{ "openapi": "3.0.1", "info": { "title": "Hugging Face Datasets Catalog — ML Training Data Intel", "description": "Hugging Face dataset registry: downloads, likes, last_modified, task_categories, language, size_categories, license, tags, author. Filter by task/language/size. Sort by downloads/likes/trending/modified. ML researchers, MLOps, AI compliance.", "version": "0.0", "x-build-id": "BtyoeLlVYatQZgz5j" }, "servers": [ { "url": "https://api.apify.com/v2" } ], "paths": { "/acts/nexgendata~huggingface-datasets-catalog/run-sync-get-dataset-items": { "post": { "operationId": "run-sync-get-dataset-items-nexgendata-huggingface-datasets-catalog", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for its completion, and returns Actor's dataset items in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } }, "/acts/nexgendata~huggingface-datasets-catalog/runs": { "post": { "operationId": "runs-sync-nexgendata-huggingface-datasets-catalog", "x-openai-isConsequential": false, "summary": "Executes an Actor and returns information about the initiated run in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK", "content": { "application/json": { "schema": { "$ref": "#/components/schemas/runsResponseSchema" } } } } } } }, "/acts/nexgendata~huggingface-datasets-catalog/run-sync": { "post": { "operationId": "run-sync-nexgendata-huggingface-datasets-catalog", "x-openai-isConsequential": false, "summary": "Executes an Actor, waits for completion, and returns the OUTPUT from Key-value store in response.", "tags": [ "Run Actor" ], "requestBody": { "required": true, "content": { "application/json": { "schema": { "$ref": "#/components/schemas/inputSchema" } } } }, "parameters": [ { "name": "token", "in": "query", "required": true, "schema": { "type": "string" }, "description": "Enter your Apify token here" } ], "responses": { "200": { "description": "OK" } } } } }, "components": { "schemas": { "inputSchema": { "type": "object", "properties": { "task": { "title": "Task category filter (optional)", "type": "string", "description": "Restrict to a single Hugging Face task category — the canonical taxonomy that labels what each dataset is built to train. Common slugs: `text-classification`, `token-classification`, `question-answering`, `translation`, `summarization`, `text-generation`, `text2text-generation`, `fill-mask`, `sentence-similarity`, `feature-extraction`, `automatic-speech-recognition`, `audio-classification`, `text-to-speech`, `image-classification`, `image-segmentation`, `object-detection`, `image-to-text`, `text-to-image`, `visual-question-answering`, `document-question-answering`, `video-classification`, `reinforcement-learning`, `tabular-classification`, `tabular-regression`, `time-series-forecasting`. The API filter is `task_categories:{slug}` — passed verbatim. Browse the full taxonomy at https://huggingface.co/datasets?task_categories=task_categories. Empty = no task filter (return all datasets).", "default": "" }, "language": { "title": "Language filter (optional)", "type": "string", "description": "Restrict to datasets tagged with a specific ISO 639-1 language code — e.g. `en`, `fr`, `de`, `es`, `zh`, `ja`, `ko`, `ar`, `ru`, `pt`, `hi`, `bn`, `vi`, `id`, `th`, `tr`, `pl`, `it`, `nl`, `sv`. Hugging Face uses the `language:{code}` tag — passed verbatim to the API as `filter=language:{code}`. Multi-language datasets carry one `language:` tag per language they cover, so a filter of `en` will also return multilingual datasets that include English. Use `multilingual` for datasets explicitly tagged as cross-lingual. Empty = no language filter.", "default": "" }, "size": { "title": "Size bucket filter (optional)", "enum": [ "", "n<1K", "1K1T" ], "type": "string", "description": "Restrict to datasets in a single size bucket — Hugging Face's standard `size_categories` tag. Valid slugs: `n<1K`, `1K1T`. The size refers to the row count of the largest split. Useful for filtering toy/eval datasets (`n<1K`, `1K1T`). The API filter is `size_categories:{bucket}` — passed verbatim. Empty = no size filter.", "default": "" }, "sort": { "title": "Sort order", "enum": [ "downloads", "likes", "trending", "lastModified", "createdAt" ], "type": "string", "description": "How Hugging Face should rank the listing. `downloads` is the canonical popularity signal — total all-time dataset loads via `datasets.load_dataset()` and direct file downloads, the de-facto ranking on hf.co/datasets. `likes` ranks by community upvotes — better proxy for quality/curation than raw downloads. `trending` is the rolling-7-day engagement leaderboard (currently-hot datasets — newly-released benchmarks, viral instruction-tuning corpora, fresh evaluation suites). `lastModified` surfaces actively-maintained datasets (commit-recency proxy for staleness). `createdAt` ranks by upload date — the freshest datasets first (great for catching new releases before they accumulate downloads). The actor maps each value to the correct HF API param: `downloads`, `likes`, `lastModified`, `createdAt`, and `trending` (HF's documented trending sort).", "default": "downloads" }, "search": { "title": "Search query (optional)", "type": "string", "description": "Free-text search across dataset id + description (`?search=` on the HF datasets API). Use this for keyword-based discovery — e.g. `search=instruct`, `search=code`, `search=medical`, `search=multilingual`, `search=arxiv`. Pairs well with `task`/`language`/`size` filters: combine `task=question-answering` + `search=medical` to find medical QA datasets, or `language=fr` + `search=legal` to find French legal corpora. Empty = no search (the listing returns the full HF dataset registry sorted by the chosen `sort` order, filtered by the other params).", "default": "" }, "maxResults": { "title": "Max datasets", "minimum": 1, "maximum": 1000, "type": "integer", "description": "Hard ceiling on the number of dataset rows emitted per run (1–1000). The actor fires one paginated list request per ~100 results — a run of 50 datasets is a single API call, 500 is five. Hugging Face's public API tolerates ~500 requests per 5 minutes per IP unauthenticated, so even a maxResults=1000 run (≤10 API calls) stays well below the rate ceiling. Use 10 for smoke tests, 50 for a typical trend snapshot (default), 200 for a daily-sweep of top training corpora, 1000 for a full weekly archive.", "default": 50 } } }, "runsResponseSchema": { "type": "object", "properties": { "data": { "type": "object", "properties": { "id": { "type": "string" }, "actId": { "type": "string" }, "userId": { "type": "string" }, "startedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "finishedAt": { "type": "string", "format": "date-time", "example": "2025-01-08T00:00:00.000Z" }, "status": { "type": "string", "example": "READY" }, "meta": { "type": "object", "properties": { "origin": { "type": "string", "example": "API" }, "userAgent": { "type": "string" } } }, "stats": { "type": "object", "properties": { "inputBodyLen": { "type": "integer", "example": 2000 }, "rebootCount": { "type": "integer", "example": 0 }, "restartCount": { "type": "integer", "example": 0 }, "resurrectCount": { "type": "integer", "example": 0 }, "computeUnits": { "type": "integer", "example": 0 } } }, "options": { "type": "object", "properties": { "build": { "type": "string", "example": "latest" }, "timeoutSecs": { "type": "integer", "example": 300 }, "memoryMbytes": { "type": "integer", "example": 1024 }, "diskMbytes": { "type": "integer", "example": 2048 } } }, "buildId": { "type": "string" }, "defaultKeyValueStoreId": { "type": "string" }, "defaultDatasetId": { "type": "string" }, "defaultRequestQueueId": { "type": "string" }, "buildNumber": { "type": "string", "example": "1.0.0" }, "containerUrl": { "type": "string" }, "usage": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "integer", "example": 1 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } }, "usageTotalUsd": { "type": "number", "example": 0.00005 }, "usageUsd": { "type": "object", "properties": { "ACTOR_COMPUTE_UNITS": { "type": "integer", "example": 0 }, "DATASET_READS": { "type": "integer", "example": 0 }, "DATASET_WRITES": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_READS": { "type": "integer", "example": 0 }, "KEY_VALUE_STORE_WRITES": { "type": "number", "example": 0.00005 }, "KEY_VALUE_STORE_LISTS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_READS": { "type": "integer", "example": 0 }, "REQUEST_QUEUE_WRITES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_INTERNAL_GBYTES": { "type": "integer", "example": 0 }, "DATA_TRANSFER_EXTERNAL_GBYTES": { "type": "integer", "example": 0 }, "PROXY_RESIDENTIAL_TRANSFER_GBYTES": { "type": "integer", "example": 0 }, "PROXY_SERPS": { "type": "integer", "example": 0 } } } } } } } } } }