Atlas-Coder-0.5B / last-checkpoint /trainer_state.json
Siddh07ETH's picture
Training in progress, step 3500, checkpoint
3bbea8c verified
Raw
History Blame Contribute Delete
60.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.827600141435571,
"eval_steps": 500,
"global_step": 3500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008082032631206748,
"grad_norm": 74073.046875,
"learning_rate": 0.0,
"loss": 0.5988,
"step": 1
},
{
"epoch": 0.008082032631206748,
"grad_norm": 151597.078125,
"learning_rate": 4.0540540540540545e-06,
"loss": 0.5791,
"step": 10
},
{
"epoch": 0.016164065262413495,
"grad_norm": 119755.34375,
"learning_rate": 8.558558558558558e-06,
"loss": 0.489,
"step": 20
},
{
"epoch": 0.024246097893620245,
"grad_norm": 83862.2890625,
"learning_rate": 1.3063063063063064e-05,
"loss": 0.4207,
"step": 30
},
{
"epoch": 0.03232813052482699,
"grad_norm": 47998.07421875,
"learning_rate": 1.756756756756757e-05,
"loss": 0.4906,
"step": 40
},
{
"epoch": 0.04041016315603374,
"grad_norm": 73080.90625,
"learning_rate": 2.2072072072072073e-05,
"loss": 0.6574,
"step": 50
},
{
"epoch": 0.04849219578724049,
"grad_norm": 34368.5078125,
"learning_rate": 2.6576576576576577e-05,
"loss": 0.2496,
"step": 60
},
{
"epoch": 0.05657422841844724,
"grad_norm": 9904.455078125,
"learning_rate": 3.108108108108108e-05,
"loss": 0.1602,
"step": 70
},
{
"epoch": 0.06465626104965398,
"grad_norm": 11667.8583984375,
"learning_rate": 3.558558558558558e-05,
"loss": 0.1942,
"step": 80
},
{
"epoch": 0.07273829368086074,
"grad_norm": 20582.626953125,
"learning_rate": 4.0090090090090096e-05,
"loss": 0.4106,
"step": 90
},
{
"epoch": 0.08082032631206748,
"grad_norm": 39085.125,
"learning_rate": 4.4594594594594596e-05,
"loss": 0.5921,
"step": 100
},
{
"epoch": 0.08890235894327424,
"grad_norm": 8263.34375,
"learning_rate": 4.90990990990991e-05,
"loss": 0.1584,
"step": 110
},
{
"epoch": 0.09698439157448098,
"grad_norm": 8079.8359375,
"learning_rate": 5.3603603603603604e-05,
"loss": 0.1544,
"step": 120
},
{
"epoch": 0.10506642420568774,
"grad_norm": 13801.568359375,
"learning_rate": 5.8108108108108105e-05,
"loss": 0.1987,
"step": 130
},
{
"epoch": 0.11314845683689448,
"grad_norm": 22531.623046875,
"learning_rate": 6.261261261261262e-05,
"loss": 0.4284,
"step": 140
},
{
"epoch": 0.12123048946810122,
"grad_norm": 38542.81640625,
"learning_rate": 6.711711711711713e-05,
"loss": 0.5658,
"step": 150
},
{
"epoch": 0.12931252209930796,
"grad_norm": 7678.20458984375,
"learning_rate": 7.162162162162162e-05,
"loss": 0.1602,
"step": 160
},
{
"epoch": 0.13739455473051473,
"grad_norm": 8466.353515625,
"learning_rate": 7.612612612612613e-05,
"loss": 0.1429,
"step": 170
},
{
"epoch": 0.14547658736172148,
"grad_norm": 9285.0048828125,
"learning_rate": 8.063063063063063e-05,
"loss": 0.179,
"step": 180
},
{
"epoch": 0.15355861999292822,
"grad_norm": 21297.56640625,
"learning_rate": 8.513513513513514e-05,
"loss": 0.3898,
"step": 190
},
{
"epoch": 0.16164065262413496,
"grad_norm": 32210.623046875,
"learning_rate": 8.963963963963964e-05,
"loss": 0.5816,
"step": 200
},
{
"epoch": 0.16972268525534173,
"grad_norm": 6899.923828125,
"learning_rate": 9.414414414414416e-05,
"loss": 0.1693,
"step": 210
},
{
"epoch": 0.17780471788654847,
"grad_norm": 6293.189453125,
"learning_rate": 9.864864864864865e-05,
"loss": 0.1355,
"step": 220
},
{
"epoch": 0.18588675051775522,
"grad_norm": 9546.2607421875,
"learning_rate": 9.999900851550119e-05,
"loss": 0.2025,
"step": 230
},
{
"epoch": 0.19396878314896196,
"grad_norm": 15707.43359375,
"learning_rate": 9.999415235955473e-05,
"loss": 0.4016,
"step": 240
},
{
"epoch": 0.2020508157801687,
"grad_norm": 39510.828125,
"learning_rate": 9.99852498153154e-05,
"loss": 0.5603,
"step": 250
},
{
"epoch": 0.21013284841137547,
"grad_norm": 28472.544921875,
"learning_rate": 9.997230160333048e-05,
"loss": 0.1421,
"step": 260
},
{
"epoch": 0.21821488104258221,
"grad_norm": 6706.89599609375,
"learning_rate": 9.99553087715924e-05,
"loss": 0.1321,
"step": 270
},
{
"epoch": 0.22629691367378896,
"grad_norm": 9905.5234375,
"learning_rate": 9.993427269545391e-05,
"loss": 0.208,
"step": 280
},
{
"epoch": 0.2343789463049957,
"grad_norm": 14170.8369140625,
"learning_rate": 9.99091950775168e-05,
"loss": 0.3634,
"step": 290
},
{
"epoch": 0.24246097893620244,
"grad_norm": 30831.36328125,
"learning_rate": 9.988007794749404e-05,
"loss": 0.5639,
"step": 300
},
{
"epoch": 0.2505430115674092,
"grad_norm": 5404.18115234375,
"learning_rate": 9.984692366204548e-05,
"loss": 0.135,
"step": 310
},
{
"epoch": 0.2586250441986159,
"grad_norm": 7216.95361328125,
"learning_rate": 9.980973490458728e-05,
"loss": 0.1339,
"step": 320
},
{
"epoch": 0.2667070768298227,
"grad_norm": 9833.787109375,
"learning_rate": 9.976851468507446e-05,
"loss": 0.193,
"step": 330
},
{
"epoch": 0.27478910946102947,
"grad_norm": 14580.97265625,
"learning_rate": 9.972326633975752e-05,
"loss": 0.3859,
"step": 340
},
{
"epoch": 0.2828711420922362,
"grad_norm": 24569.248046875,
"learning_rate": 9.967399353091225e-05,
"loss": 0.5424,
"step": 350
},
{
"epoch": 0.29095317472344295,
"grad_norm": 8168.17333984375,
"learning_rate": 9.962070024654338e-05,
"loss": 0.1327,
"step": 360
},
{
"epoch": 0.29903520735464967,
"grad_norm": 6062.85498046875,
"learning_rate": 9.956339080006184e-05,
"loss": 0.1147,
"step": 370
},
{
"epoch": 0.30711723998585644,
"grad_norm": 8263.6455078125,
"learning_rate": 9.950206982993555e-05,
"loss": 0.1798,
"step": 380
},
{
"epoch": 0.3151992726170632,
"grad_norm": 17083.33203125,
"learning_rate": 9.94367422993141e-05,
"loss": 0.3924,
"step": 390
},
{
"epoch": 0.3232813052482699,
"grad_norm": 23812.875,
"learning_rate": 9.936741349562697e-05,
"loss": 0.5625,
"step": 400
},
{
"epoch": 0.3313633378794767,
"grad_norm": 6225.1484375,
"learning_rate": 9.92940890301556e-05,
"loss": 0.1209,
"step": 410
},
{
"epoch": 0.33944537051068346,
"grad_norm": 4972.1669921875,
"learning_rate": 9.921677483757925e-05,
"loss": 0.1276,
"step": 420
},
{
"epoch": 0.3475274031418902,
"grad_norm": 7313.53466796875,
"learning_rate": 9.913547717549463e-05,
"loss": 0.1943,
"step": 430
},
{
"epoch": 0.35560943577309695,
"grad_norm": 13874.7763671875,
"learning_rate": 9.905020262390945e-05,
"loss": 0.3977,
"step": 440
},
{
"epoch": 0.36369146840430366,
"grad_norm": 29985.951171875,
"learning_rate": 9.896095808470985e-05,
"loss": 0.5687,
"step": 450
},
{
"epoch": 0.37177350103551043,
"grad_norm": 8002.23779296875,
"learning_rate": 9.886775078110177e-05,
"loss": 0.1187,
"step": 460
},
{
"epoch": 0.3798555336667172,
"grad_norm": 6269.2138671875,
"learning_rate": 9.877058825702636e-05,
"loss": 0.1405,
"step": 470
},
{
"epoch": 0.3879375662979239,
"grad_norm": 9276.9931640625,
"learning_rate": 9.866947837654934e-05,
"loss": 0.202,
"step": 480
},
{
"epoch": 0.3960195989291307,
"grad_norm": 14204.4365234375,
"learning_rate": 9.85644293232245e-05,
"loss": 0.3922,
"step": 490
},
{
"epoch": 0.4041016315603374,
"grad_norm": 21759.234375,
"learning_rate": 9.845544959943145e-05,
"loss": 0.5476,
"step": 500
},
{
"epoch": 0.4121836641915442,
"grad_norm": 6127.38427734375,
"learning_rate": 9.834254802568732e-05,
"loss": 0.1248,
"step": 510
},
{
"epoch": 0.42026569682275094,
"grad_norm": 5102.580078125,
"learning_rate": 9.822573373993295e-05,
"loss": 0.1246,
"step": 520
},
{
"epoch": 0.42834772945395766,
"grad_norm": 7689.44091796875,
"learning_rate": 9.810501619679324e-05,
"loss": 0.1771,
"step": 530
},
{
"epoch": 0.43642976208516443,
"grad_norm": 14029.4775390625,
"learning_rate": 9.798040516681197e-05,
"loss": 0.3893,
"step": 540
},
{
"epoch": 0.44451179471637114,
"grad_norm": 32249.396484375,
"learning_rate": 9.785191073566094e-05,
"loss": 0.5415,
"step": 550
},
{
"epoch": 0.4525938273475779,
"grad_norm": 3607.593017578125,
"learning_rate": 9.771954330332367e-05,
"loss": 0.1123,
"step": 560
},
{
"epoch": 0.4606758599787847,
"grad_norm": 8781.576171875,
"learning_rate": 9.758331358325373e-05,
"loss": 0.1139,
"step": 570
},
{
"epoch": 0.4687578926099914,
"grad_norm": 8249.1640625,
"learning_rate": 9.744323260150752e-05,
"loss": 0.1961,
"step": 580
},
{
"epoch": 0.47683992524119817,
"grad_norm": 13182.7041015625,
"learning_rate": 9.729931169585191e-05,
"loss": 0.3634,
"step": 590
},
{
"epoch": 0.4849219578724049,
"grad_norm": 22281.544921875,
"learning_rate": 9.715156251484662e-05,
"loss": 0.5675,
"step": 600
},
{
"epoch": 0.49300399050361166,
"grad_norm": 5673.654296875,
"learning_rate": 9.699999701690133e-05,
"loss": 0.1319,
"step": 610
},
{
"epoch": 0.5010860231348184,
"grad_norm": 6733.84423828125,
"learning_rate": 9.684462746930783e-05,
"loss": 0.1277,
"step": 620
},
{
"epoch": 0.5091680557660252,
"grad_norm": 7341.2841796875,
"learning_rate": 9.668546644724724e-05,
"loss": 0.1763,
"step": 630
},
{
"epoch": 0.5172500883972319,
"grad_norm": 13195.1572265625,
"learning_rate": 9.652252683277205e-05,
"loss": 0.3906,
"step": 640
},
{
"epoch": 0.5253321210284386,
"grad_norm": 20004.513671875,
"learning_rate": 9.63558218137636e-05,
"loss": 0.5521,
"step": 650
},
{
"epoch": 0.5334141536596454,
"grad_norm": 3604.01025390625,
"learning_rate": 9.618536488286465e-05,
"loss": 0.1153,
"step": 660
},
{
"epoch": 0.5414961862908522,
"grad_norm": 5774.5830078125,
"learning_rate": 9.601116983638728e-05,
"loss": 0.1423,
"step": 670
},
{
"epoch": 0.5495782189220589,
"grad_norm": 8617.830078125,
"learning_rate": 9.583325077319633e-05,
"loss": 0.1883,
"step": 680
},
{
"epoch": 0.5576602515532656,
"grad_norm": 11544.3994140625,
"learning_rate": 9.565162209356828e-05,
"loss": 0.3566,
"step": 690
},
{
"epoch": 0.5657422841844724,
"grad_norm": 21384.017578125,
"learning_rate": 9.546629849802562e-05,
"loss": 0.5487,
"step": 700
},
{
"epoch": 0.5738243168156791,
"grad_norm": 4286.736328125,
"learning_rate": 9.527729498614714e-05,
"loss": 0.142,
"step": 710
},
{
"epoch": 0.5819063494468859,
"grad_norm": 5452.06689453125,
"learning_rate": 9.508462685535389e-05,
"loss": 0.127,
"step": 720
},
{
"epoch": 0.5899883820780927,
"grad_norm": 7830.87158203125,
"learning_rate": 9.4888309699671e-05,
"loss": 0.1965,
"step": 730
},
{
"epoch": 0.5980704147092993,
"grad_norm": 13688.564453125,
"learning_rate": 9.46883594084656e-05,
"loss": 0.3926,
"step": 740
},
{
"epoch": 0.6061524473405061,
"grad_norm": 25603.287109375,
"learning_rate": 9.44847921651607e-05,
"loss": 0.5407,
"step": 750
},
{
"epoch": 0.6142344799717129,
"grad_norm": 5079.77392578125,
"learning_rate": 9.427762444592547e-05,
"loss": 0.1205,
"step": 760
},
{
"epoch": 0.6223165126029196,
"grad_norm": 5124.65478515625,
"learning_rate": 9.40668730183416e-05,
"loss": 0.1164,
"step": 770
},
{
"epoch": 0.6303985452341264,
"grad_norm": 8535.7373046875,
"learning_rate": 9.38525549400462e-05,
"loss": 0.1778,
"step": 780
},
{
"epoch": 0.6384805778653331,
"grad_norm": 12861.8642578125,
"learning_rate": 9.363468755735123e-05,
"loss": 0.3672,
"step": 790
},
{
"epoch": 0.6465626104965398,
"grad_norm": 29041.7421875,
"learning_rate": 9.341328850383955e-05,
"loss": 0.5503,
"step": 800
},
{
"epoch": 0.6546446431277466,
"grad_norm": 13424.740234375,
"learning_rate": 9.318837569893761e-05,
"loss": 0.1317,
"step": 810
},
{
"epoch": 0.6627266757589534,
"grad_norm": 5875.9970703125,
"learning_rate": 9.295996734646523e-05,
"loss": 0.1262,
"step": 820
},
{
"epoch": 0.6708087083901602,
"grad_norm": 7396.07080078125,
"learning_rate": 9.272808193316218e-05,
"loss": 0.1882,
"step": 830
},
{
"epoch": 0.6788907410213669,
"grad_norm": 13698.2109375,
"learning_rate": 9.249273822719183e-05,
"loss": 0.391,
"step": 840
},
{
"epoch": 0.6869727736525736,
"grad_norm": 21535.517578125,
"learning_rate": 9.225395527662229e-05,
"loss": 0.5424,
"step": 850
},
{
"epoch": 0.6950548062837804,
"grad_norm": 6823.61181640625,
"learning_rate": 9.201175240788452e-05,
"loss": 0.1254,
"step": 860
},
{
"epoch": 0.7031368389149871,
"grad_norm": 4432.48046875,
"learning_rate": 9.176614922420823e-05,
"loss": 0.1192,
"step": 870
},
{
"epoch": 0.7112188715461939,
"grad_norm": 7185.48876953125,
"learning_rate": 9.15171656040352e-05,
"loss": 0.1834,
"step": 880
},
{
"epoch": 0.7193009041774007,
"grad_norm": 12420.2373046875,
"learning_rate": 9.126482169941039e-05,
"loss": 0.3819,
"step": 890
},
{
"epoch": 0.7273829368086073,
"grad_norm": 23922.16796875,
"learning_rate": 9.100913793435091e-05,
"loss": 0.5452,
"step": 900
},
{
"epoch": 0.7354649694398141,
"grad_norm": 3583.359130859375,
"learning_rate": 9.075013500319283e-05,
"loss": 0.132,
"step": 910
},
{
"epoch": 0.7435470020710209,
"grad_norm": 5175.21337890625,
"learning_rate": 9.048783386891644e-05,
"loss": 0.1221,
"step": 920
},
{
"epoch": 0.7516290347022276,
"grad_norm": 7758.79443359375,
"learning_rate": 9.022225576144939e-05,
"loss": 0.1895,
"step": 930
},
{
"epoch": 0.7597110673334344,
"grad_norm": 14610.314453125,
"learning_rate": 8.995342217594854e-05,
"loss": 0.4185,
"step": 940
},
{
"epoch": 0.7677930999646411,
"grad_norm": 69954.1171875,
"learning_rate": 8.968135487106004e-05,
"loss": 0.5404,
"step": 950
},
{
"epoch": 0.7758751325958478,
"grad_norm": 6344.6826171875,
"learning_rate": 8.94060758671584e-05,
"loss": 0.1469,
"step": 960
},
{
"epoch": 0.7839571652270546,
"grad_norm": 5866.19287109375,
"learning_rate": 8.912760744456415e-05,
"loss": 0.1077,
"step": 970
},
{
"epoch": 0.7920391978582614,
"grad_norm": 7522.1220703125,
"learning_rate": 8.884597214174053e-05,
"loss": 0.1789,
"step": 980
},
{
"epoch": 0.8001212304894681,
"grad_norm": 13574.55078125,
"learning_rate": 8.856119275346928e-05,
"loss": 0.3586,
"step": 990
},
{
"epoch": 0.8082032631206748,
"grad_norm": 24247.8671875,
"learning_rate": 8.827329232900573e-05,
"loss": 0.562,
"step": 1000
},
{
"epoch": 0.8162852957518816,
"grad_norm": 4175.88427734375,
"learning_rate": 8.798229417021327e-05,
"loss": 0.1247,
"step": 1010
},
{
"epoch": 0.8243673283830883,
"grad_norm": 5754.068359375,
"learning_rate": 8.768822182967726e-05,
"loss": 0.1199,
"step": 1020
},
{
"epoch": 0.8324493610142951,
"grad_norm": 7051.59423828125,
"learning_rate": 8.739109910879888e-05,
"loss": 0.1751,
"step": 1030
},
{
"epoch": 0.8405313936455019,
"grad_norm": 13473.8896484375,
"learning_rate": 8.709095005586865e-05,
"loss": 0.3552,
"step": 1040
},
{
"epoch": 0.8486134262767085,
"grad_norm": 18274.482421875,
"learning_rate": 8.678779896411997e-05,
"loss": 0.5489,
"step": 1050
},
{
"epoch": 0.8566954589079153,
"grad_norm": 3864.3779296875,
"learning_rate": 8.648167036976302e-05,
"loss": 0.1112,
"step": 1060
},
{
"epoch": 0.8647774915391221,
"grad_norm": 5771.716796875,
"learning_rate": 8.617258904999876e-05,
"loss": 0.1185,
"step": 1070
},
{
"epoch": 0.8728595241703289,
"grad_norm": 6908.23779296875,
"learning_rate": 8.58605800210136e-05,
"loss": 0.1978,
"step": 1080
},
{
"epoch": 0.8809415568015356,
"grad_norm": 13523.759765625,
"learning_rate": 8.554566853595456e-05,
"loss": 0.3712,
"step": 1090
},
{
"epoch": 0.8890235894327423,
"grad_norm": 20850.318359375,
"learning_rate": 8.522788008288553e-05,
"loss": 0.547,
"step": 1100
},
{
"epoch": 0.8971056220639491,
"grad_norm": 3731.269287109375,
"learning_rate": 8.490724038272413e-05,
"loss": 0.1154,
"step": 1110
},
{
"epoch": 0.9051876546951558,
"grad_norm": 5460.666015625,
"learning_rate": 8.458377538716006e-05,
"loss": 0.1365,
"step": 1120
},
{
"epoch": 0.9132696873263626,
"grad_norm": 11220.3447265625,
"learning_rate": 8.425751127655463e-05,
"loss": 0.1694,
"step": 1130
},
{
"epoch": 0.9213517199575694,
"grad_norm": 14035.845703125,
"learning_rate": 8.392847445782173e-05,
"loss": 0.3908,
"step": 1140
},
{
"epoch": 0.929433752588776,
"grad_norm": 19453.658203125,
"learning_rate": 8.359669156229062e-05,
"loss": 0.5536,
"step": 1150
},
{
"epoch": 0.9375157852199828,
"grad_norm": 3479.091064453125,
"learning_rate": 8.326218944355038e-05,
"loss": 0.1218,
"step": 1160
},
{
"epoch": 0.9455978178511896,
"grad_norm": 4935.92724609375,
"learning_rate": 8.292499517527649e-05,
"loss": 0.1284,
"step": 1170
},
{
"epoch": 0.9536798504823963,
"grad_norm": 7227.61083984375,
"learning_rate": 8.258513604903956e-05,
"loss": 0.1989,
"step": 1180
},
{
"epoch": 0.9617618831136031,
"grad_norm": 13335.935546875,
"learning_rate": 8.224263957209644e-05,
"loss": 0.3673,
"step": 1190
},
{
"epoch": 0.9698439157448098,
"grad_norm": 21474.529296875,
"learning_rate": 8.189753346516387e-05,
"loss": 0.5263,
"step": 1200
},
{
"epoch": 0.9779259483760165,
"grad_norm": 5255.57177734375,
"learning_rate": 8.154984566017473e-05,
"loss": 0.1166,
"step": 1210
},
{
"epoch": 0.9860079810072233,
"grad_norm": 7392.04931640625,
"learning_rate": 8.119960429801749e-05,
"loss": 0.1669,
"step": 1220
},
{
"epoch": 0.9940900136384301,
"grad_norm": 12895.5009765625,
"learning_rate": 8.08468377262584e-05,
"loss": 0.3262,
"step": 1230
},
{
"epoch": 1.002424609789362,
"grad_norm": 5290.23193359375,
"learning_rate": 8.049157449684723e-05,
"loss": 0.4882,
"step": 1240
},
{
"epoch": 1.0105066424205689,
"grad_norm": 2962.177978515625,
"learning_rate": 8.013384336380631e-05,
"loss": 0.123,
"step": 1250
},
{
"epoch": 1.017780471788655,
"grad_norm": 5345.2529296875,
"learning_rate": 7.97736732809032e-05,
"loss": 0.13,
"step": 1260
},
{
"epoch": 1.0258625044198615,
"grad_norm": 7691.60693359375,
"learning_rate": 7.941109339930737e-05,
"loss": 0.222,
"step": 1270
},
{
"epoch": 1.0339445370510683,
"grad_norm": 14488.4716796875,
"learning_rate": 7.90461330652307e-05,
"loss": 0.4303,
"step": 1280
},
{
"epoch": 1.042026569682275,
"grad_norm": 3190.872802734375,
"learning_rate": 7.86788218175523e-05,
"loss": 0.4731,
"step": 1290
},
{
"epoch": 1.0501086023134818,
"grad_norm": 4965.77587890625,
"learning_rate": 7.830918938542775e-05,
"loss": 0.127,
"step": 1300
},
{
"epoch": 1.0581906349446886,
"grad_norm": 4486.927734375,
"learning_rate": 7.793726568588285e-05,
"loss": 0.1454,
"step": 1310
},
{
"epoch": 1.0662726675758953,
"grad_norm": 7518.36572265625,
"learning_rate": 7.756308082139221e-05,
"loss": 0.2014,
"step": 1320
},
{
"epoch": 1.074354700207102,
"grad_norm": 14143.939453125,
"learning_rate": 7.718666507744294e-05,
"loss": 0.4438,
"step": 1330
},
{
"epoch": 1.0824367328383089,
"grad_norm": 5718.95849609375,
"learning_rate": 7.680804892008329e-05,
"loss": 0.4564,
"step": 1340
},
{
"epoch": 1.0905187654695156,
"grad_norm": 3672.863525390625,
"learning_rate": 7.642726299345693e-05,
"loss": 0.1011,
"step": 1350
},
{
"epoch": 1.0986007981007224,
"grad_norm": 4599.12109375,
"learning_rate": 7.604433811732264e-05,
"loss": 0.1125,
"step": 1360
},
{
"epoch": 1.1066828307319292,
"grad_norm": 7452.79638671875,
"learning_rate": 7.56593052845598e-05,
"loss": 0.2057,
"step": 1370
},
{
"epoch": 1.1147648633631357,
"grad_norm": 13006.572265625,
"learning_rate": 7.527219565866004e-05,
"loss": 0.4531,
"step": 1380
},
{
"epoch": 1.1228468959943425,
"grad_norm": 5128.93408203125,
"learning_rate": 7.488304057120485e-05,
"loss": 0.4337,
"step": 1390
},
{
"epoch": 1.1309289286255493,
"grad_norm": 4290.82470703125,
"learning_rate": 7.449187151932969e-05,
"loss": 0.1138,
"step": 1400
},
{
"epoch": 1.139010961256756,
"grad_norm": 5360.4326171875,
"learning_rate": 7.409872016317473e-05,
"loss": 0.1189,
"step": 1410
},
{
"epoch": 1.1470929938879628,
"grad_norm": 7620.78076171875,
"learning_rate": 7.370361832332242e-05,
"loss": 0.1793,
"step": 1420
},
{
"epoch": 1.1551750265191696,
"grad_norm": 14199.1806640625,
"learning_rate": 7.330659797822186e-05,
"loss": 0.4253,
"step": 1430
},
{
"epoch": 1.1632570591503764,
"grad_norm": 5826.39306640625,
"learning_rate": 7.290769126160073e-05,
"loss": 0.4451,
"step": 1440
},
{
"epoch": 1.1713390917815831,
"grad_norm": 4373.45654296875,
"learning_rate": 7.25069304598644e-05,
"loss": 0.1064,
"step": 1450
},
{
"epoch": 1.17942112441279,
"grad_norm": 4005.251220703125,
"learning_rate": 7.210434800948271e-05,
"loss": 0.1311,
"step": 1460
},
{
"epoch": 1.1875031570439964,
"grad_norm": 8311.0283203125,
"learning_rate": 7.169997649436472e-05,
"loss": 0.2099,
"step": 1470
},
{
"epoch": 1.1955851896752032,
"grad_norm": 14924.146484375,
"learning_rate": 7.129384864322145e-05,
"loss": 0.4236,
"step": 1480
},
{
"epoch": 1.20366722230641,
"grad_norm": 5517.5380859375,
"learning_rate": 7.088599732691689e-05,
"loss": 0.4284,
"step": 1490
},
{
"epoch": 1.2117492549376168,
"grad_norm": 4764.0791015625,
"learning_rate": 7.047645555580752e-05,
"loss": 0.1192,
"step": 1500
},
{
"epoch": 1.2198312875688235,
"grad_norm": 4838.24169921875,
"learning_rate": 7.006525647707054e-05,
"loss": 0.1189,
"step": 1510
},
{
"epoch": 1.2279133202000303,
"grad_norm": 7582.78125,
"learning_rate": 6.965243337202108e-05,
"loss": 0.206,
"step": 1520
},
{
"epoch": 1.235995352831237,
"grad_norm": 16726.548828125,
"learning_rate": 6.923801965341842e-05,
"loss": 0.3862,
"step": 1530
},
{
"epoch": 1.2440773854624438,
"grad_norm": 5702.87158203125,
"learning_rate": 6.882204886276176e-05,
"loss": 0.4396,
"step": 1540
},
{
"epoch": 1.2521594180936506,
"grad_norm": 5636.6826171875,
"learning_rate": 6.840455466757534e-05,
"loss": 0.1171,
"step": 1550
},
{
"epoch": 1.2602414507248574,
"grad_norm": 6394.9111328125,
"learning_rate": 6.798557085868353e-05,
"loss": 0.1206,
"step": 1560
},
{
"epoch": 1.2683234833560642,
"grad_norm": 9082.0771484375,
"learning_rate": 6.7565131347476e-05,
"loss": 0.2004,
"step": 1570
},
{
"epoch": 1.2764055159872707,
"grad_norm": 14265.474609375,
"learning_rate": 6.714327016316287e-05,
"loss": 0.3939,
"step": 1580
},
{
"epoch": 1.2844875486184775,
"grad_norm": 3703.578125,
"learning_rate": 6.672002145002052e-05,
"loss": 0.4088,
"step": 1590
},
{
"epoch": 1.2925695812496842,
"grad_norm": 3103.66552734375,
"learning_rate": 6.629541946462817e-05,
"loss": 0.1112,
"step": 1600
},
{
"epoch": 1.300651613880891,
"grad_norm": 4975.703125,
"learning_rate": 6.586949857309503e-05,
"loss": 0.1093,
"step": 1610
},
{
"epoch": 1.3087336465120978,
"grad_norm": 7071.45849609375,
"learning_rate": 6.544229324827901e-05,
"loss": 0.1796,
"step": 1620
},
{
"epoch": 1.3168156791433046,
"grad_norm": 14340.7646484375,
"learning_rate": 6.501383806699648e-05,
"loss": 0.4226,
"step": 1630
},
{
"epoch": 1.3248977117745113,
"grad_norm": 3171.169189453125,
"learning_rate": 6.458416770722371e-05,
"loss": 0.4268,
"step": 1640
},
{
"epoch": 1.332979744405718,
"grad_norm": 5536.4130859375,
"learning_rate": 6.415331694529014e-05,
"loss": 0.1127,
"step": 1650
},
{
"epoch": 1.3410617770369249,
"grad_norm": 5350.509765625,
"learning_rate": 6.372132065306373e-05,
"loss": 0.1157,
"step": 1660
},
{
"epoch": 1.3491438096681314,
"grad_norm": 6592.888671875,
"learning_rate": 6.328821379512846e-05,
"loss": 0.1933,
"step": 1670
},
{
"epoch": 1.3572258422993384,
"grad_norm": 17171.4453125,
"learning_rate": 6.285403142595447e-05,
"loss": 0.4236,
"step": 1680
},
{
"epoch": 1.365307874930545,
"grad_norm": 4746.12451171875,
"learning_rate": 6.241880868706075e-05,
"loss": 0.4297,
"step": 1690
},
{
"epoch": 1.3733899075617517,
"grad_norm": 3841.875732421875,
"learning_rate": 6.198258080417097e-05,
"loss": 0.1022,
"step": 1700
},
{
"epoch": 1.3814719401929585,
"grad_norm": 4221.74853515625,
"learning_rate": 6.154538308436232e-05,
"loss": 0.1269,
"step": 1710
},
{
"epoch": 1.3895539728241653,
"grad_norm": 8146.25439453125,
"learning_rate": 6.110725091320799e-05,
"loss": 0.2129,
"step": 1720
},
{
"epoch": 1.397636005455372,
"grad_norm": 13520.1435546875,
"learning_rate": 6.066821975191301e-05,
"loss": 0.4078,
"step": 1730
},
{
"epoch": 1.4057180380865788,
"grad_norm": 3531.795166015625,
"learning_rate": 6.022832513444421e-05,
"loss": 0.4144,
"step": 1740
},
{
"epoch": 1.4138000707177856,
"grad_norm": 3877.4111328125,
"learning_rate": 5.978760266465415e-05,
"loss": 0.1103,
"step": 1750
},
{
"epoch": 1.4218821033489923,
"grad_norm": 5296.13427734375,
"learning_rate": 5.9346088013399504e-05,
"loss": 0.1116,
"step": 1760
},
{
"epoch": 1.4299641359801991,
"grad_norm": 9570.111328125,
"learning_rate": 5.890381691565388e-05,
"loss": 0.1852,
"step": 1770
},
{
"epoch": 1.4380461686114057,
"grad_norm": 14839.6357421875,
"learning_rate": 5.8460825167615585e-05,
"loss": 0.4161,
"step": 1780
},
{
"epoch": 1.4461282012426127,
"grad_norm": 4930.75537109375,
"learning_rate": 5.801714862381038e-05,
"loss": 0.3991,
"step": 1790
},
{
"epoch": 1.4542102338738192,
"grad_norm": 4098.77587890625,
"learning_rate": 5.757282319418949e-05,
"loss": 0.0978,
"step": 1800
},
{
"epoch": 1.462292266505026,
"grad_norm": 5240.30224609375,
"learning_rate": 5.712788484122314e-05,
"loss": 0.1172,
"step": 1810
},
{
"epoch": 1.4703742991362327,
"grad_norm": 11101.5068359375,
"learning_rate": 5.668236957698991e-05,
"loss": 0.1818,
"step": 1820
},
{
"epoch": 1.4784563317674395,
"grad_norm": 14308.4248046875,
"learning_rate": 5.623631346026197e-05,
"loss": 0.4063,
"step": 1830
},
{
"epoch": 1.4865383643986463,
"grad_norm": 5231.79150390625,
"learning_rate": 5.5789752593586555e-05,
"loss": 0.426,
"step": 1840
},
{
"epoch": 1.494620397029853,
"grad_norm": 4275.7998046875,
"learning_rate": 5.5342723120364037e-05,
"loss": 0.1099,
"step": 1850
},
{
"epoch": 1.5027024296610598,
"grad_norm": 4952.5634765625,
"learning_rate": 5.489526122192246e-05,
"loss": 0.1166,
"step": 1860
},
{
"epoch": 1.5107844622922664,
"grad_norm": 7838.9306640625,
"learning_rate": 5.444740311458915e-05,
"loss": 0.1812,
"step": 1870
},
{
"epoch": 1.5188664949234734,
"grad_norm": 15917.42578125,
"learning_rate": 5.3999185046759514e-05,
"loss": 0.4096,
"step": 1880
},
{
"epoch": 1.52694852755468,
"grad_norm": 4498.63916015625,
"learning_rate": 5.355064329596316e-05,
"loss": 0.4184,
"step": 1890
},
{
"epoch": 1.535030560185887,
"grad_norm": 6156.8642578125,
"learning_rate": 5.31018141659277e-05,
"loss": 0.1092,
"step": 1900
},
{
"epoch": 1.5431125928170935,
"grad_norm": 6385.4296875,
"learning_rate": 5.2652733983640424e-05,
"loss": 0.1191,
"step": 1910
},
{
"epoch": 1.5511946254483002,
"grad_norm": 9750.2802734375,
"learning_rate": 5.220343909640809e-05,
"loss": 0.1931,
"step": 1920
},
{
"epoch": 1.559276658079507,
"grad_norm": 14060.41796875,
"learning_rate": 5.175396586891509e-05,
"loss": 0.3732,
"step": 1930
},
{
"epoch": 1.5673586907107138,
"grad_norm": 6008.52197265625,
"learning_rate": 5.1304350680280166e-05,
"loss": 0.4266,
"step": 1940
},
{
"epoch": 1.5754407233419205,
"grad_norm": 3793.300048828125,
"learning_rate": 5.0854629921112015e-05,
"loss": 0.1171,
"step": 1950
},
{
"epoch": 1.583522755973127,
"grad_norm": 5250.0126953125,
"learning_rate": 5.0404839990563936e-05,
"loss": 0.1195,
"step": 1960
},
{
"epoch": 1.591604788604334,
"grad_norm": 8599.5751953125,
"learning_rate": 4.995501729338774e-05,
"loss": 0.1968,
"step": 1970
},
{
"epoch": 1.5996868212355406,
"grad_norm": 15399.6591796875,
"learning_rate": 4.95051982369873e-05,
"loss": 0.4164,
"step": 1980
},
{
"epoch": 1.6077688538667476,
"grad_norm": 4506.6201171875,
"learning_rate": 4.9055419228471786e-05,
"loss": 0.4034,
"step": 1990
},
{
"epoch": 1.6158508864979542,
"grad_norm": 5589.3203125,
"learning_rate": 4.860571667170903e-05,
"loss": 0.1049,
"step": 2000
},
{
"epoch": 1.623932919129161,
"grad_norm": 4892.732421875,
"learning_rate": 4.815612696437907e-05,
"loss": 0.1088,
"step": 2010
},
{
"epoch": 1.6320149517603677,
"grad_norm": 7929.884765625,
"learning_rate": 4.7706686495028196e-05,
"loss": 0.1813,
"step": 2020
},
{
"epoch": 1.6400969843915745,
"grad_norm": 15299.4853515625,
"learning_rate": 4.725743164012382e-05,
"loss": 0.3857,
"step": 2030
},
{
"epoch": 1.6481790170227812,
"grad_norm": 6454.0263671875,
"learning_rate": 4.680839876111024e-05,
"loss": 0.4309,
"step": 2040
},
{
"epoch": 1.656261049653988,
"grad_norm": 4587.6982421875,
"learning_rate": 4.63596242014656e-05,
"loss": 0.1062,
"step": 2050
},
{
"epoch": 1.6643430822851948,
"grad_norm": 7168.818359375,
"learning_rate": 4.5911144283760424e-05,
"loss": 0.1226,
"step": 2060
},
{
"epoch": 1.6724251149164013,
"grad_norm": 8726.30859375,
"learning_rate": 4.546299530671772e-05,
"loss": 0.1861,
"step": 2070
},
{
"epoch": 1.6805071475476083,
"grad_norm": 14318.248046875,
"learning_rate": 4.501521354227506e-05,
"loss": 0.4043,
"step": 2080
},
{
"epoch": 1.6885891801788149,
"grad_norm": 3904.837646484375,
"learning_rate": 4.456783523264888e-05,
"loss": 0.415,
"step": 2090
},
{
"epoch": 1.6966712128100219,
"grad_norm": 4419.138671875,
"learning_rate": 4.412089658740108e-05,
"loss": 0.115,
"step": 2100
},
{
"epoch": 1.7047532454412284,
"grad_norm": 5125.00830078125,
"learning_rate": 4.367443378050834e-05,
"loss": 0.1115,
"step": 2110
},
{
"epoch": 1.7128352780724352,
"grad_norm": 7952.75927734375,
"learning_rate": 4.322848294743431e-05,
"loss": 0.1883,
"step": 2120
},
{
"epoch": 1.720917310703642,
"grad_norm": 14957.6865234375,
"learning_rate": 4.278308018220488e-05,
"loss": 0.3962,
"step": 2130
},
{
"epoch": 1.7289993433348487,
"grad_norm": 4730.97802734375,
"learning_rate": 4.2338261534486844e-05,
"loss": 0.4158,
"step": 2140
},
{
"epoch": 1.7370813759660555,
"grad_norm": 4476.10693359375,
"learning_rate": 4.189406300667018e-05,
"loss": 0.1186,
"step": 2150
},
{
"epoch": 1.7451634085972623,
"grad_norm": 6663.09033203125,
"learning_rate": 4.145052055095405e-05,
"loss": 0.1131,
"step": 2160
},
{
"epoch": 1.753245441228469,
"grad_norm": 9589.1103515625,
"learning_rate": 4.1007670066436956e-05,
"loss": 0.194,
"step": 2170
},
{
"epoch": 1.7613274738596756,
"grad_norm": 15578.5654296875,
"learning_rate": 4.056554739621116e-05,
"loss": 0.4351,
"step": 2180
},
{
"epoch": 1.7694095064908826,
"grad_norm": 5586.77734375,
"learning_rate": 4.012418832446167e-05,
"loss": 0.4127,
"step": 2190
},
{
"epoch": 1.7774915391220891,
"grad_norm": 3721.786376953125,
"learning_rate": 3.968362857356992e-05,
"loss": 0.1226,
"step": 2200
},
{
"epoch": 1.7855735717532961,
"grad_norm": 5573.1240234375,
"learning_rate": 3.9243903801222567e-05,
"loss": 0.1096,
"step": 2210
},
{
"epoch": 1.7936556043845027,
"grad_norm": 7120.84423828125,
"learning_rate": 3.880504959752537e-05,
"loss": 0.1725,
"step": 2220
},
{
"epoch": 1.8017376370157094,
"grad_norm": 15186.1171875,
"learning_rate": 3.836710148212271e-05,
"loss": 0.3867,
"step": 2230
},
{
"epoch": 1.8098196696469162,
"grad_norm": 3603.533447265625,
"learning_rate": 3.793009490132266e-05,
"loss": 0.4266,
"step": 2240
},
{
"epoch": 1.817901702278123,
"grad_norm": 6406.8896484375,
"learning_rate": 3.7494065225228095e-05,
"loss": 0.1177,
"step": 2250
},
{
"epoch": 1.8259837349093297,
"grad_norm": 5019.619140625,
"learning_rate": 3.705904774487396e-05,
"loss": 0.1021,
"step": 2260
},
{
"epoch": 1.8340657675405363,
"grad_norm": 9019.923828125,
"learning_rate": 3.662507766937084e-05,
"loss": 0.1747,
"step": 2270
},
{
"epoch": 1.8421478001717433,
"grad_norm": 15076.8681640625,
"learning_rate": 3.6192190123055295e-05,
"loss": 0.395,
"step": 2280
},
{
"epoch": 1.8502298328029498,
"grad_norm": 4530.0439453125,
"learning_rate": 3.576042014264695e-05,
"loss": 0.4086,
"step": 2290
},
{
"epoch": 1.8583118654341568,
"grad_norm": 3448.111572265625,
"learning_rate": 3.5329802674412736e-05,
"loss": 0.0989,
"step": 2300
},
{
"epoch": 1.8663938980653634,
"grad_norm": 6909.45947265625,
"learning_rate": 3.490037257133847e-05,
"loss": 0.1141,
"step": 2310
},
{
"epoch": 1.8744759306965701,
"grad_norm": 9935.5,
"learning_rate": 3.447216459030789e-05,
"loss": 0.2029,
"step": 2320
},
{
"epoch": 1.882557963327777,
"grad_norm": 15530.99609375,
"learning_rate": 3.404521338928958e-05,
"loss": 0.3809,
"step": 2330
},
{
"epoch": 1.8906399959589837,
"grad_norm": 4696.42919921875,
"learning_rate": 3.3619553524531804e-05,
"loss": 0.4201,
"step": 2340
},
{
"epoch": 1.8987220285901905,
"grad_norm": 4922.81494140625,
"learning_rate": 3.319521944776567e-05,
"loss": 0.1117,
"step": 2350
},
{
"epoch": 1.9068040612213972,
"grad_norm": 4787.01171875,
"learning_rate": 3.2772245503416645e-05,
"loss": 0.1128,
"step": 2360
},
{
"epoch": 1.914886093852604,
"grad_norm": 8945.86328125,
"learning_rate": 3.23506659258249e-05,
"loss": 0.1763,
"step": 2370
},
{
"epoch": 1.9229681264838105,
"grad_norm": 15152.3671875,
"learning_rate": 3.1930514836474357e-05,
"loss": 0.4296,
"step": 2380
},
{
"epoch": 1.9310501591150175,
"grad_norm": 3111.07568359375,
"learning_rate": 3.151182624123109e-05,
"loss": 0.4068,
"step": 2390
},
{
"epoch": 1.939132191746224,
"grad_norm": 4133.765625,
"learning_rate": 3.109463402759094e-05,
"loss": 0.1128,
"step": 2400
},
{
"epoch": 1.947214224377431,
"grad_norm": 6019.08984375,
"learning_rate": 3.0678971961936764e-05,
"loss": 0.1175,
"step": 2410
},
{
"epoch": 1.9552962570086376,
"grad_norm": 7155.51318359375,
"learning_rate": 3.0264873686805473e-05,
"loss": 0.1991,
"step": 2420
},
{
"epoch": 1.9633782896398444,
"grad_norm": 14954.6650390625,
"learning_rate": 2.98523727181651e-05,
"loss": 0.3817,
"step": 2430
},
{
"epoch": 1.9714603222710512,
"grad_norm": 4551.48974609375,
"learning_rate": 2.944150244270213e-05,
"loss": 0.4094,
"step": 2440
},
{
"epoch": 1.979542354902258,
"grad_norm": 6931.7294921875,
"learning_rate": 2.903229611511925e-05,
"loss": 0.1074,
"step": 2450
},
{
"epoch": 1.9876243875334647,
"grad_norm": 9782.849609375,
"learning_rate": 2.8624786855443787e-05,
"loss": 0.1707,
"step": 2460
},
{
"epoch": 1.9957064201646713,
"grad_norm": 15994.1611328125,
"learning_rate": 2.8219007646347096e-05,
"loss": 0.3575,
"step": 2470
},
{
"epoch": 2.0040410163156035,
"grad_norm": 5267.82275390625,
"learning_rate": 2.781499133047508e-05,
"loss": 0.3699,
"step": 2480
},
{
"epoch": 2.01212304894681,
"grad_norm": 4731.33203125,
"learning_rate": 2.7412770607789888e-05,
"loss": 0.105,
"step": 2490
},
{
"epoch": 2.020205081578017,
"grad_norm": 6523.8291015625,
"learning_rate": 2.7012378032923346e-05,
"loss": 0.1258,
"step": 2500
},
{
"epoch": 2.0282871142092236,
"grad_norm": 10412.5693359375,
"learning_rate": 2.6613846012542042e-05,
"loss": 0.2103,
"step": 2510
},
{
"epoch": 2.0363691468404306,
"grad_norm": 16831.271484375,
"learning_rate": 2.6217206802724464e-05,
"loss": 0.4632,
"step": 2520
},
{
"epoch": 2.044451179471637,
"grad_norm": 5685.7470703125,
"learning_rate": 2.5822492506350228e-05,
"loss": 0.3155,
"step": 2530
},
{
"epoch": 2.0525332121028437,
"grad_norm": 6857.482421875,
"learning_rate": 2.5429735070501792e-05,
"loss": 0.1164,
"step": 2540
},
{
"epoch": 2.0606152447340507,
"grad_norm": 5623.13720703125,
"learning_rate": 2.503896628387874e-05,
"loss": 0.134,
"step": 2550
},
{
"epoch": 2.068697277365257,
"grad_norm": 9535.6455078125,
"learning_rate": 2.4650217774224875e-05,
"loss": 0.2421,
"step": 2560
},
{
"epoch": 2.076779309996464,
"grad_norm": 17088.423828125,
"learning_rate": 2.426352100576844e-05,
"loss": 0.4628,
"step": 2570
},
{
"epoch": 2.0848613426276708,
"grad_norm": 6173.42333984375,
"learning_rate": 2.387890727667535e-05,
"loss": 0.2935,
"step": 2580
},
{
"epoch": 2.0929433752588777,
"grad_norm": 5329.630859375,
"learning_rate": 2.349640771651611e-05,
"loss": 0.1388,
"step": 2590
},
{
"epoch": 2.1010254078900843,
"grad_norm": 7430.10009765625,
"learning_rate": 2.3116053283746248e-05,
"loss": 0.1143,
"step": 2600
},
{
"epoch": 2.1091074405212913,
"grad_norm": 10057.865234375,
"learning_rate": 2.2737874763200616e-05,
"loss": 0.2347,
"step": 2610
},
{
"epoch": 2.117189473152498,
"grad_norm": 16924.0546875,
"learning_rate": 2.2361902763601746e-05,
"loss": 0.4846,
"step": 2620
},
{
"epoch": 2.1252715057837044,
"grad_norm": 4677.1181640625,
"learning_rate": 2.198816771508247e-05,
"loss": 0.3046,
"step": 2630
},
{
"epoch": 2.1333535384149114,
"grad_norm": 5968.64208984375,
"learning_rate": 2.1616699866723028e-05,
"loss": 0.1027,
"step": 2640
},
{
"epoch": 2.141435571046118,
"grad_norm": 5689.19677734375,
"learning_rate": 2.1247529284102723e-05,
"loss": 0.1247,
"step": 2650
},
{
"epoch": 2.149517603677325,
"grad_norm": 9622.009765625,
"learning_rate": 2.0880685846866533e-05,
"loss": 0.2225,
"step": 2660
},
{
"epoch": 2.1575996363085315,
"grad_norm": 18100.33984375,
"learning_rate": 2.0516199246306728e-05,
"loss": 0.4436,
"step": 2670
},
{
"epoch": 2.1656816689397385,
"grad_norm": 6390.056640625,
"learning_rate": 2.0154098982959745e-05,
"loss": 0.3035,
"step": 2680
},
{
"epoch": 2.173763701570945,
"grad_norm": 5377.6103515625,
"learning_rate": 1.979441436421849e-05,
"loss": 0.1145,
"step": 2690
},
{
"epoch": 2.181845734202152,
"grad_norm": 5494.45556640625,
"learning_rate": 1.9437174501960303e-05,
"loss": 0.1337,
"step": 2700
},
{
"epoch": 2.1899277668333585,
"grad_norm": 11127.666015625,
"learning_rate": 1.9082408310190686e-05,
"loss": 0.2241,
"step": 2710
},
{
"epoch": 2.1980097994645655,
"grad_norm": 16851.986328125,
"learning_rate": 1.8730144502703096e-05,
"loss": 0.4828,
"step": 2720
},
{
"epoch": 2.206091832095772,
"grad_norm": 4164.68603515625,
"learning_rate": 1.838041159075493e-05,
"loss": 0.2991,
"step": 2730
},
{
"epoch": 2.2141738647269786,
"grad_norm": 5838.01806640625,
"learning_rate": 1.803323788075992e-05,
"loss": 0.1175,
"step": 2740
},
{
"epoch": 2.2222558973581856,
"grad_norm": 6435.92626953125,
"learning_rate": 1.768865147199706e-05,
"loss": 0.1222,
"step": 2750
},
{
"epoch": 2.230337929989392,
"grad_norm": 9642.9208984375,
"learning_rate": 1.7346680254336367e-05,
"loss": 0.235,
"step": 2760
},
{
"epoch": 2.238419962620599,
"grad_norm": 16851.630859375,
"learning_rate": 1.7007351905981512e-05,
"loss": 0.507,
"step": 2770
},
{
"epoch": 2.2465019952518057,
"grad_norm": 5162.921875,
"learning_rate": 1.667069389122965e-05,
"loss": 0.3097,
"step": 2780
},
{
"epoch": 2.2545840278830127,
"grad_norm": 4561.232421875,
"learning_rate": 1.63367334582485e-05,
"loss": 0.1021,
"step": 2790
},
{
"epoch": 2.2626660605142193,
"grad_norm": 6248.9111328125,
"learning_rate": 1.6005497636871035e-05,
"loss": 0.1316,
"step": 2800
},
{
"epoch": 2.2707480931454262,
"grad_norm": 10710.337890625,
"learning_rate": 1.5677013236407666e-05,
"loss": 0.2236,
"step": 2810
},
{
"epoch": 2.278830125776633,
"grad_norm": 16736.646484375,
"learning_rate": 1.535130684347642e-05,
"loss": 0.5029,
"step": 2820
},
{
"epoch": 2.28691215840784,
"grad_norm": 4309.60888671875,
"learning_rate": 1.5028404819851083e-05,
"loss": 0.3004,
"step": 2830
},
{
"epoch": 2.2949941910390463,
"grad_norm": 3661.167236328125,
"learning_rate": 1.4708333300327554e-05,
"loss": 0.0932,
"step": 2840
},
{
"epoch": 2.303076223670253,
"grad_norm": 7402.5595703125,
"learning_rate": 1.4391118190608566e-05,
"loss": 0.1322,
"step": 2850
},
{
"epoch": 2.31115825630146,
"grad_norm": 11427.3955078125,
"learning_rate": 1.4076785165206962e-05,
"loss": 0.2263,
"step": 2860
},
{
"epoch": 2.3192402889326664,
"grad_norm": 16985.4453125,
"learning_rate": 1.376535966536764e-05,
"loss": 0.4712,
"step": 2870
},
{
"epoch": 2.3273223215638734,
"grad_norm": 3604.2158203125,
"learning_rate": 1.3456866897008435e-05,
"loss": 0.2926,
"step": 2880
},
{
"epoch": 2.33540435419508,
"grad_norm": 5294.6220703125,
"learning_rate": 1.3151331828680019e-05,
"loss": 0.1115,
"step": 2890
},
{
"epoch": 2.343486386826287,
"grad_norm": 7055.0693359375,
"learning_rate": 1.284877918954499e-05,
"loss": 0.1368,
"step": 2900
},
{
"epoch": 2.3515684194574935,
"grad_norm": 10659.4345703125,
"learning_rate": 1.2549233467376431e-05,
"loss": 0.241,
"step": 2910
},
{
"epoch": 2.3596504520887005,
"grad_norm": 17015.447265625,
"learning_rate": 1.225271890657585e-05,
"loss": 0.4833,
"step": 2920
},
{
"epoch": 2.367732484719907,
"grad_norm": 4985.89501953125,
"learning_rate": 1.1959259506210952e-05,
"loss": 0.3094,
"step": 2930
},
{
"epoch": 2.375814517351114,
"grad_norm": 4372.19775390625,
"learning_rate": 1.166887901807322e-05,
"loss": 0.1128,
"step": 2940
},
{
"epoch": 2.3838965499823206,
"grad_norm": 6398.43115234375,
"learning_rate": 1.1381600944755494e-05,
"loss": 0.1314,
"step": 2950
},
{
"epoch": 2.391978582613527,
"grad_norm": 9847.6669921875,
"learning_rate": 1.1097448537749766e-05,
"loss": 0.2315,
"step": 2960
},
{
"epoch": 2.400060615244734,
"grad_norm": 17311.58984375,
"learning_rate": 1.0816444795565229e-05,
"loss": 0.4817,
"step": 2970
},
{
"epoch": 2.4081426478759407,
"grad_norm": 4639.50732421875,
"learning_rate": 1.0538612461866875e-05,
"loss": 0.3051,
"step": 2980
},
{
"epoch": 2.4162246805071477,
"grad_norm": 5695.07763671875,
"learning_rate": 1.0263974023634665e-05,
"loss": 0.1216,
"step": 2990
},
{
"epoch": 2.424306713138354,
"grad_norm": 5502.41748046875,
"learning_rate": 9.992551709343512e-06,
"loss": 0.1066,
"step": 3000
},
{
"epoch": 2.432388745769561,
"grad_norm": 10350.740234375,
"learning_rate": 9.724367487164143e-06,
"loss": 0.2252,
"step": 3010
},
{
"epoch": 2.4404707784007678,
"grad_norm": 16755.990234375,
"learning_rate": 9.459443063185103e-06,
"loss": 0.4636,
"step": 3020
},
{
"epoch": 2.4485528110319743,
"grad_norm": 4946.88916015625,
"learning_rate": 9.197799879655872e-06,
"loss": 0.3093,
"step": 3030
},
{
"epoch": 2.4566348436631813,
"grad_norm": 5786.74072265625,
"learning_rate": 8.939459113251408e-06,
"loss": 0.1142,
"step": 3040
},
{
"epoch": 2.464716876294388,
"grad_norm": 5667.3583984375,
"learning_rate": 8.684441673358191e-06,
"loss": 0.1319,
"step": 3050
},
{
"epoch": 2.472798908925595,
"grad_norm": 10215.568359375,
"learning_rate": 8.432768200381824e-06,
"loss": 0.2403,
"step": 3060
},
{
"epoch": 2.4808809415568014,
"grad_norm": 18044.140625,
"learning_rate": 8.18445906407651e-06,
"loss": 0.4518,
"step": 3070
},
{
"epoch": 2.4889629741880084,
"grad_norm": 4549.96630859375,
"learning_rate": 7.939534361896345e-06,
"loss": 0.2893,
"step": 3080
},
{
"epoch": 2.497045006819215,
"grad_norm": 4263.11328125,
"learning_rate": 7.69801391736869e-06,
"loss": 0.1106,
"step": 3090
},
{
"epoch": 2.505127039450422,
"grad_norm": 7342.88232421875,
"learning_rate": 7.459917278489736e-06,
"loss": 0.1295,
"step": 3100
},
{
"epoch": 2.5132090720816285,
"grad_norm": 10624.1884765625,
"learning_rate": 7.2252637161423e-06,
"loss": 0.2455,
"step": 3110
},
{
"epoch": 2.5212911047128355,
"grad_norm": 16986.208984375,
"learning_rate": 6.994072222536113e-06,
"loss": 0.4854,
"step": 3120
},
{
"epoch": 2.529373137344042,
"grad_norm": 4643.2197265625,
"learning_rate": 6.766361509670688e-06,
"loss": 0.3066,
"step": 3130
},
{
"epoch": 2.5374551699752486,
"grad_norm": 4168.1787109375,
"learning_rate": 6.542150007820735e-06,
"loss": 0.0954,
"step": 3140
},
{
"epoch": 2.5455372026064556,
"grad_norm": 5714.84326171875,
"learning_rate": 6.321455864044529e-06,
"loss": 0.1267,
"step": 3150
},
{
"epoch": 2.5536192352376625,
"grad_norm": 13147.404296875,
"learning_rate": 6.104296940715121e-06,
"loss": 0.2399,
"step": 3160
},
{
"epoch": 2.561701267868869,
"grad_norm": 17294.47265625,
"learning_rate": 5.890690814074595e-06,
"loss": 0.4733,
"step": 3170
},
{
"epoch": 2.5697833005000756,
"grad_norm": 4740.875,
"learning_rate": 5.680654772811528e-06,
"loss": 0.2935,
"step": 3180
},
{
"epoch": 2.5778653331312826,
"grad_norm": 7176.07177734375,
"learning_rate": 5.474205816661643e-06,
"loss": 0.125,
"step": 3190
},
{
"epoch": 2.585947365762489,
"grad_norm": 5670.5400390625,
"learning_rate": 5.271360655031943e-06,
"loss": 0.1339,
"step": 3200
},
{
"epoch": 2.594029398393696,
"grad_norm": 9239.3994140625,
"learning_rate": 5.072135705648268e-06,
"loss": 0.2228,
"step": 3210
},
{
"epoch": 2.6021114310249027,
"grad_norm": 16892.029296875,
"learning_rate": 4.876547093226513e-06,
"loss": 0.4666,
"step": 3220
},
{
"epoch": 2.6101934636561097,
"grad_norm": 5871.22998046875,
"learning_rate": 4.684610648167503e-06,
"loss": 0.2983,
"step": 3230
},
{
"epoch": 2.6182754962873163,
"grad_norm": 3902.05712890625,
"learning_rate": 4.496341905275786e-06,
"loss": 0.1114,
"step": 3240
},
{
"epoch": 2.626357528918523,
"grad_norm": 6525.04296875,
"learning_rate": 4.311756102502213e-06,
"loss": 0.1199,
"step": 3250
},
{
"epoch": 2.633631358286609,
"grad_norm": 8627.8505859375,
"learning_rate": 4.130868179710668e-06,
"loss": 0.1848,
"step": 3260
},
{
"epoch": 2.6417133909178157,
"grad_norm": 14247.7802734375,
"learning_rate": 3.953692777468876e-06,
"loss": 0.416,
"step": 3270
},
{
"epoch": 2.6497954235490226,
"grad_norm": 4759.4443359375,
"learning_rate": 3.7802442358634115e-06,
"loss": 0.3177,
"step": 3280
},
{
"epoch": 2.657877456180229,
"grad_norm": 4269.89892578125,
"learning_rate": 3.6105365933390754e-06,
"loss": 0.0936,
"step": 3290
},
{
"epoch": 2.665959488811436,
"grad_norm": 6439.9130859375,
"learning_rate": 3.4445835855626608e-06,
"loss": 0.1237,
"step": 3300
},
{
"epoch": 2.6740415214426427,
"grad_norm": 10297.259765625,
"learning_rate": 3.2823986443111854e-06,
"loss": 0.1953,
"step": 3310
},
{
"epoch": 2.6821235540738497,
"grad_norm": 15543.5859375,
"learning_rate": 3.1239948963848407e-06,
"loss": 0.4216,
"step": 3320
},
{
"epoch": 2.6902055867050563,
"grad_norm": 3996.56396484375,
"learning_rate": 2.9693851625444635e-06,
"loss": 0.3027,
"step": 3330
},
{
"epoch": 2.698287619336263,
"grad_norm": 4716.9501953125,
"learning_rate": 2.8185819564739037e-06,
"loss": 0.1063,
"step": 3340
},
{
"epoch": 2.70636965196747,
"grad_norm": 5304.83935546875,
"learning_rate": 2.671597483767191e-06,
"loss": 0.1064,
"step": 3350
},
{
"epoch": 2.714451684598677,
"grad_norm": 8085.703125,
"learning_rate": 2.528443640940653e-06,
"loss": 0.1923,
"step": 3360
},
{
"epoch": 2.7225337172298834,
"grad_norm": 16287.6484375,
"learning_rate": 2.38913201447003e-06,
"loss": 0.4251,
"step": 3370
},
{
"epoch": 2.73061574986109,
"grad_norm": 4544.6015625,
"learning_rate": 2.253673879852708e-06,
"loss": 0.3017,
"step": 3380
},
{
"epoch": 2.738697782492297,
"grad_norm": 4039.9560546875,
"learning_rate": 2.1220802006951203e-06,
"loss": 0.1068,
"step": 3390
},
{
"epoch": 2.7467798151235034,
"grad_norm": 5422.513671875,
"learning_rate": 1.994361627825364e-06,
"loss": 0.1124,
"step": 3400
},
{
"epoch": 2.7548618477547104,
"grad_norm": 10583.4833984375,
"learning_rate": 1.8705284984311677e-06,
"loss": 0.2164,
"step": 3410
},
{
"epoch": 2.762943880385917,
"grad_norm": 16692.498046875,
"learning_rate": 1.7505908352232037e-06,
"loss": 0.4409,
"step": 3420
},
{
"epoch": 2.771025913017124,
"grad_norm": 5393.38427734375,
"learning_rate": 1.634558345623921e-06,
"loss": 0.3108,
"step": 3430
},
{
"epoch": 2.7791079456483305,
"grad_norm": 5366.48095703125,
"learning_rate": 1.5224404209818076e-06,
"loss": 0.1037,
"step": 3440
},
{
"epoch": 2.787189978279537,
"grad_norm": 5171.4765625,
"learning_rate": 1.414246135811309e-06,
"loss": 0.1039,
"step": 3450
},
{
"epoch": 2.795272010910744,
"grad_norm": 9813.6513671875,
"learning_rate": 1.309984247058349e-06,
"loss": 0.1846,
"step": 3460
},
{
"epoch": 2.803354043541951,
"grad_norm": 15926.455078125,
"learning_rate": 1.2096631933915802e-06,
"loss": 0.4189,
"step": 3470
},
{
"epoch": 2.8114360761731576,
"grad_norm": 3362.9609375,
"learning_rate": 1.1132910945193698e-06,
"loss": 0.3116,
"step": 3480
},
{
"epoch": 2.819518108804364,
"grad_norm": 4051.453125,
"learning_rate": 1.0208757505326016e-06,
"loss": 0.1044,
"step": 3490
},
{
"epoch": 2.827600141435571,
"grad_norm": 5556.53857421875,
"learning_rate": 9.324246412733939e-07,
"loss": 0.1006,
"step": 3500
}
],
"logging_steps": 10,
"max_steps": 3714,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 250,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.34267192081452e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}