Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
auto frontend
游雁834a8fd9e2e9
diff --git a/funasr/models/llm_asr/model.py b/funasr/models/llm_asr/model.pyindex a6a713481..82ad134ff 100644--- a/funasr/models/llm_asr/model.py+++ b/funasr/models/llm_asr/model.py@@ -488,6 +488,8 @@ class LLMASR2(nn.Module):fbank_fake_len = fbank_fake_lens[batch_idx].item()fbank_beg_idx = fbank_beg[batch_idx, 0].item()min_len = min(fbank_fake_len, inputs_embeds.shape[1] - fbank_beg_idx)+ fbank_fake_len = encoder_out_lens[batch_idx].item()+ min_len = min(fbank_fake_len, inputs_embeds.shape[1] - fbank_beg_idx)try:inputs_embeds[batch_idx, fbank_beg_idx : fbank_beg_idx + min_len, :] = encoder_out[batch_idx, :min_len, :