Wisen Technical Portfolio
Our Python and data-science instructors also work as applied ML and computer-vision researchers, and the datasets that work produces are not kept private — each is published with a DOI wherever the host supports one, so it can be cited and reused outside Wisen.
Below is the running record: what we have released on Zenodo, Kaggle and Mendeley Data, and the corrections filed against the open-source libraries this curriculum is built on. It doubles as a public, checkable account of who runs pythontrainingcourses.com.
- Operated by
- Wisen IT Solutions Private Limited
- CIN
- U72900TN2015PTC100050
- GST
- 33AABCW6355J2Z4
- Open datasets
- 5 published
- Hosts
- Zenodo · Kaggle · Mendeley Data
- Research focus
- Image restoration, remote sensing, supply-chain security, applied GenAI
- Location
- H8, Housing Board, Tamil Nagar, Ramapuram, Chennai, TN, India
- Enquiries
- +91 900 31 31 555 · info@wisenitsolutions.com
Datasets published by Wisen IT Solutions
Each dataset links to its canonical record — a DOI landing page where one exists — and to any mirror. All are released for free use under the licence shown.
Malicious npm Package Detection Dataset
Labelled JavaScript source files drawn from malicious and benign npm packages, for training static-analysis and neural models that flag compromised packages in the software supply chain.
- Software supply-chain security
- Static code analysis
- Machine learning
UC Merced Super-Resolution Dataset (LR–HR Pairs, Blur + Downsample)
Paired low- and high-resolution satellite tiles from the UC Merced land-use collection, generated with x4 bicubic downsampling and split into train, validation and test sets for super-resolution benchmarking.
- Image super-resolution
- Remote sensing
- Computer vision
ParisStreetView-RandomMasks: Urban Image Inpainting Dataset
22,601 urban street-view images paired with irregular free-form masks and corrupted versions, for training image-inpainting and diffusion models such as LaMa, EdgeConnect and Stable Diffusion Inpainting.
- Image inpainting
- Generative scene completion
- Computer vision
AID-Night: Low-Light Satellite Image Dataset
Satellite scenes under low-light and night-time conditions, for developing and evaluating low-light enhancement, denoising and object-detection models in poor illumination.
- Low-light image enhancement
- Remote sensing
- Computer vision
AgroGPT Agriculture Knowledge Dataset
A curated agriculture knowledge corpus for instruction-tuning and retrieval, aimed at building farming-domain question-answering assistants.
- Generative AI
- Domain knowledge base
- Agriculture
Issues we have reported upstream
Our trainers are practising engineers, and they hold official documentation to the same standard we hold our own course material. These are corrections and bugs we have reported to pandas, PyTorch, Plotly, Anthropic, Angular, TypeScript and the other projects we teach and build with. They are reports filed, not patches merged — we would not describe them as anything more.
- Anthropic CoursesDuplicate image displayed in the "Response streaming" lesson
anthropics/courses - Anthropic SDK (Python)AttributeError: 'WindowsPath' object has no attribute 'read' on files.upload
anthropics/anthropic-sdk-python - PyTorchDocs: nn.functional.unfold accepts 3-D input, not only 4-D
pytorch/pytorch - pandasDocs: a scalar Series does not in fact require an explicit index
pandas-dev/pandas - Reed–Solomon article (tomverbeure)Suspected error in 'Reed-Solomon Error Correcting Codes from the Bottom Up'
tomverbeure/tomverbeure.github.io - AngularDocs: forms guide references a non-existent ng-submitted class
angular/angular - Plotly (plotly.py)Docs: make_subplots print_grid default is False, not True
plotly/plotly.py - AngularDocs: local-variable technique restriction is overstated
angular/angular - TypeScriptDocs vs compiler: constant enum expressions evaluating to NaN/Infinity
microsoft/TypeScript - TypeScriptReverse mapping does not work for string-based enums
microsoft/TypeScript
