-
Notifications
You must be signed in to change notification settings - Fork 0
/
project_notes.txt
504 lines (497 loc) · 33.8 KB
/
project_notes.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
Word2Vec bot-only tweets output:
Manes-MacBook-Pro:Project manewilliams$ python3 tweets_word2vec.py
WARNING:tensorflow:From /Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/contrib/learn/python/learn/datasets/base.py:198: retry (from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will be removed in a future version.
Instructions for updating:
Use the retry module or similar alternatives.
Data size 3217426
Most common words (+UNK) [['UNK', 278521], ('t', 139784), ('https', 123449), ('co', 118547), ('RT', 59371)]
Sample data [0, 382, 250, 16, 379, 6, 134, 20, 125, 174] ['UNK', 'IslamKills', 'Are', 'you', 'trying', 'to', 'say', 'that', 'there', 'were']
382 IslamKills -> 250 Are
382 IslamKills -> 0 UNK
250 Are -> 16 you
250 Are -> 382 IslamKills
16 you -> 250 Are
16 you -> 379 trying
379 trying -> 16 you
379 trying -> 6 to
WARNING:tensorflow:From tweets_word2vec.py:223: calling reduce_sum (from tensorflow.python.ops.math_ops) with keep_dims is deprecated and will be removed in a future version.
Instructions for updating:
keep_dims is deprecated, use keepdims instead
2018-05-03 13:59:41.810913: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
Initialized
Average loss at step 0 : 245.0008544921875
Nearest to and: Rot, OWNERS, Zen, ridicules, TheRun2016, cerenomri, simostransit, meetthepress,
Nearest to To: Amtszeit, T_Miller_Harper, NUCLEAR, Patricia, Bargains, Singer, integration, pans,
Nearest to I: seasonal, Toure, Perfectly, Carville, intellihubnewsRT, occasional, spirits, reassurance,
Nearest to If: epidemic, LIE, TheJusticeDept, BOY, WALLS, faux, InnSichORT, 74,
Nearest to co: RapinBill, une, Korea, xXQjErgbyQRT, Crybaby, cloud, drayzze, french,
Nearest to The: theory, testament, AlexaKaySam, DJBabyAce, PenceScience, Freddie, exist, Abby,
Nearest to their: saves, USNavy, SankofaBrown, safely, ReasonWhyWeAreNotFriends, INC, OrovilleDam, Places,
Nearest to What: Youngest, goofy, e9D4TB4pLS, silencing, Rachetgrrl, wipe, re, AtlantaFalcons,
Nearest to she: HillarysHandSignals, Mccaskill, Generals, BeachCity55, rankin, UDaWeVM8TZ, Murder, easily,
Nearest to would: tten, JillStein2016, Yells, blo, fave, Wagner, Staging, AnnCoulter,
Nearest to like: classified, Frau, ogkeyser, ElevationForAll, overgrown, Jethro, TweepAQuote, Snow,
Nearest to be: peach, modesty, lauriecrosswell, CNNgo, Denounce, tube, boot, peeps,
Nearest to A: 0upLzbJy8E, Trey_Hannah617, PUTTING, revelations, livebroadcast, fu, penny, Zukunft,
Nearest to s: BWI, curious, ces, Followed, sorority, DorisTruong, Sheree, TrumpTape,
Nearest to they: teenager, ding, TimJGraham, ballo, ice, LukeRussert, heels, Pander,
Nearest to Trump: wolfblitzer, Meddling, sehe, mirror, tripping, rigged, menschlicher, fit,
Average loss at step 2000 : 118.86855748176575
Average loss at step 4000 : 57.551151961803434
Average loss at step 6000 : 35.35826615667343
Average loss at step 8000 : 25.254896528959275
Average loss at step 10000 : 17.77904777121544
Nearest to and: to, for, of, https, in, from, amp, is,
Nearest to To: Amtszeit, Singer, CrookedHillary, IfMyFriendsSawMeNow, Join, impossible, between, Ja,
Nearest to I: he, you, they, fundamental, democratic, Nationalist, Tweets, blocks,
Nearest to If: LIE, BOY, fool, too, instructions, TheJusticeDept, useless, quality,
Nearest to co: About, BIZPACReview, HillaryForPrison2016, via, warning, https, Liberal, mackhanelesa,
Nearest to The: the, https, AlexaKaySam, BONG, turn, testament, Ficklin, exist,
Nearest to their: saves, SankofaBrown, half, POTUSLastTweet, USNavy, taking, Self, censored,
Nearest to What: re, wipe, Expert, Same, Hacking, Politico, Nevada, Propaganda,
Nearest to she: he, All, BONG, Generals, easily, bed, phrase, employee,
Nearest to would: AnnCoulter, fave, had, tten, elites, Yeah, hosts, Donna,
Nearest to like: classified, ogkeyser, glitter, shut, own, cannot, couldn, hide,
Nearest to be: some, Bricks, BONG, lauriecrosswell, TheFoundingSon, Rose, consider, peach,
Nearest to A: Trey_Hannah617, UNK, needs, Early, revelations, Valley, stood, Zukunft,
Nearest to s: is, are, of, briefings, being, with, has, TopNews,
Nearest to they: I, you, teenager, not, basically, HA, ice, ding,
Nearest to Trump: the, wolfblitzer, Sanders, bra, Callisto1947, Beat, conclusions, couldn,
Average loss at step 12000 : 13.929169276952743
Average loss at step 14000 : 11.582980965137482
Average loss at step 16000 : 9.57473588824272
Average loss at step 18000 : 8.580857915639877
Average loss at step 20000 : 7.366018606901169
Nearest to and: amp, for, to, in, of, from, on, is,
Nearest to To: Singer, Amtszeit, Ja, CrookedHillary, eavesdropann, impossible, between, IfMyFriendsSawMeNow,
Nearest to I: he, you, they, we, We, You, fundamental, UNK,
Nearest to If: epidemic, LIE, BOY, instructions, 74, useless, mothers, UNK,
Nearest to co: via, About, WHAA, Goldman, https, Angeles, zijn, RT,
Nearest to The: the, AlexaKaySam, https, BONG, testament, Activists, UNK, Ficklin,
Nearest to their: WHAA, Rushs_Dad, USNavy, half, saves, Unfortunately, your, not,
Nearest to What: wipe, Comm, We, designed, Youngest, re, Expert, Politico,
Nearest to she: he, Reinstate, it, SteinBaraka, Generals, I, Risky, easily,
Nearest to would: AnnCoulter, fave, elites, can, ll, civilrights_rr, WHAA, had,
Nearest to like: classified, ogkeyser, couldn, mercedesschlapp, glitter, shut, own, unknown,
Nearest to be: have, lauriecrosswell, Bricks, Rose, consider, peach, change, some,
Nearest to A: UNK, stood, FOX5Atlanta, Early, Trey_Hannah617, needs, Your, jm_meunier,
Nearest to s: is, has, and, was, are, with, amp, being,
Nearest to they: you, I, we, it, not, ding, We, accurately,
Nearest to Trump: Obama, Hillary, Clinton, UNK, WHAA, Sanders, injures, It,
Average loss at step 22000 : 6.83914078938961
Average loss at step 24000 : 6.297456268072128
Average loss at step 26000 : 5.946202147245407
Average loss at step 28000 : 5.709687164783478
Average loss at step 30000 : 5.386937029480934
Nearest to and: amp, for, of, from, in, to, that, on,
Nearest to To: Ja, Singer, to, In, Amtszeit, eavesdropann, CrookedHillary, between,
Nearest to I: they, you, he, we, We, You, she, fundamental,
Nearest to If: epidemic, LIE, BOY, UNK, instructions, mothers, useless, JamFlint,
Nearest to co: via, https, WHAA, Chainsaw, About, Goldman, Angeles, lasts,
Nearest to The: the, AlexaKaySam, BONG, Activists, testament, DailyCaller, UNK, https,
Nearest to their: your, our, the, Rushs_Dad, WHAA, half, USNavy, Unfortunately,
Nearest to What: silencing, designed, Why, It, We, wipe, Comm, tagging,
Nearest to she: he, it, Reinstate, that, I, SteinBaraka, Generals, there,
Nearest to would: ll, can, could, AnnCoulter, had, elites, will, fave,
Nearest to like: ogkeyser, classified, mercedesschlapp, own, haven, couldn, lane, shut,
Nearest to be: have, peach, lauriecrosswell, Bricks, change, consider, Rose, tube,
Nearest to A: Bias, welcoming, stood, needs, Your, revelations, Wasn, SOS,
Nearest to s: is, has, was, and, amp, Determine, being, Anti,
Nearest to they: you, I, we, We, it, accurately, he, not,
Nearest to Trump: Obama, Hillary, Clinton, Sanders, Cruz, it, WHAA, injures,
Average loss at step 32000 : 5.199194256544113
Average loss at step 34000 : 5.101616035699845
Average loss at step 36000 : 4.904428316950798
Average loss at step 38000 : 4.8290396772623065
Average loss at step 40000 : 4.759026542305946
Nearest to and: amp, for, from, but, of, or, to, in,
Nearest to To: to, Ja, In, Singer, eavesdropann, Amtszeit, For, IG,
Nearest to I: they, we, you, he, You, We, i, she,
Nearest to If: epidemic, LIE, instructions, mothers, BOY, UNK, useless, JamFlint,
Nearest to co: via, RT, Chainsaw, Goldman, Angeles, concussion, https, lasts,
Nearest to The: the, AlexaKaySam, Activists, BONG, A, testament, DailyCaller, DaysRT,
Nearest to their: your, our, the, his, Rushs_Dad, my, WHAA, USNavy,
Nearest to What: silencing, Why, It, designed, tagging, BehindTheRant, We, UNK,
Nearest to she: he, it, Reinstate, I, that, SteinBaraka, there, It,
Nearest to would: ll, can, could, will, may, s, AnnCoulter, hillarysemail,
Nearest to like: haven, ogkeyser, hide, fatalmocking, Everywhere, shut, lane, MikeRumbo66,
Nearest to be: have, Bricks, get, peach, is, change, consider, lauriecrosswell,
Nearest to A: Bias, The, welcoming, Langston, EXPIRATION, DocThompsonShow, revelations, needs,
Nearest to s: is, was, has, and, Determine, Is, shd, Missed,
Nearest to they: you, we, I, We, They, accurately, he, it,
Nearest to Trump: Obama, Hillary, Clinton, Sanders, Cruz, realDonaldTrump, injures, POTUS,
Average loss at step 42000 : 4.691221898317337
Average loss at step 44000 : 4.628633765101433
Average loss at step 46000 : 4.6018352398872375
Average loss at step 48000 : 4.5429726408720015
Average loss at step 50000 : 4.501960401535034
Nearest to and: amp, but, from, or, for, to, of, that,
Nearest to To: to, In, Ja, For, On, eavesdropann, Singer, integration,
Nearest to I: they, we, he, you, i, You, We, she,
Nearest to If: epidemic, LIE, mothers, instructions, So, BOY, When, UNK,
Nearest to co: via, RT, Chainsaw, Goldman, Sh, concussion, WHAA, lasts,
Nearest to The: the, AlexaKaySam, A, testament, DaysRT, BONG, Activists, Astounding,
Nearest to their: your, our, his, the, my, her, Rushs_Dad, WHAA,
Nearest to What: Why, It, silencing, designed, tagging, This, BehindTheRant, what,
Nearest to she: he, it, that, I, Reinstate, they, there, She,
Nearest to would: could, will, ll, can, may, RedRightRepubRT, hillarysemail, s,
Nearest to like: haven, hide, database, TrumpWeiners4, ogkeyser, fatalmocking, MikeRumbo66, shut,
Nearest to be: have, get, is, Bricks, kill, peach, consider, not,
Nearest to A: Bias, a, The, Langston, welcoming, DocThompsonShow, pantsRT, needs,
Nearest to s: is, was, has, and, Determine, says, Is, Missed,
Nearest to they: you, we, I, We, he, They, it, she,
Nearest to Trump: Obama, Hillary, Clinton, Sanders, Cruz, realDonaldTrump, injures, Youngstown,
Average loss at step 52000 : 4.48813548374176
Average loss at step 54000 : 4.4395130319595335
Average loss at step 56000 : 4.4136402312517165
Average loss at step 58000 : 4.387754002332687
Average loss at step 60000 : 4.340440725684166
Nearest to and: amp, but, or, from, that, as, for, when,
Nearest to To: to, In, For, On, Ja, eavesdropann, With, beginning,
Nearest to I: they, we, he, you, i, You, We, she,
Nearest to If: UNK, epidemic, When, LIE, So, instructions, mothers, Thank,
Nearest to co: via, Sh, Chainsaw, concussion, lasts, t, WHAA, politicsRT,
Nearest to The: the, AlexaKaySam, DaysRT, testament, LarsWienand, Astounding, agendaRT, A,
Nearest to their: your, our, his, my, the, her, WHAA, Rushs_Dad,
Nearest to What: Why, It, silencing, designed, what, BehindTheRant, This, tagging,
Nearest to she: he, it, that, Reinstate, they, She, there, I,
Nearest to would: could, will, ll, can, may, should, RedRightRepubRT, hillarysemail,
Nearest to like: haven, hide, database, fatalmocking, Everywhere, shut, MikeRumbo66, TrumpWeiners4,
Nearest to be: have, get, Bricks, was, not, is, been, Mesmerizing,
Nearest to A: a, Bias, Langston, The, DocThompsonShow, welcoming, Rican, DJ,
Nearest to s: is, was, has, Determine, says, Missed, shd, Is,
Nearest to they: you, we, I, We, he, They, she, accurately,
Nearest to Trump: Hillary, Obama, Clinton, Sanders, Cruz, realDonaldTrump, injures, Sucks,
Average loss at step 62000 : 4.351320527672768
Average loss at step 64000 : 4.3232111774683
Average loss at step 66000 : 4.3200337978601455
Average loss at step 68000 : 4.305860041022301
Average loss at step 70000 : 4.30841260433197
Nearest to and: amp, but, or, from, that, when, as, of,
Nearest to To: to, In, For, Ja, On, Should, Will, With,
Nearest to I: we, they, i, he, you, You, We, she,
Nearest to If: When, So, epidemic, LIE, Thank, instructions, mothers, UNK,
Nearest to co: via, Chainsaw, Sh, concussion, politicsRT, lasts, WHAA, RT,
Nearest to The: the, AlexaKaySam, testament, agendaRT, DaysRT, LarsWienand, Astounding, Activists,
Nearest to their: your, our, his, her, my, the, Rushs_Dad, WHAA,
Nearest to What: Why, It, silencing, what, designed, BehindTheRant, This, Debra,
Nearest to she: he, it, they, She, that, He, Reinstate, there,
Nearest to would: will, could, can, ll, may, should, RedRightRepubRT, must,
Nearest to like: hide, haven, database, Everywhere, fatalmocking, TrumpWeiners4, Consumer, MikeRumbo66,
Nearest to be: have, get, been, was, is, not, Bricks, kill,
Nearest to A: a, Bias, Langston, DocThompsonShow, DJ, welcoming, The, Rican,
Nearest to s: is, was, has, Determine, says, Is, Missed, Illegally,
Nearest to they: we, you, I, he, They, We, she, accurately,
Nearest to Trump: Obama, Hillary, Clinton, Sanders, Cruz, realDonaldTrump, rtifyouare, HillaryClinton,
Average loss at step 72000 : 4.303540543317795
Average loss at step 74000 : 4.2762672786712645
Average loss at step 76000 : 4.291633704900741
Average loss at step 78000 : 4.279861431837082
Average loss at step 80000 : 4.279908079624176
Nearest to and: amp, but, or, from, when, that, as, of,
Nearest to To: to, For, In, Should, Will, On, With, Ja,
Nearest to I: we, they, i, he, you, You, We, They,
Nearest to If: When, Thank, So, LIE, epidemic, mothers, instructions, And,
Nearest to co: via, Chainsaw, Sh, concussion, politicsRT, WHAA, Oz, lasts,
Nearest to The: the, AlexaKaySam, testament, agendaRT, DaysRT, Astounding, LarsWienand, BONG,
Nearest to their: your, his, our, her, my, the, WHAA, Rushs_Dad,
Nearest to What: Why, It, silencing, what, designed, BehindTheRant, This, Debra,
Nearest to she: he, they, She, it, He, there, that, Reinstate,
Nearest to would: will, could, ll, can, may, should, d, must,
Nearest to like: hide, haven, fatalmocking, TrumpWeiners4, database, Consumer, Everywhere, Motor,
Nearest to be: have, get, been, was, is, not, Bricks, kill,
Nearest to A: a, Bias, DocThompsonShow, Langston, DJ, welcoming, needs, Rican,
Nearest to s: is, was, has, Determine, says, would, Missed, shd,
Nearest to they: we, you, I, he, They, she, We, there,
Nearest to Trump: Obama, Hillary, Sanders, Clinton, Cruz, trump, realDonaldTrump, rtifyouare,
Average loss at step 82000 : 4.271103955745697
Average loss at step 84000 : 4.266893053650856
Average loss at step 86000 : 4.257736647248268
Average loss at step 88000 : 4.241679049134254
Average loss at step 90000 : 4.245081868171692
Nearest to and: amp, but, or, from, when, that, for, as,
Nearest to To: to, For, In, Will, Should, On, With, Pauvres,
Nearest to I: they, we, i, he, You, you, she, We,
Nearest to If: When, Thank, So, And, LIE, instructions, mothers, epidemic,
Nearest to co: via, Sh, Chainsaw, politicsRT, concussion, WHAA, Oz, TimBartenderRT,
Nearest to The: the, AlexaKaySam, DaysRT, Astounding, testament, agendaRT, LarsWienand, No,
Nearest to their: your, his, our, her, my, the, WHAA, stolen,
Nearest to What: Why, It, silencing, what, BehindTheRant, designed, This, Debra,
Nearest to she: he, She, they, He, it, there, I, that,
Nearest to would: will, could, can, ll, may, should, must, d,
Nearest to like: hide, haven, fatalmocking, database, Everywhere, couldn, Consumer, TrumpWeiners4,
Nearest to be: have, get, been, was, is, Bricks, m, not,
Nearest to A: a, Bias, Langston, DocThompsonShow, DJ, welcoming, Rican, OPPOSE,
Nearest to s: is, was, has, Determine, says, Missed, would, takes,
Nearest to they: we, you, I, he, They, she, We, there,
Nearest to Trump: Obama, Hillary, Clinton, Sanders, trump, Cruz, realDonaldTrump, rtifyouare,
Average loss at step 92000 : 4.242088904261589
Average loss at step 94000 : 4.247817409873009
Average loss at step 96000 : 4.244096995711327
Average loss at step 98000 : 4.2387411421537395
Average loss at step 100000 : 4.226129394173622
Nearest to and: amp, but, or, from, when, as, thought, for,
Nearest to To: to, For, In, Will, On, Should, With, Pauvres,
Nearest to I: i, they, we, he, You, you, We, They,
Nearest to If: When, Thank, So, And, LIE, epidemic, mothers, UNK,
Nearest to co: via, Sh, Chainsaw, politicsRT, concussion, shore, Oz, zijn,
Nearest to The: the, DaysRT, Astounding, testament, AlexaKaySam, agendaRT, standardnews, No,
Nearest to their: your, his, our, her, my, the, WHAA, stolen,
Nearest to What: Why, It, what, silencing, This, designed, When, BehindTheRant,
Nearest to she: he, they, She, it, He, there, I, that,
Nearest to would: will, could, ll, can, may, should, must, d,
Nearest to like: hide, haven, database, TrumpWeiners4, fatalmocking, Motor, Consumer, Everywhere,
Nearest to be: have, get, been, was, is, m, kill, Mesmerizing,
Nearest to A: a, Langston, Bias, DocThompsonShow, LAWS, DJ, OPPOSE, Rican,
Nearest to s: is, was, has, takes, Determine, says, Missed, Is,
Nearest to they: we, you, he, I, she, They, We, there,
Nearest to Trump: Obama, Hillary, Sanders, trump, Clinton, Cruz, realDonaldTrump, amwriting,
Word2vec total output:
Manes-MacBook-Pro:Project manewilliams$ python3 tweets_word2vec.py
WARNING:tensorflow:From /Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/contrib/learn/python/learn/datasets/base.py:198: retry (from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will be removed in a future version.
Instructions for updating:
Use the retry module or similar alternatives.
Data size 4421054
Most common words (+UNK) [['UNK', 329446], ('the', 124586), ('to', 97262), ('a', 69242), ('of', 62113)]
Sample data [0, 20, 7, 86, 7428, 61, 1360, 548, 0, 38] ['UNK', 'clinton', 'trump', 'should', 'apologized', 'more', 'attacked', 'less', 'UNK', 'who']
20 clinton -> 0 UNK
20 clinton -> 7 trump
7 trump -> 86 should
7 trump -> 20 clinton
86 should -> 7428 apologized
86 should -> 7 trump
7428 apologized -> 61 more
7428 apologized -> 86 should
WARNING:tensorflow:From tweets_word2vec.py:289: calling reduce_sum (from tensorflow.python.ops.math_ops) with keep_dims is deprecated and will be removed in a future version.
Instructions for updating:
keep_dims is deprecated, use keepdims instead
2018-05-05 01:32:47.029220: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
Initialized
Average loss at step 0 : 296.1373291015625
Nearest to will: 100b, magaselfie, fútbol, lace, 60, caring, appt, bure,
Nearest to us: complicating, bahnhof, paintings, prerequisite, emeka, stor, assad, iota,
Nearest to says: wembley, beat, developer, sevenlayercake, attacked, propose, clintoncash, hong,
Nearest to up: outlaw, indveng, recognizes, daveweigel, 3pm, jeanie, mitarbeiter, radicalislam,
Nearest to never: tierrah46, lacy0902, wooohooo, statewide, clan, cabinet, amounted, dho,
Nearest to would: richmovies, addcartoonstohistory, betcha, anarchy, hatchet, o3ik907v7c, n, trumpsfirstorder,
Nearest to so: fixer, fanatical, türkische, kenny, fetches, 8319pswn6d, a7, acct,
Nearest to should: c0dnfytha1, tuberculosis, ignacio, cun, gifts, opadqqow53, wildly, smiley,
Nearest to good: untappd, swing, bert, reacting, scratch, leftistviolence, vivaces, bill,
Nearest to of: winery, dating, jtwwytojk9, trishintel, claimed, grandfather, 487, afterbirth,
Nearest to like: sig, goldenshowers, notwithher, operative, khanoisseur, toconservatives, sabéis, realdonaldtru,
Nearest to realdonaldtrump: baiter, bauer, providence, disappearing, bombard, dougschoen, icihgbrter, birmingham,
Nearest to at: ap8bfcwbd4, ज, sally, legislatives2017, fathersday, anyone, marryhandy, obamas,
Nearest to me: kashmir, 1776trump, btw, halves, zahra, yep, s1osjv4ftq, mcmurray,
Nearest to more: melanie, bárcenas, keithscott, ibelieveinipa, florid, wher, wahlprogrammen, regierungserklärung,
Nearest to his: win, onstage, qim, sli, reactions, chnvhliuzb, mores, heute,
Average loss at step 2000 : 117.58672986125946
Average loss at step 4000 : 54.86665396404266
Average loss at step 6000 : 34.74959244060516
Average loss at step 8000 : 24.08619478774071
Average loss at step 10000 : 18.083574054956436
Nearest to will: can, and, to, gonna, lou, jewish, installment, ads,
Nearest to us: bigger, gohomederay, the, whaa, brian, worked, pleased, demand,
Nearest to says: clintoncash, beat, 3qdisjyhb9, univision, stay, wembley, lie, attacked,
Nearest to up: outlaw, oct, community, behind, gifts, lick, intentionally, laughs,
Nearest to never: bong, cabinet, letting, bound, editorial, lacy0902, finds, makeamericagreatagain,
Nearest to would: might, n, bong, richmovies, addcartoonstohistory, dictator, 16, icelebratetrumpwith,
Nearest to so: nap, got, managed, imnotopposedto, trending, years, fixer, seth,
Nearest to should: bong, baked, vegasgopdebate, gifts, republic, fakes, visitors, toss,
Nearest to good: swing, fundraising, lion, doctor, ceremony, trump2016, stamp, scratch,
Nearest to of: in, for, on, is, with, to, and, smiles,
Nearest to like: operative, lawyers, ge, differences, shld, toconservatives, testimony, scared,
Nearest to realdonaldtrump: defund, disappearing, whaa, 2011, topnews, fame, assange, ruling,
Nearest to at: brings, ज, rebootliberty, obamas, of, fusion, curse, ap8bfcwbd4,
Nearest to me: whaa, financial, yep, btw, fixes, you, fish, notmypresident,
Nearest to more: whaa, grandpa, wheniamonmyowni, olympia, king, 84, melanie, cheat,
Nearest to his: play, the, win, heute, obama, bras, dx1xhel8tk, usarubric,
Average loss at step 12000 : 14.359438372850418
Average loss at step 14000 : 11.906547516822815
Average loss at step 16000 : 10.24623385643959
Average loss at step 18000 : 9.032378518104553
Average loss at step 20000 : 8.143683428764342
Nearest to will: can, and, to, gonna, should, cookout, jaws, would,
Nearest to us: bigger, the, warm, lunatics, brian, worked, paintings, pleased,
Nearest to says: clintoncash, 3qdisjyhb9, and, univision, doesn, beat, is, clarify,
Nearest to up: outlaw, oct, gifts, community, behind, wasteamillionin3words, intentionally, onewordoffbooks,
Nearest to never: tierrah46, you, may, bound, asylum, wont, lacy0902, editorial,
Nearest to would: might, n, richmovies, should, will, thehashpod, bong, addcartoonstohistory,
Nearest to so: got, nap, managed, but, that, fixer, trending, UNK,
Nearest to should: will, baked, don, would, toss, and, bong, vegasgopdebate,
Nearest to good: swing, scratch, stamp, fundraising, closet, sore, horribly, ceremony,
Nearest to of: for, in, and, on, buried, with, at, newshour,
Nearest to like: ge, operative, lawyers, were, than, optimize, scared, deemed,
Nearest to realdonaldtrump: defund, disappearing, topnews, whaa, liberry, fame, neverhillary, baiter,
Nearest to at: in, brings, ज, of, for, and, curse, obamas,
Nearest to me: you, financial, yep, whaa, fixes, btw, fish, notmypresident,
Nearest to more: whaa, keithscott, grandpa, these, heating, wher, 84, nest,
Nearest to his: your, her, the, dx1xhel8tk, play, obama, trump, my,
Average loss at step 22000 : 7.3056180014610295
Average loss at step 24000 : 6.8485847411155705
Average loss at step 26000 : 6.434102659702301
Average loss at step 28000 : 6.038018933057785
Average loss at step 30000 : 5.762489190578461
Nearest to will: can, to, gonna, and, should, would, could, cookout,
Nearest to us: me, bigger, lunatics, warm, our, the, reputation, brian,
Nearest to says: clintoncash, 3qdisjyhb9, doesn, and, univision, maxtu2qlpm, has, is,
Nearest to up: outlaw, nohostagestaken, behind, wasteamillionin3words, gifts, them, f35, min,
Nearest to never: tierrah46, may, i, you, bound, wont, asylum, armani,
Nearest to would: might, will, should, can, n, richmovies, to, thehashpod,
Nearest to so: that, but, managed, nap, got, too, fixer, kenny,
Nearest to should: will, would, don, baked, and, toss, terencecrutcher, vegasgopdebate,
Nearest to good: closet, sore, yr, stamp, scratch, ceremony, daytona500, horribly,
Nearest to of: for, in, and, buried, with, from, is, at,
Nearest to like: ge, operative, were, optimize, lawyers, deemed, than, jong,
Nearest to realdonaldtrump: defund, disappearing, UNK, neverhillary, liberry, topnews, now, whaa,
Nearest to at: in, brings, ज, for, on, and, of, obamas,
Nearest to me: you, fixes, financial, them, whaa, yep, us, pedogate,
Nearest to more: whaa, wher, ign, these, nest, trashed, keithscott, selma,
Nearest to his: your, her, my, the, their, dx1xhel8tk, obama, trump,
Average loss at step 32000 : 5.567626395225525
/'
"Average loss at step 34000 : 5.455491584062576
"Average loss at step 36000 : 5.2876623101234435
Average loss at step 38000 : 5.237178799510002
Average loss at step 40000 : 6.847827666759491
Nearest to will: can, should, to, would, gonna, could, and, won,
Nearest to us: me, freedomcities, lunatics, our, bigger, them, warm, america,
Nearest to says: doesn, 3qdisjyhb9, clintoncash, has, maxtu2qlpm, and, univision, but,
Nearest to up: outlaw, nohostagestaken, them, min, behind, wasteamillionin3words, defenseless, aft,
Nearest to never: tierrah46, may, i, mormonism, you, wont, bound, not,
Nearest to would: will, should, might, can, to, richmovies, don, doesn,
Nearest to so: too, but, that, got, UNK, managed, nap, studies,
Nearest to should: will, would, don, can, baked, and, toss, to,
Nearest to good: great, sore, jnr, yr, closet, stamp, freedomcities, scratch,
Nearest to of: for, in, and, buried, from, by, newshour, with,
Nearest to like: ge, mean, were, operative, optimize, psychedelic, 305, feminist,
Nearest to realdonaldtrump: UNK, defund, neverhillary, disappearing, trump, whaa, liberry, now,
Nearest to at: in, brings, for, and, on, after, ज, obamas,
Nearest to me: them, you, us, him, financial, fixes, whaa, yep,
Nearest to more: whaa, wher, better, ign, these, caves, trashed, selma,
Nearest to his: her, your, my, their, the, trump, dx1xhel8tk, obama,
Average loss at step 42000 : 6.2939271893501285
Average loss at step 44000 : 5.825182631015777
Average loss at step 46000 : 5.508351857423782
Average loss at step 48000 : 5.3009319543838505
Average loss at step 50000 : 5.192287965536117
Nearest to will: can, should, would, could, to, gonna, and, won,
Nearest to us: me, them, our, u, lunatics, america, freedomcities, cleats,
Nearest to says: doesn, has, 3qdisjyhb9, clintoncash, and, maxtu2qlpm, but, univision,
Nearest to up: outlaw, comebacks, them, nohostagestaken, dianne, decisive, off, min,
Nearest to never: ever, mormonism, tierrah46, i, you, not, may, we,
Nearest to would: will, should, can, might, to, don, doesn, may,
Nearest to so: too, but, that, UNK, how, managed, as, nap,
Nearest to should: will, would, can, must, don, and, to, may,
Nearest to good: great, sore, jnr, small, yr, scratch, closet, this,
Nearest to of: for, in, and, buried, from, is, by, therealnews,
Nearest to like: ge, mean, were, psychedelic, 305, make, operative, optimize,
Nearest to realdonaldtrump: trump, UNK, now, neverhillary, disappearing, hillaryclinton, liberry, defund,
Nearest to at: in, after, brings, for, and, on, ज, obamas,
Nearest to me: them, you, him, us, fixes, financial, whaa, pedogate,
Nearest to more: whaa, better, wher, ign, caves, these, nest, bengals,
Nearest to his: your, her, my, their, the, dx1xhel8tk, trump, our,
Average loss at step 52000 : 5.063740147829056
Average loss at step 54000 : 4.971848911881446
Average loss at step 56000 : 4.888473462224007
Average loss at step 58000 : 4.8398058959245684
Average loss at step 60000 : 4.789416933655739
Nearest to will: can, should, would, could, gonna, to, and, must,
Nearest to us: me, them, u, our, lunatics, freedomcities, america, cleats,
Nearest to says: doesn, has, 3qdisjyhb9, but, clintoncash, and, maxtu2qlpm, dun,
Nearest to up: outlaw, comebacks, nohostagestaken, them, dianne, ano, down, off,
Nearest to never: ever, not, tierrah46, mormonism, i, just, kcstar, that,
Nearest to would: will, should, can, might, to, don, may, and,
Nearest to so: too, but, how, as, really, managed, UNK, that,
Nearest to should: will, would, can, must, don, to, may, could,
Nearest to good: great, small, sore, scratch, this, jnr, meaningless, lizs_heart,
Nearest to of: in, for, from, buried, therealnews, ʷʰʸ, and, about,
Nearest to like: mean, psychedelic, make, ge, than, optimize, 305, were,
Nearest to realdonaldtrump: UNK, trump, now, hillaryclinton, yes, neverhillary, liberry, gaylovematters,
Nearest to at: in, brings, after, for, on, and, ज, obamas,
Nearest to me: them, him, us, you, fixes, financial, whaa, sca,
Nearest to more: better, whaa, most, ign, wher, caves, bengals, dawnellemurray,
Nearest to his: your, her, my, their, the, its, our, trump,
Average loss at step 62000 : 4.773062491297722
Average loss at step 64000 : 4.733378041744232
Average loss at step 66000 : 4.707176345109939
Average loss at step 68000 : 4.679180788755417
Average loss at step 70000 : 4.726253412604332
Nearest to will: can, would, should, could, gonna, must, may, to,
Nearest to us: u, me, them, our, america, lunatics, cleats, freedomcities,
Nearest to says: doesn, has, but, 3qdisjyhb9, and, did, clintoncash, maxtu2qlpm,
Nearest to up: outlaw, comebacks, down, nohostagestaken, them, decisive, off, ano,
Nearest to never: ever, just, not, i, tierrah46, mormonism, that, kcstar,
Nearest to would: will, should, can, might, to, could, may, must,
Nearest to so: too, but, how, as, not, really, that, ʷʰʸ,
Nearest to should: will, would, must, can, don, could, may, to,
Nearest to good: great, small, sore, scratch, lizs_heart, meaningless, jnr, yr,
Nearest to of: for, ʷʰʸ, in, from, buried, therealnews, and, newshour,
Nearest to like: mean, psychedelic, optimize, ge, make, ielvisgomes, allergic, 305,
Nearest to realdonaldtrump: trump, hillaryclinton, now, neverhillary, gaylovematters, liberry, potus, yes,
Nearest to at: in, brings, after, on, for, obamas, curse, picked,
Nearest to me: them, him, us, you, fixes, her, pedogate, whaa,
Nearest to more: better, most, whaa, wher, ign, caves, less, bengals,
Nearest to his: your, her, their, my, its, the, our, trump,
Average loss at step 72000 : 4.7247595744133
Average loss at step 74000 : 4.6746429134607315
Average loss at step 76000 : 4.650019417047501
Average loss at step 78000 : 4.642708010435104
Average loss at step 80000 : 4.6193217248916625
Nearest to will: would, can, should, could, gonna, may, must, won,
Nearest to us: u, them, me, our, america, freedomcities, cleats, him,
Nearest to says: doesn, has, 3qdisjyhb9, but, would, clintoncash, beyoncé, said,
Nearest to up: outlaw, comebacks, down, nohostagestaken, decisive, off, them, ano,
Nearest to never: ever, i, just, tierrah46, mormonism, not, kcstar, always,
Nearest to would: will, should, can, might, may, could, must, doesn,
Nearest to so: too, but, how, really, as, ʷʰʸ, very, trending,
Nearest to should: will, would, must, can, don, may, could, to,
Nearest to good: great, small, lizs_heart, bad, meaningless, sore, scratch, yr,
Nearest to of: for, from, buried, freebiafra, therealnews, ʷʰʸ, ll90iekoet, in,
Nearest to like: mean, optimize, psychedelic, feminist, allergic, ge, ielvisgomes, make,
Nearest to realdonaldtrump: trump, hillaryclinton, gaylovematters, now, potus, neverhillary, yes, liberry,
Nearest to at: brings, in, after, obamas, legislatives2017, fathersday, pecker, curse,
Nearest to me: them, him, us, you, her, myself, fixes, mkd9wroe9e,
Nearest to more: better, most, whaa, less, ign, caves, ʷʰʸ, bengals,
Nearest to his: her, your, their, my, its, the, trump, our,
Average loss at step 82000 : 4.610889441967011
Average loss at step 84000 : 4.607703956007957
Average loss at step 86000 : 4.589995376706123
Average loss at step 88000 : 4.583386127948761
Average loss at step 90000 : 4.576929998755455
Nearest to will: should, would, can, could, gonna, may, won, must,
Nearest to us: u, them, me, our, america, cleats, lunatics, him,
Nearest to says: doesn, has, 3qdisjyhb9, said, beyoncé, knows, wants, would,
Nearest to up: outlaw, down, nohostagestaken, comebacks, f35, ano, decisive, off,
Nearest to never: ever, just, i, always, not, mormonism, tierrah46, wont,
Nearest to would: will, should, can, might, could, may, must, doesn,
Nearest to so: too, but, how, as, really, very, ʷʰʸ, carpoolkaraoke,
Nearest to should: will, would, must, can, could, don, may, to,
Nearest to good: great, bad, small, lizs_heart, meaningless, scratch, sore, disowning,
Nearest to of: for, buried, ʷʰʸ, from, imperative, freebiafra, tbyg, newshour,
Nearest to like: mean, optimize, psychedelic, feminist, make, ge, ielvisgomes, allergic,
Nearest to realdonaldtrump: trump, hillaryclinton, potus, now, neverhillary, yes, gaylovematters, liberry,
Nearest to at: brings, in, after, fathersday, obamas, pecker, sellbeats, legislatives2017,
Nearest to me: them, him, us, her, you, myself, mkd9wroe9e, fixes,
Nearest to more: better, most, less, whaa, caves, ign, ʷʰʸ, worse,
Nearest to his: her, your, their, my, its, the, our, ur,
Average loss at step 92000 : 4.577942245841026
Average loss at step 94000 : 4.570807527422905
Average loss at step 96000 : 4.5589815218448635
Average loss at step 98000 : 4.546109932661056
Average loss at step 100000 : 4.539864033818245
Nearest to will: should, would, can, could, gonna, may, must, won,
Nearest to us: u, me, them, our, him, america, lunatics, american,
Nearest to says: doesn, said, has, would, 3qdisjyhb9, dun, did, beyoncé,
Nearest to up: outlaw, nohostagestaken, comebacks, down, off, ano, f35, decisive,
Nearest to never: ever, always, just, wont, finally, not, tierrah46, kcstar,
Nearest to would: will, should, can, might, could, may, doesn, must,
Nearest to so: too, but, how, as, really, ʷʰʸ, very, carpoolkaraoke,
Nearest to should: will, would, must, can, could, don, may, didn,
Nearest to good: great, bad, small, lizs_heart, meaningless, scratch, disowning, real,
Nearest to of: buried, ʷʰʸ, for, tbyg, ll90iekoet, in, imperative, cathedral,
Nearest to like: mean, psychedelic, optimize, feminist, ge, stre, ielvisgomes, want,
Nearest to realdonaldtrump: trump, hillaryclinton, potus, neverhillary, now, yes, UNK, gaylovematters,
Nearest to at: brings, 200m, fathersday, in, legislatives2017, pecker, obamas, presidentbannon,
Nearest to me: them, him, us, myself, her, you, fixes, mkd9wroe9e,
Nearest to more: less, better, most, caves, ign, whaa, worse, wiretapped,
Nearest to his: her, their, your, my, its, the, dx1xhel8tk, our,