fd0563c34f
[feat] Update model configurations and dependencies for ASR and TTS components
2026-04-10 17:54:47 +08:00
196ba7571c
[feat] Enhance LLMPostprocess with default configurations and request handling for sampling parameters
2026-04-09 16:40:49 +08:00
13f08ef899
update: change subproject commit reference in SDK
2026-04-09 10:31:05 +08:00
cce686fa89
update: change subproject commit reference in SDK
2026-04-08 19:08:07 +08:00
9398776fdc
feat: Add kokoro-1.0-axcl model configuration and update llm_pack.py for packaging
2026-04-07 17:59:02 +08:00
dffe3fa0a4
feat: Add Qwen3 model configurations and update llm_pack.py for new models
2026-04-02 18:13:32 +08:00
cdc4b66e28
feat: Add PaddleOCR-VL image processor and enhance logger for UTF-8 support
...
- Implemented PaddleOCR-VL image processor to generate patches in [N, C, pH, pW] format.
- Enhanced logger initialization to support UTF-8 encoding on Windows terminals.
- Updated mrope index handling to support shared image/video token IDs.
- Introduced port availability check function in net_utils for better error handling.
- Improved disk cache loading with additional validation for corrupted files.
- Added support for PaddleOCR-VL in vision module, including encoding and processing of image and video frames.
- Updated TTS build configuration to include new dependencies and libraries.
2026-04-02 17:39:23 +08:00
425c425782
[update] update tokenizer commit reference and add requirements for ax_msp in TTS
2026-04-02 09:48:26 +08:00
ce2596ce97
[update] axcl supported qwen3 via llm_llm2
2026-03-27 18:04:14 +08:00
20079ff13d
[update] update llm-asr
2026-01-30 14:30:58 +08:00
f15cc9a96e
[update] update llm-kws version
2026-01-24 15:58:54 +08:00
1adb2a19b3
[update] update llm-audio, llm-melotts
2026-01-24 15:50:50 +08:00
854a31f2b3
[update] update llm-asr
2026-01-24 10:53:00 +08:00
0cfeeb6e52
[update] update llm-audio supported mutil channel audio input
2026-01-15 11:14:02 +08:00
5571e73992
[update] update llm-camera, update llm-kws
2026-01-13 16:06:16 +08:00
f7dd499527
[update] Delete ax630/650 model, update melotts model
2026-01-13 10:33:49 +08:00
c2dc98389c
[update] lib -llm llm-cosyvoice llm-llm llm-vlm version
2026-01-12 10:12:48 +08:00
656407c861
[update] perf AXCL CMM consume
2026-01-08 15:13:04 +08:00
a975cac644
[update] llm_asr supported sensevoie & zipformer, update llm-audio & llm-openai-api
2025-12-31 09:43:35 +08:00
4e82b32801
[update] add qwen2.5-7b-int4 model
2025-11-25 09:53:37 +08:00
9916fb59c6
[fix] Fix CosyVoice Deinit bug
2025-11-21 16:27:25 +08:00
c4d8a338f8
[update] update OpenAI-Plugin version
2025-11-20 16:31:34 +08:00
468a819dee
[fix] Fix LLM load model bug
2025-11-20 14:49:27 +08:00
a716274044
[fix] Fix pzqm & Add tokenizer launch timeout.
2025-11-19 12:04:07 +08:00
164479ae67
[update] update cosy_voice, llm, vlm version
2025-11-19 09:54:07 +08:00
LittleMouse