{"id":13467997,"url":"https://github.com/HazyResearch/meerkat","last_synced_at":"2025-03-26T03:31:24.957Z","repository":{"id":37268245,"uuid":"365070826","full_name":"HazyResearch/meerkat","owner":"HazyResearch","description":"Creative interactive views of any dataset. ","archived":false,"fork":false,"pushed_at":"2024-12-24T15:06:30.000Z","size":69728,"stargazers_count":836,"open_issues_count":11,"forks_count":43,"subscribers_count":12,"default_branch":"main","last_synced_at":"2025-03-24T17:09:44.521Z","etag":null,"topics":["data-science","foundation-models","machine-learning","ml","pandas"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/HazyResearch.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE.md","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2021-05-07T00:26:35.000Z","updated_at":"2025-03-22T10:27:16.000Z","dependencies_parsed_at":"2024-11-16T17:31:34.775Z","dependency_job_id":null,"html_url":"https://github.com/HazyResearch/meerkat","commit_stats":{"total_commits":1404,"total_committers":24,"mean_commits":58.5,"dds":0.7051282051282051,"last_synced_commit":"bfd44e0c6943d374e6034cb50c6e531067749ff9"},"previous_names":["robustness-gym/meerkat"],"tags_count":42,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HazyResearch%2Fmeerkat","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HazyResearch%2Fmeerkat/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HazyResearch%2Fmeerkat/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HazyResearch%2Fmeerkat/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/HazyResearch","download_url":"https://codeload.github.com/HazyResearch/meerkat/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245315291,"owners_count":20595218,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["data-science","foundation-models","machine-learning","ml","pandas"],"created_at":"2024-07-31T15:01:03.953Z","updated_at":"2025-03-26T03:31:24.949Z","avatar_url":"https://github.com/HazyResearch.png","language":"Python","funding_links":[],"categories":["Python"],"sub_categories":[],"readme":"\u003cdiv align=\"center\"\u003e\n    \u003cimg src=\"docs/assets/meerkat_banner.png\" height=100 alt=\"Meerkat logo\"/\u003e\n\n---\n\n\u003c!-- ![GitHub Workflow Status](https://github.com/HazyResearch/meerkat/actions/workflows/.github/workflows/ci.yml/badge.svg) --\u003e\n[![GitHub](https://img.shields.io/github/license/HazyResearch/meerkat)](https://img.shields.io/github/license/HazyResearch/meerkat)\n[![pre-commit](https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit\u0026logoColor=white)](https://github.com/pre-commit/pre-commit)\n\nCreate interactive views of any dataset.\n\n[**Website**](http://meerkat.wiki)\n| [**Quickstart**](http://meerkat.wiki/docs/start/quickstart-df.html)\n| [**Docs**](http://meerkat.wiki/docs/index.html)\n| [**Contributing**](CONTRIBUTING.md)\n| [**Discord**](https://discord.gg/pw8E4Q26Tq)\n| [**Blogpost**](https://hazyresearch.stanford.edu/blog/2023-03-01-meerkat)\n\n\u003c/div\u003e\n\n\n## ⚡️ Quickstart\n\n```bash\npip install meerkat-ml\n```\n\u003c!-- \n\u003e **_GPU Install_**: If you want to use Meerkat with a GPU, you will need to install PyTorch with GPU support. See [here](https://pytorch.org/get-started/locally/) for more details. --\u003e\n\n\u003c!-- ```bash\npip install \"meerkat-ml @ git+https://github.com/robustness-gym/meerkat@clever-dev\"\n```  --\u003e\n\u003c!-- \n\u003e **_Optional Dependencies_**: some parts of Meerkat rely on optional dependencies e.g. audio processing may rely on utilities from `torchaudio`. See  --\u003e\n\u003c!-- \nThen try one of our demos,\n\n```bash\nmk demo tutorial-image-gallery --copy\n```\n\nExplore the code for this demo in `tutorial-image-gallery.py`. To see a full list of demos, use `mk demo --help`.  --\u003e\n\n**Next Steps**.\nCheck out our [Getting Started page](http://meerkat.wiki/docs/start/quickstart-df.html) and our [documentation](http://meerkat.wiki/docs/index.html) to start building with Meerkat.\n\n## Why Meerkat?\n\nMeerkat is an open-source Python library that helps users visualize, explore, and annotate any dataset. It is especially useful when processing unstructured data types (_e.g._ free text, PDFs, images, video) with machine learning models. \n\n### ✏️ Features and Design Principles\n\nHere are four principles that inform Meerkat's design.\n\n**(1) Low overhead.**  With four lines of Python, start interacting with any dataset. \n- Zero-copy integrations with your preferred data abstractions: Pandas, Arrow, HF Datasets, Ibis, SQL.\n- Limited data movement. With Meerkat, you interact with your data where it already lives: no uploads to external databases and no reformatting.\n\n```python\nimport meerkat as mk\ndf = mk.from_csv(\"paintings.csv\")\ndf[\"image\"] = mk.files(\"image_url\")\ndf\n```\n\n\u003cdiv align=\"center\"\u003e\n  \u003cimg src=\"website/static/dataframe-demo.gif\" height=300 alt=\"Meerkat logo\"/\u003e\n\u003c/div\u003e\n\n\n**(2) Diverse data types.** Visualize and annotate almost any data type in Meerkat interfaces: text, images, audio, video, MRI scans, PDFs, HTML, JSON. \n\n\u003cdiv align=\"center\"\u003e\n\t\u003cimg src=\"website/static/data-types.gif\" height=300 alt=\"\"/\u003e\n\u003c/div\u003e\n\n\n**(3) \"Intelligent\" user interfaces.** Meerkat makes it easy to embed **machine learning models** (e.g. LLMs) within user interfaces to enable intelligent functionality such as searching, grouping and autocomplete. \n\n```python\ndf[\"embedding\"] = mk.embed(df[\"img\"], engine=\"clip\")\nmatch = mk.gui.Match(df,\n\tagainst=\"embedding\",\n\tengine=\"clip\"\n)\nsorted_df = mk.sort(df,\n\tby=match.criterion.name,\n\tascending=False\n)\ngallery = mk.gui.Gallery(sorted_df)\nmk.gui.html.div([match, gallery])\n```\n\n\u003cdiv align=\"center\"\u003e\n\t\u003cimg src=\"website/static/interact-demo.gif\" height=300 alt=\"Meerkat logo\"/\u003e\n\u003c/div\u003e\n\n**(4) Declarative (think: Seaborn), but also infinitely customizable and composable.**\nMeerkat visualization components can be composed and customized to create new interfaces. \n\n```python\nplot = mk.gui.plotly.Scatter(df=plot_df, x=\"umap_1\", y=\"umap_2\",)\n\n@mk.gui.reactive\ndef filter(selected: list, df: mk.DataFrame):\n    return df[df.primary_key.isin(selected)]\n\nfiltered_df = filter(plot.selected, plot_df)\ntable = mk.gui.Table(filtered_df, classes=\"h-full\")\n\nmk.gui.html.flex([plot, table], classes=\"h-[600px]\") \n```\n\n\u003cdiv align=\"center\"\u003e\n\t\u003cimg src=\"website/static/compose.gif\" height=300 alt=\"Meerkat logo\"/\u003e\n\u003c/div\u003e\n\n\n### ✨ Use cases where Meerkat shines\n- _Exploratory analysis over unstructured data types._ [Demo](https://www.youtube.com/watch?v=a8FBT33QACQ)\n- _Spot-checking the behavior of large language models (e.g. GPT-3)._  [Demo](https://www.youtube.com/watch?v=3ItA70qoe-o)\n- _Identifying systematic errors made by machine learning models._ [Demo](https://youtu.be/4Kk_LZbNWNs)\n- _Rapid labeling of validation data._\n\n### 🤔 Use cases where Meerkat may not be the right fit\n\n- _Are you only working with structured data (e.g. numerical and categorical variables)?_ Popular data visualization libraries (_e.g._ [Seaborn](https://seaborn.pydata.org/), [Matplotlib](https://matplotlib.org/)) are often sufficient. If you're looking for interactivity, [Plotly](https://plotly.com/) and [Streamlit](https://streamlit.io/) work well with structured data. Meerkat is differentiated in how it visualizes unstructured data types: long-form text, PDFs, HTML, images, video, audio...  \n- _Are you trying to make a straightforward demo of a machine learning model (single input/output, chatbot) and share with the world?_ [Gradio](https://gradio.app/) is likely a better fit! Though, if your demo involves visualizing lots of data, you may find Meerkat useful.\n- _Are you trying to manually label tens of thousands of data points?_  If you are looking for a data labeling tool to use with a labeling team, there are great open source labeling solutions designed for this (_e.g._ [LabelStudio](https://labelstud.io/)). In contrast, Meerkat is great fit for teams/individuals without access to a large labeling workforce who are using pretrained models (_e.g._ GPT-3) and need to label validation data or in-context examples.\n\n\n\n\n\u003c!-- Our goal is to make foundation models a more reliable\nsoftware abstraction for processing unstructured datasets.\n[Read our blogpost to learn more.](https://hazyresearch.stanford.edu/blog/2023-03-01-meerkat)\n --\u003e\n\n\n\n## ✉️ About\n\nMeerkat is being built by Machine Learning PhD students in the [Hazy Research](https://hazyresearch.stanford.edu) lab at Stanford. We're excited to build for a future where models will make it easier for teams to sift and reason through large volumes of unstructtured data effortlessly. \n\nPlease reach out to `kgoel [at] cs [dot] stanford [dot] edu, eyuboglu [at] stanford [dot] edu, and arjundd [at] stanford [dot] edu` if you would like to use Meerkat for a project, at your company or if you have any questions.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FHazyResearch%2Fmeerkat","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FHazyResearch%2Fmeerkat","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FHazyResearch%2Fmeerkat/lists"}