ihebMissaoui
commited on
Commit
•
6706286
1
Parent(s):
37f08dc
ihebMissaoui/layoutlmv3-fine-tuned-funsd-kie
Browse files- README.md +94 -0
- config.json +58 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- preprocessor_config.json +43 -0
- special_tokens_map.json +51 -0
- tokenizer.json +0 -0
- tokenizer_config.json +79 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
@@ -0,0 +1,94 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: cc-by-nc-sa-4.0
|
3 |
+
base_model: microsoft/layoutlmv3-base
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- precision
|
8 |
+
- recall
|
9 |
+
- f1
|
10 |
+
- accuracy
|
11 |
+
model-index:
|
12 |
+
- name: test
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/elyadata/Ft%20layoutlmv3%20funsd%20max%20epochs%20100%20%2Cearlystop%3D4%2Cbatch%3D2%2Clr%3D1e-5%20adamw%2CFULL%20models%20params/runs/6zlgssbd)
|
20 |
+
# test
|
21 |
+
|
22 |
+
This model is a fine-tuned version of [microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) on an unknown dataset.
|
23 |
+
It achieves the following results on the evaluation set:
|
24 |
+
- Loss: 0.8476
|
25 |
+
- Precision: 0.8955
|
26 |
+
- Recall: 0.9071
|
27 |
+
- F1: 0.9013
|
28 |
+
- Accuracy: 0.8691
|
29 |
+
|
30 |
+
## Model description
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Intended uses & limitations
|
35 |
+
|
36 |
+
More information needed
|
37 |
+
|
38 |
+
## Training and evaluation data
|
39 |
+
|
40 |
+
More information needed
|
41 |
+
|
42 |
+
## Training procedure
|
43 |
+
|
44 |
+
### Training hyperparameters
|
45 |
+
|
46 |
+
The following hyperparameters were used during training:
|
47 |
+
- learning_rate: 1e-05
|
48 |
+
- train_batch_size: 2
|
49 |
+
- eval_batch_size: 2
|
50 |
+
- seed: 42
|
51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
52 |
+
- lr_scheduler_type: linear
|
53 |
+
- num_epochs: 100
|
54 |
+
|
55 |
+
### Training results
|
56 |
+
|
57 |
+
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
58 |
+
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
59 |
+
| No log | 1.0 | 75 | 0.9234 | 0.6139 | 0.7526 | 0.6762 | 0.7416 |
|
60 |
+
| No log | 2.0 | 150 | 0.6101 | 0.7549 | 0.8341 | 0.7925 | 0.7940 |
|
61 |
+
| No log | 3.0 | 225 | 0.5135 | 0.8332 | 0.8882 | 0.8598 | 0.8091 |
|
62 |
+
| No log | 4.0 | 300 | 0.5467 | 0.8189 | 0.8624 | 0.8401 | 0.8109 |
|
63 |
+
| No log | 5.0 | 375 | 0.4879 | 0.8660 | 0.9051 | 0.8851 | 0.8504 |
|
64 |
+
| No log | 6.0 | 450 | 0.5352 | 0.8787 | 0.9180 | 0.8980 | 0.8480 |
|
65 |
+
| 0.5752 | 7.0 | 525 | 0.5900 | 0.8730 | 0.8982 | 0.8854 | 0.8343 |
|
66 |
+
| 0.5752 | 8.0 | 600 | 0.6014 | 0.8832 | 0.9016 | 0.8923 | 0.8506 |
|
67 |
+
| 0.5752 | 9.0 | 675 | 0.6173 | 0.8883 | 0.9126 | 0.9003 | 0.8538 |
|
68 |
+
| 0.5752 | 10.0 | 750 | 0.6278 | 0.8787 | 0.9141 | 0.8960 | 0.8571 |
|
69 |
+
| 0.5752 | 11.0 | 825 | 0.6573 | 0.8612 | 0.9155 | 0.8876 | 0.8326 |
|
70 |
+
| 0.5752 | 12.0 | 900 | 0.7333 | 0.8818 | 0.9006 | 0.8911 | 0.8387 |
|
71 |
+
| 0.5752 | 13.0 | 975 | 0.7489 | 0.8888 | 0.9136 | 0.9010 | 0.8502 |
|
72 |
+
| 0.1263 | 14.0 | 1050 | 0.7719 | 0.8908 | 0.8997 | 0.8952 | 0.8318 |
|
73 |
+
| 0.1263 | 15.0 | 1125 | 0.8295 | 0.8945 | 0.9101 | 0.9022 | 0.8438 |
|
74 |
+
| 0.1263 | 16.0 | 1200 | 0.8447 | 0.8798 | 0.9126 | 0.8959 | 0.8465 |
|
75 |
+
| 0.1263 | 17.0 | 1275 | 0.8359 | 0.9090 | 0.8932 | 0.9010 | 0.8486 |
|
76 |
+
| 0.1263 | 18.0 | 1350 | 0.8430 | 0.8966 | 0.9091 | 0.9028 | 0.8414 |
|
77 |
+
| 0.1263 | 19.0 | 1425 | 0.8179 | 0.8854 | 0.9021 | 0.8937 | 0.8400 |
|
78 |
+
| 0.0482 | 20.0 | 1500 | 0.8950 | 0.8968 | 0.8982 | 0.8975 | 0.8475 |
|
79 |
+
| 0.0482 | 21.0 | 1575 | 0.8790 | 0.9053 | 0.9121 | 0.9087 | 0.8565 |
|
80 |
+
| 0.0482 | 22.0 | 1650 | 0.7915 | 0.9056 | 0.9101 | 0.9078 | 0.8595 |
|
81 |
+
| 0.0482 | 23.0 | 1725 | 0.8760 | 0.8938 | 0.8952 | 0.8945 | 0.8504 |
|
82 |
+
| 0.0482 | 24.0 | 1800 | 0.8320 | 0.9113 | 0.9086 | 0.9100 | 0.8625 |
|
83 |
+
| 0.0482 | 25.0 | 1875 | 0.8880 | 0.9017 | 0.9021 | 0.9019 | 0.8538 |
|
84 |
+
| 0.0482 | 26.0 | 1950 | 0.8611 | 0.9083 | 0.9101 | 0.9092 | 0.8499 |
|
85 |
+
| 0.0163 | 27.0 | 2025 | 0.8747 | 0.9068 | 0.9086 | 0.9077 | 0.8600 |
|
86 |
+
| 0.0163 | 28.0 | 2100 | 0.8476 | 0.8955 | 0.9071 | 0.9013 | 0.8691 |
|
87 |
+
|
88 |
+
|
89 |
+
### Framework versions
|
90 |
+
|
91 |
+
- Transformers 4.42.0.dev0
|
92 |
+
- Pytorch 2.3.0+cu121
|
93 |
+
- Datasets 2.19.1
|
94 |
+
- Tokenizers 0.19.1
|
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": "4.42.0.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
|
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2c268eafa2fcad50f6d6ca4a6a17eb6c8989732d9a0c4330dd4b804f6f3d36d9
|
3 |
+
size 501355532
|
preprocessor_config.json
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_valid_processor_keys": [
|
3 |
+
"images",
|
4 |
+
"do_resize",
|
5 |
+
"size",
|
6 |
+
"resample",
|
7 |
+
"do_rescale",
|
8 |
+
"rescale_factor",
|
9 |
+
"do_normalize",
|
10 |
+
"image_mean",
|
11 |
+
"image_std",
|
12 |
+
"apply_ocr",
|
13 |
+
"ocr_lang",
|
14 |
+
"tesseract_config",
|
15 |
+
"return_tensors",
|
16 |
+
"data_format",
|
17 |
+
"input_data_format"
|
18 |
+
],
|
19 |
+
"apply_ocr": false,
|
20 |
+
"do_normalize": true,
|
21 |
+
"do_rescale": true,
|
22 |
+
"do_resize": true,
|
23 |
+
"image_mean": [
|
24 |
+
0.5,
|
25 |
+
0.5,
|
26 |
+
0.5
|
27 |
+
],
|
28 |
+
"image_processor_type": "LayoutLMv3ImageProcessor",
|
29 |
+
"image_std": [
|
30 |
+
0.5,
|
31 |
+
0.5,
|
32 |
+
0.5
|
33 |
+
],
|
34 |
+
"ocr_lang": null,
|
35 |
+
"processor_class": "LayoutLMv3Processor",
|
36 |
+
"resample": 2,
|
37 |
+
"rescale_factor": 0.00392156862745098,
|
38 |
+
"size": {
|
39 |
+
"height": 224,
|
40 |
+
"width": 224
|
41 |
+
},
|
42 |
+
"tesseract_config": ""
|
43 |
+
}
|
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 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_prefix_space": true,
|
3 |
+
"added_tokens_decoder": {
|
4 |
+
"0": {
|
5 |
+
"content": "<s>",
|
6 |
+
"lstrip": false,
|
7 |
+
"normalized": true,
|
8 |
+
"rstrip": false,
|
9 |
+
"single_word": false,
|
10 |
+
"special": true
|
11 |
+
},
|
12 |
+
"1": {
|
13 |
+
"content": "<pad>",
|
14 |
+
"lstrip": false,
|
15 |
+
"normalized": true,
|
16 |
+
"rstrip": false,
|
17 |
+
"single_word": false,
|
18 |
+
"special": true
|
19 |
+
},
|
20 |
+
"2": {
|
21 |
+
"content": "</s>",
|
22 |
+
"lstrip": false,
|
23 |
+
"normalized": true,
|
24 |
+
"rstrip": false,
|
25 |
+
"single_word": false,
|
26 |
+
"special": true
|
27 |
+
},
|
28 |
+
"3": {
|
29 |
+
"content": "<unk>",
|
30 |
+
"lstrip": false,
|
31 |
+
"normalized": true,
|
32 |
+
"rstrip": false,
|
33 |
+
"single_word": false,
|
34 |
+
"special": true
|
35 |
+
},
|
36 |
+
"50264": {
|
37 |
+
"content": "<mask>",
|
38 |
+
"lstrip": true,
|
39 |
+
"normalized": true,
|
40 |
+
"rstrip": false,
|
41 |
+
"single_word": false,
|
42 |
+
"special": true
|
43 |
+
}
|
44 |
+
},
|
45 |
+
"apply_ocr": false,
|
46 |
+
"bos_token": "<s>",
|
47 |
+
"clean_up_tokenization_spaces": true,
|
48 |
+
"cls_token": "<s>",
|
49 |
+
"cls_token_box": [
|
50 |
+
0,
|
51 |
+
0,
|
52 |
+
0,
|
53 |
+
0
|
54 |
+
],
|
55 |
+
"eos_token": "</s>",
|
56 |
+
"errors": "replace",
|
57 |
+
"mask_token": "<mask>",
|
58 |
+
"model_max_length": 512,
|
59 |
+
"only_label_first_subword": true,
|
60 |
+
"pad_token": "<pad>",
|
61 |
+
"pad_token_box": [
|
62 |
+
0,
|
63 |
+
0,
|
64 |
+
0,
|
65 |
+
0
|
66 |
+
],
|
67 |
+
"pad_token_label": -100,
|
68 |
+
"processor_class": "LayoutLMv3Processor",
|
69 |
+
"sep_token": "</s>",
|
70 |
+
"sep_token_box": [
|
71 |
+
0,
|
72 |
+
0,
|
73 |
+
0,
|
74 |
+
0
|
75 |
+
],
|
76 |
+
"tokenizer_class": "LayoutLMv3Tokenizer",
|
77 |
+
"trim_offsets": true,
|
78 |
+
"unk_token": "<unk>"
|
79 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d0a10385dd78382bbdf1d2925aa46442694230a28c0581ddda1608dddb32e9b5
|
3 |
+
size 5048
|
vocab.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|