Hugging Face's logo

not-lain
/
matanyone

MatAnyone
Safetensors
model_hub_mixin
pytorch_model_hub_mixin
Model card Files Community
not-lain commited on
Commit
aa15ceb
·
verified ·
1 Parent(s): 653e39a

Push model using huggingface_hub.

Browse files
Files changed (3) hide show
  1. README.md +10 -0
  2. config.json +139 -0
  3. model.safetensors +3 -0
README.md ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: matanyone
3
+ tags:
4
+ - model_hub_mixin
5
+ - pytorch_model_hub_mixin
6
+ ---
7
+
8
+ This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
9
+ - Library: https://github.com/pq-yang/MatAnyone
10
+ - Docs: [More Information Needed]
config.json ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cfg": {
3
+ "amp": false,
4
+ "chunk_size": -1,
5
+ "flip_aug": false,
6
+ "long_term": {
7
+ "buffer_tokens": 2000,
8
+ "count_usage": true,
9
+ "max_mem_frames": 10,
10
+ "max_num_tokens": 10000,
11
+ "min_mem_frames": 5,
12
+ "num_prototypes": 128
13
+ },
14
+ "max_internal_size": -1,
15
+ "max_mem_frames": 5,
16
+ "mem_every": 5,
17
+ "model": {
18
+ "aux_loss": {
19
+ "query": {
20
+ "enabled": true,
21
+ "weight": 0.01
22
+ },
23
+ "sensory": {
24
+ "enabled": true,
25
+ "weight": 0.01
26
+ }
27
+ },
28
+ "embed_dim": 256,
29
+ "key_dim": 64,
30
+ "mask_decoder": {
31
+ "up_dims": [
32
+ 256,
33
+ 128,
34
+ 128,
35
+ 64,
36
+ 16
37
+ ]
38
+ },
39
+ "mask_encoder": {
40
+ "final_dim": 256,
41
+ "type": "resnet18"
42
+ },
43
+ "object_summarizer": {
44
+ "add_pe": true,
45
+ "embed_dim": "${model.object_transformer.embed_dim}",
46
+ "num_summaries": "${model.object_transformer.num_queries}"
47
+ },
48
+ "object_transformer": {
49
+ "embed_dim": "${model.embed_dim}",
50
+ "ff_dim": 2048,
51
+ "num_blocks": 3,
52
+ "num_heads": 8,
53
+ "num_queries": 16,
54
+ "pixel_self_attention": {
55
+ "add_pe_to_qkv": [
56
+ true,
57
+ true,
58
+ false
59
+ ]
60
+ },
61
+ "query_self_attention": {
62
+ "add_pe_to_qkv": [
63
+ true,
64
+ true,
65
+ false
66
+ ]
67
+ },
68
+ "read_from_memory": {
69
+ "add_pe_to_qkv": [
70
+ true,
71
+ true,
72
+ false
73
+ ]
74
+ },
75
+ "read_from_past": {
76
+ "add_pe_to_qkv": [
77
+ true,
78
+ true,
79
+ false
80
+ ]
81
+ },
82
+ "read_from_pixel": {
83
+ "add_pe_to_qkv": [
84
+ true,
85
+ true,
86
+ false
87
+ ],
88
+ "input_add_pe": false,
89
+ "input_norm": false
90
+ },
91
+ "read_from_query": {
92
+ "add_pe_to_qkv": [
93
+ true,
94
+ true,
95
+ false
96
+ ],
97
+ "output_norm": false
98
+ }
99
+ },
100
+ "pixel_dim": 256,
101
+ "pixel_encoder": {
102
+ "ms_dims": [
103
+ 1024,
104
+ 512,
105
+ 256,
106
+ 64,
107
+ 3
108
+ ],
109
+ "type": "resnet50"
110
+ },
111
+ "pixel_mean": [
112
+ 0.485,
113
+ 0.456,
114
+ 0.406
115
+ ],
116
+ "pixel_pe_scale": 32,
117
+ "pixel_pe_temperature": 128,
118
+ "pixel_std": [
119
+ 0.229,
120
+ 0.224,
121
+ 0.225
122
+ ],
123
+ "sensory_dim": 256,
124
+ "value_dim": 256
125
+ },
126
+ "output_dir": null,
127
+ "save_all": true,
128
+ "save_aux": false,
129
+ "save_scores": false,
130
+ "stagger_updates": 5,
131
+ "top_k": 30,
132
+ "use_all_masks": false,
133
+ "use_long_term": false,
134
+ "visualize": false,
135
+ "weights": "pretrained_models/matanyone.pth"
136
+ },
137
+ "pretrained_resnet": false,
138
+ "single_object": true
139
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07f2a00683e6aec3d62f1cb0b13b7bbe45622d06355443f2861767074a6ec7b1
3
+ size 141199924