Hugging Face's logo

hf-tiny-v2
/
tiny-random-FlavaModel

Safetensors
flava
Model card Files
xet
 Community
hf-transformers-bot commited on
Commit
3d06db5
·
verified ·
1 Parent(s): 195521e

Update tiny models for FlavaModel

Browse files
Files changed (5) hide show
  1. config.json +86 -0
  2. model.safetensors +3 -0
  3. preprocessor_config.json +58 -0
  4. tokenizer.json +0 -0
  5. tokenizer_config.json +16 -0
config.json ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "FlavaModel"
4
+ ],
5
+ "ce_ignore_index": -100,
6
+ "dtype": "float32",
7
+ "global_backprop_contrastive": true,
8
+ "global_contrastive_weight": 1.0,
9
+ "hidden_size": 32,
10
+ "image_codebook_config": {
11
+ "freeze": true,
12
+ "hidden_size": 32,
13
+ "initializer_range": 0.02,
14
+ "input_channels": 3,
15
+ "model_type": "",
16
+ "num_blocks_per_group": 2,
17
+ "num_groups": 2,
18
+ "vocab_size": 99
19
+ },
20
+ "image_config": {
21
+ "attention_probs_dropout_prob": 0.0,
22
+ "hidden_act": "gelu",
23
+ "hidden_dropout_prob": 0.0,
24
+ "hidden_size": 32,
25
+ "image_size": 30,
26
+ "initializer_range": 0.02,
27
+ "intermediate_size": 37,
28
+ "layer_norm_eps": 1e-12,
29
+ "mask_token": true,
30
+ "model_type": "flava_image_model",
31
+ "num_attention_heads": 4,
32
+ "num_channels": 3,
33
+ "num_hidden_layers": 2,
34
+ "patch_size": 2,
35
+ "qkv_bias": true,
36
+ "vocab_size": 99
37
+ },
38
+ "init_codebook": true,
39
+ "initializer_factor": 1.0,
40
+ "initializer_range": 0.02,
41
+ "itm_weight": 1.0,
42
+ "layer_norm_eps": 1e-12,
43
+ "logit_scale_init_value": 2.6592,
44
+ "mim_weight": 1.0,
45
+ "mlm_weight": 1.0,
46
+ "mmm_image_weight": 1.0,
47
+ "mmm_text_weight": 1.0,
48
+ "model_type": "flava",
49
+ "multimodal_config": {
50
+ "attention_probs_dropout_prob": 0.0,
51
+ "ce_ignore_index": -100,
52
+ "hidden_act": "gelu",
53
+ "hidden_dropout_prob": 0.0,
54
+ "hidden_size": 32,
55
+ "initializer_range": 0.02,
56
+ "intermediate_size": 37,
57
+ "layer_norm_eps": 1e-12,
58
+ "model_type": "flava_multimodal_model",
59
+ "num_attention_heads": 4,
60
+ "num_hidden_layers": 2,
61
+ "qkv_bias": true,
62
+ "use_cls_token": true
63
+ },
64
+ "projection_dim": 32,
65
+ "return_loss": true,
66
+ "skip_unmasked_multimodal_encoder": true,
67
+ "text_config": {
68
+ "attention_probs_dropout_prob": 0.0,
69
+ "hidden_act": "gelu",
70
+ "hidden_dropout_prob": 0.0,
71
+ "hidden_size": 32,
72
+ "initializer_range": 0.02,
73
+ "intermediate_size": 37,
74
+ "layer_norm_eps": 1e-12,
75
+ "max_position_embeddings": 512,
76
+ "model_type": "flava_text_model",
77
+ "num_attention_heads": 4,
78
+ "num_hidden_layers": 2,
79
+ "pad_token_id": 0,
80
+ "qkv_bias": true,
81
+ "type_vocab_size": 2,
82
+ "vocab_size": 30522
83
+ },
84
+ "tie_word_embeddings": true,
85
+ "transformers_version": "5.16.0.dev0"
86
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7f0a55aa8ed252685ceb364b036b27692303cd2b5862f797f04cf8100fcac2a
3
+ size 4211644
preprocessor_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "codebook_crop_size": {
3
+ "height": 112,
4
+ "width": 112
5
+ },
6
+ "codebook_do_center_crop": true,
7
+ "codebook_do_map_pixels": true,
8
+ "codebook_do_normalize": true,
9
+ "codebook_do_rescale": true,
10
+ "codebook_do_resize": true,
11
+ "codebook_image_mean": [
12
+ 0.0,
13
+ 0.0,
14
+ 0.0
15
+ ],
16
+ "codebook_image_std": [
17
+ 1.0,
18
+ 1.0,
19
+ 1.0
20
+ ],
21
+ "codebook_resample": 1,
22
+ "codebook_rescale_factor": 0.00392156862745098,
23
+ "codebook_size": {
24
+ "height": 112,
25
+ "width": 112
26
+ },
27
+ "crop_size": {
28
+ "height": 224,
29
+ "width": 224
30
+ },
31
+ "do_center_crop": true,
32
+ "do_normalize": true,
33
+ "do_rescale": true,
34
+ "do_resize": true,
35
+ "image_mean": [
36
+ 0.48145466,
37
+ 0.4578275,
38
+ 0.40821073
39
+ ],
40
+ "image_processor_type": "FlavaImageProcessor",
41
+ "image_std": [
42
+ 0.26862954,
43
+ 0.26130258,
44
+ 0.27577711
45
+ ],
46
+ "input_size_patches": 14,
47
+ "mask_group_min_aspect_ratio": 0.3,
48
+ "mask_group_min_patches": 16,
49
+ "resample": 3,
50
+ "rescale_factor": 0.00392156862745098,
51
+ "return_codebook_pixels": false,
52
+ "return_image_mask": false,
53
+ "size": {
54
+ "height": 224,
55
+ "width": 224
56
+ },
57
+ "total_mask_patches": 75
58
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_lower_case": true,
5
+ "is_local": true,
6
+ "local_files_only": false,
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 512,
9
+ "pad_token": "[PAD]",
10
+ "processor_class": "FlavaProcessor",
11
+ "sep_token": "[SEP]",
12
+ "strip_accents": null,
13
+ "tokenize_chinese_chars": true,
14
+ "tokenizer_class": "BertTokenizer",
15
+ "unk_token": "[UNK]"
16
+ }