ZachW commited on
Commit
6507c9f
·
verified ·
1 Parent(s): a265173

Upload extrapolation artifact: rl_ckpt18809_op11_14_step200_merged_hf far-OOD metrics

Browse files
020305/evals/far_ood/rl_ckpt18809_op11_14_step200_merged_hf/rl_ckpt18809_op11_14_step200_merged_hf_metrics.json ADDED
@@ -0,0 +1,904 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "checkpoint": "rl_ckpt18809_op11_14_step200_merged_hf",
3
+ "id": {
4
+ "answer_accuracy": 0.0,
5
+ "avg_response_len": 0.0,
6
+ "avg_loss": 0.0,
7
+ "count": 0,
8
+ "validation_metrics": {},
9
+ "pass_at_k": {},
10
+ "per_op_pass_at_k": {},
11
+ "per_op_accuracy": {},
12
+ "per_op_avg_response_len": {},
13
+ "per_op_avg_loss": {},
14
+ "per_op_length": {},
15
+ "per_op_loss": {},
16
+ "per_template": {
17
+ "crazy_zootopia": {
18
+ "count": 0,
19
+ "correct": 0,
20
+ "answer_accuracy": 0.0,
21
+ "avg_response_len": 0.0,
22
+ "resp_tokens_sum": 0,
23
+ "pass_at_k": {},
24
+ "per_op_pass_at_k": {}
25
+ },
26
+ "teachers_in_school": {
27
+ "count": 0,
28
+ "correct": 0,
29
+ "answer_accuracy": 0.0,
30
+ "avg_response_len": 0.0,
31
+ "resp_tokens_sum": 0,
32
+ "pass_at_k": {},
33
+ "per_op_pass_at_k": {}
34
+ },
35
+ "movie_festival_awards": {
36
+ "count": 0,
37
+ "correct": 0,
38
+ "answer_accuracy": 0.0,
39
+ "avg_response_len": 0.0,
40
+ "resp_tokens_sum": 0,
41
+ "pass_at_k": {},
42
+ "per_op_pass_at_k": {}
43
+ }
44
+ }
45
+ },
46
+ "ood": {
47
+ "answer_accuracy": 0.0025,
48
+ "avg_response_len": 298.277109375,
49
+ "avg_loss": 0.0,
50
+ "count": 64000,
51
+ "validation_metrics": {},
52
+ "pass_at_k": {
53
+ "pass@1": 0.0025,
54
+ "pass@2": 0.00463608870967742,
55
+ "pass@4": 0.008184218576195775,
56
+ "pass@8": 0.013711085203882761,
57
+ "pass@16": 0.022133159099567364,
58
+ "pass@32": 0.0345
59
+ },
60
+ "per_op_pass_at_k": {
61
+ "21": {
62
+ "pass@1": 0.0034375,
63
+ "pass@2": 0.006653225806451612,
64
+ "pass@4": 0.01250194660734149,
65
+ "pass@8": 0.022337999011247067,
66
+ "pass@16": 0.037052589296749476,
67
+ "pass@32": 0.055
68
+ },
69
+ "22": {
70
+ "pass@1": 0.00359375,
71
+ "pass@2": 0.0070665322580645154,
72
+ "pass@4": 0.013669215795328142,
73
+ "pass@8": 0.02563542825361513,
74
+ "pass@16": 0.045553392658509446,
75
+ "pass@32": 0.075
76
+ },
77
+ "23": {
78
+ "pass@1": 0.001875,
79
+ "pass@2": 0.0033870967741935483,
80
+ "pass@4": 0.005643770856507231,
81
+ "pass@8": 0.00851692336214027,
82
+ "pass@16": 0.01249796069707082,
83
+ "pass@32": 0.02
84
+ },
85
+ "24": {
86
+ "pass@1": 0.005,
87
+ "pass@2": 0.008981854838709677,
88
+ "pass@4": 0.01523567853170189,
89
+ "pass@8": 0.024872691879866515,
90
+ "pass@16": 0.03990823009547857,
91
+ "pass@32": 0.06
92
+ },
93
+ "25": {
94
+ "pass@1": 0.00328125,
95
+ "pass@2": 0.005897177419354839,
96
+ "pass@4": 0.009723998887652947,
97
+ "pass@8": 0.014333853854710364,
98
+ "pass@16": 0.019982385151510268,
99
+ "pass@32": 0.03
100
+ },
101
+ "26": {
102
+ "pass@1": 0.0025,
103
+ "pass@2": 0.004213709677419355,
104
+ "pass@4": 0.006336067853170189,
105
+ "pass@8": 0.008714071190211345,
106
+ "pass@16": 0.012499991939514114,
107
+ "pass@32": 0.02
108
+ },
109
+ "27": {
110
+ "pass@1": 0.0015625,
111
+ "pass@2": 0.0030645161290322577,
112
+ "pass@4": 0.0058951612903225805,
113
+ "pass@8": 0.010919354838709676,
114
+ "pass@16": 0.018870967741935484,
115
+ "pass@32": 0.03
116
+ },
117
+ "28": {
118
+ "pass@1": 0.00171875,
119
+ "pass@2": 0.0031451612903225807,
120
+ "pass@4": 0.005337041156840933,
121
+ "pass@8": 0.008118126978694276,
122
+ "pass@16": 0.011284204671857619,
123
+ "pass@32": 0.015
124
+ },
125
+ "29": {
126
+ "pass@1": 0.0015625,
127
+ "pass@2": 0.0030241935483870967,
128
+ "pass@4": 0.005684788654060067,
129
+ "pass@8": 0.010194660734149055,
130
+ "pass@16": 0.01739154616240267,
131
+ "pass@32": 0.03
132
+ },
133
+ "30": {
134
+ "pass@1": 0.00046875,
135
+ "pass@2": 0.0009274193548387094,
136
+ "pass@4": 0.0018145161290322581,
137
+ "pass@8": 0.0034677419354838712,
138
+ "pass@16": 0.006290322580645161,
139
+ "pass@32": 0.01
140
+ }
141
+ },
142
+ "per_op_accuracy": {
143
+ "21": 0.0034375,
144
+ "22": 0.00359375,
145
+ "23": 0.001875,
146
+ "24": 0.005,
147
+ "25": 0.00328125,
148
+ "26": 0.0025,
149
+ "27": 0.0015625,
150
+ "28": 0.00171875,
151
+ "29": 0.0015625,
152
+ "30": 0.00046875
153
+ },
154
+ "per_op_avg_response_len": {
155
+ "21": 292.32015625,
156
+ "22": 296.5990625,
157
+ "23": 294.07046875,
158
+ "24": 299.96671875,
159
+ "25": 298.75125,
160
+ "26": 304.253125,
161
+ "27": 305.4503125,
162
+ "28": 295.45109375,
163
+ "29": 299.44328125,
164
+ "30": 296.465625
165
+ },
166
+ "per_op_avg_loss": {},
167
+ "per_op_length": {
168
+ "21": 292.32015625,
169
+ "22": 296.5990625,
170
+ "23": 294.07046875,
171
+ "24": 299.96671875,
172
+ "25": 298.75125,
173
+ "26": 304.253125,
174
+ "27": 305.4503125,
175
+ "28": 295.45109375,
176
+ "29": 299.44328125,
177
+ "30": 296.465625
178
+ },
179
+ "per_op_loss": {},
180
+ "per_template": {
181
+ "crazy_zootopia": {
182
+ "count": 23072,
183
+ "correct": 60,
184
+ "answer_accuracy": 0.0026005547850208046,
185
+ "avg_response_len": 276.871012482663,
186
+ "resp_tokens_sum": 6387968,
187
+ "pass_at_k": {
188
+ "pass@1": 0.0026005547850208046,
189
+ "pass@2": 0.004667017135698627,
190
+ "pass@4": 0.007791173425859214,
191
+ "pass@8": 0.012087750689001269,
192
+ "pass@16": 0.018247633171405726,
193
+ "pass@32": 0.027739251040221916
194
+ },
195
+ "per_op_pass_at_k": {
196
+ "21": {
197
+ "pass@1": 0.0011867088607594937,
198
+ "pass@2": 0.0023734177215189874,
199
+ "pass@4": 0.004746835443037975,
200
+ "pass@8": 0.00949367088607595,
201
+ "pass@16": 0.0189873417721519,
202
+ "pass@32": 0.0379746835443038
203
+ },
204
+ "22": {
205
+ "pass@1": 0.0026041666666666665,
206
+ "pass@2": 0.005124327956989246,
207
+ "pass@4": 0.009923835125448028,
208
+ "pass@8": 0.018637992831541217,
209
+ "pass@16": 0.033154121863799284,
210
+ "pass@32": 0.05555555555555555
211
+ },
212
+ "23": {
213
+ "pass@1": 0.0,
214
+ "pass@2": 0.0,
215
+ "pass@4": 0.0,
216
+ "pass@8": 0.0,
217
+ "pass@16": 0.0,
218
+ "pass@32": 0.0
219
+ },
220
+ "24": {
221
+ "pass@1": 0.008802816901408451,
222
+ "pass@2": 0.014964788732394365,
223
+ "pass@4": 0.023108618339626186,
224
+ "pass@8": 0.033562487270676344,
225
+ "pass@16": 0.049522940531216975,
226
+ "pass@32": 0.07042253521126761
227
+ },
228
+ "25": {
229
+ "pass@1": 0.009424603174603174,
230
+ "pass@2": 0.016737071172555042,
231
+ "pass@4": 0.026901583770326816,
232
+ "pass@8": 0.037567790014953535,
233
+ "pass@16": 0.04756312746511196,
234
+ "pass@32": 0.06349206349206349
235
+ },
236
+ "26": {
237
+ "pass@1": 0.0,
238
+ "pass@2": 0.0,
239
+ "pass@4": 0.0,
240
+ "pass@8": 0.0,
241
+ "pass@16": 0.0,
242
+ "pass@32": 0.0
243
+ },
244
+ "27": {
245
+ "pass@1": 0.00043402777777777775,
246
+ "pass@2": 0.0008680555555555555,
247
+ "pass@4": 0.001736111111111111,
248
+ "pass@8": 0.003472222222222222,
249
+ "pass@16": 0.006944444444444444,
250
+ "pass@32": 0.013888888888888888
251
+ },
252
+ "28": {
253
+ "pass@1": 0.0035919540229885057,
254
+ "pass@2": 0.006511865035224323,
255
+ "pass@4": 0.010832278521473412,
256
+ "pass@8": 0.015788797652170748,
257
+ "pass@16": 0.020193573958293378,
258
+ "pass@32": 0.022988505747126436
259
+ },
260
+ "29": {
261
+ "pass@1": 0.0004166666666666667,
262
+ "pass@2": 0.0008333333333333334,
263
+ "pass@4": 0.0016666666666666668,
264
+ "pass@8": 0.0033333333333333335,
265
+ "pass@16": 0.006666666666666667,
266
+ "pass@32": 0.013333333333333334
267
+ },
268
+ "30": {
269
+ "pass@1": 0.0,
270
+ "pass@2": 0.0,
271
+ "pass@4": 0.0,
272
+ "pass@8": 0.0,
273
+ "pass@16": 0.0,
274
+ "pass@32": 0.0
275
+ }
276
+ }
277
+ },
278
+ "teachers_in_school": {
279
+ "count": 19936,
280
+ "correct": 32,
281
+ "answer_accuracy": 0.0016051364365971107,
282
+ "avg_response_len": 290.62841091492777,
283
+ "resp_tokens_sum": 5793968,
284
+ "pass_at_k": {
285
+ "pass@1": 0.0016051364365971107,
286
+ "pass@2": 0.0031520219541241647,
287
+ "pass@4": 0.006083761697052369,
288
+ "pass@8": 0.01137611435570466,
289
+ "pass@16": 0.020132946005638506,
290
+ "pass@32": 0.03210272873194221
291
+ },
292
+ "per_op_pass_at_k": {
293
+ "21": {
294
+ "pass@1": 0.0033653846153846156,
295
+ "pass@2": 0.006637717121588087,
296
+ "pass@4": 0.012903225806451613,
297
+ "pass@8": 0.024317617866004965,
298
+ "pass@16": 0.042679900744416875,
299
+ "pass@32": 0.06153846153846154
300
+ },
301
+ "22": {
302
+ "pass@1": 0.005122950819672131,
303
+ "pass@2": 0.010146747752511896,
304
+ "pass@4": 0.019896879957694343,
305
+ "pass@8": 0.03820729772607087,
306
+ "pass@16": 0.07006874669487044,
307
+ "pass@32": 0.11475409836065574
308
+ },
309
+ "23": {
310
+ "pass@1": 0.0008928571428571428,
311
+ "pass@2": 0.0017857142857142857,
312
+ "pass@4": 0.0035714285714285713,
313
+ "pass@8": 0.007142857142857143,
314
+ "pass@16": 0.014285714285714285,
315
+ "pass@32": 0.02857142857142857
316
+ },
317
+ "24": {
318
+ "pass@1": 0.0033482142857142855,
319
+ "pass@2": 0.006624423963133639,
320
+ "pass@4": 0.012960829493087559,
321
+ "pass@8": 0.02476958525345622,
322
+ "pass@16": 0.044930875576036866,
323
+ "pass@32": 0.07142857142857142
324
+ },
325
+ "25": {
326
+ "pass@1": 0.0,
327
+ "pass@2": 0.0,
328
+ "pass@4": 0.0,
329
+ "pass@8": 0.0,
330
+ "pass@16": 0.0,
331
+ "pass@32": 0.0
332
+ },
333
+ "26": {
334
+ "pass@1": 0.0,
335
+ "pass@2": 0.0,
336
+ "pass@4": 0.0,
337
+ "pass@8": 0.0,
338
+ "pass@16": 0.0,
339
+ "pass@32": 0.0
340
+ },
341
+ "27": {
342
+ "pass@1": 0.0,
343
+ "pass@2": 0.0,
344
+ "pass@4": 0.0,
345
+ "pass@8": 0.0,
346
+ "pass@16": 0.0,
347
+ "pass@32": 0.0
348
+ },
349
+ "28": {
350
+ "pass@1": 0.0,
351
+ "pass@2": 0.0,
352
+ "pass@4": 0.0,
353
+ "pass@8": 0.0,
354
+ "pass@16": 0.0,
355
+ "pass@32": 0.0
356
+ },
357
+ "29": {
358
+ "pass@1": 0.0033653846153846156,
359
+ "pass@2": 0.006420595533498759,
360
+ "pass@4": 0.011722426627877128,
361
+ "pass@8": 0.019829725335843245,
362
+ "pass@16": 0.030435526653546673,
363
+ "pass@32": 0.046153846153846156
364
+ },
365
+ "30": {
366
+ "pass@1": 0.0,
367
+ "pass@2": 0.0,
368
+ "pass@4": 0.0,
369
+ "pass@8": 0.0,
370
+ "pass@16": 0.0,
371
+ "pass@32": 0.0
372
+ }
373
+ }
374
+ },
375
+ "movie_festival_awards": {
376
+ "count": 20992,
377
+ "correct": 68,
378
+ "answer_accuracy": 0.0032393292682926828,
379
+ "avg_response_len": 329.0681688262195,
380
+ "resp_tokens_sum": 6907799,
381
+ "pass_at_k": {
382
+ "pass@1": 0.0032393292682926828,
383
+ "pass@2": 0.006011506687647521,
384
+ "pass@4": 0.010611002401041809,
385
+ "pass@8": 0.017712778837487186,
386
+ "pass@16": 0.028303276404021957,
387
+ "pass@32": 0.04420731707317073
388
+ },
389
+ "per_op_pass_at_k": {
390
+ "21": {
391
+ "pass@1": 0.006696428571428571,
392
+ "pass@2": 0.012708813364055297,
393
+ "pass@4": 0.0229764222151597,
394
+ "pass@8": 0.0381599043028409,
395
+ "pass@16": 0.056005791267192816,
396
+ "pass@32": 0.07142857142857142
397
+ },
398
+ "22": {
399
+ "pass@1": 0.0032649253731343282,
400
+ "pass@2": 0.006349301877708233,
401
+ "pass@4": 0.012024139591253964,
402
+ "pass@8": 0.021709029933757243,
403
+ "pass@16": 0.036558032971958886,
404
+ "pass@32": 0.05970149253731343
405
+ },
406
+ "23": {
407
+ "pass@1": 0.0048828125,
408
+ "pass@2": 0.008631552419354838,
409
+ "pass@4": 0.013730533926585095,
410
+ "pass@8": 0.018802885506688342,
411
+ "pass@16": 0.023431127178346313,
412
+ "pass@32": 0.03125
413
+ },
414
+ "24": {
415
+ "pass@1": 0.0025684931506849314,
416
+ "pass@2": 0.004971277065841803,
417
+ "pass@4": 0.009323525378274185,
418
+ "pass@8": 0.01650006856933884,
419
+ "pass@16": 0.02670394807015405,
420
+ "pass@32": 0.0410958904109589
421
+ },
422
+ "25": {
423
+ "pass@1": 0.0009765625,
424
+ "pass@2": 0.001953125,
425
+ "pass@4": 0.00390625,
426
+ "pass@8": 0.0078125,
427
+ "pass@16": 0.015625,
428
+ "pass@32": 0.03125
429
+ },
430
+ "26": {
431
+ "pass@1": 0.006024096385542169,
432
+ "pass@2": 0.010153517294986397,
433
+ "pass@4": 0.015267633381132984,
434
+ "pass@8": 0.020997761904123724,
435
+ "pass@16": 0.03012046250485329,
436
+ "pass@32": 0.04819277108433735
437
+ },
438
+ "27": {
439
+ "pass@1": 0.0040178571428571425,
440
+ "pass@2": 0.007862903225806451,
441
+ "pass@4": 0.015057603686635944,
442
+ "pass@8": 0.027626728110599075,
443
+ "pass@16": 0.0467741935483871,
444
+ "pass@32": 0.07142857142857142
445
+ },
446
+ "28": {
447
+ "pass@1": 0.0005896226415094339,
448
+ "pass@2": 0.0011792452830188679,
449
+ "pass@4": 0.0023584905660377358,
450
+ "pass@8": 0.0047169811320754715,
451
+ "pass@16": 0.009433962264150943,
452
+ "pass@32": 0.018867924528301886
453
+ },
454
+ "29": {
455
+ "pass@1": 0.0010416666666666667,
456
+ "pass@2": 0.0020833333333333333,
457
+ "pass@4": 0.004166666666666667,
458
+ "pass@8": 0.008333333333333333,
459
+ "pass@16": 0.016666666666666666,
460
+ "pass@32": 0.03333333333333333
461
+ },
462
+ "30": {
463
+ "pass@1": 0.0014204545454545455,
464
+ "pass@2": 0.0028103616813294225,
465
+ "pass@4": 0.005498533724340176,
466
+ "pass@8": 0.01050830889540567,
467
+ "pass@16": 0.01906158357771261,
468
+ "pass@32": 0.030303030303030304
469
+ }
470
+ }
471
+ }
472
+ }
473
+ },
474
+ "total": {
475
+ "answer_accuracy": 0.0025,
476
+ "avg_loss": 0.0,
477
+ "count": 64000,
478
+ "validation_metrics": {
479
+ "id": {},
480
+ "ood": {}
481
+ },
482
+ "pass_at_k": {
483
+ "pass@1": 0.0025,
484
+ "pass@2": 0.00463608870967742,
485
+ "pass@4": 0.008184218576195775,
486
+ "pass@8": 0.013711085203882761,
487
+ "pass@16": 0.022133159099567364,
488
+ "pass@32": 0.0345
489
+ },
490
+ "per_op_pass_at_k": {
491
+ "21": {
492
+ "pass@1": 0.0034375,
493
+ "pass@2": 0.006653225806451612,
494
+ "pass@4": 0.01250194660734149,
495
+ "pass@8": 0.022337999011247067,
496
+ "pass@16": 0.037052589296749476,
497
+ "pass@32": 0.055
498
+ },
499
+ "22": {
500
+ "pass@1": 0.00359375,
501
+ "pass@2": 0.0070665322580645154,
502
+ "pass@4": 0.013669215795328142,
503
+ "pass@8": 0.02563542825361513,
504
+ "pass@16": 0.045553392658509446,
505
+ "pass@32": 0.075
506
+ },
507
+ "23": {
508
+ "pass@1": 0.001875,
509
+ "pass@2": 0.0033870967741935483,
510
+ "pass@4": 0.005643770856507231,
511
+ "pass@8": 0.00851692336214027,
512
+ "pass@16": 0.01249796069707082,
513
+ "pass@32": 0.02
514
+ },
515
+ "24": {
516
+ "pass@1": 0.005,
517
+ "pass@2": 0.008981854838709677,
518
+ "pass@4": 0.01523567853170189,
519
+ "pass@8": 0.024872691879866515,
520
+ "pass@16": 0.03990823009547857,
521
+ "pass@32": 0.06
522
+ },
523
+ "25": {
524
+ "pass@1": 0.00328125,
525
+ "pass@2": 0.005897177419354839,
526
+ "pass@4": 0.009723998887652947,
527
+ "pass@8": 0.014333853854710364,
528
+ "pass@16": 0.019982385151510268,
529
+ "pass@32": 0.03
530
+ },
531
+ "26": {
532
+ "pass@1": 0.0025,
533
+ "pass@2": 0.004213709677419355,
534
+ "pass@4": 0.006336067853170189,
535
+ "pass@8": 0.008714071190211345,
536
+ "pass@16": 0.012499991939514114,
537
+ "pass@32": 0.02
538
+ },
539
+ "27": {
540
+ "pass@1": 0.0015625,
541
+ "pass@2": 0.0030645161290322577,
542
+ "pass@4": 0.0058951612903225805,
543
+ "pass@8": 0.010919354838709676,
544
+ "pass@16": 0.018870967741935484,
545
+ "pass@32": 0.03
546
+ },
547
+ "28": {
548
+ "pass@1": 0.00171875,
549
+ "pass@2": 0.0031451612903225807,
550
+ "pass@4": 0.005337041156840933,
551
+ "pass@8": 0.008118126978694276,
552
+ "pass@16": 0.011284204671857619,
553
+ "pass@32": 0.015
554
+ },
555
+ "29": {
556
+ "pass@1": 0.0015625,
557
+ "pass@2": 0.0030241935483870967,
558
+ "pass@4": 0.005684788654060067,
559
+ "pass@8": 0.010194660734149055,
560
+ "pass@16": 0.01739154616240267,
561
+ "pass@32": 0.03
562
+ },
563
+ "30": {
564
+ "pass@1": 0.00046875,
565
+ "pass@2": 0.0009274193548387094,
566
+ "pass@4": 0.0018145161290322581,
567
+ "pass@8": 0.0034677419354838712,
568
+ "pass@16": 0.006290322580645161,
569
+ "pass@32": 0.01
570
+ }
571
+ },
572
+ "per_op_accuracy": {
573
+ "21": 0.0034375,
574
+ "22": 0.00359375,
575
+ "23": 0.001875,
576
+ "24": 0.005,
577
+ "25": 0.00328125,
578
+ "26": 0.0025,
579
+ "27": 0.0015625,
580
+ "28": 0.00171875,
581
+ "29": 0.0015625,
582
+ "30": 0.00046875
583
+ },
584
+ "per_op_avg_response_len": {
585
+ "21": 292.32015625,
586
+ "22": 296.5990625,
587
+ "23": 294.07046875,
588
+ "24": 299.96671875,
589
+ "25": 298.75125,
590
+ "26": 304.253125,
591
+ "27": 305.4503125,
592
+ "28": 295.45109375,
593
+ "29": 299.44328125,
594
+ "30": 296.465625
595
+ },
596
+ "per_op_avg_loss": {},
597
+ "per_op_length": {
598
+ "21": 292.32015625,
599
+ "22": 296.5990625,
600
+ "23": 294.07046875,
601
+ "24": 299.96671875,
602
+ "25": 298.75125,
603
+ "26": 304.253125,
604
+ "27": 305.4503125,
605
+ "28": 295.45109375,
606
+ "29": 299.44328125,
607
+ "30": 296.465625
608
+ },
609
+ "per_op_loss": {},
610
+ "per_template": {
611
+ "crazy_zootopia": {
612
+ "count": 23072,
613
+ "correct": 60,
614
+ "answer_accuracy": 0.0026005547850208046,
615
+ "avg_response_len": 276.871012482663,
616
+ "resp_tokens_sum": 6387968,
617
+ "pass_at_k": {
618
+ "pass@1": 0.0026005547850208046,
619
+ "pass@2": 0.004667017135698627,
620
+ "pass@4": 0.007791173425859214,
621
+ "pass@8": 0.012087750689001269,
622
+ "pass@16": 0.018247633171405726,
623
+ "pass@32": 0.027739251040221916
624
+ },
625
+ "per_op_pass_at_k": {
626
+ "21": {
627
+ "pass@1": 0.0011867088607594937,
628
+ "pass@2": 0.0023734177215189874,
629
+ "pass@4": 0.004746835443037975,
630
+ "pass@8": 0.00949367088607595,
631
+ "pass@16": 0.0189873417721519,
632
+ "pass@32": 0.0379746835443038
633
+ },
634
+ "22": {
635
+ "pass@1": 0.0026041666666666665,
636
+ "pass@2": 0.005124327956989246,
637
+ "pass@4": 0.009923835125448028,
638
+ "pass@8": 0.018637992831541217,
639
+ "pass@16": 0.033154121863799284,
640
+ "pass@32": 0.05555555555555555
641
+ },
642
+ "23": {
643
+ "pass@1": 0.0,
644
+ "pass@2": 0.0,
645
+ "pass@4": 0.0,
646
+ "pass@8": 0.0,
647
+ "pass@16": 0.0,
648
+ "pass@32": 0.0
649
+ },
650
+ "24": {
651
+ "pass@1": 0.008802816901408451,
652
+ "pass@2": 0.014964788732394365,
653
+ "pass@4": 0.023108618339626186,
654
+ "pass@8": 0.033562487270676344,
655
+ "pass@16": 0.049522940531216975,
656
+ "pass@32": 0.07042253521126761
657
+ },
658
+ "25": {
659
+ "pass@1": 0.009424603174603174,
660
+ "pass@2": 0.016737071172555042,
661
+ "pass@4": 0.026901583770326816,
662
+ "pass@8": 0.037567790014953535,
663
+ "pass@16": 0.04756312746511196,
664
+ "pass@32": 0.06349206349206349
665
+ },
666
+ "26": {
667
+ "pass@1": 0.0,
668
+ "pass@2": 0.0,
669
+ "pass@4": 0.0,
670
+ "pass@8": 0.0,
671
+ "pass@16": 0.0,
672
+ "pass@32": 0.0
673
+ },
674
+ "27": {
675
+ "pass@1": 0.00043402777777777775,
676
+ "pass@2": 0.0008680555555555555,
677
+ "pass@4": 0.001736111111111111,
678
+ "pass@8": 0.003472222222222222,
679
+ "pass@16": 0.006944444444444444,
680
+ "pass@32": 0.013888888888888888
681
+ },
682
+ "28": {
683
+ "pass@1": 0.0035919540229885057,
684
+ "pass@2": 0.006511865035224323,
685
+ "pass@4": 0.010832278521473412,
686
+ "pass@8": 0.015788797652170748,
687
+ "pass@16": 0.020193573958293378,
688
+ "pass@32": 0.022988505747126436
689
+ },
690
+ "29": {
691
+ "pass@1": 0.0004166666666666667,
692
+ "pass@2": 0.0008333333333333334,
693
+ "pass@4": 0.0016666666666666668,
694
+ "pass@8": 0.0033333333333333335,
695
+ "pass@16": 0.006666666666666667,
696
+ "pass@32": 0.013333333333333334
697
+ },
698
+ "30": {
699
+ "pass@1": 0.0,
700
+ "pass@2": 0.0,
701
+ "pass@4": 0.0,
702
+ "pass@8": 0.0,
703
+ "pass@16": 0.0,
704
+ "pass@32": 0.0
705
+ }
706
+ }
707
+ },
708
+ "teachers_in_school": {
709
+ "count": 19936,
710
+ "correct": 32,
711
+ "answer_accuracy": 0.0016051364365971107,
712
+ "avg_response_len": 290.62841091492777,
713
+ "resp_tokens_sum": 5793968,
714
+ "pass_at_k": {
715
+ "pass@1": 0.0016051364365971107,
716
+ "pass@2": 0.0031520219541241647,
717
+ "pass@4": 0.006083761697052369,
718
+ "pass@8": 0.01137611435570466,
719
+ "pass@16": 0.020132946005638506,
720
+ "pass@32": 0.03210272873194221
721
+ },
722
+ "per_op_pass_at_k": {
723
+ "21": {
724
+ "pass@1": 0.0033653846153846156,
725
+ "pass@2": 0.006637717121588087,
726
+ "pass@4": 0.012903225806451613,
727
+ "pass@8": 0.024317617866004965,
728
+ "pass@16": 0.042679900744416875,
729
+ "pass@32": 0.06153846153846154
730
+ },
731
+ "22": {
732
+ "pass@1": 0.005122950819672131,
733
+ "pass@2": 0.010146747752511896,
734
+ "pass@4": 0.019896879957694343,
735
+ "pass@8": 0.03820729772607087,
736
+ "pass@16": 0.07006874669487044,
737
+ "pass@32": 0.11475409836065574
738
+ },
739
+ "23": {
740
+ "pass@1": 0.0008928571428571428,
741
+ "pass@2": 0.0017857142857142857,
742
+ "pass@4": 0.0035714285714285713,
743
+ "pass@8": 0.007142857142857143,
744
+ "pass@16": 0.014285714285714285,
745
+ "pass@32": 0.02857142857142857
746
+ },
747
+ "24": {
748
+ "pass@1": 0.0033482142857142855,
749
+ "pass@2": 0.006624423963133639,
750
+ "pass@4": 0.012960829493087559,
751
+ "pass@8": 0.02476958525345622,
752
+ "pass@16": 0.044930875576036866,
753
+ "pass@32": 0.07142857142857142
754
+ },
755
+ "25": {
756
+ "pass@1": 0.0,
757
+ "pass@2": 0.0,
758
+ "pass@4": 0.0,
759
+ "pass@8": 0.0,
760
+ "pass@16": 0.0,
761
+ "pass@32": 0.0
762
+ },
763
+ "26": {
764
+ "pass@1": 0.0,
765
+ "pass@2": 0.0,
766
+ "pass@4": 0.0,
767
+ "pass@8": 0.0,
768
+ "pass@16": 0.0,
769
+ "pass@32": 0.0
770
+ },
771
+ "27": {
772
+ "pass@1": 0.0,
773
+ "pass@2": 0.0,
774
+ "pass@4": 0.0,
775
+ "pass@8": 0.0,
776
+ "pass@16": 0.0,
777
+ "pass@32": 0.0
778
+ },
779
+ "28": {
780
+ "pass@1": 0.0,
781
+ "pass@2": 0.0,
782
+ "pass@4": 0.0,
783
+ "pass@8": 0.0,
784
+ "pass@16": 0.0,
785
+ "pass@32": 0.0
786
+ },
787
+ "29": {
788
+ "pass@1": 0.0033653846153846156,
789
+ "pass@2": 0.006420595533498759,
790
+ "pass@4": 0.011722426627877128,
791
+ "pass@8": 0.019829725335843245,
792
+ "pass@16": 0.030435526653546673,
793
+ "pass@32": 0.046153846153846156
794
+ },
795
+ "30": {
796
+ "pass@1": 0.0,
797
+ "pass@2": 0.0,
798
+ "pass@4": 0.0,
799
+ "pass@8": 0.0,
800
+ "pass@16": 0.0,
801
+ "pass@32": 0.0
802
+ }
803
+ }
804
+ },
805
+ "movie_festival_awards": {
806
+ "count": 20992,
807
+ "correct": 68,
808
+ "answer_accuracy": 0.0032393292682926828,
809
+ "avg_response_len": 329.0681688262195,
810
+ "resp_tokens_sum": 6907799,
811
+ "pass_at_k": {
812
+ "pass@1": 0.0032393292682926828,
813
+ "pass@2": 0.006011506687647521,
814
+ "pass@4": 0.010611002401041809,
815
+ "pass@8": 0.017712778837487186,
816
+ "pass@16": 0.028303276404021957,
817
+ "pass@32": 0.04420731707317073
818
+ },
819
+ "per_op_pass_at_k": {
820
+ "21": {
821
+ "pass@1": 0.006696428571428571,
822
+ "pass@2": 0.012708813364055297,
823
+ "pass@4": 0.0229764222151597,
824
+ "pass@8": 0.0381599043028409,
825
+ "pass@16": 0.056005791267192816,
826
+ "pass@32": 0.07142857142857142
827
+ },
828
+ "22": {
829
+ "pass@1": 0.0032649253731343282,
830
+ "pass@2": 0.006349301877708233,
831
+ "pass@4": 0.012024139591253964,
832
+ "pass@8": 0.021709029933757243,
833
+ "pass@16": 0.036558032971958886,
834
+ "pass@32": 0.05970149253731343
835
+ },
836
+ "23": {
837
+ "pass@1": 0.0048828125,
838
+ "pass@2": 0.008631552419354838,
839
+ "pass@4": 0.013730533926585095,
840
+ "pass@8": 0.018802885506688342,
841
+ "pass@16": 0.023431127178346313,
842
+ "pass@32": 0.03125
843
+ },
844
+ "24": {
845
+ "pass@1": 0.0025684931506849314,
846
+ "pass@2": 0.004971277065841803,
847
+ "pass@4": 0.009323525378274185,
848
+ "pass@8": 0.01650006856933884,
849
+ "pass@16": 0.02670394807015405,
850
+ "pass@32": 0.0410958904109589
851
+ },
852
+ "25": {
853
+ "pass@1": 0.0009765625,
854
+ "pass@2": 0.001953125,
855
+ "pass@4": 0.00390625,
856
+ "pass@8": 0.0078125,
857
+ "pass@16": 0.015625,
858
+ "pass@32": 0.03125
859
+ },
860
+ "26": {
861
+ "pass@1": 0.006024096385542169,
862
+ "pass@2": 0.010153517294986397,
863
+ "pass@4": 0.015267633381132984,
864
+ "pass@8": 0.020997761904123724,
865
+ "pass@16": 0.03012046250485329,
866
+ "pass@32": 0.04819277108433735
867
+ },
868
+ "27": {
869
+ "pass@1": 0.0040178571428571425,
870
+ "pass@2": 0.007862903225806451,
871
+ "pass@4": 0.015057603686635944,
872
+ "pass@8": 0.027626728110599075,
873
+ "pass@16": 0.0467741935483871,
874
+ "pass@32": 0.07142857142857142
875
+ },
876
+ "28": {
877
+ "pass@1": 0.0005896226415094339,
878
+ "pass@2": 0.0011792452830188679,
879
+ "pass@4": 0.0023584905660377358,
880
+ "pass@8": 0.0047169811320754715,
881
+ "pass@16": 0.009433962264150943,
882
+ "pass@32": 0.018867924528301886
883
+ },
884
+ "29": {
885
+ "pass@1": 0.0010416666666666667,
886
+ "pass@2": 0.0020833333333333333,
887
+ "pass@4": 0.004166666666666667,
888
+ "pass@8": 0.008333333333333333,
889
+ "pass@16": 0.016666666666666666,
890
+ "pass@32": 0.03333333333333333
891
+ },
892
+ "30": {
893
+ "pass@1": 0.0014204545454545455,
894
+ "pass@2": 0.0028103616813294225,
895
+ "pass@4": 0.005498533724340176,
896
+ "pass@8": 0.01050830889540567,
897
+ "pass@16": 0.01906158357771261,
898
+ "pass@32": 0.030303030303030304
899
+ }
900
+ }
901
+ }
902
+ }
903
+ }
904
+ }