Update bioclip-2_float16: metadata backfill + UUID dedup

.gitattributes

@@ -58,4 +58,3 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
-bioclip-2_float16/*.parquet filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,44 +1,9 @@
 ---
-license: cc0-1.0
-language:
-- en
-pretty_name: TreeOfLife-200M Embeddings
-task_categories:
-- feature-extraction
-tags:
-- biology
-- biodiversity
-- embeddings
-- bioclip
-- tree-of-life
-- imageomics
-- parquet
-- duckdb
-- taxonomy
-- plants
-- animals
-- fungi
-- image
-size_categories:
-- 100M<n<1B
-datasets:
-- imageomics/TreeOfLife-200M
-description: >-
-  Pre-computed image embeddings for 200M+ images from the TreeOfLife-200M
-  dataset, sorted by taxonomic hierarchy for efficient filtered access via
-  DuckDB. Multiple embedding model configs supported. No images stored.
 dataset_info:
   - config_name: bioclip-2_float16
     features:
       - name: uuid
         dtype: string
-      - name: emb
-        sequence: float16
-        length: 768
-      - name: source_dataset
-        dtype: string
-      - name: source_id
-        dtype: string
       - name: kingdom
         dtype: string
       - name: phylum
@@ -53,260 +18,87 @@ dataset_info:
         dtype: string
       - name: species
         dtype: string
-      - name: scientific_name
+      - name: img_type
         dtype: string
-      - name: common_name
+      - name: basisOfRecord
         dtype: string
       - name: publisher
         dtype: string
-      - name: basisOfRecord
-        dtype: string
       - name: identifier
         dtype: string
-      - name: img_type
-        dtype: string
+      - name: emb
+        sequence: float16
+        length: 768
     splits:
       - name: train
-        num_examples: 233055986
+        num_examples: 239580103
 configs:
   - config_name: bioclip-2_float16
     data_files:
       - split: train
         path: bioclip-2_float16/*.parquet
-  - config_name: darwins_finches
-    data_files:
-      - split: train
-        path: demo_subset/darwins_finches/bioclip-2_float16/*.parquet
-  - config_name: wolf_sample
-    data_files:
-      - split: train
-        path: demo_subset/wolf_sample/bioclip-2_float16/*.parquet
+license: mit
 ---
 
 # TreeOfLife-200M Embeddings
 
-<!-- TODO: Remove after reviewed -->
-> **Work in progress.** This dataset and card are under active development.
-
-Pre-computed image embeddings for 200M+ images from the [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) dataset, sorted by taxonomic hierarchy for efficient filtered access.
-
-This repository hosts embedding **configs** for TreeOfLife-200M. Each config corresponds to a different embedding model and/or precision. Currently available: [BioCLIP 2](https://huggingface.co/imageomics/bioclip-2) (float16). Additional configs (e.g., [BioCLIP 2.5 Huge](https://huggingface.co/imageomics/bioclip-2.5-vith14)) will be added as new embeddings are generated.
-
-We recommend using [DuckDB](https://duckdb.org/) to access this dataset.
+> **Work in progress** — this dataset card is under active development.
 
-**This repository does not contain or redistribute any images.** It contains only pre-computed embedding vectors and metadata derived from [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M).
+239 million BioCLIP 2 embeddings (768-dim, float16) for [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) images, sorted by taxonomic hierarchy. We recommend using [DuckDB](https://duckdb.org/) to access this dataset.
 
-## Dataset Details
+## Query Remotely
 
-### Dataset Description
+Query directly from HuggingFace without downloading anything. DuckDB uses Parquet page indexes to skip irrelevant files, making filtered queries fast over the network.
 
-<!-- TODO: confirm curated by list -->
-- **Curated by:** Net Zhang, Elizabeth Campolongo, Matthew Thompson, Jianyang Gu
-- **Repository:** <!-- TODO: link to project repo -->
-- **Paper:** [BioCLIP 2: Emergent Properties from Scaling Hierarchical Contrastive Learning](https://arxiv.org/abs/2505.23883)
-- **License:** [CC0 1.0](https://creativecommons.org/publicdomain/zero/1.0/) for embeddings; metadata inherits upstream licenses (see [Licensing Information](#licensing-information)).
-- **Source dataset:** [imageomics/TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M)
-- **Embedding models:** [BioCLIP 2](https://huggingface.co/imageomics/bioclip-2), [BioCLIP 2.5 Huge](https://huggingface.co/imageomics/bioclip-2.5-vith14) (more configs coming)
+> `order` and `class` are SQL reserved words — always quote them as `"order"` and `"class"`.
 
-This repository contains pre-computed embedding vectors derived from the [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) dataset. Each row pairs a TreeOfLife-200M image (identified by `uuid`) with its embedding vector and associated taxonomic metadata. The data is sorted by taxonomy and stored in Parquet format with page indexes, enabling fast filtered access through Hugging Face. 
-
-### Supported Tasks
-
-- **Feature extraction:** Retrieve pre-computed embedding vectors for any (taxonomic) subset without re-running the image embedding model.
-- **Taxonomy-filtered retrieval:** Query embeddings by kingdom, order, family, genus, or species using DuckDB with page-index predicate pushdown.
-- **Downstream training:** Use filtered embedding subsets for fine-tuning classifiers, training probes, or building custom indices.
-- **Exact similarity computation:** Compute similarity between embeddings.
-
-
-## Dataset Structure
-
-### Configs
-
-Each config corresponds to a different embedding model and precision. Users load a specific config:
-
-```python
-# Python
-from datasets import load_dataset
-ds = load_dataset("imageomics/TreeOfLife-200M-Embeddings", "bioclip-2_float16")
-```
-
-| Config | Model | Precision | Dimensions | Files | Size | Rows |
-|--------|-------|-----------|------------|-------|------|------|
-| `bioclip-2_float16` | [BioCLIP 2](https://huggingface.co/imageomics/bioclip-2) (ViT-L/14) | float16 | 768 | 666 | ~346 GB | 233,055,986 |
-| `darwins_finches` | [BioCLIP 2](https://huggingface.co/imageomics/bioclip-2) (ViT-L/14) | float16 | 768 | 1 | ~1 MB | 677 |
-| `wolf_sample` | [BioCLIP 2](https://huggingface.co/imageomics/bioclip-2) (ViT-L/14) | float16 | 768 | 1 | ~1.5 MB | 960 |
-
-The `demo_subset/` directory contains small **curated subsets** of the full
-`bioclip-2_float16` TreeOfLife-200M embeddings, such as `darwins_finches` and
-`wolf_sample`. These subsets were curated for quick experiments, prototyping,
-and demos—for example the
-[Image Embedding Explorer](https://github.com/Imageomics/emb-explorer) demo—without
-downloading the full 200M embedding dataset.
-
-### Data Fields
-
-| Column | Type | Description |
-|--------|------|-------------|
-| `uuid` | `string` | Unique image identifier (matches [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M)) |
-| `emb` | `fixed_size_list<float>[N]` | Image embedding vector (dimensions depend on config) |
-| `source_dataset` | `string` | Data source: `gbif`, `eol`, `bioscan`, or `fathomnet` |
-| `source_id` | `string` | Unique identifier from source (e.g., GBIF occurrence ID) |
-| `kingdom` | `string` | Taxonomic kingdom |
-| `phylum` | `string` | Taxonomic phylum |
-| `class` | `string` | Taxonomic class |
-| `order` | `string` | Taxonomic order |
-| `family` | `string` | Taxonomic family |
-| `genus` | `string` | Taxonomic genus |
-| `species` | `string` | Species epithet (e.g., `plexippus`; use `scientific_name` for the full binomial `Danaus plexippus`) |
-| `scientific_name` | `string` | Full scientific name |
-| `common_name` | `string` | Vernacular/common name |
-| `publisher` | `string` | Organization that published the data (GBIF records only; NULL for non-GBIF) |
-| `basisOfRecord` | `string` | GBIF basis of record (GBIF records only; NULL for non-GBIF) |
-| `identifier` | `string` | Source image URL |
-| `img_type` | `string` | Image type (e.g., `Citizen Science`, `Museum Specimen: Fungi`, `Camera-trap`); GBIF records only, NULL for non-GBIF |
-
-For more background on metadata columns, see the [TreeOfLife-200M data field descriptions](https://huggingface.co/datasets/imageomics/TreeOfLife-200M#data-fields).
-
-### Data Organization
-
-- **Sort order:** `source_dataset > kingdom > phylum > class > order > family > genus > species > common_name`
-- **Row groups:** 50,000 rows each with column statistics and page indexes
-- **Compression:** ZSTD level 3
-- **Precision:** config-dependent (see [Configs](#configs) table)
-- **File naming:** `train-00000-of-NNNNN.parquet` (count depends on config)
-
-Repository layout:
-
-```text
-TreeOfLife-200M-Embeddings/
-├── bioclip-2_float16/                         # full dataset — 666 shards
-│   ├── train-00000-of-00666.parquet
-│   └── ...
-└── demo_subset/                               # curated subsets for quick experiments / demos
-    ├── darwins_finches/
-    │   └── bioclip-2_float16/
-    │       └── emb_darwin_finches.parquet     # 677 rows
-    └── wolf_sample/
-        └── bioclip-2_float16/
-            └── emb_wolf_sample.parquet        # 960 rows
-```
-
-### Data Splits
-
-Single `train` split containing all 233,055,986 rows.
-
-
-## Usage
-
-We recommend using [DuckDB](https://duckdb.org/) to access this dataset. All examples below use the `bioclip-2_float16` config; replace with the appropriate config name for other embeddings.
-
-> `order` and `class` are SQL reserved words: always quote them as `"order"` and `"class"`.
-
-**Choosing between download vs. remote query.** Downloading the full config is the preferred path. Once local, queries run in seconds, read access is deterministic, and there are no network failure modes. The tradeoffs: it requires ~320 GB of disk, and users must occasionally re-sync as the dataset is updated on Hugging Face (`hf download` skips unchanged files based on LFS hashes, so subsequent syncs only transfer what changed). Remote queries over `hf://` are still useful for quick checks such as row counts, metadata browsing, aggregate statistics, and fetching a single species' embeddings. However, at the time of writing, `COPY` of larger slices (more than ~20 MB of `emb` data spanning multiple files) is unstable and frequently fails mid-transfer with `TProtocolException`. For any repeated batch access to embeddings at scale, download first.
-
-### Download, then Query Locally (Recommended)
-
-The fastest and most reliable workflow is to download the config once, then run DuckDB queries against the local files. All filtered slices return in seconds thanks to ZSTD decompression and page-index predicate pushdown.
-
-**Benchmark** ([Cardinal](https://www.osc.edu/resources/technical_support/supercomputers/cardinal) login node, single session): full download ~6 min at ~1 GB/s. Once local, a Felidae slice (152,867 rows, 466 MB) is extracted in ~6 s.
-
-#### 1. Download the config
-
-```bash
-hf download imageomics/TreeOfLife-200M-Embeddings \
-    --repo-type dataset \
-    --include "bioclip-2_float16/*" \
-    --local-dir ./data
-```
-
-**Check if your local copy is in sync.** Re-run the same `hf download` command. It performs an incremental sync: files whose LFS hashes already match the remote are skipped with no bytes transferred, and only new or changed files are downloaded. A fully in-sync local copy completes in about 1 second.
-
-#### 2. Query locally with DuckDB
+### Python
 
 ```python
 import duckdb
 
 con = duckdb.connect()
-glob = "./data/bioclip-2_float16/*.parquet"
+# Authenticate for higher rate limits (recommended)
+con.execute("CREATE SECRET (TYPE HUGGINGFACE, TOKEN 'hf_...')")
 
-# Extract a species slice to a new file (fast: ~6s for 150K rows)
-con.sql(f"""
-    COPY (
-        SELECT * FROM read_parquet('{glob}')
-        WHERE family = 'Felidae'
-    ) TO 'felidae.parquet'
-""")
+glob = "hf://datasets/imageomics/TreeOfLife-200M-Embeddings/bioclip-2_float16/*.parquet"
 
-# Aggregate across the full config (fast: page index + dictionary encoding)
+# Get all embeddings for a species
+df = con.sql(f"""
+    SELECT uuid, emb
+    FROM read_parquet('{glob}')
+    WHERE species = 'Danaus plexippus'
+""").df()
+
+# Count by taxonomic order
 con.sql(f"""
-    SELECT "order", COUNT(*) AS n
+    SELECT "order", COUNT(*) as n
     FROM read_parquet('{glob}')
     WHERE "order" IS NOT NULL
     GROUP BY "order"
     ORDER BY n DESC
     LIMIT 10
 """).show()
-```
-
-```bash
-# CLI equivalent
-duckdb -c "
-    COPY (SELECT * FROM read_parquet('./data/bioclip-2_float16/*.parquet')
-          WHERE \"order\" = 'Primates')
-    TO 'primates.parquet';
-"
-```
-
-#### Slice sizes (after local extraction)
-
-| Slice | Rows | Size on disk | Local `COPY` time* |
-|---|---|---|---|
-| `scientific_name = 'Puma concolor'` | 11,811 | 36 MB | 1.7 s |
-| `family = 'Felidae'` | 152,867 | 466 MB | 5.9 s |
-| `"order" = 'Primates'` | 164,239 | 501 MB | 5.8 s |
-
-*Measured on the [Cardinal](https://www.osc.edu/resources/technical_support/supercomputers/cardinal) cluster login node after the full config is already downloaded locally.*
-
-### Query Remotely
-
-For **aggregate queries** and **small embedding slices**, you can query directly from Hugging Face with DuckDB. The `hf://` protocol streams only the page ranges needed, so filtered aggregations typically return in ~10 s and a single-species `COPY` finishes in under a minute.
-
-> **Limitations of remote `COPY`:** sustained multi-file reads from `hf://` (e.g., extracting >~20 MB of embeddings across many files) can be interrupted by the Hugging Face CDN and fail with `TProtocolException: Invalid data`. If this happens, fall back to the [download-then-query workflow](#download-then-query-locally-recommended) above.
-
-#### Python
-
-```python
-import duckdb
 
-con = duckdb.connect()
-# Authenticate for higher rate limits (recommended)
-con.execute("CREATE SECRET (TYPE HUGGINGFACE, TOKEN 'hf_...')")
-
-glob = "hf://datasets/imageomics/TreeOfLife-200M-Embeddings/bioclip-2_float16/*.parquet"
-
-# Count by taxonomic order (aggregate: fast, ~10s)
+# Filter by family
 con.sql(f"""
-    SELECT "order", COUNT(*) AS n
+    SELECT genus, species, COUNT(*) as n
     FROM read_parquet('{glob}')
-    WHERE "order" IS NOT NULL
-    GROUP BY "order"
+    WHERE family = 'Felidae'
+    GROUP BY genus, species
     ORDER BY n DESC
     LIMIT 10
 """).show()
-
-# Fetch embeddings for a single species (~18 MB, safe slice size)
-df = con.sql(f"""
-    SELECT uuid, emb
-    FROM read_parquet('{glob}')
-    WHERE scientific_name = 'Puma concolor'
-""").df()
 ```
 
-#### CLI
+### CLI
 
 ```bash
+# Install DuckDB: https://duckdb.org/docs/installation
+# Then query directly from the command line:
+
 duckdb -c "
-    SELECT species, COUNT(*) AS n
+    SELECT species, COUNT(*) as n
     FROM read_parquet('hf://datasets/imageomics/TreeOfLife-200M-Embeddings/bioclip-2_float16/*.parquet')
     WHERE family = 'Nymphalidae'
     GROUP BY species
@@ -315,135 +107,98 @@ duckdb -c "
 "
 ```
 
-*First query in a session includes a one-time ~3 min glob resolution overhead: DuckDB expands the `*.parquet` wildcard by making HTTP requests to Hugging Face to discover all 666 filenames. Subsequent queries in the same session reuse the cached file list.*
-
-## Dataset Creation
-
-### Curation Rationale
-
-The [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) dataset contains 200M+ organism images totaling ~92 TB. Running an embedding model over this corpus is expensive. This repository provides pre-computed embeddings to enable downstream tasks such as:
-
-- Training taxonomic classifiers or linear probes on embedding subsets
-- Computing exact pairwise similarity between organisms
-- Building custom search indices or retrieval systems
-- Analyzing embedding space structure across taxonomic groups
-
-### Source Data
-
-#### Data Collection and Processing
-
-1. **Embedding generation.** All images in [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) are embedded using a vision foundation model (see [Configs](#configs) for model details). Each image produces a fixed-dimensional vector.
-2. **Global sort.** All rows are sorted by `source_dataset > kingdom > phylum > class > order > family > genus > species > common_name`, so taxonomically similar rows are adjacent. This enables page-index predicate pushdown for fast filtered access.
-3. **Precision.** Embedding precision is config-dependent (e.g., float16 for `bioclip-2_float16`).
-4. **Parquet write.** PyArrow writes files targeting ~500 MB each, with ZSTD level 3 compression, [page indexes](https://huggingface.co/docs/hub/en/datasets-data-files-configuration#dataset-viewer-size-limit-errors-toobigcontenterror), and 50,000-row row groups for optimal streaming and query performance.
-
-#### Source Data Producers
-
-- **Images:** [TreeOfLife-200M revision 94bbc0b](https://huggingface.co/datasets/imageomics/TreeOfLife-200M/commit/94bbc0b90cd26cbb15f94d3144941026f4b41213), aggregated from [GBIF](https://www.gbif.org/), [EOL](https://eol.org/), [BIOSCAN-5M](https://github.com/bioscan-ml/BIOSCAN-5M), and [FathomNet](https://www.fathomnet.org/).
-- **Taxonomic metadata:** Standardized using [TaxonoPy](https://github.com/Imageomics/TaxonoPy).
-- **Other metadata:** For instance, publisher, basis of record, image URLs, source IDs are inherited from the respective data sources via the [TreeOfLife-200M catalog revision 94bbc0b](https://huggingface.co/datasets/imageomics/TreeOfLife-200M/commit/94bbc0b90cd26cbb15f94d3144941026f4b41213).
-
-### Annotations
-
-This dataset does not include annotations created specifically for this repository. All taxonomic labels, common names, and provenance metadata are inherited directly from the TreeOfLife-200M catalog, which aligned the taxonomic names provided by [GBIF](https://www.gbif.org/), [EOL](https://eol.org/), [BIOSCAN-5M](https://github.com/bioscan-ml/BIOSCAN-5M), and [FathomNet](https://www.fathomnet.org/) using [TaxonoPy](https://imageomics.github.io/TaxonoPy/). See the [TreeOfLife-200M dataset card](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) for details on annotation processes and provenance.
-
-### Personal and Sensitive Information
-
-This repository does not contain or redistribute any images. However, the metadata includes URLs (`identifier` column) pointing to source images that may occasionally contain humans in the background (e.g., citizen science observations). The upstream TreeOfLife-200M dataset applies human face detection filtering to minimize such occurrences. See the [TreeOfLife-200M dataset card](https://huggingface.co/datasets/imageomics/TreeOfLife-200M#data-curation-and-processing) for details.
-
-
-## Considerations for Using the Data
-
-### Bias, Risks, and Limitations
-
-This dataset inherits biases from [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M#considerations-for-using-the-data):
+## Download a Taxonomic Slice
 
-- **Taxonomic coverage is uneven.** Citizen science observations (primarily iNaturalist) comprise ~60% of the data, skewing representation toward charismatic species and regions where citizen science is most active.
-- **Incomplete taxonomic labels.** Not all records have complete taxonomy at every rank. Records may have NULL values at lower ranks and *incertae sedis* at kingdom level due to biodiversity data complexities (unresolved taxonomy, ambiguous identifications, etc.).
-- **Embedding bias.** Similarity is determined by the embedding model, which may encode biases from its training data.
+Instead of downloading the full 331 GB dataset, extract only the taxa you need. The sorted data and page indexes ensure only the relevant row groups are fetched over the network.
 
-### Recommendations
+### Python
 
-- When using results for research, verify taxonomic labels against authoritative sources, since labels are inherited from community-contributed data.
-- Be aware of geographic and taxonomic sampling biases when interpreting embedding-based analyses.
-- For issues with specific records, report via the [Community tab](https://huggingface.co/datasets/imageomics/TreeOfLife-200M-Embeddings/discussions).
-
-
-## Related Datasets
+```python
+import duckdb
 
-- [TreeOfLife-200M](https://huggingface.co/datasets/imageomics/TreeOfLife-200M) (source images and metadata).
-- [bioclip-image-search-lite](https://huggingface.co/datasets/imageomics/bioclip-image-search-lite) (FAISS index and DuckDB metadata for BioCLIP 2 image similarity search over TreeOfLife-200M).
+con = duckdb.connect()
+con.execute("CREATE SECRET (TYPE HUGGINGFACE, TOKEN 'hf_...')")
 
+glob = "hf://datasets/imageomics/TreeOfLife-200M-Embeddings/bioclip-2_float16/*.parquet"
 
+# Save a single family to a local file
+con.sql(f"""
+    COPY (
+        SELECT * FROM read_parquet('{glob}')
+        WHERE family = 'Felidae'
+    ) TO 'felidae.parquet'
+""")
 
-## Licensing Information
+# Now query locally — instant
+con.sql("SELECT species, COUNT(*) as n FROM 'felidae.parquet' GROUP BY species ORDER BY n DESC").show()
+```
 
-The embedding vectors in this repository are dedicated to the public domain under the [CC0 1.0 Universal Public Domain Dedication](https://creativecommons.org/publicdomain/zero/1.0/). The metadata inherits licensing terms from its upstream sources (GBIF, EOL, BIOSCAN-5M, FathomNet); see the [TreeOfLife-200M licensing information](https://huggingface.co/datasets/imageomics/TreeOfLife-200M#licensing-information) for per-record details.
+### CLI
 
-**Important:** This repository does not contain or redistribute any images. The metadata includes URLs pointing to images hosted by their original sources. Individual images retain their original source licenses. Users must respect each image's original license terms when accessing images via the provided URLs.
+```bash
+duckdb -c "
+    COPY (
+        SELECT * FROM read_parquet('hf://datasets/imageomics/TreeOfLife-200M-Embeddings/bioclip-2_float16/*.parquet')
+        WHERE \"order\" = 'Primates'
+    ) TO 'primates.parquet';
+"
+```
 
-We ask that you cite this dataset and associated papers if you make use of it in your research.
+| Slice Example | Rows | Size | Time |
+|---------------|------|------|------|
+| Species: *Danaus plexippus* | 299,874 | 913 MB | ~1 min* |
+| Family: Felidae | 149,890 | 457 MB | ~1 min* |
+| Order: Primates | 160,307 | 489 MB | ~1 min* |
 
+*\*First query in a session includes a one-time ~3 min glob resolution overhead: DuckDB expands the `*.parquet` wildcard by making HTTP requests to HuggingFace to discover all 685 filenames. Subsequent queries in the same session reuse the cached file list.*
 
-## Citation
+## Download Everything
 
-<!-- TODO: generate DOI and finalize citation -->
-<!-- TODO: Check Citation Format -->
-**Data:**
-```bibtex
-@misc{treeoflife_200m_embeddings,
-  author = {Zhang, Net and Campolongo, Elizabeth and Thompson, Matthew and Gu, Jianyang},
-  title = {{TreeOfLife-200M Embeddings}},
-  year = {2026},
-  url = {https://huggingface.co/datasets/imageomics/TreeOfLife-200M-Embeddings},
-  publisher = {Hugging Face}
-}
-```
+For full-dataset analytics, download all files first.
 
-Please also cite the source dataset and embedding model (include data sources as appropriate):
-
-**TreeOfLife-200M:**
-```bibtex
-@dataset{treeoflife_200m,
-  title = {{T}ree{O}f{L}ife-200{M} (Revision 94bbc0b)}, 
-  author = {Jianyang Gu and Samuel Stevens and Elizabeth G Campolongo and Matthew J Thompson and Net Zhang and Jiaman Wu and Andrei Kopanev and Zheda Mai and Alexander E. White and James Balhoff and Wasila M Dahdul and Daniel Rubenstein and Hilmar Lapp and Tanya Berger-Wolf and Wei-Lun Chao and Yu Su},
-  year = {2025},
-  url = {https://huggingface.co/datasets/imageomics/TreeOfLife-200M},
-  doi = {10.57967/hf/8980},
-  publisher = {Hugging Face}
-}
+```bash
+huggingface-cli download imageomics/TreeOfLife-200M-Embeddings \
+  --repo-type dataset --include "bioclip-2_float16/*" --local-dir ./data
 ```
 
-**BioCLIP 2 Paper:**
-```bibtex
-@inproceedings{NEURIPS2025_94da80cb,
- author = {Gu, Jianyang and Stevens, Sam and Campolongo, Elizabeth and Thompson, Matthew and Zhang, Net and Wu, Jiaman and Kopanev, Andrei and Mai, Zheda and White, Alexander and Balhoff, James and Dahdul, Wasila and Rubenstein, Daniel and Lapp, Hilmar and Berger-Wolf, Tanya and Chao, Wei-Lun (Harry) and Su, Yu},
- booktitle = {Advances in Neural Information Processing Systems},
- editor = {D. Belgrave and C. Zhang and H. Lin and R. Pascanu and P. Koniusz and M. Ghassemi and N. Chen},
- pages = {102778--102811},
- publisher = {Curran Associates, Inc.},
- title = {{BioCLIP 2:} Emergent Properties from Scaling Hierarchical Contrastive Learning},
- url = {https://proceedings.neurips.cc/paper_files/paper/2025/file/94da80cbfe870c1db958c88a8a27018c-Paper-Conference.pdf},
- volume = {38},
- year = {2025}
-}
+```bash
+duckdb -c "
+    SELECT \"order\", COUNT(*) as n
+    FROM read_parquet('./data/bioclip-2_float16/*.parquet')
+    WHERE \"order\" IS NOT NULL
+    GROUP BY \"order\"
+    ORDER BY n DESC
+    LIMIT 10;
+"
 ```
 
-<!--https://arxiv.org/abs/2505.23883-->
-
-
-## Acknowledgements
-
-This work was supported by the [Imageomics Institute](https://imageomics.org), which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under [Award #2118240](https://www.nsf.gov/awardsearch/showAward?AWD_ID=2118240) (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
-
-This work used resources of the [Ohio Supercomputer Center (OSC)](https://www.osc.edu/): Ohio Supercomputer Center. 1987. Ohio Supercomputer Center. Columbus OH: Ohio Supercomputer Center. https://ror.org/01apna436.
-
+## Schema
 
-## Dataset Card Authors
+| Column | Type | Description |
+|--------|------|-------------|
+| `uuid` | `string` | Unique image identifier |
+| `kingdom` | `string` | Taxonomic kingdom |
+| `phylum` | `string` | Taxonomic phylum |
+| `class` | `string` | Taxonomic class |
+| `order` | `string` | Taxonomic order |
+| `family` | `string` | Taxonomic family |
+| `genus` | `string` | Taxonomic genus |
+| `species` | `string` | Taxonomic species |
+| `img_type` | `string` | Image source type |
+| `basisOfRecord` | `string` | Basis of record |
+| `publisher` | `string` | Data publisher |
+| `identifier` | `string` | Source URI |
+| `emb` | `fixed_size_list<float16>[768]` | BioCLIP 2 embedding |
 
-Net Zhang, Elizabeth Campolongo
+## Data Organization
 
+- **Sort order:** `kingdom > phylum > class > order > family > genus > species`
+- **Row groups:** 50,000 rows each with column statistics and page indexes
+- **Compression:** ZSTD level 3
+- **Precision:** float16 (lossless cosine similarity vs float32 source)
 
-## Dataset Card Contact
+## Configs
 
-For questions or issues, please use the [Community tab](https://huggingface.co/datasets/imageomics/TreeOfLife-200M-Embeddings/discussions) on this repository.
+| Config | Model | Precision | Files | Size | Rows |
+|--------|-------|-----------|-------|------|------|
+| `bioclip-2_float16` | BioCLIP 2 | float16 | 685 | 331 GB | 239,580,103 |

bioclip-2_float16/{train-00000-of-00666.parquet → train-00000-of-00685.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c0346c6a35b4cef66a367736d2a0925691e32172268d87e6785bce0377c55bf
-size 516095035
+oid sha256:183a8bdfb3a165620cd76ed5a3bf57bf6719497639f3e6161b0e210140afc6bb
+size 516207075

bioclip-2_float16/{train-00001-of-00666.parquet → train-00001-of-00685.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4ca3831481dd32718b1ac280a6b022fa6f361f2790a2897b6eaf0b60054c1b31
-size 515960496
+oid sha256:9c160ef8d9d31043aebbab4e7b1dab7336835d3a7e2ffb3518531bdd4d958fc2
+size 515949331

bioclip-2_float16/{train-00002-of-00666.parquet → train-00002-of-00685.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8432bf4fd9f4f6d54b950bba5f7c22484ebbdf88cbcb5eafbb7ef7e99d21b29
-size 515970027
+oid sha256:d1715962304c44d7b636cb447df5eefd3e59513117ed9ffdb5bbffb3904e187b
+size 516056737

bioclip-2_float16/train-00003-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:557987ef559a1911eaecca0efc535e63ca3cada33e3d1f08733dfaecb4c968bd
+size 515924706

bioclip-2_float16/train-00004-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f2dc954bde0bc54f1bfb09106415b59f24f9af921f221d5186d7a6bfe09c1a6e
-size 515594979

bioclip-2_float16/train-00004-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4ab9e777b9585722bb31452101afca4a1720b0a7b30ab753fc0a895b3b6ce5b
+size 515690472

bioclip-2_float16/train-00005-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c88f59595067e0ab43422a514db11e1c00af1fd7ecaa0167bb6e6421b8e4c330
-size 515593322

bioclip-2_float16/train-00005-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44a69a5fd4088446d7659e9d9f93a9fff6215133caacbb82b3bd2896dd609fec
+size 515636371

bioclip-2_float16/train-00006-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:850191756024186f3746ad6430ae1113456efe2cc16103434f7d0dabedddb2ee
-size 515707578

bioclip-2_float16/train-00006-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9375750765351800377d352b016d38ce9acbb39a8327ad7a6433a68d9a63f552
+size 515986844

bioclip-2_float16/train-00007-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2ad75f6d0e5ee50d349303eb16e078586d983646029e09c76a9eb2883f577dd9
-size 515782116

bioclip-2_float16/train-00007-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68ba79dd1a9b2a2ef8aa47d452c2e82f3fcf2032a699b9f05e93aa36d3af94fa
+size 515850649

bioclip-2_float16/train-00008-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b152ea1d13089f963ca244abe1ebf20787f4f190503e1239631f8ddd6a0628fa
-size 516011227

bioclip-2_float16/train-00008-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a390d898f071d589dc3549c1bffab70621dd72027773634440bbc8c6600f164
+size 516054789

bioclip-2_float16/train-00009-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fb5378d7e30b53b9148ff3ae84f2d9957112754423bff39b9b1e1d9e70c9ce34
-size 515800588

bioclip-2_float16/train-00009-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c5ea22d85a6f348b4b7fc3c2125e2b14688c66f087e0f25c5de7b13b217b375
+size 515811580

bioclip-2_float16/train-00010-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:449da8379944b36fd8a360460ab359b270fb677378cac14e2d8e0eace025bc43
-size 515644298

bioclip-2_float16/{train-00498-of-00666.parquet → train-00010-of-00685.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3773862ce29ddb6cbee97e3d06f153e8491a5d45d1d5c66cbac40aff27fee3d0
-size 519737334
+oid sha256:5a545aacd2f1213ff7df395f3ad23b6c926396817521ca697b27e63bb5eb2a3f
+size 515722505

bioclip-2_float16/train-00011-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0e9dac52edecb81092160a2689afff6bba79415a92a7a39dd17016774571c52a
-size 515995941

bioclip-2_float16/train-00011-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1dfe4d36b146e175ea0640da3b818a1ecaaeab8b4fec951809fa7d73f7d4153
+size 515900155

bioclip-2_float16/train-00012-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:18a27b506631d27b589b15aa31834f1d0da63f2ee36e87a317c8b317889324d8
-size 516289135

bioclip-2_float16/train-00012-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7994b2969192861251db0499080b11aac18019dacf1c2b7b9cd7c1791aa4740
+size 516563402

bioclip-2_float16/train-00013-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:df9354eb2130fd6bda40b4afc02acb644ce95c673bc8e1d584600c28ce87585a
-size 516254905

bioclip-2_float16/train-00013-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62bf6fc5efb353a6c35d488ce1c44f3c47095c427a65964bb6d40ca1f21d2f03
+size 516372796

bioclip-2_float16/train-00014-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1b4e421ce6b3de1a1248214c7d5befbba52b85ce6e311dbb33a8d6997f61c1a3
-size 518313981

bioclip-2_float16/train-00014-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa70f7b55633294a57001f4af0bf2b52d10f66a65d73f1fbe0e8b7877aeca555
+size 518418043

bioclip-2_float16/train-00015-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:91448b6166139e5d7d5f09e25186bcc78ec63f02a1b752be78cea1f3ef53eeae
-size 523406380

bioclip-2_float16/train-00015-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e11829a8b6ab0c52919d909ea32eeb86eec736370cf59081c2becff5c48d695
+size 523931985

bioclip-2_float16/train-00016-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d1c93740d24ebaab83f6aeb675b53450e39538ba1f916e0f8415316a44cd2b16
-size 522944272

bioclip-2_float16/train-00016-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d06f0ad36f2ea2ba42a1a2d735b2992505fad770d09e3c48acbaf4666344d6ff
+size 523055588

bioclip-2_float16/train-00017-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d64691a2824750726416c8f1b5ae5f81be72bcfa691cfd42e76229b358b2f31f
-size 522917503

bioclip-2_float16/train-00017-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8ca55c493b955f3d3ef2d6f719068880a6b4412debfa93cbcf5e854077701dd
+size 523309019

bioclip-2_float16/train-00018-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e12bde9d1c0eeb6cf3aace190e2300f4826abeb17fc08744895a77e4d92c54c4
-size 523555997

bioclip-2_float16/train-00018-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a938a2257df2598c01a8c36ae55aa2d5b77c44d9050989cd102cf67bf131db6
+size 523784694

bioclip-2_float16/train-00019-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1355c4ebe7e1ecdf6fc9755346ff67e7075bb0af2d6b8daf0becb30aace81f66
-size 523530710

bioclip-2_float16/train-00019-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9becc65c11561b2359e1829efd41162a030a2a370a5f998893f7679e894d9254
+size 523743136

bioclip-2_float16/train-00020-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9ed5fc667c1ad1dfba6926b35032dbfd1078277ad2c22f57ba663a2b83738248
-size 523568628

bioclip-2_float16/train-00020-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70b21acc7e1d07077be97da3d07493df846c421d948bafea9659a6bbe3a96d55
+size 523689519

bioclip-2_float16/train-00021-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c5a2efcb25fd00369358d138cb44396a976762f20ed966890a77383df1d5e672
-size 523515961

bioclip-2_float16/train-00021-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bfbf0077b1dfd696c7dca27183821d0f0dbabc35240a63c81b505bb7cfe5a854
+size 523936175

bioclip-2_float16/train-00022-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7b4ef55cfe23682a1c4065cbccf8fe61f649326210c06d9833e37de8497950cf
-size 524052786

bioclip-2_float16/train-00022-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ab859419eeccfd10ad1b64e9af9d95edc33485fd5ae0a624df6a280c79356c8
+size 524470131

bioclip-2_float16/train-00023-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4fb721f4a2fd2454494d47177585b6b077286b5f5723c19ec76cd321bdb17b02
-size 522819044

bioclip-2_float16/train-00023-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:370c10cd6595bdec34b3010f729e0811d7313398a945bba785bcfe92e7651e6c
+size 521729135

bioclip-2_float16/train-00024-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:627aaf9938760d21b95ea03b8212ec5b15351f8b004ef3951736ce6abf497638
-size 523592090

bioclip-2_float16/train-00024-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3ab223570f578e9cd9ead4156d4bb73cf7fd65fa3bebc24e59cdd8036d0531b
+size 523817147

bioclip-2_float16/train-00025-of-00666.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:82e462ceeaa4b7a94101655a0379c91f432a37c1bb80e3c91e31f657b9bc9c98
-size 523567614

bioclip-2_float16/train-00025-of-00685.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ce7dfde2bf83b01a06bd24186833fbe5195bfa4ef0ead2db589815c58bb3570
+size 523713086