fix: Whisper generate, move text_prompt_ids trim up for max_new_tokens calculation (#23724)

move text_prompt_ids trimming to top

src/transformers/models/whisper/modeling_whisper.py

@@ -1633,6 +1633,9 @@ class WhisperForConditionalGeneration(WhisperPreTrainedModel):
                 )
             prompt_ids = prompt_ids.tolist()
             decoder_start_token_id, *text_prompt_ids = prompt_ids
+            # Slicing the text prompt ids in a manner consistent with the OpenAI implementation
+            # to accomodate context space for the prefix (see https://github.com/openai/whisper/blob/c09a7ae299c4c34c5839a76380ae407e7d785914/whisper/decoding.py#L599)
+            text_prompt_ids = text_prompt_ids[-self.config.max_length // 2 - 1 :]
             # Set the decoder_start_token_id to <|startofprev|>
             kwargs.update({"decoder_start_token_id": decoder_start_token_id})
 
@@ -1647,9 +1650,7 @@ class WhisperForConditionalGeneration(WhisperPreTrainedModel):
                 kwargs.pop("forced_decoder_ids", None) or generation_config.forced_decoder_ids
             )
             forced_decoder_ids = [
-                # Slicing the text prompt ids in a manner consistent with the OpenAI implementation
-                # to accomodate context space for the prefix (see https://github.com/openai/whisper/blob/c09a7ae299c4c34c5839a76380ae407e7d785914/whisper/decoding.py#L599)
-                *text_prompt_ids[-self.config.max_length // 2 - 1 :],
+                *text_prompt_ids,
                 generation_config.decoder_start_token_id,
                 *[token for _rank, token in non_prompt_forced_decoder_ids],
             ]