.

GPT_SoVITS/Accelerate/MLX/__init__.py

@@ -4,7 +4,7 @@ if importlib.util.find_spec("mlx") is not None:
     from .sample_funcs_mlx import sample_naive as sample_naive_mlx
     from .t2s_engine_mlx import T2SEngine as T2SEngineMLX
 
-    backends = ["mlx_static", "mlx_quantized", "mlx_varlen"]
+    backends = ["mlx_static", "mlx_quantized_mxfp4", "mlx_quantized_affine", "mlx_varlen"]
 else:
     backends = []
 

GPT_SoVITS/Accelerate/MLX/backends/mlx_quantized.py

@@ -164,9 +164,18 @@ class T2SDecoder(T2SDecoderABC):
         self.kv_class = KVCacheHND
         self.group_size = 32
         self.bits = 8
+        self.mode = "affine"
+
+    def set_mode(self, mode: str):
+        assert mode in ["affine", "mxfp4"]
+        self.mode = mode
+        if self.mode == "mxfp4":
+            self.bits = 4
+        else:
+            self.bits = 8
 
     def quantized(self):
-        nn.quantize(self, self.group_size, self.bits)
+        nn.quantize(self, self.group_size, self.bits, mode=self.mode)
         # for layer in self.h.layers:
         #     nn.quantize(layer.feed_forward, self.group_size, self.bits)
         #     nn.quantize(layer.attention, self.group_size, self.bits)

GPT_SoVITS/Accelerate/MLX/t2s_engine_mlx.py

@@ -213,7 +213,7 @@ class T2SEngine(T2SEngineProtocol):
                 decoder_cls: type[T2SDecoderABC] = mlx_varlen.T2SDecoder
             case "MLX-Static":
                 decoder_cls = mlx_static.T2SDecoder
-            case "MLX-Quantized":
+            case "MLX-Quantized-Affine" | "MLX-Quantized-MXFP4":
                 decoder_cls = mlx_quantized.T2SDecoder
             case _:
                 raise RuntimeError(f"Backend {backend} Not Found")
@@ -226,6 +226,12 @@ class T2SEngine(T2SEngineProtocol):
         mx.eval(decoder)
 
         if "Quantized" in backend and isinstance(decoder, mlx_quantized.T2SDecoder):
+            if backend == "MLX-Quantized-Affine":
+                decoder.set_mode("affine")
+            elif backend == "MLX-Quantized-MXFP4":
+                decoder.set_mode("mxfp4")
+            else:
+                raise RuntimeError(f"Quantized Backend {backend} Not Supported")
             decoder.quantized()
             mx.eval(decoder)
 

GPT_SoVITS/Accelerate/__init__.py

@@ -6,7 +6,13 @@ from .PyTorch.structs import T2SEngineProtocol
 backends = PyTorch.backends + MLX.backends
 
 backends = [
-    b.replace("_", "-").title().replace("Mlx", "MLX").replace("Mps", "MPS").replace("Cuda", "CUDA") for b in backends
+    b.replace("_", "-")
+    .title()
+    .replace("Mlx", "MLX")
+    .replace("Mps", "MPS")
+    .replace("Cuda", "CUDA")
+    .replace("Mxfp4", "MXFP4")
+    for b in backends
 ]
 
 

config.py

@@ -80,9 +80,6 @@ def get_weights_names(i18n):
     for key, value in name2sovits_path.items():
         if os.path.exists(value):
             SoVITS_names.append((i18n(key), value))
-    for key, value in pretrained_sovits_name.items():
-        if key in {"v3", "v4", "v1"}:
-            SoVITS_names.append((value, value))
     for path in SoVITS_weight_root:
         if not os.path.exists(path):
             continue
@@ -93,15 +90,20 @@ def get_weights_names(i18n):
     for key, value in name2gpt_path.items():
         if os.path.exists(value):
             GPT_names.append((i18n(key), value))
-    GPT_names.append((pretrained_gpt_name["v1"], pretrained_gpt_name["v1"]))
     for path in GPT_weight_root:
         if not os.path.exists(path):
             continue
         for name in os.listdir(path):
             if name.endswith(".ckpt"):
                 GPT_names.append((f"{path}/{name}", f"{path}/{name}"))
+
     SoVITS_names = sorted(SoVITS_names, key=custom_sort_key)
     GPT_names = sorted(GPT_names, key=custom_sort_key)
+
+    for key, value in pretrained_sovits_name.items():
+        if key in {"v3", "v4", "v1"}:
+            SoVITS_names.append((value, value))
+    GPT_names.append((pretrained_gpt_name["v1"], pretrained_gpt_name["v1"]))
     return SoVITS_names, GPT_names