Dev #4

dianjixz · 2025-01-21T02:09:18Z

No description provided.

…release llm_depth_anything.

…pose model. update: yolo11n-pose model, internvl2.5-1B model.

…per static lib. update libsherpa-onnx-core.a

Abandon-ht and others added 30 commits December 17, 2024 10:13

[update] add vad model.update sherpa-onnx head file.

2452f8b

[update] add depth_anything model.

dd3279d

[fix] image format.

795e74f

[update] update depth_anything.

6833b93

[update] depth_anything supports uvc input & base64 jpg output.

801d692

[update] perf kws & melotts & vad model setup.

095343a

[update] perf yolo seg & pose result.update vlm supports internvl2.5.…

71ba7e4

…release llm_depth_anything.

[update] add point number function parameter to YOLO-Pose

324de59

[fix] Increase tokenizer init timeout.

089bfaf

[add] add mode_internvl2.5, mode_yolo11n-hand-pose.

125038e

[fix] camera frame error

3d843bf

[update] ax-samples && [update] version

e8d3b36

[update] llm-sys version

db8d051

[update] add: depth_anything model, qwen2.5-1.5B model, yolo11n-hand-…

faf5506

…pose model. update: yolo11n-pose model, internvl2.5-1B model.

update static_lib version

41aab10

update whisper

d942981

update whisper runner

d91c62c

[update] llm_sys add sys_push and sys_pull code

acf245e

Merge branch 'dev' of github.com:m5stack/StackFlow into dev

e13ab7a

upload qwen2.5-1.5B model json

b678240

[update] llm_pack.py ThreadPool

0b387ca

upload whisper-tiny.json

c2156b2

[fix] fix whisper bug

169a595

[update] update vad & whisper

e8e4cc9

[update] update static_lib version v0.0.6 to v0.0.7. update main_whis…

c8ed5ec

…per static lib. update libsherpa-onnx-core.a

[update] update whisper

0ed9518

[update] update llm_pack.py add llm_vad & llm_whisper.

ee276a3

[update] update vad & whisper, vad support kws, whisper suport kws.

8a4e7e0

[fix] static_lib version auto update

29ad0ee

[update] llm supports input from whisper. perf whisper endpoint method.

28d65f4

[clean]

a830d44

dianjixz merged commit c490c4d into main Jan 21, 2025

Provide feedback