Upload to official revision
Browse files- added_tokens.json +76 -0
- artifacts.ckpt +3 -0
- config.json +24 -0
- config.yaml +29 -0
- events.out.tfevents.1750924743.a5d21858b030.709.0 +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 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s_account_name>": 57587,
|
| 3 |
+
"</s_account_number>": 57585,
|
| 4 |
+
"</s_address>": 57547,
|
| 5 |
+
"</s_attention>": 57553,
|
| 6 |
+
"</s_authorized_by>": 57589,
|
| 7 |
+
"</s_bank_name>": 57583,
|
| 8 |
+
"</s_bast_number>": 57593,
|
| 9 |
+
"</s_bill_to>": 57555,
|
| 10 |
+
"</s_company>": 57549,
|
| 11 |
+
"</s_credit_term>": 57537,
|
| 12 |
+
"</s_customer>": 57543,
|
| 13 |
+
"</s_customer_gfb_id>": 57533,
|
| 14 |
+
"</s_customer_id>": 57531,
|
| 15 |
+
"</s_customer_number>": 57529,
|
| 16 |
+
"</s_description>": 57561,
|
| 17 |
+
"</s_due_date>": 57535,
|
| 18 |
+
"</s_grand_total>": 57579,
|
| 19 |
+
"</s_invoice_date>": 57527,
|
| 20 |
+
"</s_invoice_number>": 57525,
|
| 21 |
+
"</s_item_number>": 57559,
|
| 22 |
+
"</s_items>": 57557,
|
| 23 |
+
"</s_name>": 57545,
|
| 24 |
+
"</s_npwp>": 57551,
|
| 25 |
+
"</s_payment_info>": 57581,
|
| 26 |
+
"</s_po_number>": 57539,
|
| 27 |
+
"</s_position>": 57591,
|
| 28 |
+
"</s_quantity>": 57563,
|
| 29 |
+
"</s_quotation_number>": 57541,
|
| 30 |
+
"</s_subtotal>": 57575,
|
| 31 |
+
"</s_totals>": 57573,
|
| 32 |
+
"</s_unit_price>": 57565,
|
| 33 |
+
"</s_value>": 57567,
|
| 34 |
+
"</s_vat>": 57569,
|
| 35 |
+
"</s_vat_amount>": 57571,
|
| 36 |
+
"</s_vat_total>": 57577,
|
| 37 |
+
"<s_account_name>": 57588,
|
| 38 |
+
"<s_account_number>": 57586,
|
| 39 |
+
"<s_address>": 57548,
|
| 40 |
+
"<s_attention>": 57554,
|
| 41 |
+
"<s_authorized_by>": 57590,
|
| 42 |
+
"<s_bank_name>": 57584,
|
| 43 |
+
"<s_bast_number>": 57594,
|
| 44 |
+
"<s_bill_to>": 57556,
|
| 45 |
+
"<s_company>": 57550,
|
| 46 |
+
"<s_credit_term>": 57538,
|
| 47 |
+
"<s_customer>": 57544,
|
| 48 |
+
"<s_customer_gfb_id>": 57534,
|
| 49 |
+
"<s_customer_id>": 57532,
|
| 50 |
+
"<s_customer_number>": 57530,
|
| 51 |
+
"<s_description>": 57562,
|
| 52 |
+
"<s_due_date>": 57536,
|
| 53 |
+
"<s_grab_invoices>": 57595,
|
| 54 |
+
"<s_grand_total>": 57580,
|
| 55 |
+
"<s_iitcdip>": 57523,
|
| 56 |
+
"<s_invoice_date>": 57528,
|
| 57 |
+
"<s_invoice_number>": 57526,
|
| 58 |
+
"<s_item_number>": 57560,
|
| 59 |
+
"<s_items>": 57558,
|
| 60 |
+
"<s_name>": 57546,
|
| 61 |
+
"<s_npwp>": 57552,
|
| 62 |
+
"<s_payment_info>": 57582,
|
| 63 |
+
"<s_po_number>": 57540,
|
| 64 |
+
"<s_position>": 57592,
|
| 65 |
+
"<s_quantity>": 57564,
|
| 66 |
+
"<s_quotation_number>": 57542,
|
| 67 |
+
"<s_subtotal>": 57576,
|
| 68 |
+
"<s_synthdog>": 57524,
|
| 69 |
+
"<s_totals>": 57574,
|
| 70 |
+
"<s_unit_price>": 57566,
|
| 71 |
+
"<s_value>": 57568,
|
| 72 |
+
"<s_vat>": 57570,
|
| 73 |
+
"<s_vat_amount>": 57572,
|
| 74 |
+
"<s_vat_total>": 57578,
|
| 75 |
+
"<sep/>": 57522
|
| 76 |
+
}
|
artifacts.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a93ca3ad86cf60b40c3299385eb8f6e43d05b6302228637094a26817326846d
|
| 3 |
+
size 1611526480
|
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": 1024,
|
| 19 |
+
"max_position_embeddings": 1024,
|
| 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/grab_invoices'
|
| 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: 1024
|
| 15 |
+
align_long_axis: False
|
| 16 |
+
num_nodes: 1
|
| 17 |
+
seed: 2022
|
| 18 |
+
lr: 3e-05
|
| 19 |
+
warmup_steps: 100
|
| 20 |
+
num_training_samples_per_epoch: 80
|
| 21 |
+
max_epochs: 10
|
| 22 |
+
max_steps: -1
|
| 23 |
+
num_workers: 8
|
| 24 |
+
val_check_interval: 1.0
|
| 25 |
+
check_val_every_n_epoch: 5
|
| 26 |
+
gradient_clip_val: 1.0
|
| 27 |
+
verbose: True
|
| 28 |
+
exp_name: 'train_sroie'
|
| 29 |
+
exp_version: '20250626_075818'
|
events.out.tfevents.1750924743.a5d21858b030.709.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:515bfe5de4fca6ce3c7a7dcbad178105f6f2282e71ede70a2975cda732f76cd6
|
| 3 |
+
size 3396
|
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:b73d274322a46048bf49bdd8e1503e1cf11546a29b7ca2f1e9704bc8c105e228
|
| 3 |
+
size 859665150
|
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_grab_invoices>"
|
| 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 |
+
}
|