Skip to content

Commit e779703

Browse files
sophon-leevisophon-leevi
committed
release(v0.2.4): Release v0.2.4.
Change-Id: I8a1b69fb3d7b49d5007811694998733886026021
1 parent 0aee192 commit e779703

File tree

3 files changed

+27
-17
lines changed

3 files changed

+27
-17
lines changed

‎README.md

+13-8
Original file line numberDiff line numberDiff line change
@@ -39,6 +39,7 @@ SOPHON-DEMO提供的例子从易到难分为`tutorial`、`sample`、`application
3939
| [YOLOv9_seg](./sample/YOLOv9_seg/README.md) | 实例分割 | C++/Python | FP32/FP16/INT8 |
4040
| [PP-OCR](./sample/PP-OCR/README.md) | OCR | C++/Python | FP32/FP16 |
4141
| [OpenPose](./sample/OpenPose/README.md) | 人体关键点检测 | C++/Python | FP32/FP16/INT8 |
42+
| [YOLOv8_pose](./sample/YOLOv8_pose/README.md) | 人体关键点检测 | C++/Python | FP32/FP16/INT8 |
4243
| [C3D](./sample/C3D/README.md) | 视频动作识别 | C++/Python | FP32/FP16/INT8 |
4344
| [DeepSORT](./sample/DeepSORT/README.md) | 多目标跟踪 | C++/Python | FP32/FP16/INT8 |
4445
| [ByteTrack](./sample/ByteTrack/README.md) | 多目标跟踪 | C++/Python | FP32/FP16/INT8 |
@@ -59,17 +60,17 @@ SOPHON-DEMO提供的例子从易到难分为`tutorial`、`sample`、`application
5960
| [Whisper](./sample/Whisper/README.md) | 语音识别 | Python | FP16 |
6061
| [Seamless](./sample/Seamless/README.md) | 语音识别 | Python | FP32/FP16 |
6162
| [BERT](./sample/BERT/README.md) | 语言模型 | C++/Python | FP32/FP16 |
62-
| [ChatGLM2](./sample/ChatGLM2/README.md) | 语言模型 | C++/Python | FP16/INT8/INT4 |
63-
| [Llama2](./sample/Llama2/README.md) | 语言模型 | C++/Python | FP16/INT8/INT4 |
64-
| [ChatGLM3](./sample/ChatGLM3/README.md) | 语言模型 | Python | FP16/INT8/INT4 |
65-
| [Qwen](./sample/Qwen/README.md) | 语言模型 | Python | FP16/INT8/INT4 |
66-
| [Qwen1_5](./sample/Qwen1_5/README.md) | 语言模型 | Python | FP16/INT8/INT4 |
67-
| [MiniCPM](./sample/MiniCPM/README.md) | 语言模型 | C++ | INT8/INT4 |
68-
| [Baichuan2](./sample/Baichuan2/README.md) | 语言模型 | Python | INT8/INT4 |
69-
| [ChatGLM4](./sample/ChatGLM4/README.md) | 语言模型 | Python | INT8/INT4 |
63+
| [ChatGLM2](./sample/ChatGLM2/README.md) | 大规模语言模型 | C++/Python | FP16/INT8/INT4 |
64+
| [Llama2](./sample/Llama2/README.md) | 大规模语言模型 | C++/Python | FP16/INT8/INT4 |
65+
| [ChatGLM3](./sample/ChatGLM3/README.md) | 大规模语言模型 | Python | FP16/INT8/INT4 |
66+
| [Qwen](./sample/Qwen/README.md) | 大规模语言模型 | Python | FP16/INT8/INT4 |
67+
| [MiniCPM](./sample/MiniCPM/README.md) | 大规模语言模型 | C++ | INT8/INT4 |
68+
| [Baichuan2](./sample/Baichuan2/README.md) | 大规模语言模型 | Python | INT8/INT4 |
69+
| [ChatGLM4](./sample/ChatGLM4/README.md) | 大规模语言模型 | Python | INT8/INT4 |
7070
| [StableDiffusionV1.5](./sample/StableDiffusionV1_5/README.md) | 图像生成 | Python | FP32/FP16 |
7171
| [StableDiffusionXL](./sample/StableDiffusionXL/README.md) | 图像生成 | Python | FP32/FP16 |
7272
| [GroundingDINO](./sample/GroundingDINO/README.md) | 多模态目标检测 | Python | FP16 |
73+
| [Qwen-VL-Chat](./sample/Qwen-VL-Chat/README.md) | 大规模视觉语言模型 | Python | FP16/INT8 |
7374
| [Real-ESRGAN](./sample/Real-ESRGAN/README.md) | 超分辨 | C++/Python | FP32/FP16/INT8 |
7475
| [P2PNet](./sample/P2PNet/README.md) | 人群计数 | C++/Python | FP32/FP16/INT8 |
7576
| [CLIP](./sample/CLIP/README.md) | 图文生成 | Python | FP16 |
@@ -80,10 +81,12 @@ SOPHON-DEMO提供的例子从易到难分为`tutorial`、`sample`、`application
8081
| [VLPR](./application/VLPR/README.md) | 多路车牌检测+识别 | C++/Python |
8182
| [YOLOv5_multi](./application/YOLOv5_multi/README.md) | 多路目标检测 | C++ |
8283
| [YOLOv5_multi_QT](./application/YOLOv5_multi_QT/README.md) | 多路目标检测+QT_HDMI显示 | C++ |
84+
| [Grounded-sam](./application/Grounded-sam/README.md) | 自动化图像检测和分割系统 | Python |
8385

8486
## 版本说明
8587
| 版本 | 说明 |
8688
|--- |--- |
89+
| 0.2.4 | 完善和修复文档、代码问题，**修复VideoDecFFM系统内存泄漏问题**，sample模块新增例程YOLOv8_pose、Qwen-VL-Chat例程，application新增Grounded-sam例程。 |
8790
| 0.2.3 | 完善和修复文档、代码问题，sample模块新增例程StableDiffusionXL、ChatGLM4、Seamless、YOLOv10，tutorial模块新增mmap、video_encode例程。 |
8891
| 0.2.2 | 完善和修复文档、代码问题，部分例程补充CV186X支持，sample模块新增例程Whisper、Real-ESRGAN、SCRFD、P2PNet、MiniCPM、CLIP、SuperGlue、YOLOv5_fuse、YOLOv8_seg、YOLOv9_seg、Baichuan2等例程，tutorial模块新增avframe_ocv、ocv_avframe、bm1688_2core2task_yolov5例程。 |
8992
| 0.2.1 | 完善和修复文档、代码问题，部分例程补充CV186X支持，YOLOv5适配SG2042，sample模块新增例程GroundingDINO、Qwen1_5，StableDiffusionV1_5新支持多种分辨率，Qwen、Llama2、ChatGLM3添加web和多会话模式。tutorial模块新增blend和stitch例程 |
@@ -104,6 +107,7 @@ SOPHON-DEMO提供的例子从易到难分为`tutorial`、`sample`、`application
104107
SOPHON-DEMO主要依赖TPU-MLIR、TPU-NNTC、LIBSOPHON、SOPHON-FFMPEG、SOPHON-OPENCV、SOPHON-SAIL，对于BM1684/BM1684X SOPHONSDK，其版本要求如下：
105108
|SOPHON-DEMO|TPU-MLIR |TPU-NNTC |LIBSOPHON|SOPHON-FFMPEG|SOPHON-OPENCV|SOPHON-SAIL| SOPHONSDK |
106109
|-------- |------------| --------|---------|--------- |---------- | ------ | -------- |
110+
| 0.2.4 | >=1.9 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v24.04.01|
107111
| 0.2.3 | >=1.8 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v24.04.01|
108112
| 0.2.2 | >=1.8 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v23.10.01|
109113
| 0.2.1 | >=1.7 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v23.10.01|
@@ -123,6 +127,7 @@ SOPHON-DEMO主要依赖TPU-MLIR、TPU-NNTC、LIBSOPHON、SOPHON-FFMPEG、SOPHON-
123127
对于BM1688/CV186AH SOPHONSDK，其版本要求如下：
124128
|SOPHON-DEMO|TPU-MLIR |LIBSOPHON|SOPHON-FFMPEG|SOPHON-OPENCV|SOPHON-SAIL| SOPHONSDK |
125129
|-------- |------------|---------|--------- |---------- | ------ | -------- |
130+
| 0.2.4 | >=1.9 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.7.0 |
126131
| 0.2.3 | >=1.8 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.7.0 |
127132
| 0.2.2 | >=1.8 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.6.0 |
128133
| 0.2.1 | >=1.7 | >=0.4.9 | >=1.5.0 | >=1.5.0 | >=3.8.0 | >=v1.5.0 |

‎README_EN.md

+13-8
Original file line numberDiff line numberDiff line change
@@ -39,6 +39,7 @@ The examples provided by SOPHON-DEMO are divided into three modules: `tutorial`,
3939
| [YOLOv9_seg](./sample/YOLOv9_seg/README.md) | Instance segmentation | C++/Python | FP32/FP16/INT8 |
4040
| [PP-OCR](./sample/PP-OCR/README.md) | OCR | C++/Python | FP32/FP16 |
4141
| [OpenPose](./sample/OpenPose/README.md) | Keypoint detection | C++/Python | FP32/FP16/INT8 |
42+
| [YOLOv8_pose](./sample/YOLOv8_pose/README.md) | Keypoint detection | C++/Python | FP32/FP16/INT8 |
4243
| [C3D](./sample/C3D/README.md) | Video recognition | C++/Python | FP32/FP16/INT8 |
4344
| [DeepSORT](./sample/DeepSORT/README.md) | Object tracking | C++/Python | FP32/FP16/INT8 |
4445
| [ByteTrack](./sample/ByteTrack/README.md) | Object tracking | C++/Python | FP32/FP16/INT8 |
@@ -59,17 +60,17 @@ The examples provided by SOPHON-DEMO are divided into three modules: `tutorial`,
5960
| [Whisper](./sample/Whisper/README.md) | Speech Recognition | Python | FP16 |
6061
| [Seamless](./sample/Seamless/README.md) | Speech Recognition | Python | FP32/FP16 |
6162
| [BERT](./sample/BERT/README.md) | Language | C++/Python | FP32/FP16 |
62-
| [ChatGLM2](./sample/ChatGLM2/README.md) | Language | C++/Python | FP16/INT8/INT4 |
63-
| [Llama2](./sample/Llama2/README.md) | Language | C++ | FP16/INT8/INT4 |
64-
| [ChatGLM3](./sample/ChatGLM3/README.md) | Language | Python | FP16/INT8/INT4 |
65-
| [Qwen](./sample/Qwen/README.md) | Language | Python | FP16/INT8/INT4 |
66-
| [Qwen1_5](./sample/Qwen1_5/README.md) | Language | Python | FP16/INT8/INT4 |
67-
| [MiniCPM](./sample/MiniCPM/README.md) | Language | C++ | INT8/INT4 |
68-
| [Baichuan2](./sample/Baichuan2/README.md) | Language | Python | INT8/INT4 |
69-
| [ChatGLM4](./sample/ChatGLM4/README.md) | Language | Python | FP16/INT8/INT4 |
63+
| [ChatGLM2](./sample/ChatGLM2/README.md) | Large Language Model | C++/Python | FP16/INT8/INT4 |
64+
| [Llama2](./sample/Llama2/README.md) | Large Language Model | C++ | FP16/INT8/INT4 |
65+
| [ChatGLM3](./sample/ChatGLM3/README.md) | Large Language Model | Python | FP16/INT8/INT4 |
66+
| [Qwen](./sample/Qwen/README.md) | Large Language Model | Python | FP16/INT8/INT4 |
67+
| [MiniCPM](./sample/MiniCPM/README.md) | Large Language Model | C++ | INT8/INT4 |
68+
| [Baichuan2](./sample/Baichuan2/README.md) | Large Language Model | Python | INT8/INT4 |
69+
| [ChatGLM4](./sample/ChatGLM4/README.md) | Large Language Model | Python | FP16/INT8/INT4 |
7070
| [StableDiffusionV1.5](./sample/StableDiffusionV1_5/README.md) | Image Generation | Python | FP32/FP16 |
7171
| [StableDiffusionXL](./sample/StableDiffusionXL/README.md) | Image Generation | Python | FP32/FP16 |
7272
| [GroundingDINO](./sample/GroundingDINO/README.md) | MultiModal Object Detection | Python | FP16 |
73+
| [Qwen-VL-Chat](./sample/Qwen-VL-Chat/README.md) | Large Vision Language Model | Python | FP16/INT8 |
7374
| [Real-ESRGAN](./sample/Real-ESRGAN/README.md) | Super Resolution | C++/Python | FP32/FP16/INT8 |
7475
| [P2PNet](./sample/P2PNet/README.md) | Crowd Counting | C++/Python | FP32/FP16/INT8 |
7576
| [CLIP](./sample/CLIP/README.md) | Image Captioning | Python | FP16 |
@@ -80,10 +81,12 @@ The examples provided by SOPHON-DEMO are divided into three modules: `tutorial`,
8081
| [VLPR](./application/VLPR/README.md) | Multi-streams Vehicle License Plate Recognition | C++/Python |
8182
| [YOLOv5_multi](./application/YOLOv5_multi/README.md) | Multi-streams Object Detection | C++ |
8283
| [YOLOv5_multi_QT](./application/YOLOv5_multi_QT/README.md) | Multi-streams Object Detection + QT_HDMI display | C++ |
84+
| [Grounded-sam](./application/Grounded-sam/README.md) | Automatic image detection and segmentation system | Python |
8385

8486
## Release Notes
8587
| version | description |
8688
|---|---|
89+
| 0.2.4 | Fix documentation and other issues, **Fix host memory leak in VideoDecFFM**，Release new samples including YOLOv8_pose/Qwen-VL-Chat, new application Grounded-sam. |
8790
| 0.2.3 | Fix documentation and other issues, Release new samples including StableDiffusionXL/ChatGLM4/Seamless/YOLOv10, new tutorials including mmap/video_encode. |
8891
| 0.2.2 | Fix documentation and other issues, some examples support CV186X. Release new samples including Whisper/Real-ESRGAN/SCRFD/P2PNet/MiniCPM/CLIP/SuperGlue/YOLOv5_fuse/YOLOv8_seg/YOLOv9_seg/Baichuan2, new tutorials including avframe_ocv/ocv_avframe/bm1688_2core2task_yolov5. |
8992
| 0.2.1 | Fix documentation and other issues, some examples support CV186X, sample/YOLOv5 support SG2042, release new samples GroundingDINO and Qwen1_5, StableDiffusionV1_5 newly support multilize resolution models, Qwen/Llama2/ChatGLM3 add web and multi-session support. tutorial module add blend and stitch examples. |
@@ -104,6 +107,7 @@ The examples provided by SOPHON-DEMO are divided into three modules: `tutorial`,
104107
SOPHON-DEMO mainly depends on TPU-MLIR, TPU-NNTC, LIBSOPHON, SOPHON-FFMPEG, SOPHON-OPENCV, SOPHON-SAIL, for BM1684/BM1684X SOPHONSDK, version requirements are as follows:
105108
|SOPHON-DEMO|TPU-MLIR |TPU-NNTC |LIBSOPHON|SOPHON-FFMPEG|SOPHON-OPENCV|SOPHON-SAIL| SOPHONSDK |
106109
|-------- |------------| --------|---------|--------- |---------- | ------ | -------- |
110+
| 0.2.4 | >=1.9 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v24.04.01|
107111
| 0.2.3 | >=1.8 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v24.04.01|
108112
| 0.2.2 | >=1.8 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v23.10.01|
109113
| 0.2.1 | >=1.7 | >=3.1.7 | >=0.5.0 | >=0.7.3 | >=0.7.3 | >=3.7.0 | >=v23.10.01|
@@ -123,6 +127,7 @@ SOPHON-DEMO mainly depends on TPU-MLIR, TPU-NNTC, LIBSOPHON, SOPHON-FFMPEG, SOPH
123127
For BM1688/CV186AH SOPHONSDK, version requirements are as follows:
124128
|SOPHON-DEMO|TPU-MLIR |LIBSOPHON|SOPHON-FFMPEG|SOPHON-OPENCV|SOPHON-SAIL| SOPHONSDK |
125129
|-------- |------------|---------|--------- |---------- | ------ | -------- |
130+
| 0.2.4 | >=1.9 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.7.0 |
126131
| 0.2.3 | >=1.8 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.7.0 |
127132
| 0.2.2 | >=1.8 | >=0.4.9 | >=1.6.0 | >=1.6.0 | >=3.8.0 | >=v1.6.0 |
128133
| 0.2.1 | >=1.7 | >=0.4.9 | >=1.5.0 | >=1.5.0 | >=3.8.0 | >=v1.5.0 |

‎git_version

+1-1
Original file line numberDiff line numberDiff line change
@@ -1 +1 @@
1-
0.2.3
1+
0.2.4

0 commit comments

Comments
 (0)