Instructions to use unum-cloud/uform-vl-english with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use unum-cloud/uform-vl-english with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="unum-cloud/uform-vl-english")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("unum-cloud/uform-vl-english", dtype="auto") - Notebooks
- Google Colab
- Kaggle
Upload torch_config.json
Browse files- torch_config.json +26 -0
torch_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"text_encoder": {
|
| 3 |
+
"model_type": "bert",
|
| 4 |
+
"dim": 768,
|
| 5 |
+
"context_dim": 768,
|
| 6 |
+
"vocab_size": 30522,
|
| 7 |
+
"padding_idx": 0,
|
| 8 |
+
"num_layers": 4,
|
| 9 |
+
"num_heads": 12,
|
| 10 |
+
"embedding_dim": 256,
|
| 11 |
+
"multimodal_layers_ids": [2, 3],
|
| 12 |
+
"head_one_neuron": false,
|
| 13 |
+
"pooling": "cls",
|
| 14 |
+
"max_position_embeddings": 77,
|
| 15 |
+
"dropout_prob": 0.1
|
| 16 |
+
},
|
| 17 |
+
"image_encoder": {
|
| 18 |
+
"dim": 768,
|
| 19 |
+
"patch_size": 16,
|
| 20 |
+
"image_size": 224,
|
| 21 |
+
"num_layers": 12,
|
| 22 |
+
"num_heads": 12,
|
| 23 |
+
"embedding_dim": 256,
|
| 24 |
+
"pooling": "cls"
|
| 25 |
+
}
|
| 26 |
+
}
|