UniqAI/mini-omni
1
0
Fork 0
mirror of https://github.com/gpt-omni/mini-omni synced 2024-11-16 05:03:47 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge branch 'main' of github.com:gpt-omni/mini-omni

Browse Source
This commit is contained in:
mini-omni 2024-09-09 11:42:06 +08:00
commit 2b08f18735
3 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@ -24,7 +24,7 @@ Mini-Omni is an open-source multimodal large language model that can **hear, tal
✅ **Talking while thinking**, with the ability to generate text and audio at the same time. ✅ **Talking while thinking**, with the ability to generate text and audio at the same time.
✅ **Streaming audio outupt** capabilities. ✅ **Streaming audio output** capabilities.
✅ With "Audio-to-Text" and "Audio-to-Audio" **batch inference** to further boost the performance. ✅ With "Audio-to-Text" and "Audio-to-Audio" **batch inference** to further boost the performance.

2
inference.py
View File

@ -399,7 +399,7 @@ class OmniInference:
model = self.model model = self.model
with self.fabric.init_tensor(): with self.fabric.init_tensor():
model.set_kv_cache(batch_size=2) model.set_kv_cache(batch_size=2,device=self.device)
mel, leng = load_audio(audio_path) mel, leng = load_audio(audio_path)
audio_feature, input_ids = get_input_ids_whisper_ATBatch(mel, leng, self.whispermodel, self.device) audio_feature, input_ids = get_input_ids_whisper_ATBatch(mel, leng, self.whispermodel, self.device)

4
server.py
View File

@ -46,9 +46,9 @@ def create_app():
return server.server return server.server
def serve(ip='0.0.0.0', port=60808): def serve(ip='0.0.0.0', port=60808, device='cuda:0'):
OmniChatServer(ip, port=port, run_app=True) OmniChatServer(ip, port=port,run_app=True, device=device)
if __name__ == "__main__": if __name__ == "__main__":