Zero-Shot Image Classification
OpenCLIP
Safetensors
English
biology
CV
images
imageomics
clip
species-classification
biological visual task
multimodal
animals
plants
fungi
species
taxonomy
rare species
endangered species
evolutionary biology
knowledge-guided
Instructions to use imageomics/bioclip-2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- OpenCLIP
How to use imageomics/bioclip-2 with OpenCLIP:
import open_clip model, preprocess_train, preprocess_val = open_clip.create_model_and_transforms('hf-hub:imageomics/bioclip-2') tokenizer = open_clip.get_tokenizer('hf-hub:imageomics/bioclip-2') - Notebooks
- Google Colab
- Kaggle
update replay description
Browse files
README.md
CHANGED
|
@@ -113,7 +113,7 @@ The dataset consists of nearly 214M images covering 952K taxa.
|
|
| 113 |
The scale of TreeOfLife-200M fosters the emergent properties of BioCLIP 2.
|
| 114 |
|
| 115 |
In addition, we also used a subset of LAION-2B that consists of 26M samples for experience replay.
|
| 116 |
-
This part of data was downloaded from the first three parquet metadata files of LAION-2B, and the first
|
| 117 |
|
| 118 |
### Training Procedure
|
| 119 |
|
|
|
|
| 113 |
The scale of TreeOfLife-200M fosters the emergent properties of BioCLIP 2.
|
| 114 |
|
| 115 |
In addition, we also used a subset of LAION-2B that consists of 26M samples for experience replay.
|
| 116 |
+
This part of data was downloaded from the first three parquet metadata files of LAION-2B, and the first 4,000 tar files were used.
|
| 117 |
|
| 118 |
### Training Procedure
|
| 119 |
|