max_source_window >= cond.size(1)
         cat_condition = torch.cat([prompt_condition, chunk_cond], dim=1)
-        # Voice Conversion
-        vc_target = inference_module.cfm.inference(cat_condition,
-                                                   torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
-                                                   mel2, style2, None, diffusion_steps,
-                                                   inference_cfg_rate=inference_cfg_rate)
-        vc_target = vc_target[:, :, mel2.size(-1):]
-        vc_wave = bigvgan_fn(vc_target)[0]
+        with torch.autocast(device_type=device.type, dtype=torch.float16):
+            # Voice Conversion
+            vc_target = inference_module.cfm.inference(cat_condition,
+                                                       torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
+                                                       mel2, style2, None, diffusion_steps,
+                                                       inference_cfg_rate=inference_cfg_rate)
+            vc_target = vc_target[:, :, mel2.size(-1):]
+            vc_wave = bigvgan_fn(vc_target)[0]
         if processed_frames == 0:
             if is_last_chunk:
                 output_wave = vc_wave[0].cpu().numpy()

diff --git a/app_svc.py b/app_svc.py
@@ -4,7 +4,7 @@
 import torch
 import torchaudio
 import librosa
-from modules.commons import build_model, load_checkpoint, recursive_munch
+from modules.commons import build_model, load_checkpoint, recursive_munch, str2bool
 import yaml
 from hf_utils import load_custom_model_from_hf
 import numpy as np
@@ -13,28 +13,21 @@
 # Load model and configuration
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
-# Load additional modules
-from modules.campplus.DTDNN import CAMPPlus
-
-campplus_ckpt_path = load_custom_model_from_hf("funasr/campplus", "campplus_cn_common.bin", config_filename=None)
-campplus_model = CAMPPlus(feat_dim=80, embedding_size=192)
-campplus_model.load_state_dict(torch.load(campplus_ckpt_path, map_location="cpu"))
-campplus_model.eval()
-campplus_model.to(device)
-
-from modules.audio import mel_spectrogram
-
+fp16 = False
 def load_models(args):
-    global sr, hop_length
+    global sr, hop_length, fp16
+    fp16 = args.fp16
+    print(f"Using device: {device}")
+    print(f"Using fp16: {fp16}")
     # f0 conditioned model
     if args.checkpoint_path is None or args.checkpoint_path == "":
         dit_checkpoint_path, dit_config_path = load_custom_model_from_hf("Plachta/Seed-VC",
                                                                          "DiT_seed_v2_uvit_whisper_base_f0_44k_bigvgan_pruned_ft_ema_v2.pth",
                                                                          "config_dit_mel_seed_uvit_whisper_base_f0_44k.yml")
     else:
+        print(f"Using custom checkpoint: {args.checkpoint_path}")
         dit_checkpoint_path = args.checkpoint_path
         dit_config_path = args.config_path
-
     config = yaml.safe_load(open(dit_config_path, "r"))
     model_params = recursive_munch(config["model_params"])
     model_params.dit_type = 'DiT'
@@ -336,13 +329,14 @@ def voice_conversion(source, target, diffusion_steps, length_adjust, inference_c
         chunk_f0 = interpolated_shifted_f0_alt[:, processed_frames:processed_frames + max_source_window]
         is_last_chunk = processed_frames + max_source_window >= cond.size(1)
         cat_condition = torch.cat([prompt_condition, chunk_cond], dim=1)
-        # Voice Conversion
-        vc_target = inference_module.cfm.inference(cat_condition,
-                                                   torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
-                                                   mel2, style2, None, diffusion_steps,
-                                                   inference_cfg_rate=inference_cfg_rate)
-        vc_target = vc_target[:, :, mel2.size(-1):]
-        vc_wave = vocoder_fn(vc_target).squeeze().cpu()
+        with torch.autocast(device_type=device.type, dtype=torch.float16 if fp16 else torch.float32):
+            # Voice Conversion
+            vc_target = inference_module.cfm.inference(cat_condition,
+                                                       torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
+                                                       mel2, style2, None, diffusion_steps,
+                                                       inference_cfg_rate=inference_cfg_rate)
+            vc_target = vc_target[:, :, mel2.size(-1):]
+            vc_wave = vocoder_fn(vc_target).squeeze().cpu()
         if vc_wave.ndim == 1:
             vc_wave = vc_wave.unsqueeze(0)
         if processed_frames == 0:
@@ -437,6 +431,7 @@ def main(args):
     parser = argparse.ArgumentParser()
     parser.add_argument("--checkpoint-path", type=str, help="Path to the checkpoint file", default=None)
     parser.add_argument("--config-path", type=str, help="Path to the config file", default=None)
-    parser.add_argument("--share", type=bool, help="Whether to share url link", default=False)
+    parser.add_argument("--share", type=str2bool, nargs="?", const=True, default=False, help="Whether to share the app")
+    parser.add_argument("--fp16", type=str2bool, nargs="?", const=True, help="Whether to use fp16", default=True)
     args = parser.parse_args()
     main(args)
diff --git a/app_vc.py b/app_vc.py
@@ -4,7 +4,7 @@
 import torch
 import torchaudio
 import librosa
-from modules.commons import build_model, load_checkpoint, recursive_munch
+from modules.commons import build_model, load_checkpoint, recursive_munch, str2bool
 import yaml
 from hf_utils import load_custom_model_from_hf
 import numpy as np
@@ -13,9 +13,12 @@
 
 # Load model and configuration
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
+fp16 = False
 def load_models(args):
-    global sr, hop_length
+    global sr, hop_length, fp16
+    fp16 = args.fp16
+    print(f"Using device: {device}")
+    print(f"Using fp16: {fp16}")
     if args.checkpoint_path is None or args.checkpoint_path == "":
         dit_checkpoint_path, dit_config_path = load_custom_model_from_hf("Plachta/Seed-VC",
                                                                          "DiT_seed_v2_uvit_whisper_small_wavenet_bigvgan_pruned.pth",
@@ -285,13 +288,14 @@ def voice_conversion(source, target, diffusion_steps, length_adjust, inference_c
         chunk_cond = cond[:, processed_frames:processed_frames + max_source_window]
         is_last_chunk = processed_frames + max_source_window >= cond.size(1)
         cat_condition = torch.cat([prompt_condition, chunk_cond], dim=1)
-        # Voice Conversion
-        vc_target = inference_module.cfm.inference(cat_condition,
-                                                   torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
-                                                   mel2, style2, None, diffusion_steps,
-                                                   inference_cfg_rate=inference_cfg_rate)
-        vc_target = vc_target[:, :, mel2.size(-1):]
-        vc_wave = vocoder_fn(vc_target)[0]
+        with torch.autocast(device_type=device.type, dtype=torch.float16 if fp16 else torch.float32):
+            # Voice Conversion
+            vc_target = inference_module.cfm.inference(cat_condition,
+                                                       torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
+                                                       mel2, style2, None, diffusion_steps,
+                                                       inference_cfg_rate=inference_cfg_rate)
+            vc_target = vc_target[:, :, mel2.size(-1):]
+            vc_wave = vocoder_fn(vc_target)[0]
         if vc_wave.ndim == 1:
             vc_wave = vc_wave.unsqueeze(0)
         if processed_frames == 0:
@@ -380,6 +384,7 @@ def main(args):
     parser = argparse.ArgumentParser()
     parser.add_argument("--checkpoint-path", type=str, help="Path to the checkpoint file", default=None)
     parser.add_argument("--config-path", type=str, help="Path to the config file", default=None)
-    parser.add_argument("--share", type=bool, help="Whether to share url link", default=False)
+    parser.add_argument("--share", type=str2bool, nargs="?", const=True, default=False, help="Whether to share the app")
+    parser.add_argument("--fp16", type=str2bool, nargs="?", const=True, help="Whether to use fp16", default=True)
     args = parser.parse_args()
     main(args)
diff --git a/inference.py b/inference.py
@@ -16,15 +16,17 @@
 
 import torchaudio
 import librosa
-import torchaudio.compliance.kaldi as kaldi
+from modules.commons import str2bool
 
 from hf_utils import load_custom_model_from_hf
 
 
 # Load model and configuration
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
+fp16 = False
 def load_models(args):
+    global fp16
+    fp16 = args.fp16
     if not args.f0_condition:
         dit_checkpoint_path, dit_config_path = load_custom_model_from_hf("Plachta/Seed-VC",
                                                                          "DiT_seed_v2_uvit_whisper_small_wavenet_bigvgan_pruned.pth",
@@ -304,17 +306,17 @@ def main(args):
     cat_condition = torch.cat([prompt_condition, cond], dim=1)
 
     time_vc_start = time.time()
-    vc_target = model.cfm.inference(
-            cat_condition,
-            torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
-            mel2, style2, None, diffusion_steps,
-            inference_cfg_rate=inference_cfg_rate)
-    vc_target = vc_target[:, :, mel2.size(-1):]
-
-
-    # Convert to waveform
-    vc_wave = vocoder_fn(vc_target).squeeze()  # wav_gen is FloatTensor with shape [B(1), 1, T_time] and values in [-1, 1]
-    vc_wave = vc_wave[None, :]
+    with torch.autocast(device_type=device.type, dtype=torch.float16 if fp16 else torch.float32):
+        vc_target = model.cfm.inference(
+                cat_condition,
+                torch.LongTensor([cat_condition.size(1)]).to(mel2.device),
+                mel2, style2, None, diffusion_steps,
+                inference_cfg_rate=inference_cfg_rate)
+        vc_target = vc_target[:, :, mel2.size(-1):]
+
+        # Convert to waveform
+        vc_wave = vocoder_fn(vc_target).squeeze()  # wav_gen is FloatTensor with shape [B(1), 1, T_time] and values in [-1, 1]
+    vc_wave = vc_wave[None, :].float()
     time_vc_end = time.time()
     print(f"RTF: {(time_vc_end - time_vc_start) / vc_wave.size(-1) * sr}")
 
@@ -332,10 +334,11 @@ def main(args):
     parser.add_argument("--diffusion-steps", type=int, default=30)
     parser.add_argument("--length-adjust", type=float, default=1.0)
     parser.add_argument("--inference-cfg-rate", type=float, default=0.7)
-    parser.add_argument("--f0-condition", type=bool, default=False)
-    parser.add_argument("--auto-f0-adjust", type=bool, default=True)
+    parser.add_argument("--f0-condition", type=str2bool, default=True)
+    parser.add_argument("--auto-f0-adjust", type=str2bool, default=True)
     parser.add_argument("--semi-tone-shift", type=int, default=0)
     parser.add_argument("--checkpoint-path", type=str, help="Path to the checkpoint file", default=None)
     parser.add_argument("--config-path", type=str, help="Path to the config file", default=None)
+    parser.add_argument("--fp16", type=str2bool, default=True)
     args = parser.parse_args()
     main(args)
diff --git a/modules/commons.py b/modules/commons.py
@@ -5,7 +5,17 @@
 from torch.nn import functional as F
 from munch import Munch
 import json
-
+import argparse
+
+def str2bool(v):
+    if isinstance(v, bool):
+        return v
+    if v.lower() in ("yes", "true", "t", "y", "1"):
+        return True
+    elif v.lower() in ("no", "false", "f", "n", "0"):
+        return False
+    else:
+        raise argparse.ArgumentTypeError("Boolean value expected.")
 
 class AttrDict(dict):
     def __init__(self, *args, **kwargs):