Instructions to use danelcsb/rtdetr_v2_r18vd with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use danelcsb/rtdetr_v2_r18vd with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("object-detection", model="danelcsb/rtdetr_v2_r18vd")# Load model directly from transformers import AutoTokenizer, AutoModelForObjectDetection tokenizer = AutoTokenizer.from_pretrained("danelcsb/rtdetr_v2_r18vd") model = AutoModelForObjectDetection.from_pretrained("danelcsb/rtdetr_v2_r18vd") - Notebooks
- Google Colab
- Kaggle
Add config from convert_rt_detr_v2_original_pytorch_checkpoint_to_pytorch.py
Browse files- config.json +3 -4
config.json
CHANGED
|
@@ -3,7 +3,7 @@
|
|
| 3 |
"activation_function": "silu",
|
| 4 |
"anchor_image_size": null,
|
| 5 |
"architectures": [
|
| 6 |
-
"
|
| 7 |
],
|
| 8 |
"attention_dropout": 0.0,
|
| 9 |
"auxiliary_loss": true,
|
|
@@ -22,7 +22,7 @@
|
|
| 22 |
512
|
| 23 |
],
|
| 24 |
"layer_type": "basic",
|
| 25 |
-
"model_type": "
|
| 26 |
"out_features": [
|
| 27 |
"stage2",
|
| 28 |
"stage3",
|
|
@@ -49,7 +49,6 @@
|
|
| 49 |
"decoder_layers": 3,
|
| 50 |
"decoder_n_points": 4,
|
| 51 |
"decoder_offset_scale": 0.5,
|
| 52 |
-
"decoder_version": "v2",
|
| 53 |
"disable_custom_kernels": true,
|
| 54 |
"dropout": 0.0,
|
| 55 |
"encode_proj_layers": [
|
|
@@ -251,7 +250,7 @@
|
|
| 251 |
"matcher_class_cost": 2.0,
|
| 252 |
"matcher_gamma": 2.0,
|
| 253 |
"matcher_giou_cost": 2.0,
|
| 254 |
-
"model_type": "
|
| 255 |
"normalize_before": false,
|
| 256 |
"num_denoising": 100,
|
| 257 |
"num_feature_levels": 3,
|
|
|
|
| 3 |
"activation_function": "silu",
|
| 4 |
"anchor_image_size": null,
|
| 5 |
"architectures": [
|
| 6 |
+
"RTDetrV2ForObjectDetection"
|
| 7 |
],
|
| 8 |
"attention_dropout": 0.0,
|
| 9 |
"auxiliary_loss": true,
|
|
|
|
| 22 |
512
|
| 23 |
],
|
| 24 |
"layer_type": "basic",
|
| 25 |
+
"model_type": "rt_detr_v2_resnet",
|
| 26 |
"out_features": [
|
| 27 |
"stage2",
|
| 28 |
"stage3",
|
|
|
|
| 49 |
"decoder_layers": 3,
|
| 50 |
"decoder_n_points": 4,
|
| 51 |
"decoder_offset_scale": 0.5,
|
|
|
|
| 52 |
"disable_custom_kernels": true,
|
| 53 |
"dropout": 0.0,
|
| 54 |
"encode_proj_layers": [
|
|
|
|
| 250 |
"matcher_class_cost": 2.0,
|
| 251 |
"matcher_gamma": 2.0,
|
| 252 |
"matcher_giou_cost": 2.0,
|
| 253 |
+
"model_type": "rt_detr_v2",
|
| 254 |
"normalize_before": false,
|
| 255 |
"num_denoising": 100,
|
| 256 |
"num_feature_levels": 3,
|