Instructions to use leilaaaaa/florence2MED with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Adapters
How to use leilaaaaa/florence2MED with Adapters:
from adapters import AutoAdapterModel model = AutoAdapterModel.from_pretrained("undefined") model.load_adapter("leilaaaaa/florence2MED", set_active=True) - Notebooks
- Google Colab
- Kaggle
Update config.json
Browse files- config.json +2 -3
config.json
CHANGED
|
@@ -11,7 +11,7 @@
|
|
| 11 |
"eos_token_id": 2,
|
| 12 |
"ignore_index": -100,
|
| 13 |
"is_encoder_decoder": true,
|
| 14 |
-
"model_type": "
|
| 15 |
"pad_token_id": 1,
|
| 16 |
"projection_dim": 768,
|
| 17 |
"text_config": {
|
|
@@ -132,7 +132,6 @@
|
|
| 132 |
"do_sample": false,
|
| 133 |
"drop_path_rate": 0.1,
|
| 134 |
"early_stopping": false,
|
| 135 |
-
"enable_checkpoint": false,
|
| 136 |
"encoder_no_repeat_ngram_size": 0,
|
| 137 |
"eos_token_id": null,
|
| 138 |
"exponential_decay_length_penalty": null,
|
|
@@ -160,7 +159,7 @@
|
|
| 160 |
"length_penalty": 1.0,
|
| 161 |
"max_length": 20,
|
| 162 |
"min_length": 0,
|
| 163 |
-
"model_type": "",
|
| 164 |
"no_repeat_ngram_size": 0,
|
| 165 |
"num_beam_groups": 1,
|
| 166 |
"num_beams": 1,
|
|
|
|
| 11 |
"eos_token_id": 2,
|
| 12 |
"ignore_index": -100,
|
| 13 |
"is_encoder_decoder": true,
|
| 14 |
+
"model_type": "florence2",
|
| 15 |
"pad_token_id": 1,
|
| 16 |
"projection_dim": 768,
|
| 17 |
"text_config": {
|
|
|
|
| 132 |
"do_sample": false,
|
| 133 |
"drop_path_rate": 0.1,
|
| 134 |
"early_stopping": false,
|
|
|
|
| 135 |
"encoder_no_repeat_ngram_size": 0,
|
| 136 |
"eos_token_id": null,
|
| 137 |
"exponential_decay_length_penalty": null,
|
|
|
|
| 159 |
"length_penalty": 1.0,
|
| 160 |
"max_length": 20,
|
| 161 |
"min_length": 0,
|
| 162 |
+
"model_type": "davit", // Changed to "davit"
|
| 163 |
"no_repeat_ngram_size": 0,
|
| 164 |
"num_beam_groups": 1,
|
| 165 |
"num_beams": 1,
|