Tag：artificial intelligence
-
Milvus 2.0 data insertion and persistence
Editor’s note: This article introduces the Milvus2.0 data insertion process and persistence scheme in detail Introduction to the overall architecture of Milvus 2.0Introduction to components related to data writing Proxy Data coord Data node Root coord & Time tick Data allocation Data allocation Data organization structure File structure and data persistence Introduction to the overall […]
-
How to choose the underlying tools in the architecture? The application practice of OpenMLDB in Akulaku data drive gives you the answer
This article is compiled from the speech of Ma Yuxiang, Director of Akulaku Algorithm in 4Paradigm Technology Day, in the sub-forum of “Efficiently Implementing AI Toolchain and Open Source Ecosystem”.Hello everyone, I am very glad to participate in the 4Paradigm Technology Day with you and share about the application practice of OpenMLDB in Akulaku data-driven. […]
-
Why is the Python language preferred for hacker programming? This is your answer at high speed!
Add Weibo to learn Python together, welcome everyone to give advice, Xiaobai to study. mengy7762 [Wei Xin (same pronunciation)] Programming is not an easy task for any novice, especially in a country where C language is basically used as an enlightenment language in China. Python is indeed a boon for anyone who wants to learn […]
-
【MindSpore Easier】Deep Learning Series – Classical Convolutional Neural Network
Last week, my friend said that our convolutional neural network is too simple, the basics, of course, must be laid first (). No, a more complex convolutional neural network is here~ The LeNet handwritten font recognition model of the classic CNN, LeNet5, was born in 1994 and is one of the earliest convolutional neural networks. […]
-
Anatomy of Milvus data processing process
Editor’s note:This article dissects the main data processing flow and access layer (Access Layer) of Milvus 2.0 in detail. Sharing Outline: Review the architecture of Milvus 2.0; Introduce the organization of Milvus 2.0 code structure; Introduce the main data processing process; Introduce the access layer in detail. Main Data Processing Process The main data processing […]
-
MindSpore error ‘half_pixel_centers’ =True only support in Ascend
1 Error description 1.1 System environment Hardware Environment(Ascend/GPU/CPU): CPUSoftware Environment:– MindSpore version (source or binary): 1.8.0– Python version (eg, Python 3.7.5): 3.7.6– OS platform and distribution (eg, Linux Ubuntu 16.04): Ubuntu 4.15.0-74-generic– GCC/Compiler version (if compiled from source): 1.2 Basic information 1.2.1 The script calls the ResizeBilinear operator and uses bilinear interpolation to adjust the […]
-
ModelArts Deploying Zidong Taichu Large-scale Model Inference Service Experience Sharing
Zero, project backgroundThe fine-tuned downstream task inference service needs to be deployed on ModelArts.Main pitfalls:1. Environmental preparationMindSpore 1.6.1Docker 20.10.17Zidong Taichu model codehttps://gitee.com/mindspore/o…ckpt weights fileModelArts Account2. Model conversion1. Refer to the test code to initialize the model and load ckpt 2. Use MindSpore.export(net,inputs, filename=“**”, file_format=”MINDIR”) exportMINDIR model Model format: The current versions of AIR and ONNX […]
-
How Android implements background segmentation for real-time audio and video conferences
1 Introduction to Background Segmentation Function ZegoEffects SDK AI Vision SDK provides background segmentation function, which is suitable for various scenarios such as online education, video conferencing, hairdressing, and short videos. 2 Prerequisites for the realization of the background segmentation function in video Before using the background segmentation function of the ZegoEffects SDK AI Vision […]
-
[Model reasoning] Tengine model conversion and quantization
Welcome to pay attention to my public account [Jizhi Vision], reply 001 to get Google programming specification O_o >_< o_O O_o ~_~ o_O This article introduces the Tengine model conversion and quantization process. Like ncnn, Tengine is also an excellent end-to-side reasoning framework. I wrote an article “ncnn model […]
-
ALL in ONE! Borui Data held a grand launch ceremony for ONE 2.0
Recently, BonreeONE 2.0, an integrated intelligent observable platform of Boree Data, was officially released. Leaders and colleagues of Boree Data gathered together to hold a ceremony for the full launch of ONE 2.0, witnessing a milestone in the development of China’s IT operation and maintenance market. Since its listing in 2020, Borui Data has focused […]
-
OpenMLDB Airflow Connector: Make MLOps workflow more free
Introduction In order to improve the ease of use and convenience of OpenMLDB, lower the threshold for developers to use, and create a smoother and more efficient MLOps workflow, OpenMLDB has been actively promoting cooperation among communities, and consciously connecting the upstream and downstream of the ecology. Today’s article will introduce to you another new […]
-
MindSpore reports ValueError: `padding_idx` in `Embedding is out of range error
1 Error description 1.1 System environment Hardware Environment(Ascend/GPU/CPU): AscendSoftware Environment:– MindSpore version (source or binary): 1.8.0– Python version (eg, Python 3.7.5): 3.7.6– OS platform and distribution (eg, Linux Ubuntu 16.04): Ubuntu 4.15.0-74-generic– GCC/Compiler version (if compiled from source): 1.2 Basic information 1.2.1 Script training script is to complete embedding by building a single operator network […]