From 21999e17025800053e3310ac9c1bb239d6c14f2b Mon Sep 17 00:00:00 2001
From: Ryan Heise <ryan@ryanheise.com>
Date: Sun, 31 Mar 2024 13:11:45 +1100
Subject: [PATCH] Omit space prefix in initial_prompt for spaceless languages.

---
 whisper/transcribe.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/whisper/transcribe.py b/whisper/transcribe.py
index 1c075a201..70d70aeb5 100644
--- a/whisper/transcribe.py
+++ b/whisper/transcribe.py
@@ -228,7 +228,8 @@ def decode_with_fallback(segment: torch.Tensor) -> DecodingResult:
     prompt_reset_since = 0
 
     if initial_prompt is not None:
-        initial_prompt_tokens = tokenizer.encode(" " + initial_prompt.strip())
+        space = "" if language in {"zh", "ja", "th", "lo", "my", "yue"} else " "
+        initial_prompt_tokens = tokenizer.encode(space + initial_prompt.strip())
         all_tokens.extend(initial_prompt_tokens)
     else:
         initial_prompt_tokens = []