Change the repository type filter
All
Repositories list
60 repositories
a-guide-to-mlops
Publicflipml-service
PublicDetects if a scanned document is upside downai-days-website
Publicdocumentation
PublicThe Swiss AI Center project documentation.core-engine
Publicydata-profiling-service
PublicThis service will produce a complete Exploratory Data Analysis (EDA) on a csv file and export it to a zipped html file.- Service that performs anomaly detection on unseen time series (zero shot) using the Moment foundation model
text-summarizer-service
PublicText summarizer service that summarizes a text using the Hugging Face library.- Service that performs optical text recognition (OCR) on animage and returns a JSON file with the detected text and bounding boxes,
- A dummy service demonstrating how to integrate MLOps guide tools (BentoML)
nlp-langid-service
PublicLanguage identification service from a piece of text- This service checks the integrity of a dataset based on a configuration file
- Detector of offending image including nudity, pornographic and hentai.
- Draws bounding boxes on an image. The image and the boxes coordinates are given as parameters, the output is a new image in the same format as the input.
- The service is used to query text-to-text AI models from the Hugging Face inference API.
- The service is used to query text-to-image AI models from the Hugging Face inference API.
- The service is used to query text-to-audio AI models from the Hugging Face inference API.
hugging-face-service
Public- The service is used to query image-to-text AI models from the Hugging Face inference API.
doodle-service
PublicThis service will analyse a black and white doodle and guess what have been drawn.- This service call Whisper via the Transformers library to transcribe an audio input and return the transcribed text.
audio-qa-service
PublicThis service transcribe and store the text of an audio file into a vector store, to do QA with an language model.pdf-extractor-service
Publicsummarization-service
PublicThis service uses langchain and LLMs to create summaries of documents.- This service uses rebalances a dataset based on a target class, it combines oversampling (SMOTE) and undersampling (ENN) to be more generalizable. In order for the service to work your dataset label column must be called "target". Finally, avoid having multiple empty lines at the end of the file.
- This service benchmarks a dataset with various models and outputs the results sorted by accuracy. In order for the service to work your dataset label column must be called "target". Also to improve the results you may want to remove uneccessary columns from the dataset. Finally, avoid having multiple empty lines at the end of the file.
- This service benchmarks a dataset with various models and outputs the results sorted by accuracy. In order for the service to work your dataset label column must be called "target". Also to improve the results you may want to remove uneccessary columns from the dataset. Finally, avoid having multiple empty lines at the end of the file.
- This service takes an image as input and outputs the same image, binarized. Zizag uses a window to estimate the average intensity around every pixel to determine whether that pixel is text (foreground) or background. Once this is done, the image is 'scanned' again with this window to make a background estimation, with which the text pixels will be.
common-code
Public