|
{ |
|
"_name_or_path": "tiny_models/data2vec-vision/Data2VecVisionForSemanticSegmentation", |
|
"architectures": [ |
|
"Data2VecVisionForSemanticSegmentation" |
|
], |
|
"attention_probs_dropout_prob": 0.1, |
|
"auxiliary_channels": 256, |
|
"auxiliary_concat_input": false, |
|
"auxiliary_loss_weight": 0.4, |
|
"auxiliary_num_convs": 1, |
|
"drop_path_rate": 0.1, |
|
"hidden_act": "gelu", |
|
"hidden_dropout_prob": 0.1, |
|
"hidden_size": 32, |
|
"image_size": 30, |
|
"initializer_range": 0.02, |
|
"intermediate_size": 37, |
|
"layer_norm_eps": 1e-12, |
|
"layer_scale_init_value": 0.1, |
|
"model_type": "data2vec-vision", |
|
"num_attention_heads": 4, |
|
"num_channels": 3, |
|
"num_hidden_layers": 4, |
|
"out_indices": [ |
|
0, |
|
1, |
|
2, |
|
3 |
|
], |
|
"patch_size": 2, |
|
"pool_scales": [ |
|
1, |
|
2, |
|
3, |
|
6 |
|
], |
|
"semantic_loss_ignore_index": 255, |
|
"torch_dtype": "float32", |
|
"transformers_version": "4.28.0.dev0", |
|
"use_absolute_position_embeddings": false, |
|
"use_auxiliary_head": true, |
|
"use_mask_token": false, |
|
"use_mean_pooling": true, |
|
"use_relative_position_bias": false, |
|
"use_shared_relative_position_bias": false, |
|
"vocab_size": 100 |
|
} |
|
|