From be53fb179c17c8db16656d017af26578633cd343 Mon Sep 17 00:00:00 2001
From: SWivid <swivid@qq.com>
Date: Mon, 11 Nov 2024 14:46:50 +0800
Subject: [PATCH] fix. force reload from custom cache file to override
 infer_gradio init loaded value in case webpage refreshed

---
 src/f5_tts/infer/infer_gradio.py | 27 +++++++++++++++------------
 src/f5_tts/infer/utils_infer.py  |  2 +-
 2 files changed, 16 insertions(+), 13 deletions(-)

diff --git a/src/f5_tts/infer/infer_gradio.py b/src/f5_tts/infer/infer_gradio.py
index 3de5b55..4e5dc63 100644
--- a/src/f5_tts/infer/infer_gradio.py
+++ b/src/f5_tts/infer/infer_gradio.py
@@ -757,17 +757,22 @@ If you're having issues, try converting your reference audio to WAV or MP3, clip
                 "hf://SWivid/F5-TTS/F5TTS_Base/vocab.txt",
             ]
 
-    def switch_tts_model(new_choice, custom_ckpt_path, custom_vocab_path):
+    def switch_tts_model(new_choice):
         global tts_model_choice
-        if new_choice == "Custom":
+        if new_choice == "Custom":  # override in case webpage is refreshed
+            custom_ckpt_path, custom_vocab_path = load_last_used_custom()
             tts_model_choice = ["Custom", custom_ckpt_path, custom_vocab_path]
-            with open(last_used_custom, "w") as f:
-                f.write(f"{custom_ckpt_path},{custom_vocab_path}")
-            return gr.update(visible=True), gr.update(visible=True)
+            return gr.update(visible=True, value=custom_ckpt_path), gr.update(visible=True, value=custom_vocab_path)
         else:
             tts_model_choice = new_choice
             return gr.update(visible=False), gr.update(visible=False)
 
+    def set_custom_model(custom_ckpt_path, custom_vocab_path):
+        global tts_model_choice
+        tts_model_choice = ["Custom", custom_ckpt_path, custom_vocab_path]
+        with open(last_used_custom, "w") as f:
+            f.write(f"{custom_ckpt_path},{custom_vocab_path}")
+
     with gr.Row():
         if not USING_SPACES:
             choose_tts_model = gr.Radio(
@@ -794,20 +799,18 @@ If you're having issues, try converting your reference audio to WAV or MP3, clip
 
     choose_tts_model.change(
         switch_tts_model,
-        inputs=[choose_tts_model, custom_ckpt_path, custom_vocab_path],
+        inputs=[choose_tts_model],
         outputs=[custom_ckpt_path, custom_vocab_path],
         show_progress="hidden",
     )
     custom_ckpt_path.change(
-        switch_tts_model,
-        inputs=[choose_tts_model, custom_ckpt_path, custom_vocab_path],
-        outputs=[custom_ckpt_path, custom_vocab_path],
+        set_custom_model,
+        inputs=[custom_ckpt_path, custom_vocab_path],
         show_progress="hidden",
     )
     custom_vocab_path.change(
-        switch_tts_model,
-        inputs=[choose_tts_model, custom_ckpt_path, custom_vocab_path],
-        outputs=[custom_ckpt_path, custom_vocab_path],
+        set_custom_model,
+        inputs=[custom_ckpt_path, custom_vocab_path],
         show_progress="hidden",
     )
 
diff --git a/src/f5_tts/infer/utils_infer.py b/src/f5_tts/infer/utils_infer.py
index f3ebc33..469855f 100644
--- a/src/f5_tts/infer/utils_infer.py
+++ b/src/f5_tts/infer/utils_infer.py
@@ -194,7 +194,7 @@ def load_model(
     tokenizer = "custom"
 
     print("\nvocab : ", vocab_file)
-    print("tokenizer : ", tokenizer)
+    print("token : ", tokenizer)
     print("model : ", ckpt_path, "\n")
 
     vocab_char_map, vocab_size = get_tokenizer(vocab_file, tokenizer)