Skip to content

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

decoding

游雁de11189e9634
diff --git a/examples/industrial_data_pretraining/llm_asr/demo_speech2text.py b/examples/industrial_data_pretraining/llm_asr/demo_speech2text.py
index e5e3e2302..41b344028 100644
--- a/examples/industrial_data_pretraining/llm_asr/demo_speech2text.py
+++ b/examples/industrial_data_pretraining/llm_asr/demo_speech2text.py
@@ -28,6 +28,9 @@ model = AutoModel(
init_param=f"{os.path.join(ckpt_dir, ckpt_id)}",
output_dir=output_dir,
device=device,
+ fp16=False,
+ bf16=False,
+ llm_dtype="bf16",
)
diff --git a/examples/industrial_data_pretraining/llm_asr/demo_speech2text.sh b/examples/industrial_data_pretraining/llm_asr/demo_speech2text.sh
index 8e22cb955..bb63d4490 100644
--- a/examples/industrial_data_pretraining/llm_asr/demo_speech2text.sh
+++ b/examples/industrial_data_pretraining/llm_asr/demo_speech2text.sh
@@ -41,7 +41,7 @@ for data_set in "aishell1_test_speech2text.jsonl" "aishell2_ios_test_speech2text
}&
done
-
+wait
#for data_set in "s2tt_en2zh.v20240605.test.jsonl"; do
# jsonl=${jsonl_dir}/${data_set}
# output_dir=${out_dir}/${data_set}