sopro 1.0.1__py3-none-any.whl → 1.0.2__py3-none-any.whl

sopro/__init__.py

@@ -3,4 +3,4 @@ from __future__ import annotations
 from .model import SoproTTS
 
 __all__ = ["SoproTTS"]
-__version__ = "1.0.1"
+__version__ = "1.0.2"

sopro/cli.py

@@ -97,79 +97,80 @@ def main() -> None:
         arr = np.load(args.ref_tokens)
         ref_tokens_tq = torch.from_numpy(arr).long()
 
-    text_ids = tts.encode_text(args.text)
-    ref = tts.encode_reference(
-        ref_audio_path=args.ref_audio,
-        ref_tokens_tq=ref_tokens_tq,
-        ref_seconds=args.ref_seconds,
-    )
+    with torch.inference_mode():
+        text_ids = tts.encode_text(args.text)
+        ref = tts.encode_reference(
+            ref_audio_path=args.ref_audio,
+            ref_tokens_tq=ref_tokens_tq,
+            ref_seconds=args.ref_seconds,
+        )
 
-    prep = tts.model.prepare_conditioning(
-        text_ids,
-        ref,
-        max_frames=args.max_frames,
-        device=tts.device,
-        style_strength=float(
-            args.style_strength
-            if args.style_strength is not None
-            else cfg.style_strength
-        ),
-    )
+        prep = tts.model.prepare_conditioning(
+            text_ids,
+            ref,
+            max_frames=args.max_frames,
+            device=tts.device,
+            style_strength=float(
+                args.style_strength
+                if args.style_strength is not None
+                else cfg.style_strength
+            ),
+        )
 
-    t_start = time.perf_counter()
+        t_start = time.perf_counter()
 
-    hist_A: list[int] = []
-    pbar = tqdm(
-        total=args.max_frames,
-        desc="AR sampling",
-        unit="frame",
-        disable=args.quiet,
-    )
+        hist_A: list[int] = []
+        pbar = tqdm(
+            total=args.max_frames,
+            desc="AR sampling",
+            unit="frame",
+            disable=args.quiet,
+        )
 
-    for _t, rvq1, p_stop in tts.model.ar_stream(
-        prep,
-        max_frames=args.max_frames,
-        top_p=args.top_p,
-        temperature=args.temperature,
-        anti_loop=(not args.no_anti_loop),
-        use_prefix=(not args.no_prefix),
-        prefix_sec_fixed=args.prefix_sec,
-        use_stop_head=(False if args.no_stop_head else None),
-        stop_patience=args.stop_patience,
-        stop_threshold=args.stop_threshold,
-    ):
-        hist_A.append(int(rvq1))
-        pbar.update(1)
-        if p_stop is None:
-            pbar.set_postfix(p_stop="off")
-        else:
-            pbar.set_postfix(p_stop=f"{float(p_stop):.2f}")
-
-    pbar.n = len(hist_A)
-    pbar.close()
-
-    t_after_sampling = time.perf_counter()
-
-    T = len(hist_A)
-    if T == 0:
-        save_audio(args.out, torch.zeros(1, 0), sr=TARGET_SR)
-        t_end = time.perf_counter()
-        if not args.quiet:
-            print(
-                f"[Timing] sampling={t_after_sampling - t_start:.2f}s, "
-                f"postproc+decode+save={t_end - t_after_sampling:.2f}s, "
-                f"total={t_end - t_start:.2f}s"
-            )
-            print(f"[Done] Wrote {args.out}")
-        return
-
-    tokens_A = torch.tensor(hist_A, device=tts.device, dtype=torch.long).unsqueeze(0)
-    cond_seq = prep["cond_all"][:, :T, :]
-    tokens_1xTQ = tts.model.nar_refine(cond_seq, tokens_A)
-    tokens_tq = tokens_1xTQ.squeeze(0)
-
-    wav = tts.codec.decode_full(tokens_tq)
-    save_audio(args.out, wav, sr=TARGET_SR)
+        for _t, rvq1, p_stop in tts.model.ar_stream(
+            prep,
+            max_frames=args.max_frames,
+            top_p=args.top_p,
+            temperature=args.temperature,
+            anti_loop=(not args.no_anti_loop),
+            use_prefix=(not args.no_prefix),
+            prefix_sec_fixed=args.prefix_sec,
+            use_stop_head=(False if args.no_stop_head else None),
+            stop_patience=args.stop_patience,
+            stop_threshold=args.stop_threshold,
+        ):
+            hist_A.append(int(rvq1))
+            pbar.update(1)
+            if p_stop is None:
+                pbar.set_postfix(p_stop="off")
+            else:
+                pbar.set_postfix(p_stop=f"{float(p_stop):.2f}")
+
+        pbar.n = len(hist_A)
+        pbar.close()
+
+        t_after_sampling = time.perf_counter()
+
+        T = len(hist_A)
+        if T == 0:
+            save_audio(args.out, torch.zeros(1, 0), sr=TARGET_SR)
+            t_end = time.perf_counter()
+            if not args.quiet:
+                print(
+                    f"[Timing] sampling={t_after_sampling - t_start:.2f}s, "
+                    f"postproc+decode+save={t_end - t_after_sampling:.2f}s, "
+                    f"total={t_end - t_start:.2f}s"
+                )
+                print(f"[Done] Wrote {args.out}")
+            return
+
+        tokens_A = torch.tensor(hist_A, device=tts.device, dtype=torch.long).unsqueeze(0)
+        cond_seq = prep["cond_all"][:, :T, :]
+        tokens_1xTQ = tts.model.nar_refine(cond_seq, tokens_A)
+        tokens_tq = tokens_1xTQ.squeeze(0)
+
+        wav = tts.codec.decode_full(tokens_tq)
+        save_audio(args.out, wav, sr=TARGET_SR)
 
     t_end = time.perf_counter()
     if not args.quiet:

sopro/model.py

@@ -793,7 +793,7 @@ class SoproTTS:
         )
         return ref
 
-    @torch.no_grad()
+    @torch.inference_mode()
     def synthesize(
         self,
         text: str,

sopro/streaming.py

@@ -145,7 +145,7 @@ class SoproTTSStreamer:
             if wav is not None:
                 yield wav
 
-
+@torch.inference_mode()
 def stream(
     tts: SoproTTS,
     text: str,

{sopro-1.0.1.dist-info → sopro-1.0.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sopro
-Version: 1.0.1
+Version: 1.0.2
 Summary: A lightweight text-to-speech model with zero-shot voice cloning.
 Author-email: Samuel Vitorino <samvitorino@gmail.com>
 License: Apache 2.0

{sopro-1.0.1.dist-info → sopro-1.0.2.dist-info}/RECORD

@@ -1,12 +1,12 @@
-sopro/__init__.py,sha256=NFZuESqdCL7bGXuTB8c61XxUJqhkHPUOSTqzH4pyUfU,110
+sopro/__init__.py,sha256=SreucP3h4V4KsUU8PskOofqKEAmL8RvXYA6Ma53zb8Y,110
 sopro/audio.py,sha256=xlp6aYzzGlOMcNZ-p9lDeeU0TUkSHMcvmLantwg_4-0,4162
-sopro/cli.py,sha256=YKfGalyhbRuvjVrGJuo1NlIC7h8CszlMxuTwhYgUSwQ,5751
+sopro/cli.py,sha256=v4mZ_zfYt6BusVFMIHo3ae32xLw2_O_v5SDkfmt_Cvc,6040
 sopro/config.py,sha256=OBD-k2z5GUdjFS545MyBXx-dAGhwnhRG11LW-zQt1-g,1063
 sopro/constants.py,sha256=wSjFKeFIcLCxyVUVb3njxMK666IuxjlNzVT4_jfPovQ,97
 sopro/hub.py,sha256=xsHfeO8X7v__FELvaQxWHYG8P39ygrgbluPs5GQjoCM,1391
-sopro/model.py,sha256=YXwcVGN3v5T0kvKttmo9WNPpewF-b5aOZoTMVypkzO8,28624
+sopro/model.py,sha256=Rj10OPdx8UEhH2reU2SQ4oTztNJCOvLS7pm84V0E_xo,28631
 sopro/sampling.py,sha256=Q5rbuef_BIuy12cv5J7v6k9ob3zQ0OFJIlMHssOkiuU,2951
-sopro/streaming.py,sha256=O5Kkl4cUBjzgjTrEwQK2ka5h6sgcYaEZmIp66-obcPM,4975
+sopro/streaming.py,sha256=AgPzaxdimeDT-8potXAMXuRi7zrWUGgxViwHJw2R2Lk,4998
 sopro/tokenizer.py,sha256=ucb86Jr-EaAyD9OHDoCmwB9Nh9AFIZK_TlZmMkv46KQ,1325
 sopro/codec/__init__.py,sha256=6D6Q0M-SUZZnq79OT1nATenEc8zIZDrhZBpm7zdPEE4,129
 sopro/codec/mimi.py,sha256=RNKnXfhWXUqHiU27C90wj18Rb3R2IZHpm5_cS_XAs9Y,5798
@@ -15,9 +15,9 @@ sopro/nn/blocks.py,sha256=zDEVUH2LXapXuQ4DyhplNh1I0iJYrNUL20IxHoz8ucs,3221
 sopro/nn/embeddings.py,sha256=7YfYKj1v1oafTV4-iucJG4fmeT43fP_rQiJ6ACRKPNI,3185
 sopro/nn/speaker.py,sha256=L2bs-bPlyxoWZyMTctBBuMTaEWm6FP7K1udrXehnTGM,2964
 sopro/nn/xattn.py,sha256=OeRo1HbRZs0AkQ6AV6Q8cqYZP9K4vI-IwT3uVn9jOqg,2939
-sopro-1.0.1.dist-info/licenses/LICENSE.txt,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-sopro-1.0.1.dist-info/METADATA,sha256=tlq9mTTsNEFgMyCtle7om5hqKRm5LwrVCFLo4olQ3_s,6470
-sopro-1.0.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-sopro-1.0.1.dist-info/entry_points.txt,sha256=OWcKgC5Syk8rzOhNzTZ3QR5GJEG88UfiShkovrwb2cI,44
-sopro-1.0.1.dist-info/top_level.txt,sha256=Tik26_lEwzSKDuwQdqwoqA_O0b7CDATzousa0Q17PBo,6
-sopro-1.0.1.dist-info/RECORD,,
+sopro-1.0.2.dist-info/licenses/LICENSE.txt,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+sopro-1.0.2.dist-info/METADATA,sha256=LPMr5tnwQx3Rq5FX9CCMq6s4IvreA-EWQvt-OzQkm7g,6470
+sopro-1.0.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+sopro-1.0.2.dist-info/entry_points.txt,sha256=OWcKgC5Syk8rzOhNzTZ3QR5GJEG88UfiShkovrwb2cI,44
+sopro-1.0.2.dist-info/top_level.txt,sha256=Tik26_lEwzSKDuwQdqwoqA_O0b7CDATzousa0Q17PBo,6
+sopro-1.0.2.dist-info/RECORD,,