josephloh commited on
Commit
ea8cb78
1 Parent(s): 32aa90c

josephloh/layoutlmv3-funsd

Browse files
README.md ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-sa-4.0
3
+ base_model: microsoft/layoutlmv3-base
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - funsd-layoutlmv3
8
+ metrics:
9
+ - precision
10
+ - recall
11
+ - f1
12
+ - accuracy
13
+ model-index:
14
+ - name: test
15
+ results:
16
+ - task:
17
+ name: Token Classification
18
+ type: token-classification
19
+ dataset:
20
+ name: funsd-layoutlmv3
21
+ type: funsd-layoutlmv3
22
+ config: funsd
23
+ split: test
24
+ args: funsd
25
+ metrics:
26
+ - name: Precision
27
+ type: precision
28
+ value: 0.8876459143968871
29
+ - name: Recall
30
+ type: recall
31
+ value: 0.9066070541480378
32
+ - name: F1
33
+ type: f1
34
+ value: 0.8970262963873188
35
+ - name: Accuracy
36
+ type: accuracy
37
+ value: 0.86009746820397
38
+ ---
39
+
40
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
41
+ should probably proofread and complete it, then remove this comment. -->
42
+
43
+ # test
44
+
45
+ This model is a fine-tuned version of [microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) on the funsd-layoutlmv3 dataset.
46
+ It achieves the following results on the evaluation set:
47
+ - Loss: 0.6568
48
+ - Precision: 0.8876
49
+ - Recall: 0.9066
50
+ - F1: 0.8970
51
+ - Accuracy: 0.8601
52
+
53
+ ## Model description
54
+
55
+ More information needed
56
+
57
+ ## Intended uses & limitations
58
+
59
+ More information needed
60
+
61
+ ## Training and evaluation data
62
+
63
+ More information needed
64
+
65
+ ## Training procedure
66
+
67
+ ### Training hyperparameters
68
+
69
+ The following hyperparameters were used during training:
70
+ - learning_rate: 1e-05
71
+ - train_batch_size: 2
72
+ - eval_batch_size: 2
73
+ - seed: 42
74
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
75
+ - lr_scheduler_type: linear
76
+ - training_steps: 1000
77
+
78
+ ### Training results
79
+
80
+ | Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
81
+ |:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
82
+ | No log | 1.33 | 100 | 0.6157 | 0.7621 | 0.8400 | 0.7991 | 0.8051 |
83
+ | No log | 2.67 | 200 | 0.4834 | 0.7915 | 0.8902 | 0.8380 | 0.8334 |
84
+ | No log | 4.0 | 300 | 0.4929 | 0.8484 | 0.8922 | 0.8697 | 0.8493 |
85
+ | No log | 5.33 | 400 | 0.5191 | 0.8746 | 0.9006 | 0.8874 | 0.8556 |
86
+ | 0.5561 | 6.67 | 500 | 0.5553 | 0.8671 | 0.9041 | 0.8852 | 0.8487 |
87
+ | 0.5561 | 8.0 | 600 | 0.5766 | 0.8723 | 0.9091 | 0.8903 | 0.8388 |
88
+ | 0.5561 | 9.33 | 700 | 0.6486 | 0.8816 | 0.8917 | 0.8866 | 0.8511 |
89
+ | 0.5561 | 10.67 | 800 | 0.6188 | 0.8861 | 0.9086 | 0.8972 | 0.8608 |
90
+ | 0.5561 | 12.0 | 900 | 0.6317 | 0.8890 | 0.9071 | 0.8980 | 0.8630 |
91
+ | 0.1298 | 13.33 | 1000 | 0.6568 | 0.8876 | 0.9066 | 0.8970 | 0.8601 |
92
+
93
+
94
+ ### Framework versions
95
+
96
+ - Transformers 4.34.0.dev0
97
+ - Pytorch 2.0.1+cpu
98
+ - Datasets 2.12.0
99
+ - Tokenizers 0.13.3
config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/layoutlmv3-base",
3
+ "architectures": [
4
+ "LayoutLMv3ForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "coordinate_size": 128,
10
+ "eos_token_id": 2,
11
+ "has_relative_attention_bias": true,
12
+ "has_spatial_attention_bias": true,
13
+ "hidden_act": "gelu",
14
+ "hidden_dropout_prob": 0.1,
15
+ "hidden_size": 768,
16
+ "id2label": {
17
+ "0": "O",
18
+ "1": "B-HEADER",
19
+ "2": "I-HEADER",
20
+ "3": "B-QUESTION",
21
+ "4": "I-QUESTION",
22
+ "5": "B-ANSWER",
23
+ "6": "I-ANSWER"
24
+ },
25
+ "initializer_range": 0.02,
26
+ "input_size": 224,
27
+ "intermediate_size": 3072,
28
+ "label2id": {
29
+ "B-ANSWER": 5,
30
+ "B-HEADER": 1,
31
+ "B-QUESTION": 3,
32
+ "I-ANSWER": 6,
33
+ "I-HEADER": 2,
34
+ "I-QUESTION": 4,
35
+ "O": 0
36
+ },
37
+ "layer_norm_eps": 1e-05,
38
+ "max_2d_position_embeddings": 1024,
39
+ "max_position_embeddings": 514,
40
+ "max_rel_2d_pos": 256,
41
+ "max_rel_pos": 128,
42
+ "model_type": "layoutlmv3",
43
+ "num_attention_heads": 12,
44
+ "num_channels": 3,
45
+ "num_hidden_layers": 12,
46
+ "pad_token_id": 1,
47
+ "patch_size": 16,
48
+ "rel_2d_pos_bins": 64,
49
+ "rel_pos_bins": 32,
50
+ "second_input_size": 112,
51
+ "shape_size": 128,
52
+ "text_embed": true,
53
+ "torch_dtype": "float32",
54
+ "transformers_version": "4.34.0.dev0",
55
+ "type_vocab_size": 1,
56
+ "visual_embed": true,
57
+ "vocab_size": 50265
58
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "apply_ocr": false,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "feature_extractor_type": "LayoutLMv3FeatureExtractor",
7
+ "image_mean": [
8
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "LayoutLMv3ImageProcessor",
13
+ "image_std": [
14
+ 0.5,
15
+ 0.5,
16
+ 0.5
17
+ ],
18
+ "ocr_lang": null,
19
+ "processor_class": "LayoutLMv3Processor",
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 224,
24
+ "width": 224
25
+ },
26
+ "tesseract_config": ""
27
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d225a1f2cf938294cc561ba6ba538f46b1737e37ea6ce587127cf1c741f714f6
3
+ size 501400357
special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": true,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": true,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": true,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "apply_ocr": false,
4
+ "bos_token": {
5
+ "__type": "AddedToken",
6
+ "content": "<s>",
7
+ "lstrip": false,
8
+ "normalized": true,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "clean_up_tokenization_spaces": true,
13
+ "cls_token": {
14
+ "__type": "AddedToken",
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": true,
18
+ "rstrip": false,
19
+ "single_word": false
20
+ },
21
+ "cls_token_box": [
22
+ 0,
23
+ 0,
24
+ 0,
25
+ 0
26
+ ],
27
+ "eos_token": {
28
+ "__type": "AddedToken",
29
+ "content": "</s>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false
34
+ },
35
+ "errors": "replace",
36
+ "mask_token": {
37
+ "__type": "AddedToken",
38
+ "content": "<mask>",
39
+ "lstrip": true,
40
+ "normalized": true,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "model_max_length": 512,
45
+ "only_label_first_subword": true,
46
+ "pad_token": {
47
+ "__type": "AddedToken",
48
+ "content": "<pad>",
49
+ "lstrip": false,
50
+ "normalized": true,
51
+ "rstrip": false,
52
+ "single_word": false
53
+ },
54
+ "pad_token_box": [
55
+ 0,
56
+ 0,
57
+ 0,
58
+ 0
59
+ ],
60
+ "pad_token_label": -100,
61
+ "processor_class": "LayoutLMv3Processor",
62
+ "sep_token": {
63
+ "__type": "AddedToken",
64
+ "content": "</s>",
65
+ "lstrip": false,
66
+ "normalized": true,
67
+ "rstrip": false,
68
+ "single_word": false
69
+ },
70
+ "sep_token_box": [
71
+ 0,
72
+ 0,
73
+ 0,
74
+ 0
75
+ ],
76
+ "tokenizer_class": "LayoutLMv3Tokenizer",
77
+ "trim_offsets": true,
78
+ "unk_token": {
79
+ "__type": "AddedToken",
80
+ "content": "<unk>",
81
+ "lstrip": false,
82
+ "normalized": true,
83
+ "rstrip": false,
84
+ "single_word": false
85
+ }
86
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f39e74d629c0f6affc1a9b976c6ff3b121c76b615ef7dbe7cc9a70026c26b216
3
+ size 4027
vocab.json ADDED
The diff for this file is too large to render. See raw diff