Image Classification
Transformers
Safetensors
mobilevit
knowledge_distillation
vision
Generated from Trainer
Instructions to use c14kevincardenas/mobilevit-small_alpha0.7_temp5.0_t3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use c14kevincardenas/mobilevit-small_alpha0.7_temp5.0_t3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-classification", model="c14kevincardenas/mobilevit-small_alpha0.7_temp5.0_t3") pipe("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png")# Load model directly from transformers import AutoImageProcessor, AutoModelForImageClassification processor = AutoImageProcessor.from_pretrained("c14kevincardenas/mobilevit-small_alpha0.7_temp5.0_t3") model = AutoModelForImageClassification.from_pretrained("c14kevincardenas/mobilevit-small_alpha0.7_temp5.0_t3", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Upload 2 files
Browse files- config.json +43 -0
- preprocessor_config.json +9 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MobileViTForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"aspp_dropout_prob": 0.1,
|
| 6 |
+
"aspp_out_channels": 256,
|
| 7 |
+
"atrous_rates": [
|
| 8 |
+
6,
|
| 9 |
+
12,
|
| 10 |
+
18
|
| 11 |
+
],
|
| 12 |
+
"attention_probs_dropout_prob": 0.0,
|
| 13 |
+
"classifier_dropout_prob": 0.1,
|
| 14 |
+
"conv_kernel_size": 3,
|
| 15 |
+
"expand_ratio": 4.0,
|
| 16 |
+
"hidden_act": "silu",
|
| 17 |
+
"hidden_dropout_prob": 0.1,
|
| 18 |
+
"hidden_sizes": [
|
| 19 |
+
144,
|
| 20 |
+
192,
|
| 21 |
+
240
|
| 22 |
+
],
|
| 23 |
+
"layer_norm_eps": 1e-05,
|
| 24 |
+
"mlp_ratio": 2.0,
|
| 25 |
+
"model_type": "mobilevit",
|
| 26 |
+
"neck_hidden_sizes": [
|
| 27 |
+
16,
|
| 28 |
+
32,
|
| 29 |
+
64,
|
| 30 |
+
96,
|
| 31 |
+
128,
|
| 32 |
+
160,
|
| 33 |
+
640
|
| 34 |
+
],
|
| 35 |
+
"num_attention_heads": 4,
|
| 36 |
+
"num_channels": 3,
|
| 37 |
+
"output_stride": 32,
|
| 38 |
+
"patch_size": 2,
|
| 39 |
+
"qkv_bias": true,
|
| 40 |
+
"semantic_loss_ignore_index": 255,
|
| 41 |
+
"torch_dtype": "float32",
|
| 42 |
+
"transformers_version": "4.20.0.dev0"
|
| 43 |
+
}
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_size": 256,
|
| 3 |
+
"do_center_crop": true,
|
| 4 |
+
"do_flip_channels": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"feature_extractor_type": "MobileViTFeatureExtractor",
|
| 7 |
+
"resample": 2,
|
| 8 |
+
"size": 288
|
| 9 |
+
}
|