Training done

added_tokens.json

@@ -1,14 +1,16 @@
 {
-  "</s_address>": 57532,
-  "</s_company>": 57530,
-  "</s_date>": 57528,
-  "</s_total>": 57526,
-  "<s_address>": 57531,
-  "<s_company>": 57529,
-  "<s_cord-v2>": 57533,
-  "<s_date>": 57527,
+  "</s_amt_in_figures>": 57528,
+  "</s_amt_in_words>": 57526,
+  "</s_bank_name>": 57534,
+  "</s_cheque_date>": 57532,
+  "</s_payee_name>": 57530,
+  "<s_amt_in_figures>": 57527,
+  "<s_amt_in_words>": 57525,
+  "<s_bank_name>": 57533,
+  "<s_cheque_date>": 57531,
+  "<s_cord-v2>": 57535,
   "<s_iitcdip>": 57523,
+  "<s_payee_name>": 57529,
   "<s_synthdog>": 57524,
-  "<s_total>": 57525,
   "<sep/>": 57522
 }

preprocessor_config.json

@@ -37,8 +37,8 @@
   "processor_class": "DonutProcessor",
   "resample": 2,
   "rescale_factor": 0.00392156862745098,
-  "size": {
-    "height": 720,
-    "width": 960
-  }
+  "size": [
+    960,
+    720
+  ]
 }

tokenizer_config.json

@@ -65,7 +65,7 @@
       "special": true
     },
     "57525": {
-      "content": "<s_total>",
+      "content": "<s_amt_in_words>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -73,7 +73,7 @@
       "special": false
     },
     "57526": {
-      "content": "</s_total>",
+      "content": "</s_amt_in_words>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -81,7 +81,7 @@
       "special": false
     },
     "57527": {
-      "content": "<s_date>",
+      "content": "<s_amt_in_figures>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -89,7 +89,7 @@
       "special": false
     },
     "57528": {
-      "content": "</s_date>",
+      "content": "</s_amt_in_figures>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -97,7 +97,7 @@
       "special": false
     },
     "57529": {
-      "content": "<s_company>",
+      "content": "<s_payee_name>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -105,7 +105,7 @@
       "special": false
     },
     "57530": {
-      "content": "</s_company>",
+      "content": "</s_payee_name>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -113,7 +113,7 @@
       "special": false
     },
     "57531": {
-      "content": "<s_address>",
+      "content": "<s_cheque_date>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -121,7 +121,7 @@
       "special": false
     },
     "57532": {
-      "content": "</s_address>",
+      "content": "</s_cheque_date>",
       "lstrip": false,
       "normalized": true,
       "rstrip": false,
@@ -129,6 +129,22 @@
       "special": false
     },
     "57533": {
+      "content": "<s_bank_name>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "57534": {
+      "content": "</s_bank_name>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "57535": {
       "content": "<s_cord-v2>",
       "lstrip": false,
       "normalized": true,
@@ -146,18 +162,11 @@
   "cls_token": "<s>",
   "eos_token": "</s>",
   "mask_token": "<mask>",
-  "max_length": 512,
   "model_max_length": 1000000000000000019884624838656,
-  "pad_to_multiple_of": null,
   "pad_token": "<pad>",
-  "pad_token_type_id": 0,
-  "padding_side": "right",
   "processor_class": "DonutProcessor",
   "sep_token": "</s>",
   "sp_model_kwargs": {},
-  "stride": 0,
   "tokenizer_class": "XLMRobertaTokenizer",
-  "truncation_side": "right",
-  "truncation_strategy": "longest_first",
   "unk_token": "<unk>"
 }