Upload 9 files
Browse files- added_tokens.json +22 -19
- config.json +5 -6
- generation_config.json +1 -1
- model.safetensors +2 -2
- preprocessor_config.json +5 -5
- tokenizer.json +0 -0
- tokenizer_config.json +43 -19
added_tokens.json
CHANGED
@@ -1,36 +1,39 @@
|
|
1 |
{
|
2 |
"</s_date>": 57532,
|
3 |
-
"</
|
4 |
-
"</s_item_key>":
|
5 |
-
"</s_item_name>":
|
6 |
-
"</s_item_quantity>":
|
7 |
-
"</s_item_value>":
|
8 |
-
"</s_line_items>":
|
|
|
|
|
9 |
"</s_store_addr>": 57528,
|
10 |
"</s_store_name>": 57526,
|
11 |
"</s_subtotal>": 57536,
|
12 |
-
"</
|
13 |
-
"</
|
14 |
"</s_time>": 57534,
|
15 |
"</s_tips>": 57544,
|
16 |
-
"</s_total>":
|
17 |
-
"<s_cord-v2>": 57555,
|
18 |
"<s_date>": 57531,
|
19 |
-
"<
|
20 |
"<s_iitcdip>": 57523,
|
21 |
-
"<s_item_key>":
|
22 |
-
"<s_item_name>":
|
23 |
-
"<s_item_quantity>":
|
24 |
-
"<s_item_value>":
|
25 |
-
"<s_line_items>":
|
|
|
|
|
26 |
"<s_store_addr>": 57527,
|
27 |
"<s_store_name>": 57525,
|
28 |
"<s_subtotal>": 57535,
|
|
|
29 |
"<s_synthdog>": 57524,
|
30 |
-
"<s_tax>":
|
31 |
-
"<s_telephone>": 57529,
|
32 |
"<s_time>": 57533,
|
33 |
"<s_tips>": 57543,
|
34 |
-
"<s_total>":
|
35 |
"<sep/>": 57522
|
36 |
}
|
|
|
1 |
{
|
2 |
"</s_date>": 57532,
|
3 |
+
"</s_discount>": 57546,
|
4 |
+
"</s_item_key>": 57550,
|
5 |
+
"</s_item_name>": 57552,
|
6 |
+
"</s_item_quantity>": 57556,
|
7 |
+
"</s_item_value>": 57554,
|
8 |
+
"</s_line_items>": 57548,
|
9 |
+
"</s_phone>": 57530,
|
10 |
+
"</s_receipt>": 57558,
|
11 |
"</s_store_addr>": 57528,
|
12 |
"</s_store_name>": 57526,
|
13 |
"</s_subtotal>": 57536,
|
14 |
+
"</s_svc>": 57538,
|
15 |
+
"</s_tax>": 57540,
|
16 |
"</s_time>": 57534,
|
17 |
"</s_tips>": 57544,
|
18 |
+
"</s_total>": 57542,
|
|
|
19 |
"<s_date>": 57531,
|
20 |
+
"<s_discount>": 57545,
|
21 |
"<s_iitcdip>": 57523,
|
22 |
+
"<s_item_key>": 57549,
|
23 |
+
"<s_item_name>": 57551,
|
24 |
+
"<s_item_quantity>": 57555,
|
25 |
+
"<s_item_value>": 57553,
|
26 |
+
"<s_line_items>": 57547,
|
27 |
+
"<s_phone>": 57529,
|
28 |
+
"<s_receipt>": 57557,
|
29 |
"<s_store_addr>": 57527,
|
30 |
"<s_store_name>": 57525,
|
31 |
"<s_subtotal>": 57535,
|
32 |
+
"<s_svc>": 57537,
|
33 |
"<s_synthdog>": 57524,
|
34 |
+
"<s_tax>": 57539,
|
|
|
35 |
"<s_time>": 57533,
|
36 |
"<s_tips>": 57543,
|
37 |
+
"<s_total>": 57541,
|
38 |
"<sep/>": 57522
|
39 |
}
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"VisionEncoderDecoderModel"
|
5 |
],
|
@@ -85,9 +85,9 @@
|
|
85 |
"typical_p": 1.0,
|
86 |
"use_bfloat16": false,
|
87 |
"use_cache": true,
|
88 |
-
"vocab_size":
|
89 |
},
|
90 |
-
"decoder_start_token_id":
|
91 |
"encoder": {
|
92 |
"_name_or_path": "",
|
93 |
"add_cross_attention": false,
|
@@ -124,8 +124,8 @@
|
|
124 |
"1": "LABEL_1"
|
125 |
},
|
126 |
"image_size": [
|
127 |
-
|
128 |
-
|
129 |
],
|
130 |
"initializer_range": 0.02,
|
131 |
"is_decoder": false,
|
@@ -187,6 +187,5 @@
|
|
187 |
"model_type": "vision-encoder-decoder",
|
188 |
"pad_token_id": 1,
|
189 |
"tie_word_embeddings": false,
|
190 |
-
"torch_dtype": "float32",
|
191 |
"transformers_version": "4.36.2"
|
192 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "C:\\Users\\adam\\Documents\\Code\\LLMs\\Trained_Models\\Donut-base\\v2",
|
3 |
"architectures": [
|
4 |
"VisionEncoderDecoderModel"
|
5 |
],
|
|
|
85 |
"typical_p": 1.0,
|
86 |
"use_bfloat16": false,
|
87 |
"use_cache": true,
|
88 |
+
"vocab_size": 57559
|
89 |
},
|
90 |
+
"decoder_start_token_id": 57557,
|
91 |
"encoder": {
|
92 |
"_name_or_path": "",
|
93 |
"add_cross_attention": false,
|
|
|
124 |
"1": "LABEL_1"
|
125 |
},
|
126 |
"image_size": [
|
127 |
+
1798,
|
128 |
+
1348
|
129 |
],
|
130 |
"initializer_range": 0.02,
|
131 |
"is_decoder": false,
|
|
|
187 |
"model_type": "vision-encoder-decoder",
|
188 |
"pad_token_id": 1,
|
189 |
"tie_word_embeddings": false,
|
|
|
190 |
"transformers_version": "4.36.2"
|
191 |
}
|
generation_config.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 0,
|
4 |
-
"decoder_start_token_id":
|
5 |
"eos_token_id": 2,
|
6 |
"forced_eos_token_id": 2,
|
7 |
"max_length": 768,
|
|
|
1 |
{
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 0,
|
4 |
+
"decoder_start_token_id": 57557,
|
5 |
"eos_token_id": 2,
|
6 |
"forced_eos_token_id": 2,
|
7 |
"max_length": 768,
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9488f552c39187659857b11e3e83b45a615e037d0c57f37f77d528a1682099cc
|
3 |
+
size 809210008
|
preprocessor_config.json
CHANGED
@@ -19,8 +19,8 @@
|
|
19 |
"processor_class": "DonutProcessor",
|
20 |
"resample": 2,
|
21 |
"rescale_factor": 0.00392156862745098,
|
22 |
-
"size":
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
}
|
|
|
19 |
"processor_class": "DonutProcessor",
|
20 |
"resample": 2,
|
21 |
"rescale_factor": 0.00392156862745098,
|
22 |
+
"size": {
|
23 |
+
"height": 1798,
|
24 |
+
"width": 1348
|
25 |
+
}
|
26 |
+
}
|
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
CHANGED
@@ -97,7 +97,7 @@
|
|
97 |
"special": false
|
98 |
},
|
99 |
"57529": {
|
100 |
-
"content": "<
|
101 |
"lstrip": false,
|
102 |
"normalized": true,
|
103 |
"rstrip": false,
|
@@ -105,7 +105,7 @@
|
|
105 |
"special": false
|
106 |
},
|
107 |
"57530": {
|
108 |
-
"content": "</
|
109 |
"lstrip": false,
|
110 |
"normalized": true,
|
111 |
"rstrip": false,
|
@@ -161,7 +161,7 @@
|
|
161 |
"special": false
|
162 |
},
|
163 |
"57537": {
|
164 |
-
"content": "<
|
165 |
"lstrip": false,
|
166 |
"normalized": true,
|
167 |
"rstrip": false,
|
@@ -169,7 +169,7 @@
|
|
169 |
"special": false
|
170 |
},
|
171 |
"57538": {
|
172 |
-
"content": "</
|
173 |
"lstrip": false,
|
174 |
"normalized": true,
|
175 |
"rstrip": false,
|
@@ -177,7 +177,7 @@
|
|
177 |
"special": false
|
178 |
},
|
179 |
"57539": {
|
180 |
-
"content": "<
|
181 |
"lstrip": false,
|
182 |
"normalized": true,
|
183 |
"rstrip": false,
|
@@ -185,7 +185,7 @@
|
|
185 |
"special": false
|
186 |
},
|
187 |
"57540": {
|
188 |
-
"content": "</
|
189 |
"lstrip": false,
|
190 |
"normalized": true,
|
191 |
"rstrip": false,
|
@@ -193,7 +193,7 @@
|
|
193 |
"special": false
|
194 |
},
|
195 |
"57541": {
|
196 |
-
"content": "<
|
197 |
"lstrip": false,
|
198 |
"normalized": true,
|
199 |
"rstrip": false,
|
@@ -201,7 +201,7 @@
|
|
201 |
"special": false
|
202 |
},
|
203 |
"57542": {
|
204 |
-
"content": "</
|
205 |
"lstrip": false,
|
206 |
"normalized": true,
|
207 |
"rstrip": false,
|
@@ -225,7 +225,7 @@
|
|
225 |
"special": false
|
226 |
},
|
227 |
"57545": {
|
228 |
-
"content": "<
|
229 |
"lstrip": false,
|
230 |
"normalized": true,
|
231 |
"rstrip": false,
|
@@ -233,7 +233,7 @@
|
|
233 |
"special": false
|
234 |
},
|
235 |
"57546": {
|
236 |
-
"content": "</
|
237 |
"lstrip": false,
|
238 |
"normalized": true,
|
239 |
"rstrip": false,
|
@@ -241,7 +241,7 @@
|
|
241 |
"special": false
|
242 |
},
|
243 |
"57547": {
|
244 |
-
"content": "<
|
245 |
"lstrip": false,
|
246 |
"normalized": true,
|
247 |
"rstrip": false,
|
@@ -249,7 +249,7 @@
|
|
249 |
"special": false
|
250 |
},
|
251 |
"57548": {
|
252 |
-
"content": "</
|
253 |
"lstrip": false,
|
254 |
"normalized": true,
|
255 |
"rstrip": false,
|
@@ -257,7 +257,7 @@
|
|
257 |
"special": false
|
258 |
},
|
259 |
"57549": {
|
260 |
-
"content": "<
|
261 |
"lstrip": false,
|
262 |
"normalized": true,
|
263 |
"rstrip": false,
|
@@ -265,7 +265,7 @@
|
|
265 |
"special": false
|
266 |
},
|
267 |
"57550": {
|
268 |
-
"content": "</
|
269 |
"lstrip": false,
|
270 |
"normalized": true,
|
271 |
"rstrip": false,
|
@@ -273,7 +273,7 @@
|
|
273 |
"special": false
|
274 |
},
|
275 |
"57551": {
|
276 |
-
"content": "<
|
277 |
"lstrip": false,
|
278 |
"normalized": true,
|
279 |
"rstrip": false,
|
@@ -281,7 +281,7 @@
|
|
281 |
"special": false
|
282 |
},
|
283 |
"57552": {
|
284 |
-
"content": "</
|
285 |
"lstrip": false,
|
286 |
"normalized": true,
|
287 |
"rstrip": false,
|
@@ -289,7 +289,7 @@
|
|
289 |
"special": false
|
290 |
},
|
291 |
"57553": {
|
292 |
-
"content": "<
|
293 |
"lstrip": false,
|
294 |
"normalized": true,
|
295 |
"rstrip": false,
|
@@ -297,7 +297,7 @@
|
|
297 |
"special": false
|
298 |
},
|
299 |
"57554": {
|
300 |
-
"content": "</
|
301 |
"lstrip": false,
|
302 |
"normalized": true,
|
303 |
"rstrip": false,
|
@@ -305,7 +305,31 @@
|
|
305 |
"special": false
|
306 |
},
|
307 |
"57555": {
|
308 |
-
"content": "<
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
309 |
"lstrip": false,
|
310 |
"normalized": true,
|
311 |
"rstrip": false,
|
|
|
97 |
"special": false
|
98 |
},
|
99 |
"57529": {
|
100 |
+
"content": "<s_phone>",
|
101 |
"lstrip": false,
|
102 |
"normalized": true,
|
103 |
"rstrip": false,
|
|
|
105 |
"special": false
|
106 |
},
|
107 |
"57530": {
|
108 |
+
"content": "</s_phone>",
|
109 |
"lstrip": false,
|
110 |
"normalized": true,
|
111 |
"rstrip": false,
|
|
|
161 |
"special": false
|
162 |
},
|
163 |
"57537": {
|
164 |
+
"content": "<s_svc>",
|
165 |
"lstrip": false,
|
166 |
"normalized": true,
|
167 |
"rstrip": false,
|
|
|
169 |
"special": false
|
170 |
},
|
171 |
"57538": {
|
172 |
+
"content": "</s_svc>",
|
173 |
"lstrip": false,
|
174 |
"normalized": true,
|
175 |
"rstrip": false,
|
|
|
177 |
"special": false
|
178 |
},
|
179 |
"57539": {
|
180 |
+
"content": "<s_tax>",
|
181 |
"lstrip": false,
|
182 |
"normalized": true,
|
183 |
"rstrip": false,
|
|
|
185 |
"special": false
|
186 |
},
|
187 |
"57540": {
|
188 |
+
"content": "</s_tax>",
|
189 |
"lstrip": false,
|
190 |
"normalized": true,
|
191 |
"rstrip": false,
|
|
|
193 |
"special": false
|
194 |
},
|
195 |
"57541": {
|
196 |
+
"content": "<s_total>",
|
197 |
"lstrip": false,
|
198 |
"normalized": true,
|
199 |
"rstrip": false,
|
|
|
201 |
"special": false
|
202 |
},
|
203 |
"57542": {
|
204 |
+
"content": "</s_total>",
|
205 |
"lstrip": false,
|
206 |
"normalized": true,
|
207 |
"rstrip": false,
|
|
|
225 |
"special": false
|
226 |
},
|
227 |
"57545": {
|
228 |
+
"content": "<s_discount>",
|
229 |
"lstrip": false,
|
230 |
"normalized": true,
|
231 |
"rstrip": false,
|
|
|
233 |
"special": false
|
234 |
},
|
235 |
"57546": {
|
236 |
+
"content": "</s_discount>",
|
237 |
"lstrip": false,
|
238 |
"normalized": true,
|
239 |
"rstrip": false,
|
|
|
241 |
"special": false
|
242 |
},
|
243 |
"57547": {
|
244 |
+
"content": "<s_line_items>",
|
245 |
"lstrip": false,
|
246 |
"normalized": true,
|
247 |
"rstrip": false,
|
|
|
249 |
"special": false
|
250 |
},
|
251 |
"57548": {
|
252 |
+
"content": "</s_line_items>",
|
253 |
"lstrip": false,
|
254 |
"normalized": true,
|
255 |
"rstrip": false,
|
|
|
257 |
"special": false
|
258 |
},
|
259 |
"57549": {
|
260 |
+
"content": "<s_item_key>",
|
261 |
"lstrip": false,
|
262 |
"normalized": true,
|
263 |
"rstrip": false,
|
|
|
265 |
"special": false
|
266 |
},
|
267 |
"57550": {
|
268 |
+
"content": "</s_item_key>",
|
269 |
"lstrip": false,
|
270 |
"normalized": true,
|
271 |
"rstrip": false,
|
|
|
273 |
"special": false
|
274 |
},
|
275 |
"57551": {
|
276 |
+
"content": "<s_item_name>",
|
277 |
"lstrip": false,
|
278 |
"normalized": true,
|
279 |
"rstrip": false,
|
|
|
281 |
"special": false
|
282 |
},
|
283 |
"57552": {
|
284 |
+
"content": "</s_item_name>",
|
285 |
"lstrip": false,
|
286 |
"normalized": true,
|
287 |
"rstrip": false,
|
|
|
289 |
"special": false
|
290 |
},
|
291 |
"57553": {
|
292 |
+
"content": "<s_item_value>",
|
293 |
"lstrip": false,
|
294 |
"normalized": true,
|
295 |
"rstrip": false,
|
|
|
297 |
"special": false
|
298 |
},
|
299 |
"57554": {
|
300 |
+
"content": "</s_item_value>",
|
301 |
"lstrip": false,
|
302 |
"normalized": true,
|
303 |
"rstrip": false,
|
|
|
305 |
"special": false
|
306 |
},
|
307 |
"57555": {
|
308 |
+
"content": "<s_item_quantity>",
|
309 |
+
"lstrip": false,
|
310 |
+
"normalized": true,
|
311 |
+
"rstrip": false,
|
312 |
+
"single_word": false,
|
313 |
+
"special": false
|
314 |
+
},
|
315 |
+
"57556": {
|
316 |
+
"content": "</s_item_quantity>",
|
317 |
+
"lstrip": false,
|
318 |
+
"normalized": true,
|
319 |
+
"rstrip": false,
|
320 |
+
"single_word": false,
|
321 |
+
"special": false
|
322 |
+
},
|
323 |
+
"57557": {
|
324 |
+
"content": "<s_receipt>",
|
325 |
+
"lstrip": false,
|
326 |
+
"normalized": true,
|
327 |
+
"rstrip": false,
|
328 |
+
"single_word": false,
|
329 |
+
"special": false
|
330 |
+
},
|
331 |
+
"57558": {
|
332 |
+
"content": "</s_receipt>",
|
333 |
"lstrip": false,
|
334 |
"normalized": true,
|
335 |
"rstrip": false,
|