Hallucinate1900-95M / training_log.csv
foss22's picture
Upload training_log.csv with huggingface_hub
1594320 verified
Raw History Blame Contribute Delete
5.72 kB
step,epoch,loss,learning_rate,grad_norm,english_ppl,russian_ppl,ppl_delta_en,ppl_delta_ru,elapsed_hours,eta_hours,timestamp
0,0.0000,0.0000,0.00e+00,0.00,604.35,156355.66,+0.00,+0.00,0.00,0.00,2026-08-20T22:09:44.208656
50,0.0032,14.6150,7.80e-05,171.00,599.40,139433.19,-4.95,-16922.47,0.73,42.91,2026-08-20T22:53:26.753891
100,0.0064,12.0384,1.00e-04,60.75,603.09,96016.37,-1.26,-60339.30,1.77,51.28,2026-08-20T23:55:54.781269
150,0.0096,9.7067,9.98e-05,34.50,625.77,56374.08,+21.42,-99981.59,2.54,48.26,2026-08-21T00:42:50.478715
200,0.0128,8.4562,9.95e-05,47.50,641.49,42435.89,+37.14,-113919.78,4.19,58.67,2026-08-21T02:21:52.568000
250,0.0160,7.8115,9.90e-05,29.75,648.78,33274.28,+44.43,-123081.38,5.76,63.32,2026-08-21T03:55:12.347823
300,0.0192,7.5262,9.84e-05,4.47,653.11,23581.47,+48.76,-132774.19,6.48,58.35,2026-08-21T04:38:50.815944
350,0.0224,7.2778,9.77e-05,4.72,653.20,20956.16,+48.85,-135399.50,7.20,54.52,2026-08-21T05:21:55.976444
400,0.0256,7.1292,9.68e-05,1.30,648.45,20145.79,+44.10,-136209.88,7.93,51.52,2026-08-21T06:05:25.277509
450,0.0288,6.9981,9.58e-05,3.25,645.85,19586.75,+41.50,-136768.91,8.63,48.92,2026-08-21T06:47:47.398825
500,0.0320,6.8856,9.46e-05,2.11,643.00,19353.95,+38.65,-137001.72,9.34,46.69,2026-08-21T07:30:04.880046
550,0.0352,6.7135,9.34e-05,1.91,644.95,19525.53,+40.60,-136830.13,10.04,44.74,2026-08-21T08:12:25.353688
600,0.0384,6.6453,9.20e-05,1.94,653.90,19469.11,+49.55,-136886.56,10.75,42.99,2026-08-21T08:54:42.957381
650,0.0416,6.5597,9.05e-05,1.71,650.49,19593.17,+46.14,-136762.49,11.45,41.41,2026-08-21T09:37:00.559053
700,0.0448,6.5004,8.89e-05,1.87,652.84,19352.53,+48.49,-137003.14,12.16,39.95,2026-08-21T10:19:18.114936
750,0.0480,6.3986,8.71e-05,3.00,647.92,19530.56,+43.57,-136825.10,12.86,38.59,2026-08-21T11:01:35.638772
800,0.0512,6.3999,8.53e-05,3.06,647.64,19316.44,+43.29,-137039.22,13.57,37.31,2026-08-21T11:43:53.288137
850,0.0544,6.3533,8.34e-05,1.86,652.00,19048.99,+47.65,-137306.67,14.27,36.10,2026-08-21T12:26:10.791842
900,0.0576,6.2817,8.13e-05,3.03,652.27,19136.03,+47.92,-137219.63,14.98,34.95,2026-08-21T13:08:28.127650
950,0.0608,6.2734,7.92e-05,2.23,650.61,19157.93,+46.26,-137197.74,15.68,33.84,2026-08-21T13:50:45.462993
1000,0.0640,6.1876,7.70e-05,2.44,656.85,18773.96,+52.50,-137581.70,16.39,32.78,2026-08-21T14:33:04.043506
1050,0.0672,6.2406,7.47e-05,2.69,660.62,18720.83,+56.27,-137634.84,17.09,31.74,2026-08-21T15:15:23.649354
1100,0.0704,6.1753,7.24e-05,3.28,660.47,18618.31,+56.12,-137737.35,17.80,30.74,2026-08-21T15:57:41.828227
1150,0.0736,6.1928,7.00e-05,3.42,659.59,18522.62,+55.24,-137833.04,18.50,29.77,2026-08-21T16:39:59.925803
1200,0.0768,6.1834,6.75e-05,3.22,655.76,18643.30,+51.41,-137712.37,19.21,28.81,2026-08-21T17:22:17.804930
1250,0.0800,6.1882,6.50e-05,2.72,660.32,18426.64,+55.97,-137929.02,19.91,27.88,2026-08-21T18:04:36.119783
1300,0.0832,6.1748,6.24e-05,3.11,660.06,18644.05,+55.71,-137711.62,20.62,26.96,2026-08-21T18:46:54.293927
1350,0.0864,6.1691,5.98e-05,2.75,661.17,18673.09,+56.82,-137682.58,21.32,26.06,2026-08-21T19:29:13.016277
1400,0.0896,6.1926,5.72e-05,3.34,661.11,18554.18,+56.76,-137801.48,22.03,25.18,2026-08-21T20:11:30.946763
1450,0.0928,6.1786,5.46e-05,2.38,663.14,18590.62,+58.79,-137765.05,22.73,24.30,2026-08-21T20:53:48.611761
1500,0.0960,6.1968,5.19e-05,2.69,661.97,18700.32,+57.62,-137655.34,23.44,23.44,2026-08-21T21:36:06.863542
1550,0.0992,6.1385,4.93e-05,2.92,661.00,18525.21,+56.65,-137830.46,24.14,22.59,2026-08-21T22:18:27.448967
1600,0.1024,6.1658,4.66e-05,3.66,662.64,18509.88,+58.29,-137845.79,24.85,21.74,2026-08-21T23:00:46.495350
1650,0.1056,6.1569,4.39e-05,4.03,663.60,18558.43,+59.25,-137797.23,25.56,20.91,2026-08-21T23:43:08.168616
1700,0.1088,6.1650,4.13e-05,2.69,661.05,18584.11,+56.70,-137771.56,26.41,20.19,2026-08-22T00:34:11.167112
1750,0.1120,6.1870,3.87e-05,3.69,663.17,18591.29,+58.82,-137764.38,27.11,19.37,2026-08-22T01:16:30.466505
1800,0.1152,6.1698,3.61e-05,4.03,659.34,18631.21,+54.99,-137724.46,27.82,18.55,2026-08-22T01:59:03.523249
50,1.0000,6.1129,0.00e+00,0.0000,641.96,7021.71,0.176,2026-08-22T05:58:50.419530
100,1.0000,5.9932,0.00e+00,0.0000,742.94,4933.64,0.351,2026-08-22T06:09:21.292202
150,1.0000,5.8891,0.00e+00,0.0000,810.60,3251.83,0.527,2026-08-22T06:19:52.347476
200,1.0000,5.7871,0.00e+00,0.0000,894.94,2845.45,0.702,2026-08-22T06:30:23.784534
250,1.0000,5.7127,0.00e+00,0.0000,974.28,2581.82,0.877,2026-08-22T06:40:54.106859
300,1.0000,5.6585,0.00e+00,0.0000,1009.54,2414.47,1.053,2026-08-22T06:51:25.469001
350,1.0000,5.6030,0.00e+00,0.0000,1020.53,2745.35,1.228,2026-08-22T07:01:56.312861
400,1.0000,5.5451,0.00e+00,0.0000,1106.05,2377.93,1.403,2026-08-22T07:12:26.786619
450,1.0000,5.4937,0.00e+00,0.0000,1143.58,2126.38,1.578,2026-08-22T07:22:57.835645
500,1.0000,5.4483,0.00e+00,0.0000,1093.74,2144.28,1.753,2026-08-22T07:33:28.827307
550,1.0000,5.4075,0.00e+00,0.0000,1225.24,2330.83,1.930,2026-08-22T07:44:04.622109
600,1.0000,5.3730,0.00e+00,0.0000,1214.07,1978.74,2.105,2026-08-22T07:54:35.664291
650,1.0000,5.3316,0.00e+00,0.0000,1232.30,2345.98,2.280,2026-08-22T08:05:06.074844
700,1.0000,5.2946,0.00e+00,0.0000,1240.91,2249.36,2.456,2026-08-22T08:15:36.887802
750,1.0000,5.2593,0.00e+00,0.0000,1187.45,2688.28,2.631,2026-08-22T08:26:08.321939
800,1.0000,5.2267,0.00e+00,0.0000,1162.14,2228.33,2.807,2026-08-22T08:36:39.990167
850,1.0000,5.1996,0.00e+00,0.0000,1219.97,2317.87,2.982,2026-08-22T08:47:11.344872
900,1.0000,5.1706,0.00e+00,0.0000,1206.07,2246.20,3.157,2026-08-22T08:57:42.705017
950,1.0000,5.1475,0.00e+00,0.0000,1206.94,2094.43,3.333,2026-08-22T09:08:14.019746
1000,1.0000,5.1224,0.00e+00,0.0000,1276.14,2665.49,3.508,2026-08-22T09:18:44.837748