tts: add speaker file support

Signed-off-by: dm4 <[email protected]>
ggml-org · Mar 1, 2025 · 0b9db0a · 0b9db0a
1 parent 06c2b15
commit 0b9db0a
Show file tree

Hide file tree

Showing 3 changed files with 208 additions and 136 deletions.
diff --git a/common/arg.cpp b/common/arg.cpp
@@ -2447,6 +2447,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
             params.vocoder.use_guide_tokens = true;
         }
     ).set_examples({LLAMA_EXAMPLE_TTS, LLAMA_EXAMPLE_SERVER}));
+    add_opt(common_arg(
+        {"--tts-speaker-file"}, "FNAME",
+        "speaker file path for audio generation",
+        [](common_params & params, const std::string & value) {
+            params.vocoder.speaker_file = value;
+        }
+    ).set_examples({LLAMA_EXAMPLE_TTS}));
 
     // model-specific
     add_opt(common_arg(

diff --git a/common/common.h b/common/common.h
@@ -200,6 +200,8 @@ struct common_params_vocoder {
     std::string model     = ""; // model path                                                // NOLINT
     std::string model_url = ""; // model url to download                                     // NOLINT
 
+    std::string speaker_file = ""; // speaker file path                                      // NOLINT
+
     bool use_guide_tokens = false; // enable guide tokens to improve TTS accuracy            // NOLINT
 };