adamadam111 commited on
Commit
7104cae
·
verified ·
1 Parent(s): 7bb8d33

Upload folder using huggingface_hub

Browse files
checkpoint-6000/config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LayoutLMv3ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "coordinate_size": 128,
9
+ "eos_token_id": 2,
10
+ "has_relative_attention_bias": true,
11
+ "has_spatial_attention_bias": true,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "form",
17
+ "1": "invoice",
18
+ "2": "budget",
19
+ "3": "file folder",
20
+ "4": "questionnaire"
21
+ },
22
+ "initializer_range": 0.02,
23
+ "input_size": 224,
24
+ "intermediate_size": 3072,
25
+ "label2id": {
26
+ "budget": 2,
27
+ "file folder": 3,
28
+ "form": 0,
29
+ "invoice": 1,
30
+ "questionnaire": 4
31
+ },
32
+ "layer_norm_eps": 1e-05,
33
+ "max_2d_position_embeddings": 1024,
34
+ "max_position_embeddings": 514,
35
+ "max_rel_2d_pos": 256,
36
+ "max_rel_pos": 128,
37
+ "model_type": "layoutlmv3",
38
+ "num_attention_heads": 12,
39
+ "num_channels": 3,
40
+ "num_hidden_layers": 12,
41
+ "pad_token_id": 1,
42
+ "patch_size": 16,
43
+ "problem_type": "single_label_classification",
44
+ "rel_2d_pos_bins": 64,
45
+ "rel_pos_bins": 32,
46
+ "second_input_size": 112,
47
+ "shape_size": 128,
48
+ "text_embed": true,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.53.0",
51
+ "type_vocab_size": 1,
52
+ "visual_embed": true,
53
+ "vocab_size": 50265
54
+ }
checkpoint-6000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:888897351408e17a05ef23c30ace359f70e9ee057f0063ad6bcb40fb5d2848d0
3
+ size 503711980
checkpoint-6000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17e334739af48f1b40f4a607df332d8c1c7625c7eeccf345fc8440f270eabac6
3
+ size 1007534138
checkpoint-6000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8f3367991c558e46e76dedf20f94300572e85c5cdf443ac1c5b98559a94124a
3
+ size 14244
checkpoint-6000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7278d15ce19a0ba738f99c206bd892d3d610b68d3519b0eb32c2dff22e2e790
3
+ size 988
checkpoint-6000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1deaa3298dfd895746f18f11ffb24ed906c98501dd64a733b2f362bb1957a467
3
+ size 1064
checkpoint-6000/trainer_state.json ADDED
@@ -0,0 +1,955 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 6000,
3
+ "best_metric": 0.9,
4
+ "best_model_checkpoint": "./layoutlmv3-docclass-finetuned/checkpoint-6000",
5
+ "epoch": 6.0,
6
+ "eval_steps": 500,
7
+ "global_step": 6000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.05,
14
+ "grad_norm": 27.367076873779297,
15
+ "learning_rate": 9.954e-06,
16
+ "loss": 0.7398,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.1,
21
+ "grad_norm": 81.9110336303711,
22
+ "learning_rate": 9.905000000000001e-06,
23
+ "loss": 0.5986,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.15,
28
+ "grad_norm": 8.801929473876953,
29
+ "learning_rate": 9.855000000000001e-06,
30
+ "loss": 0.5238,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.2,
35
+ "grad_norm": 3.4470932483673096,
36
+ "learning_rate": 9.805000000000002e-06,
37
+ "loss": 0.7008,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.25,
42
+ "grad_norm": 66.06997680664062,
43
+ "learning_rate": 9.755e-06,
44
+ "loss": 0.5785,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.3,
49
+ "grad_norm": 14.404702186584473,
50
+ "learning_rate": 9.705000000000001e-06,
51
+ "loss": 0.644,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.35,
56
+ "grad_norm": 45.596656799316406,
57
+ "learning_rate": 9.655000000000002e-06,
58
+ "loss": 0.5607,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.4,
63
+ "grad_norm": 17.859411239624023,
64
+ "learning_rate": 9.605e-06,
65
+ "loss": 0.6093,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.45,
70
+ "grad_norm": 72.80252075195312,
71
+ "learning_rate": 9.555e-06,
72
+ "loss": 0.7857,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.5,
77
+ "grad_norm": 70.62848663330078,
78
+ "learning_rate": 9.505000000000001e-06,
79
+ "loss": 0.5691,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.55,
84
+ "grad_norm": 32.26290512084961,
85
+ "learning_rate": 9.456e-06,
86
+ "loss": 0.5647,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 0.6,
91
+ "grad_norm": 0.2549276351928711,
92
+ "learning_rate": 9.406000000000001e-06,
93
+ "loss": 0.6434,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 0.65,
98
+ "grad_norm": 0.37813302874565125,
99
+ "learning_rate": 9.356e-06,
100
+ "loss": 0.6638,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 0.7,
105
+ "grad_norm": 78.35446166992188,
106
+ "learning_rate": 9.306e-06,
107
+ "loss": 0.6705,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 0.75,
112
+ "grad_norm": 54.32742691040039,
113
+ "learning_rate": 9.256e-06,
114
+ "loss": 0.5683,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 0.8,
119
+ "grad_norm": 0.2851010262966156,
120
+ "learning_rate": 9.206000000000001e-06,
121
+ "loss": 0.5826,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 0.85,
126
+ "grad_norm": 1.970667839050293,
127
+ "learning_rate": 9.156e-06,
128
+ "loss": 0.5985,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 0.9,
133
+ "grad_norm": 0.07515749335289001,
134
+ "learning_rate": 9.106e-06,
135
+ "loss": 0.5813,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 0.95,
140
+ "grad_norm": 26.057180404663086,
141
+ "learning_rate": 9.056000000000001e-06,
142
+ "loss": 0.4135,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 1.0,
147
+ "grad_norm": 0.34957027435302734,
148
+ "learning_rate": 9.006e-06,
149
+ "loss": 0.5918,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 1.0,
154
+ "eval_accuracy": 0.852,
155
+ "eval_f1": 0.8517868434541516,
156
+ "eval_loss": 0.6099042892456055,
157
+ "eval_precision": 0.8540426659530076,
158
+ "eval_recall": 0.852,
159
+ "eval_runtime": 109.9388,
160
+ "eval_samples_per_second": 9.096,
161
+ "eval_steps_per_second": 2.274,
162
+ "step": 1000
163
+ },
164
+ {
165
+ "epoch": 1.05,
166
+ "grad_norm": 37.85305404663086,
167
+ "learning_rate": 8.956e-06,
168
+ "loss": 0.4872,
169
+ "step": 1050
170
+ },
171
+ {
172
+ "epoch": 1.1,
173
+ "grad_norm": 9.535030364990234,
174
+ "learning_rate": 8.906e-06,
175
+ "loss": 0.3874,
176
+ "step": 1100
177
+ },
178
+ {
179
+ "epoch": 1.15,
180
+ "grad_norm": 0.1979241967201233,
181
+ "learning_rate": 8.856000000000001e-06,
182
+ "loss": 0.5971,
183
+ "step": 1150
184
+ },
185
+ {
186
+ "epoch": 1.2,
187
+ "grad_norm": 0.5306874513626099,
188
+ "learning_rate": 8.807e-06,
189
+ "loss": 0.4709,
190
+ "step": 1200
191
+ },
192
+ {
193
+ "epoch": 1.25,
194
+ "grad_norm": 18.236913681030273,
195
+ "learning_rate": 8.757000000000001e-06,
196
+ "loss": 0.472,
197
+ "step": 1250
198
+ },
199
+ {
200
+ "epoch": 1.3,
201
+ "grad_norm": 74.27545166015625,
202
+ "learning_rate": 8.707000000000002e-06,
203
+ "loss": 0.4007,
204
+ "step": 1300
205
+ },
206
+ {
207
+ "epoch": 1.35,
208
+ "grad_norm": 6.827157974243164,
209
+ "learning_rate": 8.657e-06,
210
+ "loss": 0.4139,
211
+ "step": 1350
212
+ },
213
+ {
214
+ "epoch": 1.4,
215
+ "grad_norm": 0.10050762444734573,
216
+ "learning_rate": 8.607e-06,
217
+ "loss": 0.5254,
218
+ "step": 1400
219
+ },
220
+ {
221
+ "epoch": 1.45,
222
+ "grad_norm": 0.05092502385377884,
223
+ "learning_rate": 8.557000000000001e-06,
224
+ "loss": 0.359,
225
+ "step": 1450
226
+ },
227
+ {
228
+ "epoch": 1.5,
229
+ "grad_norm": 14.49219036102295,
230
+ "learning_rate": 8.507e-06,
231
+ "loss": 0.4157,
232
+ "step": 1500
233
+ },
234
+ {
235
+ "epoch": 1.55,
236
+ "grad_norm": 92.9802017211914,
237
+ "learning_rate": 8.457e-06,
238
+ "loss": 0.558,
239
+ "step": 1550
240
+ },
241
+ {
242
+ "epoch": 1.6,
243
+ "grad_norm": 76.93692016601562,
244
+ "learning_rate": 8.407000000000001e-06,
245
+ "loss": 0.4579,
246
+ "step": 1600
247
+ },
248
+ {
249
+ "epoch": 1.65,
250
+ "grad_norm": 0.22004538774490356,
251
+ "learning_rate": 8.357000000000001e-06,
252
+ "loss": 0.6244,
253
+ "step": 1650
254
+ },
255
+ {
256
+ "epoch": 1.7,
257
+ "grad_norm": 0.06912333518266678,
258
+ "learning_rate": 8.307e-06,
259
+ "loss": 0.5163,
260
+ "step": 1700
261
+ },
262
+ {
263
+ "epoch": 1.75,
264
+ "grad_norm": 1.3707654476165771,
265
+ "learning_rate": 8.257e-06,
266
+ "loss": 0.4181,
267
+ "step": 1750
268
+ },
269
+ {
270
+ "epoch": 1.8,
271
+ "grad_norm": 0.1415724754333496,
272
+ "learning_rate": 8.207000000000001e-06,
273
+ "loss": 0.6508,
274
+ "step": 1800
275
+ },
276
+ {
277
+ "epoch": 1.85,
278
+ "grad_norm": 21.731523513793945,
279
+ "learning_rate": 8.157e-06,
280
+ "loss": 0.4782,
281
+ "step": 1850
282
+ },
283
+ {
284
+ "epoch": 1.9,
285
+ "grad_norm": 0.45240655541419983,
286
+ "learning_rate": 8.107e-06,
287
+ "loss": 0.5325,
288
+ "step": 1900
289
+ },
290
+ {
291
+ "epoch": 1.95,
292
+ "grad_norm": 9.7157564163208,
293
+ "learning_rate": 8.057000000000001e-06,
294
+ "loss": 0.4879,
295
+ "step": 1950
296
+ },
297
+ {
298
+ "epoch": 2.0,
299
+ "grad_norm": 61.20124816894531,
300
+ "learning_rate": 8.007e-06,
301
+ "loss": 0.5759,
302
+ "step": 2000
303
+ },
304
+ {
305
+ "epoch": 2.0,
306
+ "eval_accuracy": 0.871,
307
+ "eval_f1": 0.8709149644475538,
308
+ "eval_loss": 0.6178536415100098,
309
+ "eval_precision": 0.8759130395565256,
310
+ "eval_recall": 0.871,
311
+ "eval_runtime": 111.0284,
312
+ "eval_samples_per_second": 9.007,
313
+ "eval_steps_per_second": 2.252,
314
+ "step": 2000
315
+ },
316
+ {
317
+ "epoch": 2.05,
318
+ "grad_norm": 189.5020751953125,
319
+ "learning_rate": 7.957e-06,
320
+ "loss": 0.3242,
321
+ "step": 2050
322
+ },
323
+ {
324
+ "epoch": 2.1,
325
+ "grad_norm": 0.06589367240667343,
326
+ "learning_rate": 7.907e-06,
327
+ "loss": 0.3381,
328
+ "step": 2100
329
+ },
330
+ {
331
+ "epoch": 2.15,
332
+ "grad_norm": 6.254538536071777,
333
+ "learning_rate": 7.857e-06,
334
+ "loss": 0.2586,
335
+ "step": 2150
336
+ },
337
+ {
338
+ "epoch": 2.2,
339
+ "grad_norm": 0.035582080483436584,
340
+ "learning_rate": 7.807e-06,
341
+ "loss": 0.3091,
342
+ "step": 2200
343
+ },
344
+ {
345
+ "epoch": 2.25,
346
+ "grad_norm": 0.17389701306819916,
347
+ "learning_rate": 7.757e-06,
348
+ "loss": 0.5237,
349
+ "step": 2250
350
+ },
351
+ {
352
+ "epoch": 2.3,
353
+ "grad_norm": 33.96052169799805,
354
+ "learning_rate": 7.707000000000001e-06,
355
+ "loss": 0.4895,
356
+ "step": 2300
357
+ },
358
+ {
359
+ "epoch": 2.35,
360
+ "grad_norm": 0.2390211671590805,
361
+ "learning_rate": 7.657000000000001e-06,
362
+ "loss": 0.3044,
363
+ "step": 2350
364
+ },
365
+ {
366
+ "epoch": 2.4,
367
+ "grad_norm": 0.29711782932281494,
368
+ "learning_rate": 7.607000000000001e-06,
369
+ "loss": 0.3455,
370
+ "step": 2400
371
+ },
372
+ {
373
+ "epoch": 2.45,
374
+ "grad_norm": 0.026691796258091927,
375
+ "learning_rate": 7.557000000000001e-06,
376
+ "loss": 0.4515,
377
+ "step": 2450
378
+ },
379
+ {
380
+ "epoch": 2.5,
381
+ "grad_norm": 0.12731046974658966,
382
+ "learning_rate": 7.507000000000001e-06,
383
+ "loss": 0.3598,
384
+ "step": 2500
385
+ },
386
+ {
387
+ "epoch": 2.55,
388
+ "grad_norm": 0.042913321405649185,
389
+ "learning_rate": 7.457000000000001e-06,
390
+ "loss": 0.2815,
391
+ "step": 2550
392
+ },
393
+ {
394
+ "epoch": 2.6,
395
+ "grad_norm": 0.06428048014640808,
396
+ "learning_rate": 7.407000000000001e-06,
397
+ "loss": 0.3822,
398
+ "step": 2600
399
+ },
400
+ {
401
+ "epoch": 2.65,
402
+ "grad_norm": 0.22337721288204193,
403
+ "learning_rate": 7.357000000000001e-06,
404
+ "loss": 0.217,
405
+ "step": 2650
406
+ },
407
+ {
408
+ "epoch": 2.7,
409
+ "grad_norm": 0.5337192416191101,
410
+ "learning_rate": 7.3070000000000005e-06,
411
+ "loss": 0.4874,
412
+ "step": 2700
413
+ },
414
+ {
415
+ "epoch": 2.75,
416
+ "grad_norm": 0.03206096589565277,
417
+ "learning_rate": 7.257000000000001e-06,
418
+ "loss": 0.4564,
419
+ "step": 2750
420
+ },
421
+ {
422
+ "epoch": 2.8,
423
+ "grad_norm": 346.4653625488281,
424
+ "learning_rate": 7.207000000000001e-06,
425
+ "loss": 0.2794,
426
+ "step": 2800
427
+ },
428
+ {
429
+ "epoch": 2.85,
430
+ "grad_norm": 39.87053298950195,
431
+ "learning_rate": 7.157e-06,
432
+ "loss": 0.4392,
433
+ "step": 2850
434
+ },
435
+ {
436
+ "epoch": 2.9,
437
+ "grad_norm": 0.0478862039744854,
438
+ "learning_rate": 7.107000000000001e-06,
439
+ "loss": 0.2627,
440
+ "step": 2900
441
+ },
442
+ {
443
+ "epoch": 2.95,
444
+ "grad_norm": 0.04880313202738762,
445
+ "learning_rate": 7.057e-06,
446
+ "loss": 0.1502,
447
+ "step": 2950
448
+ },
449
+ {
450
+ "epoch": 3.0,
451
+ "grad_norm": 0.8579753041267395,
452
+ "learning_rate": 7.007000000000001e-06,
453
+ "loss": 0.3137,
454
+ "step": 3000
455
+ },
456
+ {
457
+ "epoch": 3.0,
458
+ "eval_accuracy": 0.858,
459
+ "eval_f1": 0.8568902628142059,
460
+ "eval_loss": 0.8766486048698425,
461
+ "eval_precision": 0.869630225735872,
462
+ "eval_recall": 0.858,
463
+ "eval_runtime": 112.5121,
464
+ "eval_samples_per_second": 8.888,
465
+ "eval_steps_per_second": 2.222,
466
+ "step": 3000
467
+ },
468
+ {
469
+ "epoch": 3.05,
470
+ "grad_norm": 0.09161412715911865,
471
+ "learning_rate": 6.9570000000000005e-06,
472
+ "loss": 0.2441,
473
+ "step": 3050
474
+ },
475
+ {
476
+ "epoch": 3.1,
477
+ "grad_norm": 285.05072021484375,
478
+ "learning_rate": 6.907e-06,
479
+ "loss": 0.4021,
480
+ "step": 3100
481
+ },
482
+ {
483
+ "epoch": 3.15,
484
+ "grad_norm": 0.03876444697380066,
485
+ "learning_rate": 6.857000000000001e-06,
486
+ "loss": 0.2032,
487
+ "step": 3150
488
+ },
489
+ {
490
+ "epoch": 3.2,
491
+ "grad_norm": 9.66369342803955,
492
+ "learning_rate": 6.807e-06,
493
+ "loss": 0.1972,
494
+ "step": 3200
495
+ },
496
+ {
497
+ "epoch": 3.25,
498
+ "grad_norm": 0.06236199662089348,
499
+ "learning_rate": 6.757e-06,
500
+ "loss": 0.4157,
501
+ "step": 3250
502
+ },
503
+ {
504
+ "epoch": 3.3,
505
+ "grad_norm": 2.1778385639190674,
506
+ "learning_rate": 6.707e-06,
507
+ "loss": 0.3814,
508
+ "step": 3300
509
+ },
510
+ {
511
+ "epoch": 3.35,
512
+ "grad_norm": 0.02144041284918785,
513
+ "learning_rate": 6.657e-06,
514
+ "loss": 0.1562,
515
+ "step": 3350
516
+ },
517
+ {
518
+ "epoch": 3.4,
519
+ "grad_norm": 0.4679206311702728,
520
+ "learning_rate": 6.6070000000000004e-06,
521
+ "loss": 0.2698,
522
+ "step": 3400
523
+ },
524
+ {
525
+ "epoch": 3.45,
526
+ "grad_norm": 0.019845549017190933,
527
+ "learning_rate": 6.557e-06,
528
+ "loss": 0.2024,
529
+ "step": 3450
530
+ },
531
+ {
532
+ "epoch": 3.5,
533
+ "grad_norm": 0.03582359477877617,
534
+ "learning_rate": 6.507e-06,
535
+ "loss": 0.1805,
536
+ "step": 3500
537
+ },
538
+ {
539
+ "epoch": 3.55,
540
+ "grad_norm": 0.012999696657061577,
541
+ "learning_rate": 6.457000000000001e-06,
542
+ "loss": 0.2242,
543
+ "step": 3550
544
+ },
545
+ {
546
+ "epoch": 3.6,
547
+ "grad_norm": 61.022212982177734,
548
+ "learning_rate": 6.407000000000001e-06,
549
+ "loss": 0.2972,
550
+ "step": 3600
551
+ },
552
+ {
553
+ "epoch": 3.65,
554
+ "grad_norm": 0.17647244036197662,
555
+ "learning_rate": 6.357000000000001e-06,
556
+ "loss": 0.2507,
557
+ "step": 3650
558
+ },
559
+ {
560
+ "epoch": 3.7,
561
+ "grad_norm": 0.02148498222231865,
562
+ "learning_rate": 6.307000000000001e-06,
563
+ "loss": 0.3322,
564
+ "step": 3700
565
+ },
566
+ {
567
+ "epoch": 3.75,
568
+ "grad_norm": 8.956069946289062,
569
+ "learning_rate": 6.257000000000001e-06,
570
+ "loss": 0.2174,
571
+ "step": 3750
572
+ },
573
+ {
574
+ "epoch": 3.8,
575
+ "grad_norm": 0.019011685624718666,
576
+ "learning_rate": 6.2080000000000005e-06,
577
+ "loss": 0.2118,
578
+ "step": 3800
579
+ },
580
+ {
581
+ "epoch": 3.85,
582
+ "grad_norm": 0.01807810179889202,
583
+ "learning_rate": 6.158000000000001e-06,
584
+ "loss": 0.1908,
585
+ "step": 3850
586
+ },
587
+ {
588
+ "epoch": 3.9,
589
+ "grad_norm": 0.02416962757706642,
590
+ "learning_rate": 6.108000000000001e-06,
591
+ "loss": 0.1491,
592
+ "step": 3900
593
+ },
594
+ {
595
+ "epoch": 3.95,
596
+ "grad_norm": 19.60466766357422,
597
+ "learning_rate": 6.058e-06,
598
+ "loss": 0.2031,
599
+ "step": 3950
600
+ },
601
+ {
602
+ "epoch": 4.0,
603
+ "grad_norm": 0.21647094190120697,
604
+ "learning_rate": 6.008000000000001e-06,
605
+ "loss": 0.2952,
606
+ "step": 4000
607
+ },
608
+ {
609
+ "epoch": 4.0,
610
+ "eval_accuracy": 0.873,
611
+ "eval_f1": 0.872977072862821,
612
+ "eval_loss": 0.821958065032959,
613
+ "eval_precision": 0.8834745892873105,
614
+ "eval_recall": 0.873,
615
+ "eval_runtime": 112.9225,
616
+ "eval_samples_per_second": 8.856,
617
+ "eval_steps_per_second": 2.214,
618
+ "step": 4000
619
+ },
620
+ {
621
+ "epoch": 4.05,
622
+ "grad_norm": 0.08491307497024536,
623
+ "learning_rate": 5.958e-06,
624
+ "loss": 0.1665,
625
+ "step": 4050
626
+ },
627
+ {
628
+ "epoch": 4.1,
629
+ "grad_norm": 0.10018359869718552,
630
+ "learning_rate": 5.908e-06,
631
+ "loss": 0.2381,
632
+ "step": 4100
633
+ },
634
+ {
635
+ "epoch": 4.15,
636
+ "grad_norm": 0.01517445407807827,
637
+ "learning_rate": 5.8580000000000005e-06,
638
+ "loss": 0.1407,
639
+ "step": 4150
640
+ },
641
+ {
642
+ "epoch": 4.2,
643
+ "grad_norm": 0.2072858363389969,
644
+ "learning_rate": 5.808e-06,
645
+ "loss": 0.2508,
646
+ "step": 4200
647
+ },
648
+ {
649
+ "epoch": 4.25,
650
+ "grad_norm": 76.36699676513672,
651
+ "learning_rate": 5.758000000000001e-06,
652
+ "loss": 0.2175,
653
+ "step": 4250
654
+ },
655
+ {
656
+ "epoch": 4.3,
657
+ "grad_norm": 0.0169328935444355,
658
+ "learning_rate": 5.708e-06,
659
+ "loss": 0.1878,
660
+ "step": 4300
661
+ },
662
+ {
663
+ "epoch": 4.35,
664
+ "grad_norm": 23.842693328857422,
665
+ "learning_rate": 5.658e-06,
666
+ "loss": 0.1266,
667
+ "step": 4350
668
+ },
669
+ {
670
+ "epoch": 4.4,
671
+ "grad_norm": 233.6339111328125,
672
+ "learning_rate": 5.608e-06,
673
+ "loss": 0.1279,
674
+ "step": 4400
675
+ },
676
+ {
677
+ "epoch": 4.45,
678
+ "grad_norm": 173.3831024169922,
679
+ "learning_rate": 5.558e-06,
680
+ "loss": 0.1829,
681
+ "step": 4450
682
+ },
683
+ {
684
+ "epoch": 4.5,
685
+ "grad_norm": 0.08392331004142761,
686
+ "learning_rate": 5.508e-06,
687
+ "loss": 0.1668,
688
+ "step": 4500
689
+ },
690
+ {
691
+ "epoch": 4.55,
692
+ "grad_norm": 0.020959168672561646,
693
+ "learning_rate": 5.458e-06,
694
+ "loss": 0.0815,
695
+ "step": 4550
696
+ },
697
+ {
698
+ "epoch": 4.6,
699
+ "grad_norm": 225.8976593017578,
700
+ "learning_rate": 5.408e-06,
701
+ "loss": 0.2346,
702
+ "step": 4600
703
+ },
704
+ {
705
+ "epoch": 4.65,
706
+ "grad_norm": 358.5092468261719,
707
+ "learning_rate": 5.358000000000001e-06,
708
+ "loss": 0.1021,
709
+ "step": 4650
710
+ },
711
+ {
712
+ "epoch": 4.7,
713
+ "grad_norm": 165.55499267578125,
714
+ "learning_rate": 5.308000000000001e-06,
715
+ "loss": 0.2577,
716
+ "step": 4700
717
+ },
718
+ {
719
+ "epoch": 4.75,
720
+ "grad_norm": 0.11646530777215958,
721
+ "learning_rate": 5.258000000000001e-06,
722
+ "loss": 0.2029,
723
+ "step": 4750
724
+ },
725
+ {
726
+ "epoch": 4.8,
727
+ "grad_norm": 0.03793807700276375,
728
+ "learning_rate": 5.208000000000001e-06,
729
+ "loss": 0.1465,
730
+ "step": 4800
731
+ },
732
+ {
733
+ "epoch": 4.85,
734
+ "grad_norm": 113.57576751708984,
735
+ "learning_rate": 5.158e-06,
736
+ "loss": 0.1906,
737
+ "step": 4850
738
+ },
739
+ {
740
+ "epoch": 4.9,
741
+ "grad_norm": 0.011114823631942272,
742
+ "learning_rate": 5.108000000000001e-06,
743
+ "loss": 0.1876,
744
+ "step": 4900
745
+ },
746
+ {
747
+ "epoch": 4.95,
748
+ "grad_norm": 77.72245025634766,
749
+ "learning_rate": 5.0580000000000005e-06,
750
+ "loss": 0.316,
751
+ "step": 4950
752
+ },
753
+ {
754
+ "epoch": 5.0,
755
+ "grad_norm": 0.02318619005382061,
756
+ "learning_rate": 5.008000000000001e-06,
757
+ "loss": 0.1035,
758
+ "step": 5000
759
+ },
760
+ {
761
+ "epoch": 5.0,
762
+ "eval_accuracy": 0.884,
763
+ "eval_f1": 0.8836532680104157,
764
+ "eval_loss": 0.7644630074501038,
765
+ "eval_precision": 0.8889154581905352,
766
+ "eval_recall": 0.884,
767
+ "eval_runtime": 111.6762,
768
+ "eval_samples_per_second": 8.954,
769
+ "eval_steps_per_second": 2.239,
770
+ "step": 5000
771
+ },
772
+ {
773
+ "epoch": 5.05,
774
+ "grad_norm": 0.03648075461387634,
775
+ "learning_rate": 4.958000000000001e-06,
776
+ "loss": 0.0445,
777
+ "step": 5050
778
+ },
779
+ {
780
+ "epoch": 5.1,
781
+ "grad_norm": 0.014485549181699753,
782
+ "learning_rate": 4.908e-06,
783
+ "loss": 0.0928,
784
+ "step": 5100
785
+ },
786
+ {
787
+ "epoch": 5.15,
788
+ "grad_norm": 0.019193926826119423,
789
+ "learning_rate": 4.858000000000001e-06,
790
+ "loss": 0.1156,
791
+ "step": 5150
792
+ },
793
+ {
794
+ "epoch": 5.2,
795
+ "grad_norm": 0.007525489665567875,
796
+ "learning_rate": 4.808e-06,
797
+ "loss": 0.1103,
798
+ "step": 5200
799
+ },
800
+ {
801
+ "epoch": 5.25,
802
+ "grad_norm": 0.02716565877199173,
803
+ "learning_rate": 4.758e-06,
804
+ "loss": 0.1653,
805
+ "step": 5250
806
+ },
807
+ {
808
+ "epoch": 5.3,
809
+ "grad_norm": 0.02299477905035019,
810
+ "learning_rate": 4.7080000000000005e-06,
811
+ "loss": 0.1915,
812
+ "step": 5300
813
+ },
814
+ {
815
+ "epoch": 5.35,
816
+ "grad_norm": 447.5741271972656,
817
+ "learning_rate": 4.658e-06,
818
+ "loss": 0.0903,
819
+ "step": 5350
820
+ },
821
+ {
822
+ "epoch": 5.4,
823
+ "grad_norm": 405.6177062988281,
824
+ "learning_rate": 4.608000000000001e-06,
825
+ "loss": 0.1506,
826
+ "step": 5400
827
+ },
828
+ {
829
+ "epoch": 5.45,
830
+ "grad_norm": 6.23228645324707,
831
+ "learning_rate": 4.558e-06,
832
+ "loss": 0.1342,
833
+ "step": 5450
834
+ },
835
+ {
836
+ "epoch": 5.5,
837
+ "grad_norm": 0.047868575900793076,
838
+ "learning_rate": 4.508e-06,
839
+ "loss": 0.1799,
840
+ "step": 5500
841
+ },
842
+ {
843
+ "epoch": 5.55,
844
+ "grad_norm": 115.90292358398438,
845
+ "learning_rate": 4.458e-06,
846
+ "loss": 0.0751,
847
+ "step": 5550
848
+ },
849
+ {
850
+ "epoch": 5.6,
851
+ "grad_norm": 0.1427452564239502,
852
+ "learning_rate": 4.408000000000001e-06,
853
+ "loss": 0.1132,
854
+ "step": 5600
855
+ },
856
+ {
857
+ "epoch": 5.65,
858
+ "grad_norm": 0.03756200522184372,
859
+ "learning_rate": 4.3580000000000005e-06,
860
+ "loss": 0.0219,
861
+ "step": 5650
862
+ },
863
+ {
864
+ "epoch": 5.7,
865
+ "grad_norm": 0.1948777437210083,
866
+ "learning_rate": 4.308000000000001e-06,
867
+ "loss": 0.1625,
868
+ "step": 5700
869
+ },
870
+ {
871
+ "epoch": 5.75,
872
+ "grad_norm": 0.14967912435531616,
873
+ "learning_rate": 4.2580000000000006e-06,
874
+ "loss": 0.1131,
875
+ "step": 5750
876
+ },
877
+ {
878
+ "epoch": 5.8,
879
+ "grad_norm": 2.684386730194092,
880
+ "learning_rate": 4.208e-06,
881
+ "loss": 0.1085,
882
+ "step": 5800
883
+ },
884
+ {
885
+ "epoch": 5.85,
886
+ "grad_norm": 0.037221621721982956,
887
+ "learning_rate": 4.158000000000001e-06,
888
+ "loss": 0.1957,
889
+ "step": 5850
890
+ },
891
+ {
892
+ "epoch": 5.9,
893
+ "grad_norm": 0.014206652529537678,
894
+ "learning_rate": 4.108e-06,
895
+ "loss": 0.0228,
896
+ "step": 5900
897
+ },
898
+ {
899
+ "epoch": 5.95,
900
+ "grad_norm": 0.008729885332286358,
901
+ "learning_rate": 4.058e-06,
902
+ "loss": 0.0421,
903
+ "step": 5950
904
+ },
905
+ {
906
+ "epoch": 6.0,
907
+ "grad_norm": 253.62225341796875,
908
+ "learning_rate": 4.008e-06,
909
+ "loss": 0.0865,
910
+ "step": 6000
911
+ },
912
+ {
913
+ "epoch": 6.0,
914
+ "eval_accuracy": 0.9,
915
+ "eval_f1": 0.9000352608410468,
916
+ "eval_loss": 0.7469241619110107,
917
+ "eval_precision": 0.9014304934175754,
918
+ "eval_recall": 0.9,
919
+ "eval_runtime": 112.6162,
920
+ "eval_samples_per_second": 8.88,
921
+ "eval_steps_per_second": 2.22,
922
+ "step": 6000
923
+ }
924
+ ],
925
+ "logging_steps": 50,
926
+ "max_steps": 10000,
927
+ "num_input_tokens_seen": 0,
928
+ "num_train_epochs": 10,
929
+ "save_steps": 500,
930
+ "stateful_callbacks": {
931
+ "EarlyStoppingCallback": {
932
+ "args": {
933
+ "early_stopping_patience": 2,
934
+ "early_stopping_threshold": 0.0
935
+ },
936
+ "attributes": {
937
+ "early_stopping_patience_counter": 0
938
+ }
939
+ },
940
+ "TrainerControl": {
941
+ "args": {
942
+ "should_epoch_stop": false,
943
+ "should_evaluate": false,
944
+ "should_log": false,
945
+ "should_save": true,
946
+ "should_training_stop": false
947
+ },
948
+ "attributes": {}
949
+ }
950
+ },
951
+ "total_flos": 6369958010880000.0,
952
+ "train_batch_size": 4,
953
+ "trial_name": null,
954
+ "trial_params": null
955
+ }
checkpoint-6000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f403144661d43c8a35a0a64782c635300fe17c4e900c1c22a3d8255e6ad2018
3
+ size 5304
checkpoint-8000/config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LayoutLMv3ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "coordinate_size": 128,
9
+ "eos_token_id": 2,
10
+ "has_relative_attention_bias": true,
11
+ "has_spatial_attention_bias": true,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 768,
15
+ "id2label": {
16
+ "0": "form",
17
+ "1": "invoice",
18
+ "2": "budget",
19
+ "3": "file folder",
20
+ "4": "questionnaire"
21
+ },
22
+ "initializer_range": 0.02,
23
+ "input_size": 224,
24
+ "intermediate_size": 3072,
25
+ "label2id": {
26
+ "budget": 2,
27
+ "file folder": 3,
28
+ "form": 0,
29
+ "invoice": 1,
30
+ "questionnaire": 4
31
+ },
32
+ "layer_norm_eps": 1e-05,
33
+ "max_2d_position_embeddings": 1024,
34
+ "max_position_embeddings": 514,
35
+ "max_rel_2d_pos": 256,
36
+ "max_rel_pos": 128,
37
+ "model_type": "layoutlmv3",
38
+ "num_attention_heads": 12,
39
+ "num_channels": 3,
40
+ "num_hidden_layers": 12,
41
+ "pad_token_id": 1,
42
+ "patch_size": 16,
43
+ "problem_type": "single_label_classification",
44
+ "rel_2d_pos_bins": 64,
45
+ "rel_pos_bins": 32,
46
+ "second_input_size": 112,
47
+ "shape_size": 128,
48
+ "text_embed": true,
49
+ "torch_dtype": "float32",
50
+ "transformers_version": "4.53.0",
51
+ "type_vocab_size": 1,
52
+ "visual_embed": true,
53
+ "vocab_size": 50265
54
+ }
checkpoint-8000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec6e1c6a5e7d8604e909f32bfbaeb0e40ba98946c5f5420cea3b0f4fca9df204
3
+ size 503711980
checkpoint-8000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d4694725821e140611ef97aa16f4ad915981e95f0aece165acbcfe28922267c
3
+ size 1007534138
checkpoint-8000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15f1326fd11903c8d72d967d58b235578cd2563695ef229f25c2022ef6a2432a
3
+ size 14244
checkpoint-8000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95c585bdc83f0924a7abd8d4c3964b0325ee0eff5a0c7c32476a6491bf0cccce
3
+ size 988
checkpoint-8000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd7aeb1ad1841b89f23b8cc54996197bdfcfbe9886c62934aac69a735621208c
3
+ size 1064
checkpoint-8000/trainer_state.json ADDED
@@ -0,0 +1,1259 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 6000,
3
+ "best_metric": 0.9,
4
+ "best_model_checkpoint": "./layoutlmv3-docclass-finetuned/checkpoint-6000",
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 8000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.05,
14
+ "grad_norm": 27.367076873779297,
15
+ "learning_rate": 9.954e-06,
16
+ "loss": 0.7398,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.1,
21
+ "grad_norm": 81.9110336303711,
22
+ "learning_rate": 9.905000000000001e-06,
23
+ "loss": 0.5986,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.15,
28
+ "grad_norm": 8.801929473876953,
29
+ "learning_rate": 9.855000000000001e-06,
30
+ "loss": 0.5238,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.2,
35
+ "grad_norm": 3.4470932483673096,
36
+ "learning_rate": 9.805000000000002e-06,
37
+ "loss": 0.7008,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.25,
42
+ "grad_norm": 66.06997680664062,
43
+ "learning_rate": 9.755e-06,
44
+ "loss": 0.5785,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.3,
49
+ "grad_norm": 14.404702186584473,
50
+ "learning_rate": 9.705000000000001e-06,
51
+ "loss": 0.644,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.35,
56
+ "grad_norm": 45.596656799316406,
57
+ "learning_rate": 9.655000000000002e-06,
58
+ "loss": 0.5607,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.4,
63
+ "grad_norm": 17.859411239624023,
64
+ "learning_rate": 9.605e-06,
65
+ "loss": 0.6093,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.45,
70
+ "grad_norm": 72.80252075195312,
71
+ "learning_rate": 9.555e-06,
72
+ "loss": 0.7857,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.5,
77
+ "grad_norm": 70.62848663330078,
78
+ "learning_rate": 9.505000000000001e-06,
79
+ "loss": 0.5691,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.55,
84
+ "grad_norm": 32.26290512084961,
85
+ "learning_rate": 9.456e-06,
86
+ "loss": 0.5647,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 0.6,
91
+ "grad_norm": 0.2549276351928711,
92
+ "learning_rate": 9.406000000000001e-06,
93
+ "loss": 0.6434,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 0.65,
98
+ "grad_norm": 0.37813302874565125,
99
+ "learning_rate": 9.356e-06,
100
+ "loss": 0.6638,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 0.7,
105
+ "grad_norm": 78.35446166992188,
106
+ "learning_rate": 9.306e-06,
107
+ "loss": 0.6705,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 0.75,
112
+ "grad_norm": 54.32742691040039,
113
+ "learning_rate": 9.256e-06,
114
+ "loss": 0.5683,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 0.8,
119
+ "grad_norm": 0.2851010262966156,
120
+ "learning_rate": 9.206000000000001e-06,
121
+ "loss": 0.5826,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 0.85,
126
+ "grad_norm": 1.970667839050293,
127
+ "learning_rate": 9.156e-06,
128
+ "loss": 0.5985,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 0.9,
133
+ "grad_norm": 0.07515749335289001,
134
+ "learning_rate": 9.106e-06,
135
+ "loss": 0.5813,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 0.95,
140
+ "grad_norm": 26.057180404663086,
141
+ "learning_rate": 9.056000000000001e-06,
142
+ "loss": 0.4135,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 1.0,
147
+ "grad_norm": 0.34957027435302734,
148
+ "learning_rate": 9.006e-06,
149
+ "loss": 0.5918,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 1.0,
154
+ "eval_accuracy": 0.852,
155
+ "eval_f1": 0.8517868434541516,
156
+ "eval_loss": 0.6099042892456055,
157
+ "eval_precision": 0.8540426659530076,
158
+ "eval_recall": 0.852,
159
+ "eval_runtime": 109.9388,
160
+ "eval_samples_per_second": 9.096,
161
+ "eval_steps_per_second": 2.274,
162
+ "step": 1000
163
+ },
164
+ {
165
+ "epoch": 1.05,
166
+ "grad_norm": 37.85305404663086,
167
+ "learning_rate": 8.956e-06,
168
+ "loss": 0.4872,
169
+ "step": 1050
170
+ },
171
+ {
172
+ "epoch": 1.1,
173
+ "grad_norm": 9.535030364990234,
174
+ "learning_rate": 8.906e-06,
175
+ "loss": 0.3874,
176
+ "step": 1100
177
+ },
178
+ {
179
+ "epoch": 1.15,
180
+ "grad_norm": 0.1979241967201233,
181
+ "learning_rate": 8.856000000000001e-06,
182
+ "loss": 0.5971,
183
+ "step": 1150
184
+ },
185
+ {
186
+ "epoch": 1.2,
187
+ "grad_norm": 0.5306874513626099,
188
+ "learning_rate": 8.807e-06,
189
+ "loss": 0.4709,
190
+ "step": 1200
191
+ },
192
+ {
193
+ "epoch": 1.25,
194
+ "grad_norm": 18.236913681030273,
195
+ "learning_rate": 8.757000000000001e-06,
196
+ "loss": 0.472,
197
+ "step": 1250
198
+ },
199
+ {
200
+ "epoch": 1.3,
201
+ "grad_norm": 74.27545166015625,
202
+ "learning_rate": 8.707000000000002e-06,
203
+ "loss": 0.4007,
204
+ "step": 1300
205
+ },
206
+ {
207
+ "epoch": 1.35,
208
+ "grad_norm": 6.827157974243164,
209
+ "learning_rate": 8.657e-06,
210
+ "loss": 0.4139,
211
+ "step": 1350
212
+ },
213
+ {
214
+ "epoch": 1.4,
215
+ "grad_norm": 0.10050762444734573,
216
+ "learning_rate": 8.607e-06,
217
+ "loss": 0.5254,
218
+ "step": 1400
219
+ },
220
+ {
221
+ "epoch": 1.45,
222
+ "grad_norm": 0.05092502385377884,
223
+ "learning_rate": 8.557000000000001e-06,
224
+ "loss": 0.359,
225
+ "step": 1450
226
+ },
227
+ {
228
+ "epoch": 1.5,
229
+ "grad_norm": 14.49219036102295,
230
+ "learning_rate": 8.507e-06,
231
+ "loss": 0.4157,
232
+ "step": 1500
233
+ },
234
+ {
235
+ "epoch": 1.55,
236
+ "grad_norm": 92.9802017211914,
237
+ "learning_rate": 8.457e-06,
238
+ "loss": 0.558,
239
+ "step": 1550
240
+ },
241
+ {
242
+ "epoch": 1.6,
243
+ "grad_norm": 76.93692016601562,
244
+ "learning_rate": 8.407000000000001e-06,
245
+ "loss": 0.4579,
246
+ "step": 1600
247
+ },
248
+ {
249
+ "epoch": 1.65,
250
+ "grad_norm": 0.22004538774490356,
251
+ "learning_rate": 8.357000000000001e-06,
252
+ "loss": 0.6244,
253
+ "step": 1650
254
+ },
255
+ {
256
+ "epoch": 1.7,
257
+ "grad_norm": 0.06912333518266678,
258
+ "learning_rate": 8.307e-06,
259
+ "loss": 0.5163,
260
+ "step": 1700
261
+ },
262
+ {
263
+ "epoch": 1.75,
264
+ "grad_norm": 1.3707654476165771,
265
+ "learning_rate": 8.257e-06,
266
+ "loss": 0.4181,
267
+ "step": 1750
268
+ },
269
+ {
270
+ "epoch": 1.8,
271
+ "grad_norm": 0.1415724754333496,
272
+ "learning_rate": 8.207000000000001e-06,
273
+ "loss": 0.6508,
274
+ "step": 1800
275
+ },
276
+ {
277
+ "epoch": 1.85,
278
+ "grad_norm": 21.731523513793945,
279
+ "learning_rate": 8.157e-06,
280
+ "loss": 0.4782,
281
+ "step": 1850
282
+ },
283
+ {
284
+ "epoch": 1.9,
285
+ "grad_norm": 0.45240655541419983,
286
+ "learning_rate": 8.107e-06,
287
+ "loss": 0.5325,
288
+ "step": 1900
289
+ },
290
+ {
291
+ "epoch": 1.95,
292
+ "grad_norm": 9.7157564163208,
293
+ "learning_rate": 8.057000000000001e-06,
294
+ "loss": 0.4879,
295
+ "step": 1950
296
+ },
297
+ {
298
+ "epoch": 2.0,
299
+ "grad_norm": 61.20124816894531,
300
+ "learning_rate": 8.007e-06,
301
+ "loss": 0.5759,
302
+ "step": 2000
303
+ },
304
+ {
305
+ "epoch": 2.0,
306
+ "eval_accuracy": 0.871,
307
+ "eval_f1": 0.8709149644475538,
308
+ "eval_loss": 0.6178536415100098,
309
+ "eval_precision": 0.8759130395565256,
310
+ "eval_recall": 0.871,
311
+ "eval_runtime": 111.0284,
312
+ "eval_samples_per_second": 9.007,
313
+ "eval_steps_per_second": 2.252,
314
+ "step": 2000
315
+ },
316
+ {
317
+ "epoch": 2.05,
318
+ "grad_norm": 189.5020751953125,
319
+ "learning_rate": 7.957e-06,
320
+ "loss": 0.3242,
321
+ "step": 2050
322
+ },
323
+ {
324
+ "epoch": 2.1,
325
+ "grad_norm": 0.06589367240667343,
326
+ "learning_rate": 7.907e-06,
327
+ "loss": 0.3381,
328
+ "step": 2100
329
+ },
330
+ {
331
+ "epoch": 2.15,
332
+ "grad_norm": 6.254538536071777,
333
+ "learning_rate": 7.857e-06,
334
+ "loss": 0.2586,
335
+ "step": 2150
336
+ },
337
+ {
338
+ "epoch": 2.2,
339
+ "grad_norm": 0.035582080483436584,
340
+ "learning_rate": 7.807e-06,
341
+ "loss": 0.3091,
342
+ "step": 2200
343
+ },
344
+ {
345
+ "epoch": 2.25,
346
+ "grad_norm": 0.17389701306819916,
347
+ "learning_rate": 7.757e-06,
348
+ "loss": 0.5237,
349
+ "step": 2250
350
+ },
351
+ {
352
+ "epoch": 2.3,
353
+ "grad_norm": 33.96052169799805,
354
+ "learning_rate": 7.707000000000001e-06,
355
+ "loss": 0.4895,
356
+ "step": 2300
357
+ },
358
+ {
359
+ "epoch": 2.35,
360
+ "grad_norm": 0.2390211671590805,
361
+ "learning_rate": 7.657000000000001e-06,
362
+ "loss": 0.3044,
363
+ "step": 2350
364
+ },
365
+ {
366
+ "epoch": 2.4,
367
+ "grad_norm": 0.29711782932281494,
368
+ "learning_rate": 7.607000000000001e-06,
369
+ "loss": 0.3455,
370
+ "step": 2400
371
+ },
372
+ {
373
+ "epoch": 2.45,
374
+ "grad_norm": 0.026691796258091927,
375
+ "learning_rate": 7.557000000000001e-06,
376
+ "loss": 0.4515,
377
+ "step": 2450
378
+ },
379
+ {
380
+ "epoch": 2.5,
381
+ "grad_norm": 0.12731046974658966,
382
+ "learning_rate": 7.507000000000001e-06,
383
+ "loss": 0.3598,
384
+ "step": 2500
385
+ },
386
+ {
387
+ "epoch": 2.55,
388
+ "grad_norm": 0.042913321405649185,
389
+ "learning_rate": 7.457000000000001e-06,
390
+ "loss": 0.2815,
391
+ "step": 2550
392
+ },
393
+ {
394
+ "epoch": 2.6,
395
+ "grad_norm": 0.06428048014640808,
396
+ "learning_rate": 7.407000000000001e-06,
397
+ "loss": 0.3822,
398
+ "step": 2600
399
+ },
400
+ {
401
+ "epoch": 2.65,
402
+ "grad_norm": 0.22337721288204193,
403
+ "learning_rate": 7.357000000000001e-06,
404
+ "loss": 0.217,
405
+ "step": 2650
406
+ },
407
+ {
408
+ "epoch": 2.7,
409
+ "grad_norm": 0.5337192416191101,
410
+ "learning_rate": 7.3070000000000005e-06,
411
+ "loss": 0.4874,
412
+ "step": 2700
413
+ },
414
+ {
415
+ "epoch": 2.75,
416
+ "grad_norm": 0.03206096589565277,
417
+ "learning_rate": 7.257000000000001e-06,
418
+ "loss": 0.4564,
419
+ "step": 2750
420
+ },
421
+ {
422
+ "epoch": 2.8,
423
+ "grad_norm": 346.4653625488281,
424
+ "learning_rate": 7.207000000000001e-06,
425
+ "loss": 0.2794,
426
+ "step": 2800
427
+ },
428
+ {
429
+ "epoch": 2.85,
430
+ "grad_norm": 39.87053298950195,
431
+ "learning_rate": 7.157e-06,
432
+ "loss": 0.4392,
433
+ "step": 2850
434
+ },
435
+ {
436
+ "epoch": 2.9,
437
+ "grad_norm": 0.0478862039744854,
438
+ "learning_rate": 7.107000000000001e-06,
439
+ "loss": 0.2627,
440
+ "step": 2900
441
+ },
442
+ {
443
+ "epoch": 2.95,
444
+ "grad_norm": 0.04880313202738762,
445
+ "learning_rate": 7.057e-06,
446
+ "loss": 0.1502,
447
+ "step": 2950
448
+ },
449
+ {
450
+ "epoch": 3.0,
451
+ "grad_norm": 0.8579753041267395,
452
+ "learning_rate": 7.007000000000001e-06,
453
+ "loss": 0.3137,
454
+ "step": 3000
455
+ },
456
+ {
457
+ "epoch": 3.0,
458
+ "eval_accuracy": 0.858,
459
+ "eval_f1": 0.8568902628142059,
460
+ "eval_loss": 0.8766486048698425,
461
+ "eval_precision": 0.869630225735872,
462
+ "eval_recall": 0.858,
463
+ "eval_runtime": 112.5121,
464
+ "eval_samples_per_second": 8.888,
465
+ "eval_steps_per_second": 2.222,
466
+ "step": 3000
467
+ },
468
+ {
469
+ "epoch": 3.05,
470
+ "grad_norm": 0.09161412715911865,
471
+ "learning_rate": 6.9570000000000005e-06,
472
+ "loss": 0.2441,
473
+ "step": 3050
474
+ },
475
+ {
476
+ "epoch": 3.1,
477
+ "grad_norm": 285.05072021484375,
478
+ "learning_rate": 6.907e-06,
479
+ "loss": 0.4021,
480
+ "step": 3100
481
+ },
482
+ {
483
+ "epoch": 3.15,
484
+ "grad_norm": 0.03876444697380066,
485
+ "learning_rate": 6.857000000000001e-06,
486
+ "loss": 0.2032,
487
+ "step": 3150
488
+ },
489
+ {
490
+ "epoch": 3.2,
491
+ "grad_norm": 9.66369342803955,
492
+ "learning_rate": 6.807e-06,
493
+ "loss": 0.1972,
494
+ "step": 3200
495
+ },
496
+ {
497
+ "epoch": 3.25,
498
+ "grad_norm": 0.06236199662089348,
499
+ "learning_rate": 6.757e-06,
500
+ "loss": 0.4157,
501
+ "step": 3250
502
+ },
503
+ {
504
+ "epoch": 3.3,
505
+ "grad_norm": 2.1778385639190674,
506
+ "learning_rate": 6.707e-06,
507
+ "loss": 0.3814,
508
+ "step": 3300
509
+ },
510
+ {
511
+ "epoch": 3.35,
512
+ "grad_norm": 0.02144041284918785,
513
+ "learning_rate": 6.657e-06,
514
+ "loss": 0.1562,
515
+ "step": 3350
516
+ },
517
+ {
518
+ "epoch": 3.4,
519
+ "grad_norm": 0.4679206311702728,
520
+ "learning_rate": 6.6070000000000004e-06,
521
+ "loss": 0.2698,
522
+ "step": 3400
523
+ },
524
+ {
525
+ "epoch": 3.45,
526
+ "grad_norm": 0.019845549017190933,
527
+ "learning_rate": 6.557e-06,
528
+ "loss": 0.2024,
529
+ "step": 3450
530
+ },
531
+ {
532
+ "epoch": 3.5,
533
+ "grad_norm": 0.03582359477877617,
534
+ "learning_rate": 6.507e-06,
535
+ "loss": 0.1805,
536
+ "step": 3500
537
+ },
538
+ {
539
+ "epoch": 3.55,
540
+ "grad_norm": 0.012999696657061577,
541
+ "learning_rate": 6.457000000000001e-06,
542
+ "loss": 0.2242,
543
+ "step": 3550
544
+ },
545
+ {
546
+ "epoch": 3.6,
547
+ "grad_norm": 61.022212982177734,
548
+ "learning_rate": 6.407000000000001e-06,
549
+ "loss": 0.2972,
550
+ "step": 3600
551
+ },
552
+ {
553
+ "epoch": 3.65,
554
+ "grad_norm": 0.17647244036197662,
555
+ "learning_rate": 6.357000000000001e-06,
556
+ "loss": 0.2507,
557
+ "step": 3650
558
+ },
559
+ {
560
+ "epoch": 3.7,
561
+ "grad_norm": 0.02148498222231865,
562
+ "learning_rate": 6.307000000000001e-06,
563
+ "loss": 0.3322,
564
+ "step": 3700
565
+ },
566
+ {
567
+ "epoch": 3.75,
568
+ "grad_norm": 8.956069946289062,
569
+ "learning_rate": 6.257000000000001e-06,
570
+ "loss": 0.2174,
571
+ "step": 3750
572
+ },
573
+ {
574
+ "epoch": 3.8,
575
+ "grad_norm": 0.019011685624718666,
576
+ "learning_rate": 6.2080000000000005e-06,
577
+ "loss": 0.2118,
578
+ "step": 3800
579
+ },
580
+ {
581
+ "epoch": 3.85,
582
+ "grad_norm": 0.01807810179889202,
583
+ "learning_rate": 6.158000000000001e-06,
584
+ "loss": 0.1908,
585
+ "step": 3850
586
+ },
587
+ {
588
+ "epoch": 3.9,
589
+ "grad_norm": 0.02416962757706642,
590
+ "learning_rate": 6.108000000000001e-06,
591
+ "loss": 0.1491,
592
+ "step": 3900
593
+ },
594
+ {
595
+ "epoch": 3.95,
596
+ "grad_norm": 19.60466766357422,
597
+ "learning_rate": 6.058e-06,
598
+ "loss": 0.2031,
599
+ "step": 3950
600
+ },
601
+ {
602
+ "epoch": 4.0,
603
+ "grad_norm": 0.21647094190120697,
604
+ "learning_rate": 6.008000000000001e-06,
605
+ "loss": 0.2952,
606
+ "step": 4000
607
+ },
608
+ {
609
+ "epoch": 4.0,
610
+ "eval_accuracy": 0.873,
611
+ "eval_f1": 0.872977072862821,
612
+ "eval_loss": 0.821958065032959,
613
+ "eval_precision": 0.8834745892873105,
614
+ "eval_recall": 0.873,
615
+ "eval_runtime": 112.9225,
616
+ "eval_samples_per_second": 8.856,
617
+ "eval_steps_per_second": 2.214,
618
+ "step": 4000
619
+ },
620
+ {
621
+ "epoch": 4.05,
622
+ "grad_norm": 0.08491307497024536,
623
+ "learning_rate": 5.958e-06,
624
+ "loss": 0.1665,
625
+ "step": 4050
626
+ },
627
+ {
628
+ "epoch": 4.1,
629
+ "grad_norm": 0.10018359869718552,
630
+ "learning_rate": 5.908e-06,
631
+ "loss": 0.2381,
632
+ "step": 4100
633
+ },
634
+ {
635
+ "epoch": 4.15,
636
+ "grad_norm": 0.01517445407807827,
637
+ "learning_rate": 5.8580000000000005e-06,
638
+ "loss": 0.1407,
639
+ "step": 4150
640
+ },
641
+ {
642
+ "epoch": 4.2,
643
+ "grad_norm": 0.2072858363389969,
644
+ "learning_rate": 5.808e-06,
645
+ "loss": 0.2508,
646
+ "step": 4200
647
+ },
648
+ {
649
+ "epoch": 4.25,
650
+ "grad_norm": 76.36699676513672,
651
+ "learning_rate": 5.758000000000001e-06,
652
+ "loss": 0.2175,
653
+ "step": 4250
654
+ },
655
+ {
656
+ "epoch": 4.3,
657
+ "grad_norm": 0.0169328935444355,
658
+ "learning_rate": 5.708e-06,
659
+ "loss": 0.1878,
660
+ "step": 4300
661
+ },
662
+ {
663
+ "epoch": 4.35,
664
+ "grad_norm": 23.842693328857422,
665
+ "learning_rate": 5.658e-06,
666
+ "loss": 0.1266,
667
+ "step": 4350
668
+ },
669
+ {
670
+ "epoch": 4.4,
671
+ "grad_norm": 233.6339111328125,
672
+ "learning_rate": 5.608e-06,
673
+ "loss": 0.1279,
674
+ "step": 4400
675
+ },
676
+ {
677
+ "epoch": 4.45,
678
+ "grad_norm": 173.3831024169922,
679
+ "learning_rate": 5.558e-06,
680
+ "loss": 0.1829,
681
+ "step": 4450
682
+ },
683
+ {
684
+ "epoch": 4.5,
685
+ "grad_norm": 0.08392331004142761,
686
+ "learning_rate": 5.508e-06,
687
+ "loss": 0.1668,
688
+ "step": 4500
689
+ },
690
+ {
691
+ "epoch": 4.55,
692
+ "grad_norm": 0.020959168672561646,
693
+ "learning_rate": 5.458e-06,
694
+ "loss": 0.0815,
695
+ "step": 4550
696
+ },
697
+ {
698
+ "epoch": 4.6,
699
+ "grad_norm": 225.8976593017578,
700
+ "learning_rate": 5.408e-06,
701
+ "loss": 0.2346,
702
+ "step": 4600
703
+ },
704
+ {
705
+ "epoch": 4.65,
706
+ "grad_norm": 358.5092468261719,
707
+ "learning_rate": 5.358000000000001e-06,
708
+ "loss": 0.1021,
709
+ "step": 4650
710
+ },
711
+ {
712
+ "epoch": 4.7,
713
+ "grad_norm": 165.55499267578125,
714
+ "learning_rate": 5.308000000000001e-06,
715
+ "loss": 0.2577,
716
+ "step": 4700
717
+ },
718
+ {
719
+ "epoch": 4.75,
720
+ "grad_norm": 0.11646530777215958,
721
+ "learning_rate": 5.258000000000001e-06,
722
+ "loss": 0.2029,
723
+ "step": 4750
724
+ },
725
+ {
726
+ "epoch": 4.8,
727
+ "grad_norm": 0.03793807700276375,
728
+ "learning_rate": 5.208000000000001e-06,
729
+ "loss": 0.1465,
730
+ "step": 4800
731
+ },
732
+ {
733
+ "epoch": 4.85,
734
+ "grad_norm": 113.57576751708984,
735
+ "learning_rate": 5.158e-06,
736
+ "loss": 0.1906,
737
+ "step": 4850
738
+ },
739
+ {
740
+ "epoch": 4.9,
741
+ "grad_norm": 0.011114823631942272,
742
+ "learning_rate": 5.108000000000001e-06,
743
+ "loss": 0.1876,
744
+ "step": 4900
745
+ },
746
+ {
747
+ "epoch": 4.95,
748
+ "grad_norm": 77.72245025634766,
749
+ "learning_rate": 5.0580000000000005e-06,
750
+ "loss": 0.316,
751
+ "step": 4950
752
+ },
753
+ {
754
+ "epoch": 5.0,
755
+ "grad_norm": 0.02318619005382061,
756
+ "learning_rate": 5.008000000000001e-06,
757
+ "loss": 0.1035,
758
+ "step": 5000
759
+ },
760
+ {
761
+ "epoch": 5.0,
762
+ "eval_accuracy": 0.884,
763
+ "eval_f1": 0.8836532680104157,
764
+ "eval_loss": 0.7644630074501038,
765
+ "eval_precision": 0.8889154581905352,
766
+ "eval_recall": 0.884,
767
+ "eval_runtime": 111.6762,
768
+ "eval_samples_per_second": 8.954,
769
+ "eval_steps_per_second": 2.239,
770
+ "step": 5000
771
+ },
772
+ {
773
+ "epoch": 5.05,
774
+ "grad_norm": 0.03648075461387634,
775
+ "learning_rate": 4.958000000000001e-06,
776
+ "loss": 0.0445,
777
+ "step": 5050
778
+ },
779
+ {
780
+ "epoch": 5.1,
781
+ "grad_norm": 0.014485549181699753,
782
+ "learning_rate": 4.908e-06,
783
+ "loss": 0.0928,
784
+ "step": 5100
785
+ },
786
+ {
787
+ "epoch": 5.15,
788
+ "grad_norm": 0.019193926826119423,
789
+ "learning_rate": 4.858000000000001e-06,
790
+ "loss": 0.1156,
791
+ "step": 5150
792
+ },
793
+ {
794
+ "epoch": 5.2,
795
+ "grad_norm": 0.007525489665567875,
796
+ "learning_rate": 4.808e-06,
797
+ "loss": 0.1103,
798
+ "step": 5200
799
+ },
800
+ {
801
+ "epoch": 5.25,
802
+ "grad_norm": 0.02716565877199173,
803
+ "learning_rate": 4.758e-06,
804
+ "loss": 0.1653,
805
+ "step": 5250
806
+ },
807
+ {
808
+ "epoch": 5.3,
809
+ "grad_norm": 0.02299477905035019,
810
+ "learning_rate": 4.7080000000000005e-06,
811
+ "loss": 0.1915,
812
+ "step": 5300
813
+ },
814
+ {
815
+ "epoch": 5.35,
816
+ "grad_norm": 447.5741271972656,
817
+ "learning_rate": 4.658e-06,
818
+ "loss": 0.0903,
819
+ "step": 5350
820
+ },
821
+ {
822
+ "epoch": 5.4,
823
+ "grad_norm": 405.6177062988281,
824
+ "learning_rate": 4.608000000000001e-06,
825
+ "loss": 0.1506,
826
+ "step": 5400
827
+ },
828
+ {
829
+ "epoch": 5.45,
830
+ "grad_norm": 6.23228645324707,
831
+ "learning_rate": 4.558e-06,
832
+ "loss": 0.1342,
833
+ "step": 5450
834
+ },
835
+ {
836
+ "epoch": 5.5,
837
+ "grad_norm": 0.047868575900793076,
838
+ "learning_rate": 4.508e-06,
839
+ "loss": 0.1799,
840
+ "step": 5500
841
+ },
842
+ {
843
+ "epoch": 5.55,
844
+ "grad_norm": 115.90292358398438,
845
+ "learning_rate": 4.458e-06,
846
+ "loss": 0.0751,
847
+ "step": 5550
848
+ },
849
+ {
850
+ "epoch": 5.6,
851
+ "grad_norm": 0.1427452564239502,
852
+ "learning_rate": 4.408000000000001e-06,
853
+ "loss": 0.1132,
854
+ "step": 5600
855
+ },
856
+ {
857
+ "epoch": 5.65,
858
+ "grad_norm": 0.03756200522184372,
859
+ "learning_rate": 4.3580000000000005e-06,
860
+ "loss": 0.0219,
861
+ "step": 5650
862
+ },
863
+ {
864
+ "epoch": 5.7,
865
+ "grad_norm": 0.1948777437210083,
866
+ "learning_rate": 4.308000000000001e-06,
867
+ "loss": 0.1625,
868
+ "step": 5700
869
+ },
870
+ {
871
+ "epoch": 5.75,
872
+ "grad_norm": 0.14967912435531616,
873
+ "learning_rate": 4.2580000000000006e-06,
874
+ "loss": 0.1131,
875
+ "step": 5750
876
+ },
877
+ {
878
+ "epoch": 5.8,
879
+ "grad_norm": 2.684386730194092,
880
+ "learning_rate": 4.208e-06,
881
+ "loss": 0.1085,
882
+ "step": 5800
883
+ },
884
+ {
885
+ "epoch": 5.85,
886
+ "grad_norm": 0.037221621721982956,
887
+ "learning_rate": 4.158000000000001e-06,
888
+ "loss": 0.1957,
889
+ "step": 5850
890
+ },
891
+ {
892
+ "epoch": 5.9,
893
+ "grad_norm": 0.014206652529537678,
894
+ "learning_rate": 4.108e-06,
895
+ "loss": 0.0228,
896
+ "step": 5900
897
+ },
898
+ {
899
+ "epoch": 5.95,
900
+ "grad_norm": 0.008729885332286358,
901
+ "learning_rate": 4.058e-06,
902
+ "loss": 0.0421,
903
+ "step": 5950
904
+ },
905
+ {
906
+ "epoch": 6.0,
907
+ "grad_norm": 253.62225341796875,
908
+ "learning_rate": 4.008e-06,
909
+ "loss": 0.0865,
910
+ "step": 6000
911
+ },
912
+ {
913
+ "epoch": 6.0,
914
+ "eval_accuracy": 0.9,
915
+ "eval_f1": 0.9000352608410468,
916
+ "eval_loss": 0.7469241619110107,
917
+ "eval_precision": 0.9014304934175754,
918
+ "eval_recall": 0.9,
919
+ "eval_runtime": 112.6162,
920
+ "eval_samples_per_second": 8.88,
921
+ "eval_steps_per_second": 2.22,
922
+ "step": 6000
923
+ },
924
+ {
925
+ "epoch": 6.05,
926
+ "grad_norm": 0.011558984406292439,
927
+ "learning_rate": 3.958e-06,
928
+ "loss": 0.0105,
929
+ "step": 6050
930
+ },
931
+ {
932
+ "epoch": 6.1,
933
+ "grad_norm": 0.007709025871008635,
934
+ "learning_rate": 3.9080000000000005e-06,
935
+ "loss": 0.0543,
936
+ "step": 6100
937
+ },
938
+ {
939
+ "epoch": 6.15,
940
+ "grad_norm": 0.00899557676166296,
941
+ "learning_rate": 3.858e-06,
942
+ "loss": 0.0038,
943
+ "step": 6150
944
+ },
945
+ {
946
+ "epoch": 6.2,
947
+ "grad_norm": 0.1194106861948967,
948
+ "learning_rate": 3.8090000000000003e-06,
949
+ "loss": 0.0805,
950
+ "step": 6200
951
+ },
952
+ {
953
+ "epoch": 6.25,
954
+ "grad_norm": 0.0048521412536501884,
955
+ "learning_rate": 3.7590000000000003e-06,
956
+ "loss": 0.0221,
957
+ "step": 6250
958
+ },
959
+ {
960
+ "epoch": 6.3,
961
+ "grad_norm": 463.0719299316406,
962
+ "learning_rate": 3.7090000000000004e-06,
963
+ "loss": 0.1006,
964
+ "step": 6300
965
+ },
966
+ {
967
+ "epoch": 6.35,
968
+ "grad_norm": 0.010736881755292416,
969
+ "learning_rate": 3.6590000000000005e-06,
970
+ "loss": 0.1082,
971
+ "step": 6350
972
+ },
973
+ {
974
+ "epoch": 6.4,
975
+ "grad_norm": 0.008553804829716682,
976
+ "learning_rate": 3.609e-06,
977
+ "loss": 0.106,
978
+ "step": 6400
979
+ },
980
+ {
981
+ "epoch": 6.45,
982
+ "grad_norm": 192.78749084472656,
983
+ "learning_rate": 3.559e-06,
984
+ "loss": 0.1322,
985
+ "step": 6450
986
+ },
987
+ {
988
+ "epoch": 6.5,
989
+ "grad_norm": 0.361038476228714,
990
+ "learning_rate": 3.509e-06,
991
+ "loss": 0.1171,
992
+ "step": 6500
993
+ },
994
+ {
995
+ "epoch": 6.55,
996
+ "grad_norm": 0.012542732059955597,
997
+ "learning_rate": 3.4590000000000003e-06,
998
+ "loss": 0.0034,
999
+ "step": 6550
1000
+ },
1001
+ {
1002
+ "epoch": 6.6,
1003
+ "grad_norm": 0.004605053458362818,
1004
+ "learning_rate": 3.409e-06,
1005
+ "loss": 0.0422,
1006
+ "step": 6600
1007
+ },
1008
+ {
1009
+ "epoch": 6.65,
1010
+ "grad_norm": 0.030367286875844002,
1011
+ "learning_rate": 3.359e-06,
1012
+ "loss": 0.2559,
1013
+ "step": 6650
1014
+ },
1015
+ {
1016
+ "epoch": 6.7,
1017
+ "grad_norm": 0.027533065527677536,
1018
+ "learning_rate": 3.3090000000000004e-06,
1019
+ "loss": 0.0152,
1020
+ "step": 6700
1021
+ },
1022
+ {
1023
+ "epoch": 6.75,
1024
+ "grad_norm": 0.024122724309563637,
1025
+ "learning_rate": 3.2590000000000005e-06,
1026
+ "loss": 0.0298,
1027
+ "step": 6750
1028
+ },
1029
+ {
1030
+ "epoch": 6.8,
1031
+ "grad_norm": 0.01360723003745079,
1032
+ "learning_rate": 3.2090000000000005e-06,
1033
+ "loss": 0.0255,
1034
+ "step": 6800
1035
+ },
1036
+ {
1037
+ "epoch": 6.85,
1038
+ "grad_norm": 0.010986587963998318,
1039
+ "learning_rate": 3.1590000000000006e-06,
1040
+ "loss": 0.1941,
1041
+ "step": 6850
1042
+ },
1043
+ {
1044
+ "epoch": 6.9,
1045
+ "grad_norm": 0.021083349362015724,
1046
+ "learning_rate": 3.1090000000000002e-06,
1047
+ "loss": 0.1208,
1048
+ "step": 6900
1049
+ },
1050
+ {
1051
+ "epoch": 6.95,
1052
+ "grad_norm": 0.004425186663866043,
1053
+ "learning_rate": 3.0590000000000003e-06,
1054
+ "loss": 0.0149,
1055
+ "step": 6950
1056
+ },
1057
+ {
1058
+ "epoch": 7.0,
1059
+ "grad_norm": 0.005357071291655302,
1060
+ "learning_rate": 3.0090000000000003e-06,
1061
+ "loss": 0.0314,
1062
+ "step": 7000
1063
+ },
1064
+ {
1065
+ "epoch": 7.0,
1066
+ "eval_accuracy": 0.892,
1067
+ "eval_f1": 0.8921548011209359,
1068
+ "eval_loss": 0.7714166045188904,
1069
+ "eval_precision": 0.8931565142806782,
1070
+ "eval_recall": 0.892,
1071
+ "eval_runtime": 114.8764,
1072
+ "eval_samples_per_second": 8.705,
1073
+ "eval_steps_per_second": 2.176,
1074
+ "step": 7000
1075
+ },
1076
+ {
1077
+ "epoch": 7.05,
1078
+ "grad_norm": 0.004052102565765381,
1079
+ "learning_rate": 2.9590000000000004e-06,
1080
+ "loss": 0.0531,
1081
+ "step": 7050
1082
+ },
1083
+ {
1084
+ "epoch": 7.1,
1085
+ "grad_norm": 0.0535280667245388,
1086
+ "learning_rate": 2.909e-06,
1087
+ "loss": 0.0202,
1088
+ "step": 7100
1089
+ },
1090
+ {
1091
+ "epoch": 7.15,
1092
+ "grad_norm": 0.026509596034884453,
1093
+ "learning_rate": 2.859e-06,
1094
+ "loss": 0.1402,
1095
+ "step": 7150
1096
+ },
1097
+ {
1098
+ "epoch": 7.2,
1099
+ "grad_norm": 1.2209473848342896,
1100
+ "learning_rate": 2.809e-06,
1101
+ "loss": 0.0266,
1102
+ "step": 7200
1103
+ },
1104
+ {
1105
+ "epoch": 7.25,
1106
+ "grad_norm": 26.857582092285156,
1107
+ "learning_rate": 2.759e-06,
1108
+ "loss": 0.0006,
1109
+ "step": 7250
1110
+ },
1111
+ {
1112
+ "epoch": 7.3,
1113
+ "grad_norm": 0.0071187699213624,
1114
+ "learning_rate": 2.709e-06,
1115
+ "loss": 0.0008,
1116
+ "step": 7300
1117
+ },
1118
+ {
1119
+ "epoch": 7.35,
1120
+ "grad_norm": 0.010221453383564949,
1121
+ "learning_rate": 2.6590000000000003e-06,
1122
+ "loss": 0.0487,
1123
+ "step": 7350
1124
+ },
1125
+ {
1126
+ "epoch": 7.4,
1127
+ "grad_norm": 0.003582503879442811,
1128
+ "learning_rate": 2.6090000000000003e-06,
1129
+ "loss": 0.1051,
1130
+ "step": 7400
1131
+ },
1132
+ {
1133
+ "epoch": 7.45,
1134
+ "grad_norm": 0.005786183290183544,
1135
+ "learning_rate": 2.5590000000000004e-06,
1136
+ "loss": 0.0449,
1137
+ "step": 7450
1138
+ },
1139
+ {
1140
+ "epoch": 7.5,
1141
+ "grad_norm": 0.010817117057740688,
1142
+ "learning_rate": 2.5090000000000005e-06,
1143
+ "loss": 0.0345,
1144
+ "step": 7500
1145
+ },
1146
+ {
1147
+ "epoch": 7.55,
1148
+ "grad_norm": 0.453058660030365,
1149
+ "learning_rate": 2.459e-06,
1150
+ "loss": 0.0352,
1151
+ "step": 7550
1152
+ },
1153
+ {
1154
+ "epoch": 7.6,
1155
+ "grad_norm": 0.003637389512732625,
1156
+ "learning_rate": 2.409e-06,
1157
+ "loss": 0.0392,
1158
+ "step": 7600
1159
+ },
1160
+ {
1161
+ "epoch": 7.65,
1162
+ "grad_norm": 0.005378212314099073,
1163
+ "learning_rate": 2.359e-06,
1164
+ "loss": 0.0795,
1165
+ "step": 7650
1166
+ },
1167
+ {
1168
+ "epoch": 7.7,
1169
+ "grad_norm": 0.005855490919202566,
1170
+ "learning_rate": 2.3090000000000003e-06,
1171
+ "loss": 0.0617,
1172
+ "step": 7700
1173
+ },
1174
+ {
1175
+ "epoch": 7.75,
1176
+ "grad_norm": 0.004148279316723347,
1177
+ "learning_rate": 2.259e-06,
1178
+ "loss": 0.0016,
1179
+ "step": 7750
1180
+ },
1181
+ {
1182
+ "epoch": 7.8,
1183
+ "grad_norm": 1.2424975633621216,
1184
+ "learning_rate": 2.2090000000000004e-06,
1185
+ "loss": 0.0545,
1186
+ "step": 7800
1187
+ },
1188
+ {
1189
+ "epoch": 7.85,
1190
+ "grad_norm": 0.151920348405838,
1191
+ "learning_rate": 2.1590000000000004e-06,
1192
+ "loss": 0.0003,
1193
+ "step": 7850
1194
+ },
1195
+ {
1196
+ "epoch": 7.9,
1197
+ "grad_norm": 0.0055457111448049545,
1198
+ "learning_rate": 2.109e-06,
1199
+ "loss": 0.0003,
1200
+ "step": 7900
1201
+ },
1202
+ {
1203
+ "epoch": 7.95,
1204
+ "grad_norm": 0.006306794937700033,
1205
+ "learning_rate": 2.059e-06,
1206
+ "loss": 0.0003,
1207
+ "step": 7950
1208
+ },
1209
+ {
1210
+ "epoch": 8.0,
1211
+ "grad_norm": 0.006967851426452398,
1212
+ "learning_rate": 2.009e-06,
1213
+ "loss": 0.0806,
1214
+ "step": 8000
1215
+ },
1216
+ {
1217
+ "epoch": 8.0,
1218
+ "eval_accuracy": 0.891,
1219
+ "eval_f1": 0.8907797096380854,
1220
+ "eval_loss": 0.8573828935623169,
1221
+ "eval_precision": 0.8919358819038602,
1222
+ "eval_recall": 0.891,
1223
+ "eval_runtime": 112.4936,
1224
+ "eval_samples_per_second": 8.889,
1225
+ "eval_steps_per_second": 2.222,
1226
+ "step": 8000
1227
+ }
1228
+ ],
1229
+ "logging_steps": 50,
1230
+ "max_steps": 10000,
1231
+ "num_input_tokens_seen": 0,
1232
+ "num_train_epochs": 10,
1233
+ "save_steps": 500,
1234
+ "stateful_callbacks": {
1235
+ "EarlyStoppingCallback": {
1236
+ "args": {
1237
+ "early_stopping_patience": 2,
1238
+ "early_stopping_threshold": 0.0
1239
+ },
1240
+ "attributes": {
1241
+ "early_stopping_patience_counter": 2
1242
+ }
1243
+ },
1244
+ "TrainerControl": {
1245
+ "args": {
1246
+ "should_epoch_stop": false,
1247
+ "should_evaluate": false,
1248
+ "should_log": false,
1249
+ "should_save": true,
1250
+ "should_training_stop": true
1251
+ },
1252
+ "attributes": {}
1253
+ }
1254
+ },
1255
+ "total_flos": 8493277347840000.0,
1256
+ "train_batch_size": 4,
1257
+ "trial_name": null,
1258
+ "trial_params": null
1259
+ }
checkpoint-8000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f403144661d43c8a35a0a64782c635300fe17c4e900c1c22a3d8255e6ad2018
3
+ size 5304