Upload to official revision
Browse files- added_tokens.json +60 -0
- artifacts.ckpt +3 -0
- config.json +24 -0
- config.yaml +29 -0
- events.out.tfevents.1750755636.fff78a9403bc.1524.0 +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 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s_account_number>": 57569,
|
| 3 |
+
"</s_address>": 57533,
|
| 4 |
+
"</s_attention>": 57539,
|
| 5 |
+
"</s_authorized_by>": 57573,
|
| 6 |
+
"</s_bank_name>": 57567,
|
| 7 |
+
"</s_bast_number>": 57577,
|
| 8 |
+
"</s_bill_to>": 57541,
|
| 9 |
+
"</s_city>": 57535,
|
| 10 |
+
"</s_company>": 57537,
|
| 11 |
+
"</s_credit_term>": 57543,
|
| 12 |
+
"</s_customer>": 57529,
|
| 13 |
+
"</s_description>": 57549,
|
| 14 |
+
"</s_grand_total>": 57563,
|
| 15 |
+
"</s_invoice_date>": 57527,
|
| 16 |
+
"</s_invoice_number>": 57525,
|
| 17 |
+
"</s_item_number>": 57547,
|
| 18 |
+
"</s_items>": 57545,
|
| 19 |
+
"</s_name>": 57531,
|
| 20 |
+
"</s_payment_info>": 57565,
|
| 21 |
+
"</s_position>": 57575,
|
| 22 |
+
"</s_quantity>": 57551,
|
| 23 |
+
"</s_subtotal>": 57559,
|
| 24 |
+
"</s_swift_code>": 57571,
|
| 25 |
+
"</s_totals>": 57557,
|
| 26 |
+
"</s_unit_price>": 57553,
|
| 27 |
+
"</s_value>": 57555,
|
| 28 |
+
"</s_vat_11>": 57561,
|
| 29 |
+
"<s_account_number>": 57570,
|
| 30 |
+
"<s_address>": 57534,
|
| 31 |
+
"<s_attention>": 57540,
|
| 32 |
+
"<s_authorized_by>": 57574,
|
| 33 |
+
"<s_bank_name>": 57568,
|
| 34 |
+
"<s_bast_number>": 57578,
|
| 35 |
+
"<s_bill_to>": 57542,
|
| 36 |
+
"<s_city>": 57536,
|
| 37 |
+
"<s_company>": 57538,
|
| 38 |
+
"<s_credit_term>": 57544,
|
| 39 |
+
"<s_customer>": 57530,
|
| 40 |
+
"<s_description>": 57550,
|
| 41 |
+
"<s_grand_total>": 57564,
|
| 42 |
+
"<s_iitcdip>": 57523,
|
| 43 |
+
"<s_invoice_date>": 57528,
|
| 44 |
+
"<s_invoice_number>": 57526,
|
| 45 |
+
"<s_invoices-git3>": 57579,
|
| 46 |
+
"<s_item_number>": 57548,
|
| 47 |
+
"<s_items>": 57546,
|
| 48 |
+
"<s_name>": 57532,
|
| 49 |
+
"<s_payment_info>": 57566,
|
| 50 |
+
"<s_position>": 57576,
|
| 51 |
+
"<s_quantity>": 57552,
|
| 52 |
+
"<s_subtotal>": 57560,
|
| 53 |
+
"<s_swift_code>": 57572,
|
| 54 |
+
"<s_synthdog>": 57524,
|
| 55 |
+
"<s_totals>": 57558,
|
| 56 |
+
"<s_unit_price>": 57554,
|
| 57 |
+
"<s_value>": 57556,
|
| 58 |
+
"<s_vat_11>": 57562,
|
| 59 |
+
"<sep/>": 57522
|
| 60 |
+
}
|
artifacts.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b28e292be12942671c115f4f4cfdc1406aa1ce8ab446199365f02abf11ed6da
|
| 3 |
+
size 1609298256
|
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-git3'
|
| 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: 100
|
| 20 |
+
num_training_samples_per_epoch: 41
|
| 21 |
+
max_epochs: 50
|
| 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: '20250624_085958'
|
events.out.tfevents.1750755636.fff78a9403bc.1524.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a94b2ff5058346166a0d2de860794eca73415efa70ef66dc85af6b851d56536
|
| 3 |
+
size 9328
|
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 sha256:9cec2ac1bf275a49f4b098b9f230b75baf3db9fe2fc51252ec536abae54a1d65
|
| 3 |
+
size 858551038
|
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-git3>"
|
| 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 |
+
}
|