Upload to official revision
Browse files- added_tokens.json +62 -0
- artifacts.ckpt +3 -0
- config.json +24 -0
- config.yaml +29 -0
- events.out.tfevents.1750221473.75427e6784e3.135.0 +3 -0
- hparams.yaml +1 -0
- pytorch_model.bin +3 -0
- sentencepiece.bpe.model +3 -0
- special_tokens_map.json +18 -0
- tokenizer_config.json +22 -0
added_tokens.json
ADDED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s_address>": 57549,
|
3 |
+
"</s_attention>": 57539,
|
4 |
+
"</s_city>": 57551,
|
5 |
+
"</s_credit_term>": 57537,
|
6 |
+
"</s_customer>": 57545,
|
7 |
+
"</s_customer_gfb_id>": 57535,
|
8 |
+
"</s_customer_number>": 57533,
|
9 |
+
"</s_description>": 57559,
|
10 |
+
"</s_due_date>": 57531,
|
11 |
+
"</s_extended_value>": 57565,
|
12 |
+
"</s_header>": 57525,
|
13 |
+
"</s_invoice_date>": 57529,
|
14 |
+
"</s_invoice_number>": 57527,
|
15 |
+
"</s_items>": 57555,
|
16 |
+
"</s_name>": 57547,
|
17 |
+
"</s_npwp>": 57553,
|
18 |
+
"</s_number>": 57557,
|
19 |
+
"</s_po_number>": 57541,
|
20 |
+
"</s_quantity>": 57561,
|
21 |
+
"</s_quotation_number>": 57543,
|
22 |
+
"</s_total_excl_vat>": 57573,
|
23 |
+
"</s_total_incl_vat>": 57579,
|
24 |
+
"</s_total_tax_based_value>": 57575,
|
25 |
+
"</s_total_vat_amount>": 57577,
|
26 |
+
"</s_totals>": 57571,
|
27 |
+
"</s_unit_price>": 57563,
|
28 |
+
"</s_vat_amount>": 57569,
|
29 |
+
"</s_vat_percent>": 57567,
|
30 |
+
"<s_address>": 57550,
|
31 |
+
"<s_attention>": 57540,
|
32 |
+
"<s_city>": 57552,
|
33 |
+
"<s_credit_term>": 57538,
|
34 |
+
"<s_customer>": 57546,
|
35 |
+
"<s_customer_gfb_id>": 57536,
|
36 |
+
"<s_customer_number>": 57534,
|
37 |
+
"<s_description>": 57560,
|
38 |
+
"<s_due_date>": 57532,
|
39 |
+
"<s_extended_value>": 57566,
|
40 |
+
"<s_header>": 57526,
|
41 |
+
"<s_iitcdip>": 57523,
|
42 |
+
"<s_invoice_date>": 57530,
|
43 |
+
"<s_invoice_number>": 57528,
|
44 |
+
"<s_invoices-git2>": 57581,
|
45 |
+
"<s_items>": 57556,
|
46 |
+
"<s_name>": 57548,
|
47 |
+
"<s_npwp>": 57554,
|
48 |
+
"<s_number>": 57558,
|
49 |
+
"<s_po_number>": 57542,
|
50 |
+
"<s_quantity>": 57562,
|
51 |
+
"<s_quotation_number>": 57544,
|
52 |
+
"<s_synthdog>": 57524,
|
53 |
+
"<s_total_excl_vat>": 57574,
|
54 |
+
"<s_total_incl_vat>": 57580,
|
55 |
+
"<s_total_tax_based_value>": 57576,
|
56 |
+
"<s_total_vat_amount>": 57578,
|
57 |
+
"<s_totals>": 57572,
|
58 |
+
"<s_unit_price>": 57564,
|
59 |
+
"<s_vat_amount>": 57570,
|
60 |
+
"<s_vat_percent>": 57568,
|
61 |
+
"<sep/>": 57522
|
62 |
+
}
|
artifacts.ckpt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1cbe52abeca3bd61459b353ab24df7ec44323638113e506716d00c8a6344c451
|
3 |
+
size 1609314640
|
config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "naver-clova-ix/donut-base",
|
3 |
+
"align_long_axis": false,
|
4 |
+
"architectures": [
|
5 |
+
"DonutModel"
|
6 |
+
],
|
7 |
+
"decoder_layer": 4,
|
8 |
+
"encoder_layer": [
|
9 |
+
2,
|
10 |
+
2,
|
11 |
+
14,
|
12 |
+
2
|
13 |
+
],
|
14 |
+
"input_size": [
|
15 |
+
1280,
|
16 |
+
960
|
17 |
+
],
|
18 |
+
"max_length": 768,
|
19 |
+
"max_position_embeddings": 768,
|
20 |
+
"model_type": "donut",
|
21 |
+
"torch_dtype": "float32",
|
22 |
+
"transformers_version": "4.24.0",
|
23 |
+
"window_size": 10
|
24 |
+
}
|
config.yaml
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
resume_from_checkpoint_path: None
|
2 |
+
result_path: './result'
|
3 |
+
pretrained_model_name_or_path: 'naver-clova-ix/donut-base'
|
4 |
+
dataset_name_or_paths:
|
5 |
+
- 'fauzansalim/invoices-git2'
|
6 |
+
sort_json_key: False
|
7 |
+
train_batch_sizes:
|
8 |
+
- 1
|
9 |
+
val_batch_sizes:
|
10 |
+
- 1
|
11 |
+
input_size:
|
12 |
+
- 1280
|
13 |
+
- 960
|
14 |
+
max_length: 768
|
15 |
+
align_long_axis: False
|
16 |
+
num_nodes: 1
|
17 |
+
seed: 2022
|
18 |
+
lr: 3e-05
|
19 |
+
warmup_steps: 300
|
20 |
+
num_training_samples_per_epoch: 200
|
21 |
+
max_epochs: 20
|
22 |
+
max_steps: -1
|
23 |
+
num_workers: 8
|
24 |
+
val_check_interval: 1.0
|
25 |
+
check_val_every_n_epoch: 10
|
26 |
+
gradient_clip_val: 1.0
|
27 |
+
verbose: True
|
28 |
+
exp_name: 'train_sroie'
|
29 |
+
exp_version: '20250618_043648'
|
events.out.tfevents.1750221473.75427e6784e3.135.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8ef566d6c8a5a3dc9a18421e665f56795c44159e3b16673c16697f94c60d9135
|
3 |
+
size 6592
|
hparams.yaml
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:45ab7f627288d009238ede897b37d865ef03791ad35cf4d09f3516821244dbcf
|
3 |
+
size 858559230
|
sentencepiece.bpe.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cb9e3dce4c326195d08fc3dd0f7e2eee1da8595c847bf4c1a9c78b7a82d47e2d
|
3 |
+
size 1296245
|
special_tokens_map.json
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<s_invoices-git2>"
|
4 |
+
],
|
5 |
+
"bos_token": "<s>",
|
6 |
+
"cls_token": "<s>",
|
7 |
+
"eos_token": "</s>",
|
8 |
+
"mask_token": {
|
9 |
+
"content": "<mask>",
|
10 |
+
"lstrip": true,
|
11 |
+
"normalized": true,
|
12 |
+
"rstrip": false,
|
13 |
+
"single_word": false
|
14 |
+
},
|
15 |
+
"pad_token": "<pad>",
|
16 |
+
"sep_token": "</s>",
|
17 |
+
"unk_token": "<unk>"
|
18 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"cls_token": "<s>",
|
4 |
+
"eos_token": "</s>",
|
5 |
+
"from_slow": true,
|
6 |
+
"mask_token": {
|
7 |
+
"__type": "AddedToken",
|
8 |
+
"content": "<mask>",
|
9 |
+
"lstrip": true,
|
10 |
+
"normalized": true,
|
11 |
+
"rstrip": false,
|
12 |
+
"single_word": false
|
13 |
+
},
|
14 |
+
"name_or_path": "naver-clova-ix/donut-base",
|
15 |
+
"pad_token": "<pad>",
|
16 |
+
"processor_class": "DonutProcessor",
|
17 |
+
"sep_token": "</s>",
|
18 |
+
"sp_model_kwargs": {},
|
19 |
+
"special_tokens_map_file": null,
|
20 |
+
"tokenizer_class": "XLMRobertaTokenizer",
|
21 |
+
"unk_token": "<unk>"
|
22 |
+
}
|