File size: 8,756 Bytes
52f41b8
 
 
 
 
9c07a74
 
 
 
 
52f41b8
 
 
 
 
 
9c07a74
 
52f41b8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a2eae29
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9c07a74
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
"""Pure mode handlers — one function per generation mode.

Each handler validates inputs, builds the ACE-Step kwargs for its mode, and
hands off to `backend.dispatch(...)`. Backend ownership of @spaces.GPU and
pipeline lifecycle keeps these handlers cheap to test.

The ``lyrics()`` handler is the odd one out: it does NOT touch the ACE-Step
backend at all. It calls ``lyrics_lm.generate_lyrics`` directly, since the
Qwen 2.5 7B LM is its own lazy singleton and doesn't share the DiT / 5Hz
pipeline lifecycle with the audio modes.
"""

from __future__ import annotations

from typing import Any

import lyrics_lm


def _require(params: dict[str, Any], field: str) -> Any:
    v = params.get(field)
    if v is None or (isinstance(v, str) and not v.strip()):
        raise ValueError(f"Missing required field: {field}")
    return v


def generate(backend, params: dict[str, Any]) -> tuple[str, dict[str, Any]]:
    """Text → song. Vocals + instruments in one stream."""
    prompt = _require(params, "prompt")
    lyrics = params.get("lyrics", "")
    duration_s = int(params.get("duration_s", 30))
    instrumental = bool(params.get("instrumental", False))

    return backend.dispatch(
        mode="generate",
        params={
            "prompt": prompt,
            "lyrics": lyrics,
            "duration_s": duration_s,
            "instrumental": instrumental,
            "seed": params.get("seed"),
            "loras": params.get("loras", []),
            "advanced": params.get("advanced", {}),
            "lm": params.get("lm", {}),
            "dcw": params.get("dcw", {}),
        },
    )


def cover(backend, params: dict[str, Any]) -> tuple[str, dict[str, Any]]:
    """Audio-reference cover — reference audio + new prompt -> song in that style.

    Maps to ACE-Step's ``GenerationParams(task_type="cover")`` with
    ``reference_audio`` set to the uploaded clip and ``audio_cover_strength``
    controlling how tightly the new song hugs the reference timbre/structure.
    """
    ref_audio = _require(params, "ref_audio")
    prompt = params.get("prompt", "")
    lyrics = params.get("lyrics", "")
    duration_s = int(params.get("duration_s", 30))

    return backend.dispatch(
        mode="cover",
        params={
            "prompt": prompt,
            "ref_audio": ref_audio,
            "lyrics": lyrics,
            "duration_s": duration_s,
            "audio_cover_strength": float(params.get("audio_cover_strength", 0.93)),
            "cover_noise_strength": float(params.get("cover_noise_strength", 0.0)),
            "seed": params.get("seed"),
            "loras": params.get("loras", []),
            "advanced": params.get("advanced", {}),
            "lm": params.get("lm", {}),
            "dcw": params.get("dcw", {}),
        },
    )


def extend(backend, params: dict[str, Any]) -> tuple[str, dict[str, Any]]:
    """Song continuation — seed audio + extension prompt -> extended song.

    Maps to ACE-Step's ``GenerationParams(task_type="repaint")`` with
    ``src_audio`` set to the seed and ``repainting_start``/``repainting_end``
    pointing past the end of the seed so the model paints new audio after it.
    """
    seed_audio = _require(params, "seed_audio")
    extra_prompt = params.get("extra_prompt", "")
    extra_duration_s = int(params.get("extra_duration_s", 60))

    return backend.dispatch(
        mode="extend",
        params={
            "seed_audio": seed_audio,
            "extra_prompt": extra_prompt,
            "extension_lyrics": params.get("extension_lyrics", ""),
            "extra_duration_s": extra_duration_s,
            "repaint_mode": params.get("repaint_mode", "balanced"),
            "repaint_strength": float(params.get("repaint_strength", 0.5)),
            "wav_crossfade_s": float(params.get("wav_crossfade_s", 2.0)),
            "latent_crossfade_frames": int(params.get("latent_crossfade_frames", 10)),
            "chunk_mask_mode": params.get("chunk_mask_mode", "auto"),
            "seed": params.get("seed"),
            "loras": params.get("loras", []),
            "advanced": params.get("advanced", {}),
            "lm": params.get("lm", {}),
            "dcw": params.get("dcw", {}),
        },
    )


def edit(backend, params: dict[str, Any]) -> tuple[str, dict[str, Any]]:
    """Segment-level edit — repaint a region OR morph caption-to-caption.

    Two sub-modes:

    - ``"repaint"`` (default): paint over ``[segment_start_s, segment_end_s]``
      using ACE-Step's ``task_type="repaint"`` with the segment bounds wired
      into ``repainting_start`` / ``repainting_end``.
    - ``"flow_edit"``: caption-to-caption morph. The installed ACE-Step
      ``GenerationParams`` dataclass has no native ``flow_edit_*`` fields, so
      flow-edit is implemented downstream as a ``task_type="repaint"`` pass
      with a lower ``audio_cover_strength`` to allow more style drift. The
      ``flow_source_caption`` / ``flow_n_*`` knobs are carried through the
      internal params dict so the pipeline wrapper can use them if/when the
      upstream dataclass grows native support.
    """
    source_audio = _require(params, "source_audio")
    sub_mode = params.get("sub_mode", "repaint")

    out_params: dict[str, Any] = {
        "source_audio": source_audio,
        "source_lyrics": params.get("source_lyrics", ""),
        "target_lyrics": params.get("target_lyrics", ""),
        "segment_start_s": float(params.get("segment_start_s", 0.0)),
        "segment_end_s": float(params.get("segment_end_s", 30.0)),
        "sub_mode": sub_mode,
        "seed": params.get("seed"),
        "loras": params.get("loras", []),
        "advanced": params.get("advanced", {}),
        "lm": params.get("lm", {}),
        "dcw": params.get("dcw", {}),
    }
    if sub_mode == "repaint":
        out_params.update(
            {
                "repaint_mode": params.get("repaint_mode", "balanced"),
                "repaint_strength": float(params.get("repaint_strength", 0.5)),
                "chunk_mask_mode": params.get("chunk_mask_mode", "auto"),
                "latent_crossfade_frames": int(params.get("latent_crossfade_frames", 10)),
                "wav_crossfade_s": float(params.get("wav_crossfade_s", 0.0)),
            }
        )
    elif sub_mode == "flow_edit":
        out_params.update(
            {
                "flow_source_caption": params.get("flow_source_caption", ""),
                "flow_n_min": float(params.get("flow_n_min", 0.0)),
                "flow_n_max": float(params.get("flow_n_max", 1.0)),
                "flow_n_avg": int(params.get("flow_n_avg", 1)),
            }
        )

    return backend.dispatch(mode="edit", params=out_params)


def lyrics(backend, params: dict[str, Any]) -> tuple[str, dict[str, Any]]:
    """Lyrics-only mode. Returns ``(drafted_text, metadata_dict)``.

    Does NOT touch the ACE-Step backend — Qwen 2.5 7B Instruct is owned
    by ``lyrics_lm`` as its own lazy singleton. The ``backend`` argument
    is kept in the signature for parity with the other mode handlers but
    is unused here.
    """
    del backend  # signature parity with generate/cover/extend/edit
    brief = _require(params, "brief")
    structure = params.get("structure", "intro, verse, chorus, verse, chorus, bridge, chorus, outro")
    language = params.get("language", "en")
    tone = params.get("tone", "")
    verse_lines = int(params.get("verse_lines", 6))
    chorus_lines = int(params.get("chorus_lines", 4))
    bridge_lines = int(params.get("bridge_lines", 2))
    rhyme = params.get("rhyme", "loose")
    temperature = float(params.get("temperature", 0.85))
    top_p = float(params.get("top_p", 0.9))
    top_k = int(params.get("top_k", 40))
    max_new_tokens = int(params.get("max_new_tokens", 600))
    seed = params.get("seed")

    text = lyrics_lm.generate_lyrics(
        brief=brief,
        structure=structure,
        language=language,
        tone=tone,
        verse_lines=verse_lines,
        chorus_lines=chorus_lines,
        bridge_lines=bridge_lines,
        rhyme=rhyme,
        temperature=temperature,
        top_p=top_p,
        top_k=top_k,
        max_new_tokens=max_new_tokens,
        seed=seed,
    )
    meta = {
        "mode": "lyrics",
        "model": lyrics_lm._DEFAULT_MAC_ID,
        "brief_first_line": brief.splitlines()[0] if brief else "",
        "structure": structure,
        "language": language,
        "tone": tone,
        "verse_lines": verse_lines,
        "chorus_lines": chorus_lines,
        "bridge_lines": bridge_lines,
        "rhyme": rhyme,
        "temperature": temperature,
        "top_p": top_p,
        "top_k": top_k,
        "max_new_tokens": max_new_tokens,
        "seed": seed,
    }
    return text, meta