feat: 🎸 support beans and cats_vs_dogs (#102)

See huggingface/datasets#3472
mattstern31 · Dec 22, 2021 · f83ff81 · f83ff81
1 parent bb6cf98
commit f83ff81
Show file tree

Hide file tree

Showing 5 changed files with 27 additions and 21 deletions.
diff --git a/.env.example b/.env.example
@@ -18,7 +18,7 @@
 # EXTRACT_ROWS_LIMIT=100
 
 # Git reference for the canonical datasets on https://github.com/huggingface/datasets
-# DATASETS_REVISION="1.17.0"
+# DATASETS_REVISION="c8f914473b041833fd47178fa4373cdcb56ac522"
 
 # Hugging Face token (see https://huggingface.co/settings/token)
 # HF_TOKEN=

diff --git a/README.md b/README.md
@@ -31,7 +31,7 @@ Set environment variables to configure the following aspects:
 - `APP_PORT`: the port used by the app. Defaults to `8000`.
 - `ASSETS_DIRECTORY`: directory where the asset files are stored. Defaults to empty, in which case the assets are located in the `datasets_preview_backend_assets` subdirectory inside the OS default cache directory.
 - `DATASETS_ENABLE_PRIVATE`: enable private datasets. Defaults to `False`.
-- `DATASETS_REVISION`: git reference for the canonical datasets on https://github.com/huggingface/datasets. Defaults to `1.17.0`.
+- `DATASETS_REVISION`: git reference for the canonical datasets on https://github.com/huggingface/datasets. Defaults to `c8f914473b041833fd47178fa4373cdcb56ac522`.
 - `EXTRACT_ROWS_LIMIT`: number of rows in the extract, if not specified in the API request. Defaults to `100`.
 - `LOG_LEVEL`: log level, among `DEBUG`, `INFO`, `WARNING`, `ERROR` and `CRITICAL`. Defaults to `INFO`.
 - `MAX_AGE_LONG_SECONDS`: number of seconds to set in the `max-age` header on data endpoints. Defaults to `21600` (6 hours).

diff --git a/poetry.lock b/poetry.lock
diff --git a/pyproject.toml b/pyproject.toml
@@ -2,7 +2,7 @@
 authors = ["Sylvain Lesage <severo@rednegra.net>"]
 description = "API to extract rows of 🤗 datasets"
 name = "datasets-preview-backend"
-version = "0.15.6"
+version = "0.15.7"
 
 [tool.poetry.dependencies]
 Pillow = "^8.4.0"
@@ -12,7 +12,10 @@ apache-beam = "^2.33.0"
 appdirs = "^1.4.4"
 bs4 = "^0.0.1"
 conllu = "^4.4.1"
-datasets = { extras = ["audio", "vision"], version = "^1.17.0" }
+datasets = { git = "https://github.com/huggingface/datasets.git", rev = "c8f914473b041833fd47178fa4373cdcb56ac522", extras = [
+    "audio",
+    "vision",
+] }
 diskcache = "^5.2.1"
 function-parser = "^0.0.3"
 gdown = "^4.2.0"

diff --git a/src/datasets_preview_backend/constants.py b/src/datasets_preview_backend/constants.py
@@ -4,7 +4,7 @@
 DEFAULT_APP_PORT: int = 8000
 DEFAULT_ASSETS_DIRECTORY: None = None
 DEFAULT_DATASETS_ENABLE_PRIVATE: bool = False
-DEFAULT_DATASETS_REVISION: str = "1.17.0"
+DEFAULT_DATASETS_REVISION: str = "c8f914473b041833fd47178fa4373cdcb56ac522"
 DEFAULT_EXTRACT_ROWS_LIMIT: int = 100
 DEFAULT_LOG_LEVEL: str = "INFO"
 DEFAULT_MAX_AGE_LONG_SECONDS: int = 21600  # 6 * 60 * 60 = 6 hours