Hugging Face's logo

Team-PIXEL
/
pixel-tiny-bigrams

Transformers
PyTorch
pixel
masked-auto-encoding
generated_from_trainer
Inference Endpoints
Model card Files Community
jflotz commited on
Commit
aa52c27
•
1 Parent(s): 4248b18

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +307 -1
README.md CHANGED
@@ -1,3 +1,309 @@
1
  ---
2
- license: apache-2.0
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ tags:
3
+ - generated_from_trainer
4
+ model-index:
5
+ - name: pixel-tiny-bigrams
6
+ results: []
7
  ---
8
+
9
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
+ should probably proofread and complete it, then remove this comment. -->
11
+
12
+ # pixel-tiny-bigrams
13
+
14
+ This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset.
15
+ It achieves the following results on the evaluation set:
16
+ - Loss: 0.3380
17
+
18
+ ## Model description
19
+
20
+ More information needed
21
+
22
+ ## Intended uses & limitations
23
+
24
+ More information needed
25
+
26
+ ## Training and evaluation data
27
+
28
+ More information needed
29
+
30
+ ## Training procedure
31
+
32
+ ### Training hyperparameters
33
+
34
+ The following hyperparameters were used during training:
35
+ - learning_rate: 0.0006
36
+ - train_batch_size: 128
37
+ - eval_batch_size: 8
38
+ - seed: 42
39
+ - distributed_type: multi-GPU
40
+ - num_devices: 8
41
+ - total_train_batch_size: 1024
42
+ - total_eval_batch_size: 64
43
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
+ - lr_scheduler_type: cosine
45
+ - lr_scheduler_warmup_ratio: 0.05
46
+ - training_steps: 250000
47
+
48
+ ### Training results
49
+
50
+ | Training Loss | Epoch | Step | Validation Loss |
51
+ |:-------------:|:-----:|:------:|:---------------:|
52
+ | 0.689 | 0.04 | 1000 | 0.6793 |
53
+ | 0.6802 | 0.09 | 2000 | 0.6787 |
54
+ | 0.6795 | 0.13 | 3000 | 0.6788 |
55
+ | 0.679 | 0.18 | 4000 | 0.6782 |
56
+ | 0.6787 | 0.22 | 5000 | 0.6782 |
57
+ | 0.6786 | 0.27 | 6000 | 0.6781 |
58
+ | 0.6784 | 0.31 | 7000 | 0.6781 |
59
+ | 0.6783 | 0.36 | 8000 | 0.6781 |
60
+ | 0.6781 | 0.4 | 9000 | 0.6773 |
61
+ | 0.6775 | 0.45 | 10000 | 0.6778 |
62
+ | 0.6775 | 0.49 | 11000 | 0.6769 |
63
+ | 0.6773 | 0.54 | 12000 | 0.6773 |
64
+ | 0.6774 | 0.58 | 13000 | 0.6771 |
65
+ | 0.6773 | 0.62 | 14000 | 0.6772 |
66
+ | 0.6773 | 0.67 | 15000 | 0.6772 |
67
+ | 0.6772 | 0.71 | 16000 | 0.6776 |
68
+ | 0.6773 | 0.76 | 17000 | 0.6770 |
69
+ | 0.6772 | 0.8 | 18000 | 0.6775 |
70
+ | 0.6772 | 0.85 | 19000 | 0.6770 |
71
+ | 0.6774 | 0.89 | 20000 | 0.6770 |
72
+ | 0.6772 | 0.94 | 21000 | 0.6762 |
73
+ | 0.6773 | 0.98 | 22000 | 0.6775 |
74
+ | 0.6773 | 1.03 | 23000 | 0.6764 |
75
+ | 0.6772 | 1.07 | 24000 | 0.6768 |
76
+ | 0.6772 | 1.12 | 25000 | 0.6769 |
77
+ | 0.6772 | 1.16 | 26000 | 0.6775 |
78
+ | 0.6772 | 1.2 | 27000 | 0.6776 |
79
+ | 0.6772 | 1.25 | 28000 | 0.6772 |
80
+ | 0.6772 | 1.29 | 29000 | 0.6769 |
81
+ | 0.6773 | 1.34 | 30000 | 0.6772 |
82
+ | 0.6772 | 1.38 | 31000 | 0.6777 |
83
+ | 0.6772 | 1.43 | 32000 | 0.6769 |
84
+ | 0.6773 | 1.47 | 33000 | 0.6767 |
85
+ | 0.677 | 1.52 | 34000 | 0.6766 |
86
+ | 0.6765 | 1.56 | 35000 | 0.6766 |
87
+ | 0.6763 | 1.61 | 36000 | 0.6766 |
88
+ | 0.6764 | 1.65 | 37000 | 0.6758 |
89
+ | 0.6764 | 1.7 | 38000 | 0.6762 |
90
+ | 0.6758 | 1.74 | 39000 | 0.6771 |
91
+ | 0.6772 | 1.78 | 40000 | 0.6770 |
92
+ | 0.6575 | 1.83 | 41000 | 0.6465 |
93
+ | 0.6373 | 1.87 | 42000 | 0.6318 |
94
+ | 0.6257 | 1.92 | 43000 | 0.6184 |
95
+ | 0.621 | 1.96 | 44000 | 0.6136 |
96
+ | 0.6183 | 2.01 | 45000 | 0.6127 |
97
+ | 0.6165 | 2.05 | 46000 | 0.6103 |
98
+ | 0.612 | 2.1 | 47000 | 0.6013 |
99
+ | 0.6037 | 2.14 | 48000 | 0.5943 |
100
+ | 0.6 | 2.19 | 49000 | 0.5915 |
101
+ | 0.5973 | 2.23 | 50000 | 0.5881 |
102
+ | 0.5924 | 2.28 | 51000 | 0.5799 |
103
+ | 0.5817 | 2.32 | 52000 | 0.5670 |
104
+ | 0.5719 | 2.36 | 53000 | 0.5557 |
105
+ | 0.5651 | 2.41 | 54000 | 0.5477 |
106
+ | 0.5592 | 2.45 | 55000 | 0.5408 |
107
+ | 0.5534 | 2.5 | 56000 | 0.5362 |
108
+ | 0.5446 | 2.54 | 57000 | 0.5251 |
109
+ | 0.5342 | 2.59 | 58000 | 0.5130 |
110
+ | 0.5239 | 2.63 | 59000 | 0.5024 |
111
+ | 0.5147 | 2.68 | 60000 | 0.4947 |
112
+ | 0.5061 | 2.72 | 61000 | 0.4848 |
113
+ | 0.4981 | 2.77 | 62000 | 0.4746 |
114
+ | 0.4912 | 2.81 | 63000 | 0.4681 |
115
+ | 0.4847 | 2.86 | 64000 | 0.4599 |
116
+ | 0.4792 | 2.9 | 65000 | 0.4537 |
117
+ | 0.474 | 2.94 | 66000 | 0.4491 |
118
+ | 0.4688 | 2.99 | 67000 | 0.4437 |
119
+ | 0.464 | 3.03 | 68000 | 0.4392 |
120
+ | 0.4592 | 3.08 | 69000 | 0.4324 |
121
+ | 0.4547 | 3.12 | 70000 | 0.4284 |
122
+ | 0.4507 | 3.17 | 71000 | 0.4260 |
123
+ | 0.4468 | 3.21 | 72000 | 0.4192 |
124
+ | 0.4432 | 3.26 | 73000 | 0.4161 |
125
+ | 0.44 | 3.3 | 74000 | 0.4153 |
126
+ | 0.4367 | 3.35 | 75000 | 0.4102 |
127
+ | 0.4337 | 3.39 | 76000 | 0.4062 |
128
+ | 0.4311 | 3.44 | 77000 | 0.4019 |
129
+ | 0.4286 | 3.48 | 78000 | 0.4007 |
130
+ | 0.4259 | 3.52 | 79000 | 0.3997 |
131
+ | 0.4239 | 3.57 | 80000 | 0.3968 |
132
+ | 0.4218 | 3.61 | 81000 | 0.3949 |
133
+ | 0.4201 | 3.66 | 82000 | 0.3935 |
134
+ | 0.4182 | 3.7 | 83000 | 0.3926 |
135
+ | 0.4168 | 3.75 | 84000 | 0.3879 |
136
+ | 0.4155 | 3.79 | 85000 | 0.3885 |
137
+ | 0.4136 | 3.84 | 86000 | 0.3844 |
138
+ | 0.4124 | 3.88 | 87000 | 0.3855 |
139
+ | 0.4116 | 3.93 | 88000 | 0.3830 |
140
+ | 0.4098 | 3.97 | 89000 | 0.3837 |
141
+ | 0.4087 | 4.01 | 90000 | 0.3802 |
142
+ | 0.4078 | 4.06 | 91000 | 0.3799 |
143
+ | 0.4068 | 4.1 | 92000 | 0.3794 |
144
+ | 0.4057 | 4.15 | 93000 | 0.3784 |
145
+ | 0.4047 | 4.19 | 94000 | 0.3788 |
146
+ | 0.4047 | 4.24 | 95000 | 0.3770 |
147
+ | 0.4029 | 4.28 | 96000 | 0.3750 |
148
+ | 0.4022 | 4.33 | 97000 | 0.3747 |
149
+ | 0.4015 | 4.37 | 98000 | 0.3736 |
150
+ | 0.4007 | 4.42 | 99000 | 0.3752 |
151
+ | 0.4 | 4.46 | 100000 | 0.3743 |
152
+ | 0.3995 | 4.51 | 101000 | 0.3741 |
153
+ | 0.3985 | 4.55 | 102000 | 0.3702 |
154
+ | 0.3981 | 4.59 | 103000 | 0.3800 |
155
+ | 0.3986 | 4.64 | 104000 | 0.3734 |
156
+ | 0.3966 | 4.68 | 105000 | 0.3705 |
157
+ | 0.3957 | 4.73 | 106000 | 0.3680 |
158
+ | 0.3957 | 4.77 | 107000 | 0.3663 |
159
+ | 0.3948 | 4.82 | 108000 | 0.3683 |
160
+ | 0.3943 | 4.86 | 109000 | 0.3697 |
161
+ | 0.3936 | 4.91 | 110000 | 0.3672 |
162
+ | 0.3932 | 4.95 | 111000 | 0.3649 |
163
+ | 0.3925 | 5.0 | 112000 | 0.3651 |
164
+ | 0.3919 | 5.04 | 113000 | 0.3650 |
165
+ | 0.3915 | 5.09 | 114000 | 0.3636 |
166
+ | 0.3911 | 5.13 | 115000 | 0.3655 |
167
+ | 0.3905 | 5.17 | 116000 | 0.3650 |
168
+ | 0.3905 | 5.22 | 117000 | 0.4054 |
169
+ | 0.3894 | 5.26 | 118000 | 0.3609 |
170
+ | 0.3889 | 5.31 | 119000 | 0.3599 |
171
+ | 0.3888 | 5.35 | 120000 | 0.3593 |
172
+ | 0.3887 | 5.4 | 121000 | 0.3601 |
173
+ | 0.3883 | 5.44 | 122000 | 0.3611 |
174
+ | 0.6776 | 5.49 | 123000 | 0.6769 |
175
+ | 0.3917 | 5.53 | 124000 | 0.3626 |
176
+ | 0.3897 | 5.58 | 125000 | 0.3617 |
177
+ | 0.3869 | 5.62 | 126000 | 0.3578 |
178
+ | 0.3864 | 5.67 | 127000 | 0.3578 |
179
+ | 0.3862 | 5.71 | 128000 | 0.3573 |
180
+ | 0.3855 | 5.75 | 129000 | 0.3578 |
181
+ | 0.3854 | 5.8 | 130000 | 0.3571 |
182
+ | 0.3849 | 5.84 | 131000 | 0.3566 |
183
+ | 0.3845 | 5.89 | 132000 | 0.3569 |
184
+ | 0.384 | 5.93 | 133000 | 0.3567 |
185
+ | 0.3921 | 5.98 | 134000 | 0.3628 |
186
+ | 0.3844 | 6.02 | 135000 | 0.3565 |
187
+ | 0.383 | 6.07 | 136000 | 0.3547 |
188
+ | 0.3828 | 6.11 | 137000 | 0.3586 |
189
+ | 0.3824 | 6.16 | 138000 | 0.3553 |
190
+ | 0.3825 | 6.2 | 139000 | 0.3549 |
191
+ | 0.3818 | 6.25 | 140000 | 0.3537 |
192
+ | 0.3815 | 6.29 | 141000 | 0.3550 |
193
+ | 0.3812 | 6.33 | 142000 | 0.3539 |
194
+ | 0.3806 | 6.38 | 143000 | 0.3535 |
195
+ | 0.3804 | 6.42 | 144000 | 0.3533 |
196
+ | 0.3799 | 6.47 | 145000 | 0.3539 |
197
+ | 0.3799 | 6.51 | 146000 | 0.3528 |
198
+ | 0.3794 | 6.56 | 147000 | 0.3519 |
199
+ | 0.3792 | 6.6 | 148000 | 0.3501 |
200
+ | 0.3791 | 6.65 | 149000 | 0.3513 |
201
+ | 0.3784 | 6.69 | 150000 | 0.3511 |
202
+ | 0.3833 | 6.74 | 151000 | 0.3518 |
203
+ | 0.3805 | 6.78 | 152000 | 0.3513 |
204
+ | 0.3785 | 6.83 | 153000 | 0.3522 |
205
+ | 0.3772 | 6.87 | 154000 | 0.3493 |
206
+ | 0.3772 | 6.91 | 155000 | 0.3503 |
207
+ | 0.3771 | 6.96 | 156000 | 0.3513 |
208
+ | 0.3769 | 7.0 | 157000 | 0.3505 |
209
+ | 0.3766 | 7.05 | 158000 | 0.3499 |
210
+ | 0.3762 | 7.09 | 159000 | 0.3490 |
211
+ | 0.376 | 7.14 | 160000 | 0.3465 |
212
+ | 0.3756 | 7.18 | 161000 | 0.3490 |
213
+ | 0.3753 | 7.23 | 162000 | 0.3483 |
214
+ | 0.3749 | 7.27 | 163000 | 0.3481 |
215
+ | 0.3747 | 7.32 | 164000 | 0.3470 |
216
+ | 0.375 | 7.36 | 165000 | 0.3476 |
217
+ | 0.3742 | 7.41 | 166000 | 0.3471 |
218
+ | 0.3741 | 7.45 | 167000 | 0.3462 |
219
+ | 0.3738 | 7.49 | 168000 | 0.3470 |
220
+ | 0.3735 | 7.54 | 169000 | 0.3462 |
221
+ | 0.3736 | 7.58 | 170000 | 0.3467 |
222
+ | 0.3731 | 7.63 | 171000 | 0.3457 |
223
+ | 0.3726 | 7.67 | 172000 | 0.3478 |
224
+ | 0.3725 | 7.72 | 173000 | 0.3447 |
225
+ | 0.3722 | 7.76 | 174000 | 0.3459 |
226
+ | 0.3723 | 7.81 | 175000 | 0.3462 |
227
+ | 0.3718 | 7.85 | 176000 | 0.3464 |
228
+ | 0.3716 | 7.9 | 177000 | 0.3453 |
229
+ | 0.3712 | 7.94 | 178000 | 0.3466 |
230
+ | 0.3712 | 7.99 | 179000 | 0.3456 |
231
+ | 0.3709 | 8.03 | 180000 | 0.3452 |
232
+ | 0.3709 | 8.07 | 181000 | 0.3427 |
233
+ | 0.3707 | 8.12 | 182000 | 0.3445 |
234
+ | 0.3703 | 8.16 | 183000 | 0.3452 |
235
+ | 0.3701 | 8.21 | 184000 | 0.3420 |
236
+ | 0.3699 | 8.25 | 185000 | 0.3429 |
237
+ | 0.3697 | 8.3 | 186000 | 0.3432 |
238
+ | 0.3696 | 8.34 | 187000 | 0.3425 |
239
+ | 0.3696 | 8.39 | 188000 | 0.3437 |
240
+ | 0.3694 | 8.43 | 189000 | 0.3425 |
241
+ | 0.369 | 8.48 | 190000 | 0.3429 |
242
+ | 0.369 | 8.52 | 191000 | 0.3415 |
243
+ | 0.3685 | 8.57 | 192000 | 0.3431 |
244
+ | 0.3684 | 8.61 | 193000 | 0.3415 |
245
+ | 0.3683 | 8.65 | 194000 | 0.3421 |
246
+ | 0.368 | 8.7 | 195000 | 0.3422 |
247
+ | 0.3719 | 8.74 | 196000 | 0.3433 |
248
+ | 0.3678 | 8.79 | 197000 | 0.3400 |
249
+ | 0.3675 | 8.83 | 198000 | 0.3420 |
250
+ | 0.3676 | 8.88 | 199000 | 0.3426 |
251
+ | 0.3674 | 8.92 | 200000 | 0.3396 |
252
+ | 0.3673 | 8.97 | 201000 | 0.3404 |
253
+ | 0.3671 | 9.01 | 202000 | 0.3397 |
254
+ | 0.3669 | 9.06 | 203000 | 0.3417 |
255
+ | 0.3669 | 9.1 | 204000 | 0.3413 |
256
+ | 0.3666 | 9.15 | 205000 | 0.3386 |
257
+ | 0.3666 | 9.19 | 206000 | 0.3414 |
258
+ | 0.3664 | 9.23 | 207000 | 0.3407 |
259
+ | 0.3662 | 9.28 | 208000 | 0.3401 |
260
+ | 0.3661 | 9.32 | 209000 | 0.3412 |
261
+ | 0.366 | 9.37 | 210000 | 0.3374 |
262
+ | 0.3659 | 9.41 | 211000 | 0.3400 |
263
+ | 0.3658 | 9.46 | 212000 | 0.3406 |
264
+ | 0.3658 | 9.5 | 213000 | 0.3383 |
265
+ | 0.3656 | 9.55 | 214000 | 0.3399 |
266
+ | 0.3655 | 9.59 | 215000 | 0.3385 |
267
+ | 0.3653 | 9.64 | 216000 | 0.3406 |
268
+ | 0.3652 | 9.68 | 217000 | 0.3388 |
269
+ | 0.3674 | 9.73 | 218000 | 0.3381 |
270
+ | 0.365 | 9.77 | 219000 | 0.3387 |
271
+ | 0.3648 | 9.81 | 220000 | 0.3374 |
272
+ | 0.3649 | 9.86 | 221000 | 0.3378 |
273
+ | 0.3649 | 9.9 | 222000 | 0.3379 |
274
+ | 0.3646 | 9.95 | 223000 | 0.3382 |
275
+ | 0.3647 | 9.99 | 224000 | 0.3377 |
276
+ | 0.3644 | 10.04 | 225000 | 0.3351 |
277
+ | 0.3644 | 10.08 | 226000 | 0.3374 |
278
+ | 0.3644 | 10.13 | 227000 | 0.3379 |
279
+ | 0.3651 | 10.17 | 228000 | 0.3365 |
280
+ | 0.3643 | 10.22 | 229000 | 0.3360 |
281
+ | 0.3642 | 10.26 | 230000 | 0.3371 |
282
+ | 0.364 | 10.31 | 231000 | 0.3380 |
283
+ | 0.364 | 10.35 | 232000 | 0.3375 |
284
+ | 0.364 | 10.39 | 233000 | 0.3386 |
285
+ | 0.3639 | 10.44 | 234000 | 0.3373 |
286
+ | 0.364 | 10.48 | 235000 | 0.3377 |
287
+ | 0.3636 | 10.53 | 236000 | 0.3384 |
288
+ | 0.3636 | 10.57 | 237000 | 0.3367 |
289
+ | 0.3638 | 10.62 | 238000 | 0.3374 |
290
+ | 0.3637 | 10.66 | 239000 | 0.3368 |
291
+ | 0.3635 | 10.71 | 240000 | 0.3352 |
292
+ | 0.3635 | 10.75 | 241000 | 0.3393 |
293
+ | 0.3634 | 10.8 | 242000 | 0.3344 |
294
+ | 0.3635 | 10.84 | 243000 | 0.3383 |
295
+ | 0.3633 | 10.89 | 244000 | 0.3362 |
296
+ | 0.3635 | 10.93 | 245000 | 0.3353 |
297
+ | 0.3634 | 10.97 | 246000 | 0.3357 |
298
+ | 0.3632 | 11.02 | 247000 | 0.3375 |
299
+ | 0.3633 | 11.06 | 248000 | 0.3395 |
300
+ | 0.3635 | 11.11 | 249000 | 0.3382 |
301
+ | 0.3634 | 11.15 | 250000 | 0.3380 |
302
+
303
+
304
+ ### Framework versions
305
+
306
+ - Transformers 4.17.0
307
+ - Pytorch 1.11.0
308
+ - Datasets 2.1.1.dev0
309
+ - Tokenizers 0.12.1