Feature Extraction
Transformers
clip
vision
Inference Endpoints
uform-vl-english / config.json
kimihailv's picture
Upload 2 files
1ec7966
raw
history blame contribute delete
No virus
349 Bytes
{"img_encoder": {"backbone": "deit3_base_patch16_224_in21ft1k", "dim": 768, "backbone_type": "vit", "pooling": "cls", "output_dim": 256}, "text_encoder": {"backbone": "google/bert_uncased_L-4_H-768_A-12", "backbone_type": "bert", "unimodal_n_layers": 2, "dim": 768, "pooling": "cls", "context_dim": 768, "output_dim": 256, "head_one_neuron": false}}