update
Browse files- config.json +41 -0
- pytorch_model.bin +3 -0
config.json
ADDED
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "configs/ViTLP-1920-1600",
|
3 |
+
"architectures": ["ViTLPForPreTraining"],
|
4 |
+
"model_type": "ViTLP",
|
5 |
+
"torch_dtype": "float32",
|
6 |
+
"gradient_checkpointing": true,
|
7 |
+
"init_std": 0.02,
|
8 |
+
"initializer_range": 0.02,
|
9 |
+
"layer_norm_eps": 1e-12,
|
10 |
+
"dropout": 0.1,
|
11 |
+
"hidden_dropout_prob": 0.1,
|
12 |
+
"attention_dropout": 0.1,
|
13 |
+
"activation_dropout": 0.1,
|
14 |
+
"activation_function": "gelu",
|
15 |
+
"encoder_attention_heads": 16,
|
16 |
+
"encoder_ffn_dim": 4096,
|
17 |
+
"encoder_hidden_size": 1024,
|
18 |
+
"decoder_attention_heads": 16,
|
19 |
+
"decoder_ffn_dim": 4096,
|
20 |
+
"hidden_size": 1024,
|
21 |
+
"encoder_layers": 18,
|
22 |
+
"decoder_layers": 6,
|
23 |
+
|
24 |
+
"image_height": 1920,
|
25 |
+
"image_width": 1600,
|
26 |
+
"patch_size": 32,
|
27 |
+
"patch_num": 3000,
|
28 |
+
"num_channels": 3,
|
29 |
+
"resample": 2,
|
30 |
+
"bin_size": 1001,
|
31 |
+
"load_vit": true,
|
32 |
+
|
33 |
+
"seq_length": 1280,
|
34 |
+
"decoder_start_token_id": 2,
|
35 |
+
"cls_token_id": 0,
|
36 |
+
"pad_token_id": 1,
|
37 |
+
"eos_token_id": 2,
|
38 |
+
"unk_token_id": 3,
|
39 |
+
"vocab_size": 50267,
|
40 |
+
"load_bart": true
|
41 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:153cbc7232869a17b72e56591ca751e4542ed0d44bff0e16d2e5bd1f61d7e1a5
|
3 |
+
size 1568795070
|