mtmd: add zonos3-mini TTS model (#32020)

* zonos3 init

* add audio gen pipeline

* fix string_format

* add default voice as gguf metadata

* better ref

* support voice clone

* add delivery params

* optimize speed

* nits

* nits 2

* add text normalization

* nits
This commit is contained in:
Xuan-Son Nguyen authored and GitHub committed 2026-11-15 19:44:44 +01:00
1 parent f4d0673917
commit 7ae3c6b231
38 files changed
+3532 -100

No files matched your search

+3 -3
View File
@@ -825,9 +825,9 @@ static void test_speech_convert(testing & t) {
t.assert_true ("audio streams", out.at("stream").get<bool>());
t.assert_true ("rejects sse", rejected("stream_format", "sse"));
// speed
t.assert_true ("speed 1.0", !rejected("speed", 1.0));
t.assert_true ("rejects speed 1.5", rejected("speed", 1.5));
// speed is forwarded, the model decides whether it is supported
out = convert_speech_to_tts(with("speed", 1.5));
t.assert_equal("speed forwarded", 1.5, out.at("speed").get<double>());
// instructions are ignored
t.assert_true ("instructions", !rejected("instructions", "whisper"));