math-deepseek-LORA-ArithHard / trainer_state.json
jdannem6's picture
Uploaded checkpoint-30000
a2148d5 verified
{
"best_metric": 1.262895941734314,
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-27500",
"epoch": 0.75,
"eval_steps": 500,
"global_step": 30000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 2.7229208946228027,
"learning_rate": 4.0000000000000003e-07,
"loss": 2.2066,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 5.007352828979492,
"learning_rate": 8.000000000000001e-07,
"loss": 2.1122,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 3.8536276817321777,
"learning_rate": 1.2000000000000002e-06,
"loss": 2.2073,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 6.411635875701904,
"learning_rate": 1.6000000000000001e-06,
"loss": 2.3019,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 8.24316692352295,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.1321,
"step": 50
},
{
"epoch": 0.0,
"grad_norm": 4.362759113311768,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.1916,
"step": 60
},
{
"epoch": 0.0,
"grad_norm": 4.222783088684082,
"learning_rate": 2.8000000000000003e-06,
"loss": 2.2857,
"step": 70
},
{
"epoch": 0.0,
"grad_norm": 3.4690635204315186,
"learning_rate": 3.2000000000000003e-06,
"loss": 2.2253,
"step": 80
},
{
"epoch": 0.0,
"grad_norm": 4.878602981567383,
"learning_rate": 3.6000000000000003e-06,
"loss": 2.1578,
"step": 90
},
{
"epoch": 0.0,
"grad_norm": 4.750248908996582,
"learning_rate": 4.000000000000001e-06,
"loss": 2.0044,
"step": 100
},
{
"epoch": 0.0,
"grad_norm": 7.198167324066162,
"learning_rate": 4.4e-06,
"loss": 1.8984,
"step": 110
},
{
"epoch": 0.0,
"grad_norm": 3.500101327896118,
"learning_rate": 4.800000000000001e-06,
"loss": 1.9501,
"step": 120
},
{
"epoch": 0.0,
"grad_norm": 7.465259552001953,
"learning_rate": 5.2e-06,
"loss": 1.8413,
"step": 130
},
{
"epoch": 0.0,
"grad_norm": 3.6927084922790527,
"learning_rate": 5.600000000000001e-06,
"loss": 2.0101,
"step": 140
},
{
"epoch": 0.0,
"grad_norm": 8.518953323364258,
"learning_rate": 6e-06,
"loss": 1.8778,
"step": 150
},
{
"epoch": 0.0,
"grad_norm": 3.7254889011383057,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.8361,
"step": 160
},
{
"epoch": 0.0,
"grad_norm": 4.634269714355469,
"learning_rate": 6.800000000000001e-06,
"loss": 1.6214,
"step": 170
},
{
"epoch": 0.0,
"grad_norm": 8.053196907043457,
"learning_rate": 7.2000000000000005e-06,
"loss": 1.7552,
"step": 180
},
{
"epoch": 0.0,
"grad_norm": 3.407431125640869,
"learning_rate": 7.600000000000001e-06,
"loss": 1.8447,
"step": 190
},
{
"epoch": 0.01,
"grad_norm": 8.015592575073242,
"learning_rate": 8.000000000000001e-06,
"loss": 1.736,
"step": 200
},
{
"epoch": 0.01,
"grad_norm": 10.115213394165039,
"learning_rate": 8.400000000000001e-06,
"loss": 1.5497,
"step": 210
},
{
"epoch": 0.01,
"grad_norm": 6.565807819366455,
"learning_rate": 8.8e-06,
"loss": 1.6854,
"step": 220
},
{
"epoch": 0.01,
"grad_norm": 4.260842800140381,
"learning_rate": 9.200000000000002e-06,
"loss": 1.986,
"step": 230
},
{
"epoch": 0.01,
"grad_norm": 4.239616394042969,
"learning_rate": 9.600000000000001e-06,
"loss": 1.5284,
"step": 240
},
{
"epoch": 0.01,
"grad_norm": 3.048999786376953,
"learning_rate": 1e-05,
"loss": 1.7342,
"step": 250
},
{
"epoch": 0.01,
"grad_norm": 3.2346177101135254,
"learning_rate": 1.04e-05,
"loss": 1.8604,
"step": 260
},
{
"epoch": 0.01,
"grad_norm": 4.236307144165039,
"learning_rate": 1.0800000000000002e-05,
"loss": 1.6303,
"step": 270
},
{
"epoch": 0.01,
"grad_norm": 3.784558057785034,
"learning_rate": 1.1200000000000001e-05,
"loss": 1.6309,
"step": 280
},
{
"epoch": 0.01,
"grad_norm": 3.2512731552124023,
"learning_rate": 1.16e-05,
"loss": 1.8278,
"step": 290
},
{
"epoch": 0.01,
"grad_norm": 4.695291996002197,
"learning_rate": 1.2e-05,
"loss": 1.6175,
"step": 300
},
{
"epoch": 0.01,
"grad_norm": 5.554133892059326,
"learning_rate": 1.2400000000000002e-05,
"loss": 1.6657,
"step": 310
},
{
"epoch": 0.01,
"grad_norm": 24.56378936767578,
"learning_rate": 1.2800000000000001e-05,
"loss": 1.663,
"step": 320
},
{
"epoch": 0.01,
"grad_norm": 2.6168453693389893,
"learning_rate": 1.3200000000000002e-05,
"loss": 1.811,
"step": 330
},
{
"epoch": 0.01,
"grad_norm": 3.714534282684326,
"learning_rate": 1.3600000000000002e-05,
"loss": 1.7414,
"step": 340
},
{
"epoch": 0.01,
"grad_norm": 4.220114231109619,
"learning_rate": 1.4e-05,
"loss": 1.69,
"step": 350
},
{
"epoch": 0.01,
"grad_norm": 2.1215527057647705,
"learning_rate": 1.4400000000000001e-05,
"loss": 1.6113,
"step": 360
},
{
"epoch": 0.01,
"grad_norm": 2.8771586418151855,
"learning_rate": 1.48e-05,
"loss": 1.6141,
"step": 370
},
{
"epoch": 0.01,
"grad_norm": 5.036810398101807,
"learning_rate": 1.5200000000000002e-05,
"loss": 1.7124,
"step": 380
},
{
"epoch": 0.01,
"grad_norm": 1.9079569578170776,
"learning_rate": 1.5600000000000003e-05,
"loss": 1.629,
"step": 390
},
{
"epoch": 0.01,
"grad_norm": 9.102452278137207,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.659,
"step": 400
},
{
"epoch": 0.01,
"grad_norm": 6.5280866622924805,
"learning_rate": 1.64e-05,
"loss": 1.6147,
"step": 410
},
{
"epoch": 0.01,
"grad_norm": 6.27349853515625,
"learning_rate": 1.6800000000000002e-05,
"loss": 1.5316,
"step": 420
},
{
"epoch": 0.01,
"grad_norm": 21.586790084838867,
"learning_rate": 1.72e-05,
"loss": 1.7693,
"step": 430
},
{
"epoch": 0.01,
"grad_norm": 3.198234796524048,
"learning_rate": 1.76e-05,
"loss": 1.465,
"step": 440
},
{
"epoch": 0.01,
"grad_norm": 5.132129669189453,
"learning_rate": 1.8e-05,
"loss": 1.6176,
"step": 450
},
{
"epoch": 0.01,
"grad_norm": 5.825719833374023,
"learning_rate": 1.8400000000000003e-05,
"loss": 1.6984,
"step": 460
},
{
"epoch": 0.01,
"grad_norm": 6.724545478820801,
"learning_rate": 1.88e-05,
"loss": 1.7011,
"step": 470
},
{
"epoch": 0.01,
"grad_norm": 3.723080635070801,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.6338,
"step": 480
},
{
"epoch": 0.01,
"grad_norm": 2.8718888759613037,
"learning_rate": 1.9600000000000002e-05,
"loss": 1.5002,
"step": 490
},
{
"epoch": 0.01,
"grad_norm": 11.207817077636719,
"learning_rate": 2e-05,
"loss": 1.6829,
"step": 500
},
{
"epoch": 0.01,
"eval_loss": 1.6066228151321411,
"eval_runtime": 66.1997,
"eval_samples_per_second": 15.106,
"eval_steps_per_second": 15.106,
"step": 500
},
{
"epoch": 0.01,
"grad_norm": 5.459104537963867,
"learning_rate": 1.999322033898305e-05,
"loss": 1.6216,
"step": 510
},
{
"epoch": 0.01,
"grad_norm": 2.6112444400787354,
"learning_rate": 1.9986440677966104e-05,
"loss": 1.5761,
"step": 520
},
{
"epoch": 0.01,
"grad_norm": 4.285440921783447,
"learning_rate": 1.9979661016949154e-05,
"loss": 1.6097,
"step": 530
},
{
"epoch": 0.01,
"grad_norm": 2.781306505203247,
"learning_rate": 1.9972881355932204e-05,
"loss": 1.6967,
"step": 540
},
{
"epoch": 0.01,
"grad_norm": 7.55482292175293,
"learning_rate": 1.9966101694915257e-05,
"loss": 1.5974,
"step": 550
},
{
"epoch": 0.01,
"grad_norm": 5.083601474761963,
"learning_rate": 1.9959322033898307e-05,
"loss": 1.6287,
"step": 560
},
{
"epoch": 0.01,
"grad_norm": 3.7552945613861084,
"learning_rate": 1.995254237288136e-05,
"loss": 1.4225,
"step": 570
},
{
"epoch": 0.01,
"grad_norm": 3.149674892425537,
"learning_rate": 1.994576271186441e-05,
"loss": 1.5439,
"step": 580
},
{
"epoch": 0.01,
"grad_norm": 5.270935535430908,
"learning_rate": 1.993898305084746e-05,
"loss": 1.8439,
"step": 590
},
{
"epoch": 0.01,
"grad_norm": 2.552823543548584,
"learning_rate": 1.9932203389830512e-05,
"loss": 1.5678,
"step": 600
},
{
"epoch": 0.02,
"grad_norm": 2.6956074237823486,
"learning_rate": 1.992542372881356e-05,
"loss": 1.6002,
"step": 610
},
{
"epoch": 0.02,
"grad_norm": 3.0761303901672363,
"learning_rate": 1.991864406779661e-05,
"loss": 1.7211,
"step": 620
},
{
"epoch": 0.02,
"grad_norm": 3.1744227409362793,
"learning_rate": 1.991186440677966e-05,
"loss": 1.6327,
"step": 630
},
{
"epoch": 0.02,
"grad_norm": 4.774084091186523,
"learning_rate": 1.990508474576271e-05,
"loss": 1.6871,
"step": 640
},
{
"epoch": 0.02,
"grad_norm": 3.4767563343048096,
"learning_rate": 1.9898305084745764e-05,
"loss": 1.6038,
"step": 650
},
{
"epoch": 0.02,
"grad_norm": 2.3587417602539062,
"learning_rate": 1.9891525423728814e-05,
"loss": 1.8443,
"step": 660
},
{
"epoch": 0.02,
"grad_norm": 3.177602529525757,
"learning_rate": 1.9884745762711867e-05,
"loss": 1.5371,
"step": 670
},
{
"epoch": 0.02,
"grad_norm": 2.579742193222046,
"learning_rate": 1.9877966101694917e-05,
"loss": 1.4555,
"step": 680
},
{
"epoch": 0.02,
"grad_norm": 3.631253242492676,
"learning_rate": 1.9871186440677966e-05,
"loss": 1.6794,
"step": 690
},
{
"epoch": 0.02,
"grad_norm": 2.278071641921997,
"learning_rate": 1.986440677966102e-05,
"loss": 1.477,
"step": 700
},
{
"epoch": 0.02,
"grad_norm": 3.924114227294922,
"learning_rate": 1.985762711864407e-05,
"loss": 1.6933,
"step": 710
},
{
"epoch": 0.02,
"grad_norm": 3.5087924003601074,
"learning_rate": 1.985084745762712e-05,
"loss": 1.5713,
"step": 720
},
{
"epoch": 0.02,
"grad_norm": 2.0184319019317627,
"learning_rate": 1.9844067796610172e-05,
"loss": 1.6689,
"step": 730
},
{
"epoch": 0.02,
"grad_norm": 2.6397581100463867,
"learning_rate": 1.9837288135593222e-05,
"loss": 1.5636,
"step": 740
},
{
"epoch": 0.02,
"grad_norm": 3.12296986579895,
"learning_rate": 1.9830508474576275e-05,
"loss": 1.6867,
"step": 750
},
{
"epoch": 0.02,
"grad_norm": 2.04766583442688,
"learning_rate": 1.9823728813559324e-05,
"loss": 1.6027,
"step": 760
},
{
"epoch": 0.02,
"grad_norm": 1.8128042221069336,
"learning_rate": 1.9816949152542374e-05,
"loss": 1.6354,
"step": 770
},
{
"epoch": 0.02,
"grad_norm": 2.708834171295166,
"learning_rate": 1.9810169491525427e-05,
"loss": 1.5484,
"step": 780
},
{
"epoch": 0.02,
"grad_norm": 1.9796117544174194,
"learning_rate": 1.9803389830508477e-05,
"loss": 1.4511,
"step": 790
},
{
"epoch": 0.02,
"grad_norm": 4.195634365081787,
"learning_rate": 1.9796610169491527e-05,
"loss": 1.6435,
"step": 800
},
{
"epoch": 0.02,
"grad_norm": 3.132399320602417,
"learning_rate": 1.978983050847458e-05,
"loss": 1.5196,
"step": 810
},
{
"epoch": 0.02,
"grad_norm": 2.9280669689178467,
"learning_rate": 1.9783050847457626e-05,
"loss": 1.522,
"step": 820
},
{
"epoch": 0.02,
"grad_norm": 2.2013494968414307,
"learning_rate": 1.977627118644068e-05,
"loss": 1.6589,
"step": 830
},
{
"epoch": 0.02,
"grad_norm": 9.277434349060059,
"learning_rate": 1.976949152542373e-05,
"loss": 1.5828,
"step": 840
},
{
"epoch": 0.02,
"grad_norm": 6.09869384765625,
"learning_rate": 1.9762711864406782e-05,
"loss": 1.6683,
"step": 850
},
{
"epoch": 0.02,
"grad_norm": 1.7093573808670044,
"learning_rate": 1.9755932203389832e-05,
"loss": 1.6881,
"step": 860
},
{
"epoch": 0.02,
"grad_norm": 1.834920048713684,
"learning_rate": 1.974915254237288e-05,
"loss": 1.6169,
"step": 870
},
{
"epoch": 0.02,
"grad_norm": 3.802130699157715,
"learning_rate": 1.9742372881355935e-05,
"loss": 1.5526,
"step": 880
},
{
"epoch": 0.02,
"grad_norm": 2.086676836013794,
"learning_rate": 1.9735593220338984e-05,
"loss": 1.661,
"step": 890
},
{
"epoch": 0.02,
"grad_norm": 1.9825890064239502,
"learning_rate": 1.9728813559322034e-05,
"loss": 1.6302,
"step": 900
},
{
"epoch": 0.02,
"grad_norm": 1.9660803079605103,
"learning_rate": 1.9722033898305087e-05,
"loss": 1.6146,
"step": 910
},
{
"epoch": 0.02,
"grad_norm": 3.5572495460510254,
"learning_rate": 1.9715254237288137e-05,
"loss": 1.6143,
"step": 920
},
{
"epoch": 0.02,
"grad_norm": 5.440743923187256,
"learning_rate": 1.970847457627119e-05,
"loss": 1.4907,
"step": 930
},
{
"epoch": 0.02,
"grad_norm": 2.432176351547241,
"learning_rate": 1.970169491525424e-05,
"loss": 1.6272,
"step": 940
},
{
"epoch": 0.02,
"grad_norm": 1.2422198057174683,
"learning_rate": 1.969491525423729e-05,
"loss": 1.6404,
"step": 950
},
{
"epoch": 0.02,
"grad_norm": 3.0598487854003906,
"learning_rate": 1.9688135593220342e-05,
"loss": 1.5811,
"step": 960
},
{
"epoch": 0.02,
"grad_norm": 3.399864673614502,
"learning_rate": 1.9681355932203392e-05,
"loss": 1.6221,
"step": 970
},
{
"epoch": 0.02,
"grad_norm": 1.3376494646072388,
"learning_rate": 1.9674576271186442e-05,
"loss": 1.5384,
"step": 980
},
{
"epoch": 0.02,
"grad_norm": 4.062708854675293,
"learning_rate": 1.9667796610169495e-05,
"loss": 1.5448,
"step": 990
},
{
"epoch": 0.03,
"grad_norm": 4.200625419616699,
"learning_rate": 1.9661016949152545e-05,
"loss": 1.6656,
"step": 1000
},
{
"epoch": 0.03,
"eval_loss": 1.5288029909133911,
"eval_runtime": 66.2062,
"eval_samples_per_second": 15.104,
"eval_steps_per_second": 15.104,
"step": 1000
},
{
"epoch": 0.03,
"grad_norm": 4.217132091522217,
"learning_rate": 1.9654237288135594e-05,
"loss": 1.6834,
"step": 1010
},
{
"epoch": 0.03,
"grad_norm": 1.8246206045150757,
"learning_rate": 1.9647457627118644e-05,
"loss": 1.7242,
"step": 1020
},
{
"epoch": 0.03,
"grad_norm": 3.0907254219055176,
"learning_rate": 1.9640677966101697e-05,
"loss": 1.6493,
"step": 1030
},
{
"epoch": 0.03,
"grad_norm": 2.1235597133636475,
"learning_rate": 1.9633898305084747e-05,
"loss": 1.6628,
"step": 1040
},
{
"epoch": 0.03,
"grad_norm": 4.815682411193848,
"learning_rate": 1.9627118644067796e-05,
"loss": 1.5552,
"step": 1050
},
{
"epoch": 0.03,
"grad_norm": 2.7466652393341064,
"learning_rate": 1.962033898305085e-05,
"loss": 1.7091,
"step": 1060
},
{
"epoch": 0.03,
"grad_norm": 1.9031518697738647,
"learning_rate": 1.96135593220339e-05,
"loss": 1.4593,
"step": 1070
},
{
"epoch": 0.03,
"grad_norm": 7.788964748382568,
"learning_rate": 1.960677966101695e-05,
"loss": 1.7782,
"step": 1080
},
{
"epoch": 0.03,
"grad_norm": 3.2768681049346924,
"learning_rate": 1.9600000000000002e-05,
"loss": 1.6339,
"step": 1090
},
{
"epoch": 0.03,
"grad_norm": 2.7264978885650635,
"learning_rate": 1.9593220338983052e-05,
"loss": 1.7666,
"step": 1100
},
{
"epoch": 0.03,
"grad_norm": 4.865339279174805,
"learning_rate": 1.95864406779661e-05,
"loss": 1.7127,
"step": 1110
},
{
"epoch": 0.03,
"grad_norm": 2.9894134998321533,
"learning_rate": 1.9579661016949155e-05,
"loss": 1.5671,
"step": 1120
},
{
"epoch": 0.03,
"grad_norm": 3.2608160972595215,
"learning_rate": 1.9572881355932204e-05,
"loss": 1.5396,
"step": 1130
},
{
"epoch": 0.03,
"grad_norm": 2.3654749393463135,
"learning_rate": 1.9566101694915257e-05,
"loss": 1.514,
"step": 1140
},
{
"epoch": 0.03,
"grad_norm": 1.9910569190979004,
"learning_rate": 1.9559322033898307e-05,
"loss": 1.6519,
"step": 1150
},
{
"epoch": 0.03,
"grad_norm": 3.1228861808776855,
"learning_rate": 1.9552542372881357e-05,
"loss": 1.5683,
"step": 1160
},
{
"epoch": 0.03,
"grad_norm": 2.8690907955169678,
"learning_rate": 1.954576271186441e-05,
"loss": 1.5795,
"step": 1170
},
{
"epoch": 0.03,
"grad_norm": 7.847595691680908,
"learning_rate": 1.953898305084746e-05,
"loss": 1.4718,
"step": 1180
},
{
"epoch": 0.03,
"grad_norm": 3.4904606342315674,
"learning_rate": 1.953220338983051e-05,
"loss": 1.6105,
"step": 1190
},
{
"epoch": 0.03,
"grad_norm": 2.352919340133667,
"learning_rate": 1.9525423728813562e-05,
"loss": 1.6323,
"step": 1200
},
{
"epoch": 0.03,
"grad_norm": 1.9066511392593384,
"learning_rate": 1.9518644067796612e-05,
"loss": 1.5846,
"step": 1210
},
{
"epoch": 0.03,
"grad_norm": 2.2066802978515625,
"learning_rate": 1.9511864406779665e-05,
"loss": 1.4947,
"step": 1220
},
{
"epoch": 0.03,
"grad_norm": 2.038661003112793,
"learning_rate": 1.950508474576271e-05,
"loss": 1.5167,
"step": 1230
},
{
"epoch": 0.03,
"grad_norm": 4.544153690338135,
"learning_rate": 1.9498305084745765e-05,
"loss": 1.7429,
"step": 1240
},
{
"epoch": 0.03,
"grad_norm": 4.264445781707764,
"learning_rate": 1.9491525423728814e-05,
"loss": 1.7164,
"step": 1250
},
{
"epoch": 0.03,
"grad_norm": 7.955377101898193,
"learning_rate": 1.9484745762711864e-05,
"loss": 1.4268,
"step": 1260
},
{
"epoch": 0.03,
"grad_norm": 3.6983351707458496,
"learning_rate": 1.9477966101694917e-05,
"loss": 1.5497,
"step": 1270
},
{
"epoch": 0.03,
"grad_norm": 2.3593084812164307,
"learning_rate": 1.9471186440677967e-05,
"loss": 1.4209,
"step": 1280
},
{
"epoch": 0.03,
"grad_norm": 5.2454938888549805,
"learning_rate": 1.9464406779661017e-05,
"loss": 1.7169,
"step": 1290
},
{
"epoch": 0.03,
"grad_norm": 4.33056116104126,
"learning_rate": 1.945762711864407e-05,
"loss": 1.491,
"step": 1300
},
{
"epoch": 0.03,
"grad_norm": 3.095102071762085,
"learning_rate": 1.945084745762712e-05,
"loss": 1.5366,
"step": 1310
},
{
"epoch": 0.03,
"grad_norm": 1.7006287574768066,
"learning_rate": 1.9444067796610172e-05,
"loss": 1.5085,
"step": 1320
},
{
"epoch": 0.03,
"grad_norm": 2.643651008605957,
"learning_rate": 1.9437288135593222e-05,
"loss": 1.4597,
"step": 1330
},
{
"epoch": 0.03,
"grad_norm": 1.9585516452789307,
"learning_rate": 1.9430508474576272e-05,
"loss": 1.6975,
"step": 1340
},
{
"epoch": 0.03,
"grad_norm": 2.669740676879883,
"learning_rate": 1.9423728813559325e-05,
"loss": 1.4074,
"step": 1350
},
{
"epoch": 0.03,
"grad_norm": 3.0379438400268555,
"learning_rate": 1.9416949152542375e-05,
"loss": 1.5256,
"step": 1360
},
{
"epoch": 0.03,
"grad_norm": 2.81781005859375,
"learning_rate": 1.9410169491525424e-05,
"loss": 1.5329,
"step": 1370
},
{
"epoch": 0.03,
"grad_norm": 1.2218108177185059,
"learning_rate": 1.9403389830508477e-05,
"loss": 1.5283,
"step": 1380
},
{
"epoch": 0.03,
"grad_norm": 2.4585816860198975,
"learning_rate": 1.9396610169491527e-05,
"loss": 1.4265,
"step": 1390
},
{
"epoch": 0.04,
"grad_norm": 6.117985725402832,
"learning_rate": 1.938983050847458e-05,
"loss": 1.6968,
"step": 1400
},
{
"epoch": 0.04,
"grad_norm": 1.7194918394088745,
"learning_rate": 1.938305084745763e-05,
"loss": 1.5565,
"step": 1410
},
{
"epoch": 0.04,
"grad_norm": 3.65537691116333,
"learning_rate": 1.937627118644068e-05,
"loss": 1.7526,
"step": 1420
},
{
"epoch": 0.04,
"grad_norm": 3.0874905586242676,
"learning_rate": 1.9369491525423733e-05,
"loss": 1.4787,
"step": 1430
},
{
"epoch": 0.04,
"grad_norm": 3.945019006729126,
"learning_rate": 1.936271186440678e-05,
"loss": 1.5238,
"step": 1440
},
{
"epoch": 0.04,
"grad_norm": 2.7200796604156494,
"learning_rate": 1.9355932203389832e-05,
"loss": 1.4571,
"step": 1450
},
{
"epoch": 0.04,
"grad_norm": 8.368704795837402,
"learning_rate": 1.9349152542372882e-05,
"loss": 1.5916,
"step": 1460
},
{
"epoch": 0.04,
"grad_norm": 3.0204453468322754,
"learning_rate": 1.934237288135593e-05,
"loss": 1.6997,
"step": 1470
},
{
"epoch": 0.04,
"grad_norm": 2.3051364421844482,
"learning_rate": 1.9335593220338985e-05,
"loss": 1.5315,
"step": 1480
},
{
"epoch": 0.04,
"grad_norm": 1.246631145477295,
"learning_rate": 1.9328813559322034e-05,
"loss": 1.478,
"step": 1490
},
{
"epoch": 0.04,
"grad_norm": 1.8742200136184692,
"learning_rate": 1.9322033898305087e-05,
"loss": 1.5419,
"step": 1500
},
{
"epoch": 0.04,
"eval_loss": 1.5387661457061768,
"eval_runtime": 66.1849,
"eval_samples_per_second": 15.109,
"eval_steps_per_second": 15.109,
"step": 1500
},
{
"epoch": 0.04,
"grad_norm": 1.624708652496338,
"learning_rate": 1.9315254237288137e-05,
"loss": 1.5002,
"step": 1510
},
{
"epoch": 0.04,
"grad_norm": 5.5269856452941895,
"learning_rate": 1.9308474576271187e-05,
"loss": 1.6267,
"step": 1520
},
{
"epoch": 0.04,
"grad_norm": 1.879159688949585,
"learning_rate": 1.930169491525424e-05,
"loss": 1.4951,
"step": 1530
},
{
"epoch": 0.04,
"grad_norm": 4.727252006530762,
"learning_rate": 1.929491525423729e-05,
"loss": 1.5438,
"step": 1540
},
{
"epoch": 0.04,
"grad_norm": 1.9965201616287231,
"learning_rate": 1.928813559322034e-05,
"loss": 1.6525,
"step": 1550
},
{
"epoch": 0.04,
"grad_norm": 2.61816143989563,
"learning_rate": 1.9281355932203392e-05,
"loss": 1.4746,
"step": 1560
},
{
"epoch": 0.04,
"grad_norm": 2.8157131671905518,
"learning_rate": 1.9274576271186442e-05,
"loss": 1.5454,
"step": 1570
},
{
"epoch": 0.04,
"grad_norm": 2.738640546798706,
"learning_rate": 1.9267796610169492e-05,
"loss": 1.4876,
"step": 1580
},
{
"epoch": 0.04,
"grad_norm": 3.118277072906494,
"learning_rate": 1.9261016949152545e-05,
"loss": 1.5587,
"step": 1590
},
{
"epoch": 0.04,
"grad_norm": 6.730972766876221,
"learning_rate": 1.9254237288135595e-05,
"loss": 1.6752,
"step": 1600
},
{
"epoch": 0.04,
"grad_norm": 9.461782455444336,
"learning_rate": 1.9247457627118648e-05,
"loss": 1.5409,
"step": 1610
},
{
"epoch": 0.04,
"grad_norm": 4.566100120544434,
"learning_rate": 1.9240677966101698e-05,
"loss": 1.4724,
"step": 1620
},
{
"epoch": 0.04,
"grad_norm": 3.8972136974334717,
"learning_rate": 1.9233898305084747e-05,
"loss": 1.5706,
"step": 1630
},
{
"epoch": 0.04,
"grad_norm": 2.0403244495391846,
"learning_rate": 1.92271186440678e-05,
"loss": 1.6694,
"step": 1640
},
{
"epoch": 0.04,
"grad_norm": 3.0590033531188965,
"learning_rate": 1.9220338983050847e-05,
"loss": 1.5711,
"step": 1650
},
{
"epoch": 0.04,
"grad_norm": 0.9862588047981262,
"learning_rate": 1.92135593220339e-05,
"loss": 1.486,
"step": 1660
},
{
"epoch": 0.04,
"grad_norm": 2.9184350967407227,
"learning_rate": 1.920677966101695e-05,
"loss": 1.5526,
"step": 1670
},
{
"epoch": 0.04,
"grad_norm": 3.7024717330932617,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.6754,
"step": 1680
},
{
"epoch": 0.04,
"grad_norm": 4.887876510620117,
"learning_rate": 1.9193220338983052e-05,
"loss": 1.6843,
"step": 1690
},
{
"epoch": 0.04,
"grad_norm": 2.6079001426696777,
"learning_rate": 1.9186440677966102e-05,
"loss": 1.5116,
"step": 1700
},
{
"epoch": 0.04,
"grad_norm": 3.407667875289917,
"learning_rate": 1.9179661016949155e-05,
"loss": 1.581,
"step": 1710
},
{
"epoch": 0.04,
"grad_norm": 2.201183795928955,
"learning_rate": 1.9172881355932205e-05,
"loss": 1.6449,
"step": 1720
},
{
"epoch": 0.04,
"grad_norm": 1.4738363027572632,
"learning_rate": 1.9166101694915254e-05,
"loss": 1.3959,
"step": 1730
},
{
"epoch": 0.04,
"grad_norm": 1.745976209640503,
"learning_rate": 1.9159322033898308e-05,
"loss": 1.394,
"step": 1740
},
{
"epoch": 0.04,
"grad_norm": 3.694169759750366,
"learning_rate": 1.9152542372881357e-05,
"loss": 1.6542,
"step": 1750
},
{
"epoch": 0.04,
"grad_norm": 2.3387959003448486,
"learning_rate": 1.9145762711864407e-05,
"loss": 1.6928,
"step": 1760
},
{
"epoch": 0.04,
"grad_norm": 2.8310043811798096,
"learning_rate": 1.913898305084746e-05,
"loss": 1.5955,
"step": 1770
},
{
"epoch": 0.04,
"grad_norm": 2.731762647628784,
"learning_rate": 1.913220338983051e-05,
"loss": 1.7997,
"step": 1780
},
{
"epoch": 0.04,
"grad_norm": 2.896686553955078,
"learning_rate": 1.9125423728813563e-05,
"loss": 1.5137,
"step": 1790
},
{
"epoch": 0.04,
"grad_norm": 1.8352593183517456,
"learning_rate": 1.9118644067796613e-05,
"loss": 1.5224,
"step": 1800
},
{
"epoch": 0.05,
"grad_norm": 4.474410057067871,
"learning_rate": 1.9111864406779662e-05,
"loss": 1.6659,
"step": 1810
},
{
"epoch": 0.05,
"grad_norm": 3.063828468322754,
"learning_rate": 1.9105084745762715e-05,
"loss": 1.4191,
"step": 1820
},
{
"epoch": 0.05,
"grad_norm": 1.9058163166046143,
"learning_rate": 1.9098305084745765e-05,
"loss": 1.5386,
"step": 1830
},
{
"epoch": 0.05,
"grad_norm": 2.915477991104126,
"learning_rate": 1.9091525423728815e-05,
"loss": 1.5756,
"step": 1840
},
{
"epoch": 0.05,
"grad_norm": 1.423071265220642,
"learning_rate": 1.9084745762711868e-05,
"loss": 1.5377,
"step": 1850
},
{
"epoch": 0.05,
"grad_norm": 2.5523242950439453,
"learning_rate": 1.9077966101694914e-05,
"loss": 1.5199,
"step": 1860
},
{
"epoch": 0.05,
"grad_norm": 2.609219789505005,
"learning_rate": 1.9071186440677967e-05,
"loss": 1.5402,
"step": 1870
},
{
"epoch": 0.05,
"grad_norm": 7.032884120941162,
"learning_rate": 1.9064406779661017e-05,
"loss": 1.4348,
"step": 1880
},
{
"epoch": 0.05,
"grad_norm": 3.192718029022217,
"learning_rate": 1.905762711864407e-05,
"loss": 1.5971,
"step": 1890
},
{
"epoch": 0.05,
"grad_norm": 2.0084311962127686,
"learning_rate": 1.905084745762712e-05,
"loss": 1.459,
"step": 1900
},
{
"epoch": 0.05,
"grad_norm": 1.646857738494873,
"learning_rate": 1.904406779661017e-05,
"loss": 1.507,
"step": 1910
},
{
"epoch": 0.05,
"grad_norm": 3.855058431625366,
"learning_rate": 1.9037288135593223e-05,
"loss": 1.5667,
"step": 1920
},
{
"epoch": 0.05,
"grad_norm": 1.520700216293335,
"learning_rate": 1.9030508474576272e-05,
"loss": 1.573,
"step": 1930
},
{
"epoch": 0.05,
"grad_norm": 2.6929845809936523,
"learning_rate": 1.9023728813559322e-05,
"loss": 1.4966,
"step": 1940
},
{
"epoch": 0.05,
"grad_norm": 2.283622980117798,
"learning_rate": 1.9016949152542375e-05,
"loss": 1.5018,
"step": 1950
},
{
"epoch": 0.05,
"grad_norm": 1.3686760663986206,
"learning_rate": 1.9010169491525425e-05,
"loss": 1.3841,
"step": 1960
},
{
"epoch": 0.05,
"grad_norm": 2.02280592918396,
"learning_rate": 1.9003389830508478e-05,
"loss": 1.3902,
"step": 1970
},
{
"epoch": 0.05,
"grad_norm": 2.312986135482788,
"learning_rate": 1.8996610169491528e-05,
"loss": 1.6016,
"step": 1980
},
{
"epoch": 0.05,
"grad_norm": 2.497074604034424,
"learning_rate": 1.8989830508474577e-05,
"loss": 1.6256,
"step": 1990
},
{
"epoch": 0.05,
"grad_norm": 8.351258277893066,
"learning_rate": 1.898305084745763e-05,
"loss": 1.5034,
"step": 2000
},
{
"epoch": 0.05,
"eval_loss": 1.555788516998291,
"eval_runtime": 66.1337,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 2000
},
{
"epoch": 0.05,
"grad_norm": 10.036495208740234,
"learning_rate": 1.897627118644068e-05,
"loss": 1.6968,
"step": 2010
},
{
"epoch": 0.05,
"grad_norm": 5.859747886657715,
"learning_rate": 1.896949152542373e-05,
"loss": 1.655,
"step": 2020
},
{
"epoch": 0.05,
"grad_norm": 2.809723138809204,
"learning_rate": 1.8962711864406783e-05,
"loss": 1.5661,
"step": 2030
},
{
"epoch": 0.05,
"grad_norm": 3.2033305168151855,
"learning_rate": 1.8955932203389833e-05,
"loss": 1.5528,
"step": 2040
},
{
"epoch": 0.05,
"grad_norm": 2.821791887283325,
"learning_rate": 1.8949152542372882e-05,
"loss": 1.5215,
"step": 2050
},
{
"epoch": 0.05,
"grad_norm": 3.9140360355377197,
"learning_rate": 1.8942372881355932e-05,
"loss": 1.6541,
"step": 2060
},
{
"epoch": 0.05,
"grad_norm": 1.9218617677688599,
"learning_rate": 1.8935593220338985e-05,
"loss": 1.6119,
"step": 2070
},
{
"epoch": 0.05,
"grad_norm": 1.7994463443756104,
"learning_rate": 1.8928813559322035e-05,
"loss": 1.5222,
"step": 2080
},
{
"epoch": 0.05,
"grad_norm": 2.291020631790161,
"learning_rate": 1.8922033898305085e-05,
"loss": 1.5315,
"step": 2090
},
{
"epoch": 0.05,
"grad_norm": 2.737337589263916,
"learning_rate": 1.8915254237288138e-05,
"loss": 1.4785,
"step": 2100
},
{
"epoch": 0.05,
"grad_norm": 3.1229870319366455,
"learning_rate": 1.8908474576271187e-05,
"loss": 1.6181,
"step": 2110
},
{
"epoch": 0.05,
"grad_norm": 2.7481417655944824,
"learning_rate": 1.8901694915254237e-05,
"loss": 1.4123,
"step": 2120
},
{
"epoch": 0.05,
"grad_norm": 3.3576602935791016,
"learning_rate": 1.889491525423729e-05,
"loss": 1.5839,
"step": 2130
},
{
"epoch": 0.05,
"grad_norm": 3.292433261871338,
"learning_rate": 1.888813559322034e-05,
"loss": 1.5668,
"step": 2140
},
{
"epoch": 0.05,
"grad_norm": 5.039631366729736,
"learning_rate": 1.8881355932203393e-05,
"loss": 1.3932,
"step": 2150
},
{
"epoch": 0.05,
"grad_norm": 2.142812490463257,
"learning_rate": 1.8874576271186443e-05,
"loss": 1.6613,
"step": 2160
},
{
"epoch": 0.05,
"grad_norm": 1.2970080375671387,
"learning_rate": 1.8867796610169492e-05,
"loss": 1.4982,
"step": 2170
},
{
"epoch": 0.05,
"grad_norm": 7.887810230255127,
"learning_rate": 1.8861016949152545e-05,
"loss": 1.6085,
"step": 2180
},
{
"epoch": 0.05,
"grad_norm": 3.8233675956726074,
"learning_rate": 1.8854237288135595e-05,
"loss": 1.668,
"step": 2190
},
{
"epoch": 0.06,
"grad_norm": 2.357754707336426,
"learning_rate": 1.8847457627118645e-05,
"loss": 1.6456,
"step": 2200
},
{
"epoch": 0.06,
"grad_norm": 5.471128463745117,
"learning_rate": 1.8840677966101698e-05,
"loss": 1.5198,
"step": 2210
},
{
"epoch": 0.06,
"grad_norm": 1.850195050239563,
"learning_rate": 1.8833898305084748e-05,
"loss": 1.5557,
"step": 2220
},
{
"epoch": 0.06,
"grad_norm": 2.1556105613708496,
"learning_rate": 1.8827118644067797e-05,
"loss": 1.6411,
"step": 2230
},
{
"epoch": 0.06,
"grad_norm": 3.6570167541503906,
"learning_rate": 1.882033898305085e-05,
"loss": 1.474,
"step": 2240
},
{
"epoch": 0.06,
"grad_norm": 2.989947557449341,
"learning_rate": 1.88135593220339e-05,
"loss": 1.5413,
"step": 2250
},
{
"epoch": 0.06,
"grad_norm": 1.8367825746536255,
"learning_rate": 1.8806779661016953e-05,
"loss": 1.5744,
"step": 2260
},
{
"epoch": 0.06,
"grad_norm": 2.2120065689086914,
"learning_rate": 1.88e-05,
"loss": 1.5589,
"step": 2270
},
{
"epoch": 0.06,
"grad_norm": 3.29278302192688,
"learning_rate": 1.8793220338983053e-05,
"loss": 1.4468,
"step": 2280
},
{
"epoch": 0.06,
"grad_norm": 2.6164047718048096,
"learning_rate": 1.8786440677966102e-05,
"loss": 1.5143,
"step": 2290
},
{
"epoch": 0.06,
"grad_norm": 3.008776903152466,
"learning_rate": 1.8779661016949152e-05,
"loss": 1.5692,
"step": 2300
},
{
"epoch": 0.06,
"grad_norm": 2.0924556255340576,
"learning_rate": 1.8772881355932205e-05,
"loss": 1.4559,
"step": 2310
},
{
"epoch": 0.06,
"grad_norm": 2.1632914543151855,
"learning_rate": 1.8766101694915255e-05,
"loss": 1.5876,
"step": 2320
},
{
"epoch": 0.06,
"grad_norm": 2.367845058441162,
"learning_rate": 1.8759322033898305e-05,
"loss": 1.5093,
"step": 2330
},
{
"epoch": 0.06,
"grad_norm": 2.618950843811035,
"learning_rate": 1.8752542372881358e-05,
"loss": 1.4839,
"step": 2340
},
{
"epoch": 0.06,
"grad_norm": 2.7340502738952637,
"learning_rate": 1.8745762711864407e-05,
"loss": 1.7354,
"step": 2350
},
{
"epoch": 0.06,
"grad_norm": 3.4295051097869873,
"learning_rate": 1.873898305084746e-05,
"loss": 1.4937,
"step": 2360
},
{
"epoch": 0.06,
"grad_norm": 1.078732967376709,
"learning_rate": 1.873220338983051e-05,
"loss": 1.7125,
"step": 2370
},
{
"epoch": 0.06,
"grad_norm": 3.103182554244995,
"learning_rate": 1.872542372881356e-05,
"loss": 1.4019,
"step": 2380
},
{
"epoch": 0.06,
"grad_norm": 1.760866641998291,
"learning_rate": 1.8718644067796613e-05,
"loss": 1.4383,
"step": 2390
},
{
"epoch": 0.06,
"grad_norm": 1.4431618452072144,
"learning_rate": 1.8711864406779663e-05,
"loss": 1.5302,
"step": 2400
},
{
"epoch": 0.06,
"grad_norm": 3.5087249279022217,
"learning_rate": 1.8705084745762712e-05,
"loss": 1.5693,
"step": 2410
},
{
"epoch": 0.06,
"grad_norm": 3.648737668991089,
"learning_rate": 1.8698305084745765e-05,
"loss": 1.2027,
"step": 2420
},
{
"epoch": 0.06,
"grad_norm": 4.0298357009887695,
"learning_rate": 1.8691525423728815e-05,
"loss": 1.4772,
"step": 2430
},
{
"epoch": 0.06,
"grad_norm": 4.11863899230957,
"learning_rate": 1.8684745762711868e-05,
"loss": 1.402,
"step": 2440
},
{
"epoch": 0.06,
"grad_norm": 1.1432223320007324,
"learning_rate": 1.8677966101694918e-05,
"loss": 1.5907,
"step": 2450
},
{
"epoch": 0.06,
"grad_norm": 2.0923776626586914,
"learning_rate": 1.8671186440677968e-05,
"loss": 1.569,
"step": 2460
},
{
"epoch": 0.06,
"grad_norm": 1.987396001815796,
"learning_rate": 1.866440677966102e-05,
"loss": 1.6504,
"step": 2470
},
{
"epoch": 0.06,
"grad_norm": 3.959685802459717,
"learning_rate": 1.8657627118644067e-05,
"loss": 1.4029,
"step": 2480
},
{
"epoch": 0.06,
"grad_norm": 3.1348440647125244,
"learning_rate": 1.865084745762712e-05,
"loss": 1.5063,
"step": 2490
},
{
"epoch": 0.06,
"grad_norm": 2.788222074508667,
"learning_rate": 1.864406779661017e-05,
"loss": 1.5971,
"step": 2500
},
{
"epoch": 0.06,
"eval_loss": 1.5344786643981934,
"eval_runtime": 66.1387,
"eval_samples_per_second": 15.12,
"eval_steps_per_second": 15.12,
"step": 2500
},
{
"epoch": 0.06,
"grad_norm": 1.9007914066314697,
"learning_rate": 1.863728813559322e-05,
"loss": 1.5292,
"step": 2510
},
{
"epoch": 0.06,
"grad_norm": 2.1382060050964355,
"learning_rate": 1.8630508474576273e-05,
"loss": 1.516,
"step": 2520
},
{
"epoch": 0.06,
"grad_norm": 0.914665162563324,
"learning_rate": 1.8623728813559322e-05,
"loss": 1.3994,
"step": 2530
},
{
"epoch": 0.06,
"grad_norm": 4.413509368896484,
"learning_rate": 1.8616949152542376e-05,
"loss": 1.5696,
"step": 2540
},
{
"epoch": 0.06,
"grad_norm": 1.9687520265579224,
"learning_rate": 1.8610169491525425e-05,
"loss": 1.6693,
"step": 2550
},
{
"epoch": 0.06,
"grad_norm": 3.4700605869293213,
"learning_rate": 1.8603389830508475e-05,
"loss": 1.6217,
"step": 2560
},
{
"epoch": 0.06,
"grad_norm": 2.2257847785949707,
"learning_rate": 1.8596610169491528e-05,
"loss": 1.65,
"step": 2570
},
{
"epoch": 0.06,
"grad_norm": 15.683956146240234,
"learning_rate": 1.8589830508474578e-05,
"loss": 1.5663,
"step": 2580
},
{
"epoch": 0.06,
"grad_norm": 2.2584922313690186,
"learning_rate": 1.8583050847457627e-05,
"loss": 1.5146,
"step": 2590
},
{
"epoch": 0.07,
"grad_norm": 1.900083065032959,
"learning_rate": 1.857627118644068e-05,
"loss": 1.5062,
"step": 2600
},
{
"epoch": 0.07,
"grad_norm": 2.6166887283325195,
"learning_rate": 1.856949152542373e-05,
"loss": 1.3145,
"step": 2610
},
{
"epoch": 0.07,
"grad_norm": 0.9894976019859314,
"learning_rate": 1.856271186440678e-05,
"loss": 1.6037,
"step": 2620
},
{
"epoch": 0.07,
"grad_norm": 5.897536277770996,
"learning_rate": 1.8555932203389833e-05,
"loss": 1.5846,
"step": 2630
},
{
"epoch": 0.07,
"grad_norm": 3.8714261054992676,
"learning_rate": 1.8549152542372883e-05,
"loss": 1.5008,
"step": 2640
},
{
"epoch": 0.07,
"grad_norm": 3.062786817550659,
"learning_rate": 1.8542372881355936e-05,
"loss": 1.5885,
"step": 2650
},
{
"epoch": 0.07,
"grad_norm": 4.548590660095215,
"learning_rate": 1.8535593220338986e-05,
"loss": 1.4718,
"step": 2660
},
{
"epoch": 0.07,
"grad_norm": 1.9573743343353271,
"learning_rate": 1.8528813559322035e-05,
"loss": 1.5926,
"step": 2670
},
{
"epoch": 0.07,
"grad_norm": 1.7771761417388916,
"learning_rate": 1.852203389830509e-05,
"loss": 1.5136,
"step": 2680
},
{
"epoch": 0.07,
"grad_norm": 3.312812328338623,
"learning_rate": 1.8515254237288135e-05,
"loss": 1.4674,
"step": 2690
},
{
"epoch": 0.07,
"grad_norm": 2.288461685180664,
"learning_rate": 1.8508474576271188e-05,
"loss": 1.576,
"step": 2700
},
{
"epoch": 0.07,
"grad_norm": 3.608473062515259,
"learning_rate": 1.8501694915254237e-05,
"loss": 1.5574,
"step": 2710
},
{
"epoch": 0.07,
"grad_norm": 2.6018731594085693,
"learning_rate": 1.849491525423729e-05,
"loss": 1.473,
"step": 2720
},
{
"epoch": 0.07,
"grad_norm": 2.2333261966705322,
"learning_rate": 1.848813559322034e-05,
"loss": 1.4562,
"step": 2730
},
{
"epoch": 0.07,
"grad_norm": 6.693967342376709,
"learning_rate": 1.848135593220339e-05,
"loss": 1.433,
"step": 2740
},
{
"epoch": 0.07,
"grad_norm": 2.6075336933135986,
"learning_rate": 1.8474576271186443e-05,
"loss": 1.2995,
"step": 2750
},
{
"epoch": 0.07,
"grad_norm": 2.2900924682617188,
"learning_rate": 1.8467796610169493e-05,
"loss": 1.4313,
"step": 2760
},
{
"epoch": 0.07,
"grad_norm": 5.270084381103516,
"learning_rate": 1.8461016949152542e-05,
"loss": 1.4531,
"step": 2770
},
{
"epoch": 0.07,
"grad_norm": 2.3377845287323,
"learning_rate": 1.8454237288135596e-05,
"loss": 1.4393,
"step": 2780
},
{
"epoch": 0.07,
"grad_norm": 1.9369862079620361,
"learning_rate": 1.8447457627118645e-05,
"loss": 1.5269,
"step": 2790
},
{
"epoch": 0.07,
"grad_norm": 3.107175350189209,
"learning_rate": 1.8440677966101695e-05,
"loss": 1.4581,
"step": 2800
},
{
"epoch": 0.07,
"grad_norm": 2.7586755752563477,
"learning_rate": 1.8433898305084748e-05,
"loss": 1.4577,
"step": 2810
},
{
"epoch": 0.07,
"grad_norm": 1.3124182224273682,
"learning_rate": 1.8427118644067798e-05,
"loss": 1.6085,
"step": 2820
},
{
"epoch": 0.07,
"grad_norm": 1.8919298648834229,
"learning_rate": 1.842033898305085e-05,
"loss": 1.3827,
"step": 2830
},
{
"epoch": 0.07,
"grad_norm": 2.49289870262146,
"learning_rate": 1.84135593220339e-05,
"loss": 1.4628,
"step": 2840
},
{
"epoch": 0.07,
"grad_norm": 2.832827568054199,
"learning_rate": 1.840677966101695e-05,
"loss": 1.4701,
"step": 2850
},
{
"epoch": 0.07,
"grad_norm": 1.2280476093292236,
"learning_rate": 1.8400000000000003e-05,
"loss": 1.5576,
"step": 2860
},
{
"epoch": 0.07,
"grad_norm": 2.703963279724121,
"learning_rate": 1.8393220338983053e-05,
"loss": 1.3686,
"step": 2870
},
{
"epoch": 0.07,
"grad_norm": 1.554801106452942,
"learning_rate": 1.8386440677966103e-05,
"loss": 1.5867,
"step": 2880
},
{
"epoch": 0.07,
"grad_norm": 1.4219539165496826,
"learning_rate": 1.8379661016949153e-05,
"loss": 1.696,
"step": 2890
},
{
"epoch": 0.07,
"grad_norm": 7.199713706970215,
"learning_rate": 1.8372881355932202e-05,
"loss": 1.5041,
"step": 2900
},
{
"epoch": 0.07,
"grad_norm": 2.7661526203155518,
"learning_rate": 1.8366101694915255e-05,
"loss": 1.4472,
"step": 2910
},
{
"epoch": 0.07,
"grad_norm": 6.59897518157959,
"learning_rate": 1.8359322033898305e-05,
"loss": 1.4166,
"step": 2920
},
{
"epoch": 0.07,
"grad_norm": 1.927669882774353,
"learning_rate": 1.8352542372881358e-05,
"loss": 1.4674,
"step": 2930
},
{
"epoch": 0.07,
"grad_norm": 1.0190672874450684,
"learning_rate": 1.8345762711864408e-05,
"loss": 1.4591,
"step": 2940
},
{
"epoch": 0.07,
"grad_norm": 3.63718843460083,
"learning_rate": 1.8338983050847458e-05,
"loss": 1.7469,
"step": 2950
},
{
"epoch": 0.07,
"grad_norm": 1.5168763399124146,
"learning_rate": 1.833220338983051e-05,
"loss": 1.5691,
"step": 2960
},
{
"epoch": 0.07,
"grad_norm": 1.323123574256897,
"learning_rate": 1.832542372881356e-05,
"loss": 1.385,
"step": 2970
},
{
"epoch": 0.07,
"grad_norm": 3.590506076812744,
"learning_rate": 1.831864406779661e-05,
"loss": 1.429,
"step": 2980
},
{
"epoch": 0.07,
"grad_norm": 4.562034606933594,
"learning_rate": 1.8311864406779663e-05,
"loss": 1.4169,
"step": 2990
},
{
"epoch": 0.07,
"grad_norm": 3.3229966163635254,
"learning_rate": 1.8305084745762713e-05,
"loss": 1.5123,
"step": 3000
},
{
"epoch": 0.07,
"eval_loss": 1.5320124626159668,
"eval_runtime": 66.1653,
"eval_samples_per_second": 15.114,
"eval_steps_per_second": 15.114,
"step": 3000
},
{
"epoch": 0.08,
"grad_norm": 3.8690502643585205,
"learning_rate": 1.8298305084745766e-05,
"loss": 1.2614,
"step": 3010
},
{
"epoch": 0.08,
"grad_norm": 2.8421173095703125,
"learning_rate": 1.8291525423728816e-05,
"loss": 1.3572,
"step": 3020
},
{
"epoch": 0.08,
"grad_norm": 3.3784098625183105,
"learning_rate": 1.8284745762711865e-05,
"loss": 1.4465,
"step": 3030
},
{
"epoch": 0.08,
"grad_norm": 3.611422061920166,
"learning_rate": 1.827796610169492e-05,
"loss": 1.4531,
"step": 3040
},
{
"epoch": 0.08,
"grad_norm": 2.3969950675964355,
"learning_rate": 1.8271186440677968e-05,
"loss": 1.3137,
"step": 3050
},
{
"epoch": 0.08,
"grad_norm": 4.92153263092041,
"learning_rate": 1.8264406779661018e-05,
"loss": 1.6209,
"step": 3060
},
{
"epoch": 0.08,
"grad_norm": 3.952179431915283,
"learning_rate": 1.825762711864407e-05,
"loss": 1.4923,
"step": 3070
},
{
"epoch": 0.08,
"grad_norm": 4.00267219543457,
"learning_rate": 1.825084745762712e-05,
"loss": 1.5688,
"step": 3080
},
{
"epoch": 0.08,
"grad_norm": 2.1664321422576904,
"learning_rate": 1.824406779661017e-05,
"loss": 1.4545,
"step": 3090
},
{
"epoch": 0.08,
"grad_norm": 3.1541237831115723,
"learning_rate": 1.823728813559322e-05,
"loss": 1.6623,
"step": 3100
},
{
"epoch": 0.08,
"grad_norm": 2.577681064605713,
"learning_rate": 1.8230508474576273e-05,
"loss": 1.6301,
"step": 3110
},
{
"epoch": 0.08,
"grad_norm": 2.2123563289642334,
"learning_rate": 1.8223728813559323e-05,
"loss": 1.5614,
"step": 3120
},
{
"epoch": 0.08,
"grad_norm": 1.7630640268325806,
"learning_rate": 1.8216949152542373e-05,
"loss": 1.5482,
"step": 3130
},
{
"epoch": 0.08,
"grad_norm": 1.5110881328582764,
"learning_rate": 1.8210169491525426e-05,
"loss": 1.4324,
"step": 3140
},
{
"epoch": 0.08,
"grad_norm": 2.9992825984954834,
"learning_rate": 1.8203389830508475e-05,
"loss": 1.5376,
"step": 3150
},
{
"epoch": 0.08,
"grad_norm": 2.825230836868286,
"learning_rate": 1.8196610169491525e-05,
"loss": 1.3925,
"step": 3160
},
{
"epoch": 0.08,
"grad_norm": 5.502832412719727,
"learning_rate": 1.8189830508474578e-05,
"loss": 1.5082,
"step": 3170
},
{
"epoch": 0.08,
"grad_norm": 1.258467197418213,
"learning_rate": 1.8183050847457628e-05,
"loss": 1.512,
"step": 3180
},
{
"epoch": 0.08,
"grad_norm": 6.260064601898193,
"learning_rate": 1.817627118644068e-05,
"loss": 1.5949,
"step": 3190
},
{
"epoch": 0.08,
"grad_norm": 1.3906792402267456,
"learning_rate": 1.816949152542373e-05,
"loss": 1.4887,
"step": 3200
},
{
"epoch": 0.08,
"grad_norm": 6.833672523498535,
"learning_rate": 1.816271186440678e-05,
"loss": 1.6286,
"step": 3210
},
{
"epoch": 0.08,
"grad_norm": 7.004729747772217,
"learning_rate": 1.8155932203389833e-05,
"loss": 1.6974,
"step": 3220
},
{
"epoch": 0.08,
"grad_norm": 3.688082695007324,
"learning_rate": 1.8149152542372883e-05,
"loss": 1.577,
"step": 3230
},
{
"epoch": 0.08,
"grad_norm": 1.2561973333358765,
"learning_rate": 1.8142372881355933e-05,
"loss": 1.5154,
"step": 3240
},
{
"epoch": 0.08,
"grad_norm": 2.1860358715057373,
"learning_rate": 1.8135593220338986e-05,
"loss": 1.5281,
"step": 3250
},
{
"epoch": 0.08,
"grad_norm": 1.4622399806976318,
"learning_rate": 1.8128813559322036e-05,
"loss": 1.4103,
"step": 3260
},
{
"epoch": 0.08,
"grad_norm": 3.036226511001587,
"learning_rate": 1.8122033898305085e-05,
"loss": 1.4003,
"step": 3270
},
{
"epoch": 0.08,
"grad_norm": 4.228930473327637,
"learning_rate": 1.811525423728814e-05,
"loss": 1.6889,
"step": 3280
},
{
"epoch": 0.08,
"grad_norm": 4.7660322189331055,
"learning_rate": 1.8108474576271188e-05,
"loss": 1.5185,
"step": 3290
},
{
"epoch": 0.08,
"grad_norm": 2.857886552810669,
"learning_rate": 1.810169491525424e-05,
"loss": 1.4166,
"step": 3300
},
{
"epoch": 0.08,
"grad_norm": 2.357570171356201,
"learning_rate": 1.8094915254237288e-05,
"loss": 1.6341,
"step": 3310
},
{
"epoch": 0.08,
"grad_norm": 2.1648855209350586,
"learning_rate": 1.808813559322034e-05,
"loss": 1.6001,
"step": 3320
},
{
"epoch": 0.08,
"grad_norm": 2.5360536575317383,
"learning_rate": 1.808135593220339e-05,
"loss": 1.5059,
"step": 3330
},
{
"epoch": 0.08,
"grad_norm": 3.741981267929077,
"learning_rate": 1.807457627118644e-05,
"loss": 1.5082,
"step": 3340
},
{
"epoch": 0.08,
"grad_norm": 3.4550766944885254,
"learning_rate": 1.8067796610169493e-05,
"loss": 1.6378,
"step": 3350
},
{
"epoch": 0.08,
"grad_norm": 1.9454740285873413,
"learning_rate": 1.8061016949152543e-05,
"loss": 1.6602,
"step": 3360
},
{
"epoch": 0.08,
"grad_norm": 3.708360195159912,
"learning_rate": 1.8054237288135593e-05,
"loss": 1.4584,
"step": 3370
},
{
"epoch": 0.08,
"grad_norm": 2.4115514755249023,
"learning_rate": 1.8047457627118646e-05,
"loss": 1.4053,
"step": 3380
},
{
"epoch": 0.08,
"grad_norm": 2.075867176055908,
"learning_rate": 1.8040677966101695e-05,
"loss": 1.46,
"step": 3390
},
{
"epoch": 0.09,
"grad_norm": 3.1672651767730713,
"learning_rate": 1.803389830508475e-05,
"loss": 1.4964,
"step": 3400
},
{
"epoch": 0.09,
"grad_norm": 4.5812578201293945,
"learning_rate": 1.8027118644067798e-05,
"loss": 1.4005,
"step": 3410
},
{
"epoch": 0.09,
"grad_norm": 2.583771228790283,
"learning_rate": 1.8020338983050848e-05,
"loss": 1.6024,
"step": 3420
},
{
"epoch": 0.09,
"grad_norm": 4.464291095733643,
"learning_rate": 1.80135593220339e-05,
"loss": 1.5075,
"step": 3430
},
{
"epoch": 0.09,
"grad_norm": 2.609657049179077,
"learning_rate": 1.800677966101695e-05,
"loss": 1.4527,
"step": 3440
},
{
"epoch": 0.09,
"grad_norm": 1.844227910041809,
"learning_rate": 1.8e-05,
"loss": 1.471,
"step": 3450
},
{
"epoch": 0.09,
"grad_norm": 3.5781702995300293,
"learning_rate": 1.7993220338983054e-05,
"loss": 1.5221,
"step": 3460
},
{
"epoch": 0.09,
"grad_norm": 1.8293837308883667,
"learning_rate": 1.7986440677966103e-05,
"loss": 1.4831,
"step": 3470
},
{
"epoch": 0.09,
"grad_norm": 5.273003101348877,
"learning_rate": 1.7979661016949156e-05,
"loss": 1.5131,
"step": 3480
},
{
"epoch": 0.09,
"grad_norm": 3.1019296646118164,
"learning_rate": 1.7972881355932206e-05,
"loss": 1.4346,
"step": 3490
},
{
"epoch": 0.09,
"grad_norm": 3.419081687927246,
"learning_rate": 1.7966101694915256e-05,
"loss": 1.496,
"step": 3500
},
{
"epoch": 0.09,
"eval_loss": 1.5301283597946167,
"eval_runtime": 66.1385,
"eval_samples_per_second": 15.12,
"eval_steps_per_second": 15.12,
"step": 3500
},
{
"epoch": 0.09,
"grad_norm": 4.49799919128418,
"learning_rate": 1.795932203389831e-05,
"loss": 1.5304,
"step": 3510
},
{
"epoch": 0.09,
"grad_norm": 3.267191171646118,
"learning_rate": 1.7952542372881355e-05,
"loss": 1.5103,
"step": 3520
},
{
"epoch": 0.09,
"grad_norm": 3.4655425548553467,
"learning_rate": 1.7945762711864408e-05,
"loss": 1.3131,
"step": 3530
},
{
"epoch": 0.09,
"grad_norm": 2.289358615875244,
"learning_rate": 1.7938983050847458e-05,
"loss": 1.2731,
"step": 3540
},
{
"epoch": 0.09,
"grad_norm": 1.528286337852478,
"learning_rate": 1.7932203389830508e-05,
"loss": 1.6249,
"step": 3550
},
{
"epoch": 0.09,
"grad_norm": 1.5927760601043701,
"learning_rate": 1.792542372881356e-05,
"loss": 1.5177,
"step": 3560
},
{
"epoch": 0.09,
"grad_norm": 1.5520817041397095,
"learning_rate": 1.791864406779661e-05,
"loss": 1.4523,
"step": 3570
},
{
"epoch": 0.09,
"grad_norm": 3.531754970550537,
"learning_rate": 1.7911864406779664e-05,
"loss": 1.5342,
"step": 3580
},
{
"epoch": 0.09,
"grad_norm": 2.6139304637908936,
"learning_rate": 1.7905084745762713e-05,
"loss": 1.3377,
"step": 3590
},
{
"epoch": 0.09,
"grad_norm": 5.242190837860107,
"learning_rate": 1.7898305084745763e-05,
"loss": 1.592,
"step": 3600
},
{
"epoch": 0.09,
"grad_norm": 3.10337495803833,
"learning_rate": 1.7891525423728816e-05,
"loss": 1.4516,
"step": 3610
},
{
"epoch": 0.09,
"grad_norm": 2.615224599838257,
"learning_rate": 1.7884745762711866e-05,
"loss": 1.4325,
"step": 3620
},
{
"epoch": 0.09,
"grad_norm": 5.353131294250488,
"learning_rate": 1.7877966101694916e-05,
"loss": 1.5772,
"step": 3630
},
{
"epoch": 0.09,
"grad_norm": 1.1678059101104736,
"learning_rate": 1.787118644067797e-05,
"loss": 1.572,
"step": 3640
},
{
"epoch": 0.09,
"grad_norm": 3.2298007011413574,
"learning_rate": 1.7864406779661018e-05,
"loss": 1.5214,
"step": 3650
},
{
"epoch": 0.09,
"grad_norm": 3.405092477798462,
"learning_rate": 1.785762711864407e-05,
"loss": 1.4677,
"step": 3660
},
{
"epoch": 0.09,
"grad_norm": 3.6669983863830566,
"learning_rate": 1.785084745762712e-05,
"loss": 1.2853,
"step": 3670
},
{
"epoch": 0.09,
"grad_norm": 3.2381701469421387,
"learning_rate": 1.784406779661017e-05,
"loss": 1.3501,
"step": 3680
},
{
"epoch": 0.09,
"grad_norm": 2.3621437549591064,
"learning_rate": 1.7837288135593224e-05,
"loss": 1.5105,
"step": 3690
},
{
"epoch": 0.09,
"grad_norm": 1.1474658250808716,
"learning_rate": 1.7830508474576274e-05,
"loss": 1.5107,
"step": 3700
},
{
"epoch": 0.09,
"grad_norm": 9.769930839538574,
"learning_rate": 1.7823728813559323e-05,
"loss": 1.8114,
"step": 3710
},
{
"epoch": 0.09,
"grad_norm": 1.6212892532348633,
"learning_rate": 1.7816949152542376e-05,
"loss": 1.4948,
"step": 3720
},
{
"epoch": 0.09,
"grad_norm": 3.909820795059204,
"learning_rate": 1.7810169491525423e-05,
"loss": 1.4145,
"step": 3730
},
{
"epoch": 0.09,
"grad_norm": 1.870615005493164,
"learning_rate": 1.7803389830508476e-05,
"loss": 1.4856,
"step": 3740
},
{
"epoch": 0.09,
"grad_norm": 2.5912163257598877,
"learning_rate": 1.7796610169491526e-05,
"loss": 1.5021,
"step": 3750
},
{
"epoch": 0.09,
"grad_norm": 2.303807020187378,
"learning_rate": 1.778983050847458e-05,
"loss": 1.4348,
"step": 3760
},
{
"epoch": 0.09,
"grad_norm": 2.5953946113586426,
"learning_rate": 1.778305084745763e-05,
"loss": 1.4073,
"step": 3770
},
{
"epoch": 0.09,
"grad_norm": 2.129793643951416,
"learning_rate": 1.7776271186440678e-05,
"loss": 1.4827,
"step": 3780
},
{
"epoch": 0.09,
"grad_norm": 8.186966896057129,
"learning_rate": 1.776949152542373e-05,
"loss": 1.5887,
"step": 3790
},
{
"epoch": 0.1,
"grad_norm": 2.2956864833831787,
"learning_rate": 1.776271186440678e-05,
"loss": 1.4925,
"step": 3800
},
{
"epoch": 0.1,
"grad_norm": 1.3646732568740845,
"learning_rate": 1.775593220338983e-05,
"loss": 1.5814,
"step": 3810
},
{
"epoch": 0.1,
"grad_norm": 5.3318071365356445,
"learning_rate": 1.7749152542372884e-05,
"loss": 1.6139,
"step": 3820
},
{
"epoch": 0.1,
"grad_norm": 2.53743314743042,
"learning_rate": 1.7742372881355933e-05,
"loss": 1.5198,
"step": 3830
},
{
"epoch": 0.1,
"grad_norm": 7.048913478851318,
"learning_rate": 1.7735593220338983e-05,
"loss": 1.4417,
"step": 3840
},
{
"epoch": 0.1,
"grad_norm": 2.367314100265503,
"learning_rate": 1.7728813559322036e-05,
"loss": 1.4633,
"step": 3850
},
{
"epoch": 0.1,
"grad_norm": 1.619931697845459,
"learning_rate": 1.7722033898305086e-05,
"loss": 1.3719,
"step": 3860
},
{
"epoch": 0.1,
"grad_norm": 3.264190435409546,
"learning_rate": 1.771525423728814e-05,
"loss": 1.3079,
"step": 3870
},
{
"epoch": 0.1,
"grad_norm": 4.68599796295166,
"learning_rate": 1.770847457627119e-05,
"loss": 1.6278,
"step": 3880
},
{
"epoch": 0.1,
"grad_norm": 4.21970272064209,
"learning_rate": 1.770169491525424e-05,
"loss": 1.4814,
"step": 3890
},
{
"epoch": 0.1,
"grad_norm": 2.3647706508636475,
"learning_rate": 1.769491525423729e-05,
"loss": 1.5973,
"step": 3900
},
{
"epoch": 0.1,
"grad_norm": 3.2886738777160645,
"learning_rate": 1.768813559322034e-05,
"loss": 1.5224,
"step": 3910
},
{
"epoch": 0.1,
"grad_norm": 2.6817705631256104,
"learning_rate": 1.768135593220339e-05,
"loss": 1.566,
"step": 3920
},
{
"epoch": 0.1,
"grad_norm": 1.6243913173675537,
"learning_rate": 1.767457627118644e-05,
"loss": 1.6011,
"step": 3930
},
{
"epoch": 0.1,
"grad_norm": 1.8161165714263916,
"learning_rate": 1.7667796610169494e-05,
"loss": 1.5303,
"step": 3940
},
{
"epoch": 0.1,
"grad_norm": 2.486697196960449,
"learning_rate": 1.7661016949152543e-05,
"loss": 1.4032,
"step": 3950
},
{
"epoch": 0.1,
"grad_norm": 1.8153432607650757,
"learning_rate": 1.7654237288135593e-05,
"loss": 1.5167,
"step": 3960
},
{
"epoch": 0.1,
"grad_norm": 2.186868190765381,
"learning_rate": 1.7647457627118646e-05,
"loss": 1.6066,
"step": 3970
},
{
"epoch": 0.1,
"grad_norm": 2.1040332317352295,
"learning_rate": 1.7640677966101696e-05,
"loss": 1.4142,
"step": 3980
},
{
"epoch": 0.1,
"grad_norm": 3.989042282104492,
"learning_rate": 1.7633898305084746e-05,
"loss": 1.4877,
"step": 3990
},
{
"epoch": 0.1,
"grad_norm": 3.7532989978790283,
"learning_rate": 1.76271186440678e-05,
"loss": 1.4016,
"step": 4000
},
{
"epoch": 0.1,
"eval_loss": 1.5089428424835205,
"eval_runtime": 66.2329,
"eval_samples_per_second": 15.098,
"eval_steps_per_second": 15.098,
"step": 4000
},
{
"epoch": 0.1,
"grad_norm": 6.605846405029297,
"learning_rate": 1.762033898305085e-05,
"loss": 1.6163,
"step": 4010
},
{
"epoch": 0.1,
"grad_norm": 3.1976850032806396,
"learning_rate": 1.7613559322033898e-05,
"loss": 1.4298,
"step": 4020
},
{
"epoch": 0.1,
"grad_norm": 5.412558555603027,
"learning_rate": 1.760677966101695e-05,
"loss": 1.3316,
"step": 4030
},
{
"epoch": 0.1,
"grad_norm": 5.241945266723633,
"learning_rate": 1.76e-05,
"loss": 1.6258,
"step": 4040
},
{
"epoch": 0.1,
"grad_norm": 11.647382736206055,
"learning_rate": 1.7593220338983054e-05,
"loss": 1.5479,
"step": 4050
},
{
"epoch": 0.1,
"grad_norm": 2.091214656829834,
"learning_rate": 1.7586440677966104e-05,
"loss": 1.4385,
"step": 4060
},
{
"epoch": 0.1,
"grad_norm": 2.492511034011841,
"learning_rate": 1.7579661016949153e-05,
"loss": 1.6785,
"step": 4070
},
{
"epoch": 0.1,
"grad_norm": 1.9180244207382202,
"learning_rate": 1.7572881355932206e-05,
"loss": 1.6234,
"step": 4080
},
{
"epoch": 0.1,
"grad_norm": 2.256627082824707,
"learning_rate": 1.7566101694915256e-05,
"loss": 1.3143,
"step": 4090
},
{
"epoch": 0.1,
"grad_norm": 2.2864432334899902,
"learning_rate": 1.7559322033898306e-05,
"loss": 1.4961,
"step": 4100
},
{
"epoch": 0.1,
"grad_norm": 6.091422080993652,
"learning_rate": 1.755254237288136e-05,
"loss": 1.27,
"step": 4110
},
{
"epoch": 0.1,
"grad_norm": 1.6744223833084106,
"learning_rate": 1.754576271186441e-05,
"loss": 1.6859,
"step": 4120
},
{
"epoch": 0.1,
"grad_norm": 3.741482973098755,
"learning_rate": 1.753898305084746e-05,
"loss": 1.5382,
"step": 4130
},
{
"epoch": 0.1,
"grad_norm": 3.2129416465759277,
"learning_rate": 1.7532203389830508e-05,
"loss": 1.548,
"step": 4140
},
{
"epoch": 0.1,
"grad_norm": 1.9851702451705933,
"learning_rate": 1.752542372881356e-05,
"loss": 1.5777,
"step": 4150
},
{
"epoch": 0.1,
"grad_norm": 2.7135250568389893,
"learning_rate": 1.751864406779661e-05,
"loss": 1.6374,
"step": 4160
},
{
"epoch": 0.1,
"grad_norm": 6.881992340087891,
"learning_rate": 1.751186440677966e-05,
"loss": 1.5171,
"step": 4170
},
{
"epoch": 0.1,
"grad_norm": 3.8673460483551025,
"learning_rate": 1.7505084745762714e-05,
"loss": 1.3488,
"step": 4180
},
{
"epoch": 0.1,
"grad_norm": 3.7248244285583496,
"learning_rate": 1.7498305084745763e-05,
"loss": 1.4269,
"step": 4190
},
{
"epoch": 0.1,
"grad_norm": 4.191452503204346,
"learning_rate": 1.7491525423728813e-05,
"loss": 1.591,
"step": 4200
},
{
"epoch": 0.11,
"grad_norm": 1.9477959871292114,
"learning_rate": 1.7484745762711866e-05,
"loss": 1.5194,
"step": 4210
},
{
"epoch": 0.11,
"grad_norm": 4.151371002197266,
"learning_rate": 1.7477966101694916e-05,
"loss": 1.4644,
"step": 4220
},
{
"epoch": 0.11,
"grad_norm": 3.285062551498413,
"learning_rate": 1.747118644067797e-05,
"loss": 1.3697,
"step": 4230
},
{
"epoch": 0.11,
"grad_norm": 5.2172136306762695,
"learning_rate": 1.746440677966102e-05,
"loss": 1.3542,
"step": 4240
},
{
"epoch": 0.11,
"grad_norm": 2.321295738220215,
"learning_rate": 1.745762711864407e-05,
"loss": 1.6265,
"step": 4250
},
{
"epoch": 0.11,
"grad_norm": 3.2060482501983643,
"learning_rate": 1.745084745762712e-05,
"loss": 1.5214,
"step": 4260
},
{
"epoch": 0.11,
"grad_norm": 2.697052240371704,
"learning_rate": 1.744406779661017e-05,
"loss": 1.5446,
"step": 4270
},
{
"epoch": 0.11,
"grad_norm": 2.639937162399292,
"learning_rate": 1.743728813559322e-05,
"loss": 1.5329,
"step": 4280
},
{
"epoch": 0.11,
"grad_norm": 3.5502707958221436,
"learning_rate": 1.7430508474576274e-05,
"loss": 1.4997,
"step": 4290
},
{
"epoch": 0.11,
"grad_norm": 5.514580726623535,
"learning_rate": 1.7423728813559324e-05,
"loss": 1.4641,
"step": 4300
},
{
"epoch": 0.11,
"grad_norm": 2.264512062072754,
"learning_rate": 1.7416949152542373e-05,
"loss": 1.5588,
"step": 4310
},
{
"epoch": 0.11,
"grad_norm": 4.865510940551758,
"learning_rate": 1.7410169491525427e-05,
"loss": 1.5777,
"step": 4320
},
{
"epoch": 0.11,
"grad_norm": 2.872544050216675,
"learning_rate": 1.7403389830508476e-05,
"loss": 1.4683,
"step": 4330
},
{
"epoch": 0.11,
"grad_norm": 2.1891725063323975,
"learning_rate": 1.739661016949153e-05,
"loss": 1.4572,
"step": 4340
},
{
"epoch": 0.11,
"grad_norm": 5.604435443878174,
"learning_rate": 1.7389830508474576e-05,
"loss": 1.3685,
"step": 4350
},
{
"epoch": 0.11,
"grad_norm": 1.9309135675430298,
"learning_rate": 1.738305084745763e-05,
"loss": 1.6086,
"step": 4360
},
{
"epoch": 0.11,
"grad_norm": 1.2651278972625732,
"learning_rate": 1.737627118644068e-05,
"loss": 1.5768,
"step": 4370
},
{
"epoch": 0.11,
"grad_norm": 3.4745676517486572,
"learning_rate": 1.7369491525423728e-05,
"loss": 1.4553,
"step": 4380
},
{
"epoch": 0.11,
"grad_norm": 5.904428958892822,
"learning_rate": 1.736271186440678e-05,
"loss": 1.654,
"step": 4390
},
{
"epoch": 0.11,
"grad_norm": 1.6250745058059692,
"learning_rate": 1.735593220338983e-05,
"loss": 1.5443,
"step": 4400
},
{
"epoch": 0.11,
"grad_norm": 3.289626121520996,
"learning_rate": 1.734915254237288e-05,
"loss": 1.4691,
"step": 4410
},
{
"epoch": 0.11,
"grad_norm": 5.800081729888916,
"learning_rate": 1.7342372881355934e-05,
"loss": 1.543,
"step": 4420
},
{
"epoch": 0.11,
"grad_norm": 9.145909309387207,
"learning_rate": 1.7335593220338983e-05,
"loss": 1.4814,
"step": 4430
},
{
"epoch": 0.11,
"grad_norm": 2.22583270072937,
"learning_rate": 1.7328813559322037e-05,
"loss": 1.548,
"step": 4440
},
{
"epoch": 0.11,
"grad_norm": 4.693650722503662,
"learning_rate": 1.7322033898305086e-05,
"loss": 1.5602,
"step": 4450
},
{
"epoch": 0.11,
"grad_norm": 2.1424403190612793,
"learning_rate": 1.7315254237288136e-05,
"loss": 1.5298,
"step": 4460
},
{
"epoch": 0.11,
"grad_norm": 2.8103983402252197,
"learning_rate": 1.730847457627119e-05,
"loss": 1.435,
"step": 4470
},
{
"epoch": 0.11,
"grad_norm": 2.2972493171691895,
"learning_rate": 1.730169491525424e-05,
"loss": 1.4906,
"step": 4480
},
{
"epoch": 0.11,
"grad_norm": 2.437459707260132,
"learning_rate": 1.729491525423729e-05,
"loss": 1.5637,
"step": 4490
},
{
"epoch": 0.11,
"grad_norm": 3.561988115310669,
"learning_rate": 1.728813559322034e-05,
"loss": 1.572,
"step": 4500
},
{
"epoch": 0.11,
"eval_loss": 1.4887409210205078,
"eval_runtime": 66.1778,
"eval_samples_per_second": 15.111,
"eval_steps_per_second": 15.111,
"step": 4500
},
{
"epoch": 0.11,
"grad_norm": 0.4782548248767853,
"learning_rate": 1.728135593220339e-05,
"loss": 1.606,
"step": 4510
},
{
"epoch": 0.11,
"grad_norm": 4.1626386642456055,
"learning_rate": 1.7274576271186444e-05,
"loss": 1.6311,
"step": 4520
},
{
"epoch": 0.11,
"grad_norm": 3.057555675506592,
"learning_rate": 1.7267796610169494e-05,
"loss": 1.495,
"step": 4530
},
{
"epoch": 0.11,
"grad_norm": 4.714250087738037,
"learning_rate": 1.7261016949152544e-05,
"loss": 1.5219,
"step": 4540
},
{
"epoch": 0.11,
"grad_norm": 1.1815061569213867,
"learning_rate": 1.7254237288135597e-05,
"loss": 1.5227,
"step": 4550
},
{
"epoch": 0.11,
"grad_norm": 3.602353811264038,
"learning_rate": 1.7247457627118643e-05,
"loss": 1.6756,
"step": 4560
},
{
"epoch": 0.11,
"grad_norm": 1.7794777154922485,
"learning_rate": 1.7240677966101696e-05,
"loss": 1.3716,
"step": 4570
},
{
"epoch": 0.11,
"grad_norm": 2.1229944229125977,
"learning_rate": 1.7233898305084746e-05,
"loss": 1.4612,
"step": 4580
},
{
"epoch": 0.11,
"grad_norm": 4.656471252441406,
"learning_rate": 1.7227118644067796e-05,
"loss": 1.6358,
"step": 4590
},
{
"epoch": 0.12,
"grad_norm": 2.0882911682128906,
"learning_rate": 1.722033898305085e-05,
"loss": 1.5147,
"step": 4600
},
{
"epoch": 0.12,
"grad_norm": 3.070713996887207,
"learning_rate": 1.72135593220339e-05,
"loss": 1.5253,
"step": 4610
},
{
"epoch": 0.12,
"grad_norm": 1.7038753032684326,
"learning_rate": 1.720677966101695e-05,
"loss": 1.4375,
"step": 4620
},
{
"epoch": 0.12,
"grad_norm": 2.5298070907592773,
"learning_rate": 1.72e-05,
"loss": 1.5301,
"step": 4630
},
{
"epoch": 0.12,
"grad_norm": 2.2928740978240967,
"learning_rate": 1.719322033898305e-05,
"loss": 1.2571,
"step": 4640
},
{
"epoch": 0.12,
"grad_norm": 5.660508155822754,
"learning_rate": 1.7186440677966104e-05,
"loss": 1.5168,
"step": 4650
},
{
"epoch": 0.12,
"grad_norm": 5.104097366333008,
"learning_rate": 1.7179661016949154e-05,
"loss": 1.4134,
"step": 4660
},
{
"epoch": 0.12,
"grad_norm": 4.954836368560791,
"learning_rate": 1.7172881355932204e-05,
"loss": 1.4909,
"step": 4670
},
{
"epoch": 0.12,
"grad_norm": 4.392303466796875,
"learning_rate": 1.7166101694915257e-05,
"loss": 1.5639,
"step": 4680
},
{
"epoch": 0.12,
"grad_norm": 3.9031317234039307,
"learning_rate": 1.7159322033898306e-05,
"loss": 1.4521,
"step": 4690
},
{
"epoch": 0.12,
"grad_norm": 4.280984401702881,
"learning_rate": 1.715254237288136e-05,
"loss": 1.3877,
"step": 4700
},
{
"epoch": 0.12,
"grad_norm": 1.2300760746002197,
"learning_rate": 1.714576271186441e-05,
"loss": 1.5971,
"step": 4710
},
{
"epoch": 0.12,
"grad_norm": 3.845804452896118,
"learning_rate": 1.713898305084746e-05,
"loss": 1.5987,
"step": 4720
},
{
"epoch": 0.12,
"grad_norm": 5.861780643463135,
"learning_rate": 1.7132203389830512e-05,
"loss": 1.7391,
"step": 4730
},
{
"epoch": 0.12,
"grad_norm": 4.704226970672607,
"learning_rate": 1.712542372881356e-05,
"loss": 1.3629,
"step": 4740
},
{
"epoch": 0.12,
"grad_norm": 1.362378716468811,
"learning_rate": 1.711864406779661e-05,
"loss": 1.5607,
"step": 4750
},
{
"epoch": 0.12,
"grad_norm": 3.89095139503479,
"learning_rate": 1.711186440677966e-05,
"loss": 1.6382,
"step": 4760
},
{
"epoch": 0.12,
"grad_norm": 2.8033549785614014,
"learning_rate": 1.710508474576271e-05,
"loss": 1.6113,
"step": 4770
},
{
"epoch": 0.12,
"grad_norm": 3.045806884765625,
"learning_rate": 1.7098305084745764e-05,
"loss": 1.4223,
"step": 4780
},
{
"epoch": 0.12,
"grad_norm": 7.401500701904297,
"learning_rate": 1.7091525423728814e-05,
"loss": 1.4659,
"step": 4790
},
{
"epoch": 0.12,
"grad_norm": 1.490501880645752,
"learning_rate": 1.7084745762711867e-05,
"loss": 1.4543,
"step": 4800
},
{
"epoch": 0.12,
"grad_norm": 2.137923240661621,
"learning_rate": 1.7077966101694916e-05,
"loss": 1.4117,
"step": 4810
},
{
"epoch": 0.12,
"grad_norm": 3.254201889038086,
"learning_rate": 1.7071186440677966e-05,
"loss": 1.4404,
"step": 4820
},
{
"epoch": 0.12,
"grad_norm": 2.1423888206481934,
"learning_rate": 1.706440677966102e-05,
"loss": 1.5539,
"step": 4830
},
{
"epoch": 0.12,
"grad_norm": 1.927777647972107,
"learning_rate": 1.705762711864407e-05,
"loss": 1.5285,
"step": 4840
},
{
"epoch": 0.12,
"grad_norm": 3.6136608123779297,
"learning_rate": 1.705084745762712e-05,
"loss": 1.3503,
"step": 4850
},
{
"epoch": 0.12,
"grad_norm": 4.160021781921387,
"learning_rate": 1.704406779661017e-05,
"loss": 1.5243,
"step": 4860
},
{
"epoch": 0.12,
"grad_norm": 3.0722756385803223,
"learning_rate": 1.703728813559322e-05,
"loss": 1.5306,
"step": 4870
},
{
"epoch": 0.12,
"grad_norm": 6.304449558258057,
"learning_rate": 1.703050847457627e-05,
"loss": 1.5878,
"step": 4880
},
{
"epoch": 0.12,
"grad_norm": 1.935084581375122,
"learning_rate": 1.7023728813559324e-05,
"loss": 1.5446,
"step": 4890
},
{
"epoch": 0.12,
"grad_norm": 2.6202478408813477,
"learning_rate": 1.7016949152542374e-05,
"loss": 1.6461,
"step": 4900
},
{
"epoch": 0.12,
"grad_norm": 3.2252795696258545,
"learning_rate": 1.7010169491525427e-05,
"loss": 1.445,
"step": 4910
},
{
"epoch": 0.12,
"grad_norm": 4.115199565887451,
"learning_rate": 1.7003389830508477e-05,
"loss": 1.5513,
"step": 4920
},
{
"epoch": 0.12,
"grad_norm": 2.852402448654175,
"learning_rate": 1.6996610169491526e-05,
"loss": 1.496,
"step": 4930
},
{
"epoch": 0.12,
"grad_norm": 2.5148980617523193,
"learning_rate": 1.698983050847458e-05,
"loss": 1.4281,
"step": 4940
},
{
"epoch": 0.12,
"grad_norm": 1.3501824140548706,
"learning_rate": 1.698305084745763e-05,
"loss": 1.4301,
"step": 4950
},
{
"epoch": 0.12,
"grad_norm": 3.4159467220306396,
"learning_rate": 1.697627118644068e-05,
"loss": 1.5491,
"step": 4960
},
{
"epoch": 0.12,
"grad_norm": 4.547379970550537,
"learning_rate": 1.696949152542373e-05,
"loss": 1.5137,
"step": 4970
},
{
"epoch": 0.12,
"grad_norm": 3.4466910362243652,
"learning_rate": 1.6962711864406782e-05,
"loss": 1.4767,
"step": 4980
},
{
"epoch": 0.12,
"grad_norm": 6.4336395263671875,
"learning_rate": 1.695593220338983e-05,
"loss": 1.5213,
"step": 4990
},
{
"epoch": 0.12,
"grad_norm": 1.0396947860717773,
"learning_rate": 1.694915254237288e-05,
"loss": 1.4402,
"step": 5000
},
{
"epoch": 0.12,
"eval_loss": 1.4466354846954346,
"eval_runtime": 66.2264,
"eval_samples_per_second": 15.1,
"eval_steps_per_second": 15.1,
"step": 5000
},
{
"epoch": 0.13,
"grad_norm": 3.8836324214935303,
"learning_rate": 1.6942372881355934e-05,
"loss": 1.6772,
"step": 5010
},
{
"epoch": 0.13,
"grad_norm": 4.293881893157959,
"learning_rate": 1.6935593220338984e-05,
"loss": 1.4618,
"step": 5020
},
{
"epoch": 0.13,
"grad_norm": 3.111931800842285,
"learning_rate": 1.6928813559322034e-05,
"loss": 1.5564,
"step": 5030
},
{
"epoch": 0.13,
"grad_norm": 3.7768237590789795,
"learning_rate": 1.6922033898305087e-05,
"loss": 1.5239,
"step": 5040
},
{
"epoch": 0.13,
"grad_norm": 1.8081879615783691,
"learning_rate": 1.6915254237288136e-05,
"loss": 1.5257,
"step": 5050
},
{
"epoch": 0.13,
"grad_norm": 3.2274467945098877,
"learning_rate": 1.6908474576271186e-05,
"loss": 1.3785,
"step": 5060
},
{
"epoch": 0.13,
"grad_norm": 1.4177571535110474,
"learning_rate": 1.690169491525424e-05,
"loss": 1.4504,
"step": 5070
},
{
"epoch": 0.13,
"grad_norm": 4.527300834655762,
"learning_rate": 1.689491525423729e-05,
"loss": 1.4865,
"step": 5080
},
{
"epoch": 0.13,
"grad_norm": 2.1885902881622314,
"learning_rate": 1.6888135593220342e-05,
"loss": 1.5584,
"step": 5090
},
{
"epoch": 0.13,
"grad_norm": 2.6522934436798096,
"learning_rate": 1.6881355932203392e-05,
"loss": 1.4736,
"step": 5100
},
{
"epoch": 0.13,
"grad_norm": 2.8651719093322754,
"learning_rate": 1.687457627118644e-05,
"loss": 1.6247,
"step": 5110
},
{
"epoch": 0.13,
"grad_norm": 3.211879253387451,
"learning_rate": 1.6867796610169495e-05,
"loss": 1.4822,
"step": 5120
},
{
"epoch": 0.13,
"grad_norm": 4.725778579711914,
"learning_rate": 1.6861016949152544e-05,
"loss": 1.6364,
"step": 5130
},
{
"epoch": 0.13,
"grad_norm": 2.7316319942474365,
"learning_rate": 1.6854237288135594e-05,
"loss": 1.4218,
"step": 5140
},
{
"epoch": 0.13,
"grad_norm": 3.2881827354431152,
"learning_rate": 1.6847457627118647e-05,
"loss": 1.4729,
"step": 5150
},
{
"epoch": 0.13,
"grad_norm": 4.106601238250732,
"learning_rate": 1.6840677966101697e-05,
"loss": 1.5303,
"step": 5160
},
{
"epoch": 0.13,
"grad_norm": 1.9981019496917725,
"learning_rate": 1.683389830508475e-05,
"loss": 1.4583,
"step": 5170
},
{
"epoch": 0.13,
"grad_norm": 2.5599071979522705,
"learning_rate": 1.6827118644067796e-05,
"loss": 1.5817,
"step": 5180
},
{
"epoch": 0.13,
"grad_norm": 4.673508644104004,
"learning_rate": 1.682033898305085e-05,
"loss": 1.4967,
"step": 5190
},
{
"epoch": 0.13,
"grad_norm": 7.4875922203063965,
"learning_rate": 1.68135593220339e-05,
"loss": 1.4482,
"step": 5200
},
{
"epoch": 0.13,
"grad_norm": 5.208539962768555,
"learning_rate": 1.680677966101695e-05,
"loss": 1.5031,
"step": 5210
},
{
"epoch": 0.13,
"grad_norm": 5.5398664474487305,
"learning_rate": 1.6800000000000002e-05,
"loss": 1.3661,
"step": 5220
},
{
"epoch": 0.13,
"grad_norm": 4.9544267654418945,
"learning_rate": 1.679322033898305e-05,
"loss": 1.599,
"step": 5230
},
{
"epoch": 0.13,
"grad_norm": 3.121668815612793,
"learning_rate": 1.67864406779661e-05,
"loss": 1.4801,
"step": 5240
},
{
"epoch": 0.13,
"grad_norm": 4.368370532989502,
"learning_rate": 1.6779661016949154e-05,
"loss": 1.5737,
"step": 5250
},
{
"epoch": 0.13,
"grad_norm": 4.03584098815918,
"learning_rate": 1.6772881355932204e-05,
"loss": 1.4995,
"step": 5260
},
{
"epoch": 0.13,
"grad_norm": 5.6201090812683105,
"learning_rate": 1.6766101694915257e-05,
"loss": 1.3567,
"step": 5270
},
{
"epoch": 0.13,
"grad_norm": 4.009582996368408,
"learning_rate": 1.6759322033898307e-05,
"loss": 1.4241,
"step": 5280
},
{
"epoch": 0.13,
"grad_norm": 3.4552884101867676,
"learning_rate": 1.6752542372881357e-05,
"loss": 1.3472,
"step": 5290
},
{
"epoch": 0.13,
"grad_norm": 4.6100029945373535,
"learning_rate": 1.674576271186441e-05,
"loss": 1.3696,
"step": 5300
},
{
"epoch": 0.13,
"grad_norm": 6.752505779266357,
"learning_rate": 1.673898305084746e-05,
"loss": 1.5192,
"step": 5310
},
{
"epoch": 0.13,
"grad_norm": 1.4629709720611572,
"learning_rate": 1.673220338983051e-05,
"loss": 1.5774,
"step": 5320
},
{
"epoch": 0.13,
"grad_norm": 2.904249906539917,
"learning_rate": 1.6725423728813562e-05,
"loss": 1.4843,
"step": 5330
},
{
"epoch": 0.13,
"grad_norm": 3.3308334350585938,
"learning_rate": 1.6718644067796612e-05,
"loss": 1.6006,
"step": 5340
},
{
"epoch": 0.13,
"grad_norm": 2.405097484588623,
"learning_rate": 1.671186440677966e-05,
"loss": 1.5736,
"step": 5350
},
{
"epoch": 0.13,
"grad_norm": 3.4328196048736572,
"learning_rate": 1.6705084745762715e-05,
"loss": 1.4613,
"step": 5360
},
{
"epoch": 0.13,
"grad_norm": 1.5162967443466187,
"learning_rate": 1.6698305084745764e-05,
"loss": 1.4562,
"step": 5370
},
{
"epoch": 0.13,
"grad_norm": 2.473905563354492,
"learning_rate": 1.6691525423728817e-05,
"loss": 1.4899,
"step": 5380
},
{
"epoch": 0.13,
"grad_norm": 1.3600870370864868,
"learning_rate": 1.6684745762711864e-05,
"loss": 1.4197,
"step": 5390
},
{
"epoch": 0.14,
"grad_norm": 3.9598913192749023,
"learning_rate": 1.6677966101694917e-05,
"loss": 1.4268,
"step": 5400
},
{
"epoch": 0.14,
"grad_norm": 2.6608197689056396,
"learning_rate": 1.6671186440677967e-05,
"loss": 1.2926,
"step": 5410
},
{
"epoch": 0.14,
"grad_norm": 2.557682991027832,
"learning_rate": 1.6664406779661016e-05,
"loss": 1.6283,
"step": 5420
},
{
"epoch": 0.14,
"grad_norm": 2.969449281692505,
"learning_rate": 1.665762711864407e-05,
"loss": 1.4047,
"step": 5430
},
{
"epoch": 0.14,
"grad_norm": 2.7862861156463623,
"learning_rate": 1.665084745762712e-05,
"loss": 1.4233,
"step": 5440
},
{
"epoch": 0.14,
"grad_norm": 1.7756729125976562,
"learning_rate": 1.6644067796610172e-05,
"loss": 1.4485,
"step": 5450
},
{
"epoch": 0.14,
"grad_norm": 2.082773447036743,
"learning_rate": 1.6637288135593222e-05,
"loss": 1.3878,
"step": 5460
},
{
"epoch": 0.14,
"grad_norm": 5.100788116455078,
"learning_rate": 1.663050847457627e-05,
"loss": 1.5481,
"step": 5470
},
{
"epoch": 0.14,
"grad_norm": 3.218986749649048,
"learning_rate": 1.6623728813559325e-05,
"loss": 1.3906,
"step": 5480
},
{
"epoch": 0.14,
"grad_norm": 1.727528691291809,
"learning_rate": 1.6616949152542374e-05,
"loss": 1.4824,
"step": 5490
},
{
"epoch": 0.14,
"grad_norm": 4.321578502655029,
"learning_rate": 1.6610169491525424e-05,
"loss": 1.5586,
"step": 5500
},
{
"epoch": 0.14,
"eval_loss": 1.469241976737976,
"eval_runtime": 66.1281,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 5500
},
{
"epoch": 0.14,
"grad_norm": 1.7050514221191406,
"learning_rate": 1.6603389830508477e-05,
"loss": 1.2704,
"step": 5510
},
{
"epoch": 0.14,
"grad_norm": 1.4852951765060425,
"learning_rate": 1.6596610169491527e-05,
"loss": 1.6239,
"step": 5520
},
{
"epoch": 0.14,
"grad_norm": 3.524787664413452,
"learning_rate": 1.6589830508474577e-05,
"loss": 1.45,
"step": 5530
},
{
"epoch": 0.14,
"grad_norm": 1.804272174835205,
"learning_rate": 1.658305084745763e-05,
"loss": 1.4039,
"step": 5540
},
{
"epoch": 0.14,
"grad_norm": 2.2891910076141357,
"learning_rate": 1.657627118644068e-05,
"loss": 1.3969,
"step": 5550
},
{
"epoch": 0.14,
"grad_norm": 2.775587797164917,
"learning_rate": 1.6569491525423732e-05,
"loss": 1.5571,
"step": 5560
},
{
"epoch": 0.14,
"grad_norm": 4.006053924560547,
"learning_rate": 1.6562711864406782e-05,
"loss": 1.4743,
"step": 5570
},
{
"epoch": 0.14,
"grad_norm": 2.258488416671753,
"learning_rate": 1.6555932203389832e-05,
"loss": 1.3094,
"step": 5580
},
{
"epoch": 0.14,
"grad_norm": 1.469382405281067,
"learning_rate": 1.654915254237288e-05,
"loss": 1.4247,
"step": 5590
},
{
"epoch": 0.14,
"grad_norm": 2.234476089477539,
"learning_rate": 1.654237288135593e-05,
"loss": 1.4024,
"step": 5600
},
{
"epoch": 0.14,
"grad_norm": 4.297300338745117,
"learning_rate": 1.6535593220338984e-05,
"loss": 1.3422,
"step": 5610
},
{
"epoch": 0.14,
"grad_norm": 2.316805362701416,
"learning_rate": 1.6528813559322034e-05,
"loss": 1.4723,
"step": 5620
},
{
"epoch": 0.14,
"grad_norm": 2.859321117401123,
"learning_rate": 1.6522033898305084e-05,
"loss": 1.4444,
"step": 5630
},
{
"epoch": 0.14,
"grad_norm": 5.158230304718018,
"learning_rate": 1.6515254237288137e-05,
"loss": 1.6279,
"step": 5640
},
{
"epoch": 0.14,
"grad_norm": 2.160407781600952,
"learning_rate": 1.6508474576271187e-05,
"loss": 1.4487,
"step": 5650
},
{
"epoch": 0.14,
"grad_norm": 2.0725409984588623,
"learning_rate": 1.650169491525424e-05,
"loss": 1.4906,
"step": 5660
},
{
"epoch": 0.14,
"grad_norm": 1.5659565925598145,
"learning_rate": 1.649491525423729e-05,
"loss": 1.4545,
"step": 5670
},
{
"epoch": 0.14,
"grad_norm": 8.198447227478027,
"learning_rate": 1.648813559322034e-05,
"loss": 1.5087,
"step": 5680
},
{
"epoch": 0.14,
"grad_norm": 3.8482344150543213,
"learning_rate": 1.6481355932203392e-05,
"loss": 1.5596,
"step": 5690
},
{
"epoch": 0.14,
"grad_norm": 3.828826665878296,
"learning_rate": 1.6474576271186442e-05,
"loss": 1.5452,
"step": 5700
},
{
"epoch": 0.14,
"grad_norm": 9.930662155151367,
"learning_rate": 1.646779661016949e-05,
"loss": 1.4667,
"step": 5710
},
{
"epoch": 0.14,
"grad_norm": 2.9717953205108643,
"learning_rate": 1.6461016949152545e-05,
"loss": 1.5651,
"step": 5720
},
{
"epoch": 0.14,
"grad_norm": 1.6129778623580933,
"learning_rate": 1.6454237288135594e-05,
"loss": 1.3477,
"step": 5730
},
{
"epoch": 0.14,
"grad_norm": 2.341064214706421,
"learning_rate": 1.6447457627118648e-05,
"loss": 1.5159,
"step": 5740
},
{
"epoch": 0.14,
"grad_norm": 4.6363205909729,
"learning_rate": 1.6440677966101697e-05,
"loss": 1.5524,
"step": 5750
},
{
"epoch": 0.14,
"grad_norm": 4.433897495269775,
"learning_rate": 1.6433898305084747e-05,
"loss": 1.46,
"step": 5760
},
{
"epoch": 0.14,
"grad_norm": 3.2978014945983887,
"learning_rate": 1.64271186440678e-05,
"loss": 1.5847,
"step": 5770
},
{
"epoch": 0.14,
"grad_norm": 3.325284004211426,
"learning_rate": 1.642033898305085e-05,
"loss": 1.3553,
"step": 5780
},
{
"epoch": 0.14,
"grad_norm": 1.9704495668411255,
"learning_rate": 1.64135593220339e-05,
"loss": 1.3661,
"step": 5790
},
{
"epoch": 0.14,
"grad_norm": 1.6186383962631226,
"learning_rate": 1.640677966101695e-05,
"loss": 1.5538,
"step": 5800
},
{
"epoch": 0.15,
"grad_norm": 2.1665966510772705,
"learning_rate": 1.64e-05,
"loss": 1.5188,
"step": 5810
},
{
"epoch": 0.15,
"grad_norm": 7.165323734283447,
"learning_rate": 1.6393220338983052e-05,
"loss": 1.627,
"step": 5820
},
{
"epoch": 0.15,
"grad_norm": 7.434396743774414,
"learning_rate": 1.63864406779661e-05,
"loss": 1.4751,
"step": 5830
},
{
"epoch": 0.15,
"grad_norm": 2.549182176589966,
"learning_rate": 1.6379661016949155e-05,
"loss": 1.5013,
"step": 5840
},
{
"epoch": 0.15,
"grad_norm": 2.3907253742218018,
"learning_rate": 1.6372881355932204e-05,
"loss": 1.5532,
"step": 5850
},
{
"epoch": 0.15,
"grad_norm": 2.3262014389038086,
"learning_rate": 1.6366101694915254e-05,
"loss": 1.4404,
"step": 5860
},
{
"epoch": 0.15,
"grad_norm": 2.387146234512329,
"learning_rate": 1.6359322033898307e-05,
"loss": 1.4057,
"step": 5870
},
{
"epoch": 0.15,
"grad_norm": 4.461369514465332,
"learning_rate": 1.6352542372881357e-05,
"loss": 1.4003,
"step": 5880
},
{
"epoch": 0.15,
"grad_norm": 5.3152384757995605,
"learning_rate": 1.6345762711864407e-05,
"loss": 1.6658,
"step": 5890
},
{
"epoch": 0.15,
"grad_norm": 3.839587450027466,
"learning_rate": 1.633898305084746e-05,
"loss": 1.6396,
"step": 5900
},
{
"epoch": 0.15,
"grad_norm": 1.5456769466400146,
"learning_rate": 1.633220338983051e-05,
"loss": 1.4543,
"step": 5910
},
{
"epoch": 0.15,
"grad_norm": 1.218180775642395,
"learning_rate": 1.6325423728813563e-05,
"loss": 1.6549,
"step": 5920
},
{
"epoch": 0.15,
"grad_norm": 3.9048516750335693,
"learning_rate": 1.6318644067796612e-05,
"loss": 1.4558,
"step": 5930
},
{
"epoch": 0.15,
"grad_norm": 2.0440735816955566,
"learning_rate": 1.6311864406779662e-05,
"loss": 1.392,
"step": 5940
},
{
"epoch": 0.15,
"grad_norm": 5.26229190826416,
"learning_rate": 1.6305084745762715e-05,
"loss": 1.5081,
"step": 5950
},
{
"epoch": 0.15,
"grad_norm": 2.3289761543273926,
"learning_rate": 1.6298305084745765e-05,
"loss": 1.3642,
"step": 5960
},
{
"epoch": 0.15,
"grad_norm": 2.584726333618164,
"learning_rate": 1.6291525423728814e-05,
"loss": 1.3108,
"step": 5970
},
{
"epoch": 0.15,
"grad_norm": 3.9245219230651855,
"learning_rate": 1.6284745762711868e-05,
"loss": 1.5209,
"step": 5980
},
{
"epoch": 0.15,
"grad_norm": 1.5690237283706665,
"learning_rate": 1.6277966101694917e-05,
"loss": 1.4554,
"step": 5990
},
{
"epoch": 0.15,
"grad_norm": 2.7308597564697266,
"learning_rate": 1.6271186440677967e-05,
"loss": 1.3758,
"step": 6000
},
{
"epoch": 0.15,
"eval_loss": 1.4649540185928345,
"eval_runtime": 66.131,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 6000
},
{
"epoch": 0.15,
"grad_norm": 3.45599627494812,
"learning_rate": 1.6264406779661017e-05,
"loss": 1.4816,
"step": 6010
},
{
"epoch": 0.15,
"grad_norm": 2.358900785446167,
"learning_rate": 1.625762711864407e-05,
"loss": 1.5672,
"step": 6020
},
{
"epoch": 0.15,
"grad_norm": 1.577291488647461,
"learning_rate": 1.625084745762712e-05,
"loss": 1.536,
"step": 6030
},
{
"epoch": 0.15,
"grad_norm": 1.5293314456939697,
"learning_rate": 1.624406779661017e-05,
"loss": 1.3852,
"step": 6040
},
{
"epoch": 0.15,
"grad_norm": 3.6409552097320557,
"learning_rate": 1.6237288135593222e-05,
"loss": 1.3703,
"step": 6050
},
{
"epoch": 0.15,
"grad_norm": 1.245363712310791,
"learning_rate": 1.6230508474576272e-05,
"loss": 1.4489,
"step": 6060
},
{
"epoch": 0.15,
"grad_norm": 3.5935614109039307,
"learning_rate": 1.6223728813559322e-05,
"loss": 1.3177,
"step": 6070
},
{
"epoch": 0.15,
"grad_norm": 3.89214825630188,
"learning_rate": 1.6216949152542375e-05,
"loss": 1.6123,
"step": 6080
},
{
"epoch": 0.15,
"grad_norm": 2.6698529720306396,
"learning_rate": 1.6210169491525424e-05,
"loss": 1.4598,
"step": 6090
},
{
"epoch": 0.15,
"grad_norm": 4.818114757537842,
"learning_rate": 1.6203389830508474e-05,
"loss": 1.3712,
"step": 6100
},
{
"epoch": 0.15,
"grad_norm": 5.041522979736328,
"learning_rate": 1.6196610169491527e-05,
"loss": 1.5556,
"step": 6110
},
{
"epoch": 0.15,
"grad_norm": 2.3514511585235596,
"learning_rate": 1.6189830508474577e-05,
"loss": 1.5359,
"step": 6120
},
{
"epoch": 0.15,
"grad_norm": 1.9724153280258179,
"learning_rate": 1.618305084745763e-05,
"loss": 1.5191,
"step": 6130
},
{
"epoch": 0.15,
"grad_norm": 3.0398383140563965,
"learning_rate": 1.617627118644068e-05,
"loss": 1.4751,
"step": 6140
},
{
"epoch": 0.15,
"grad_norm": 1.3548771142959595,
"learning_rate": 1.616949152542373e-05,
"loss": 1.4159,
"step": 6150
},
{
"epoch": 0.15,
"grad_norm": 3.616636276245117,
"learning_rate": 1.6162711864406783e-05,
"loss": 1.4667,
"step": 6160
},
{
"epoch": 0.15,
"grad_norm": 3.348968744277954,
"learning_rate": 1.6155932203389832e-05,
"loss": 1.4056,
"step": 6170
},
{
"epoch": 0.15,
"grad_norm": 2.5771777629852295,
"learning_rate": 1.6149152542372882e-05,
"loss": 1.4256,
"step": 6180
},
{
"epoch": 0.15,
"grad_norm": 3.0773298740386963,
"learning_rate": 1.6142372881355935e-05,
"loss": 1.2936,
"step": 6190
},
{
"epoch": 0.15,
"grad_norm": 2.114321708679199,
"learning_rate": 1.6135593220338985e-05,
"loss": 1.5362,
"step": 6200
},
{
"epoch": 0.16,
"grad_norm": 3.142282485961914,
"learning_rate": 1.6128813559322038e-05,
"loss": 1.506,
"step": 6210
},
{
"epoch": 0.16,
"grad_norm": 5.1931562423706055,
"learning_rate": 1.6122033898305084e-05,
"loss": 1.581,
"step": 6220
},
{
"epoch": 0.16,
"grad_norm": 3.0442159175872803,
"learning_rate": 1.6115254237288137e-05,
"loss": 1.4553,
"step": 6230
},
{
"epoch": 0.16,
"grad_norm": 4.841310501098633,
"learning_rate": 1.6108474576271187e-05,
"loss": 1.5559,
"step": 6240
},
{
"epoch": 0.16,
"grad_norm": 4.651193141937256,
"learning_rate": 1.6101694915254237e-05,
"loss": 1.3535,
"step": 6250
},
{
"epoch": 0.16,
"grad_norm": 3.571608543395996,
"learning_rate": 1.609491525423729e-05,
"loss": 1.5062,
"step": 6260
},
{
"epoch": 0.16,
"grad_norm": 6.008980751037598,
"learning_rate": 1.608813559322034e-05,
"loss": 1.4696,
"step": 6270
},
{
"epoch": 0.16,
"grad_norm": 11.914450645446777,
"learning_rate": 1.608135593220339e-05,
"loss": 1.4643,
"step": 6280
},
{
"epoch": 0.16,
"grad_norm": 1.8239995241165161,
"learning_rate": 1.6074576271186442e-05,
"loss": 1.6175,
"step": 6290
},
{
"epoch": 0.16,
"grad_norm": 2.543314218521118,
"learning_rate": 1.6067796610169492e-05,
"loss": 1.4362,
"step": 6300
},
{
"epoch": 0.16,
"grad_norm": 3.3625097274780273,
"learning_rate": 1.6061016949152545e-05,
"loss": 1.5615,
"step": 6310
},
{
"epoch": 0.16,
"grad_norm": 4.757710933685303,
"learning_rate": 1.6054237288135595e-05,
"loss": 1.4382,
"step": 6320
},
{
"epoch": 0.16,
"grad_norm": 2.5102877616882324,
"learning_rate": 1.6047457627118645e-05,
"loss": 1.5367,
"step": 6330
},
{
"epoch": 0.16,
"grad_norm": 3.226902484893799,
"learning_rate": 1.6040677966101698e-05,
"loss": 1.7053,
"step": 6340
},
{
"epoch": 0.16,
"grad_norm": 2.9312872886657715,
"learning_rate": 1.6033898305084747e-05,
"loss": 1.5362,
"step": 6350
},
{
"epoch": 0.16,
"grad_norm": 1.0346354246139526,
"learning_rate": 1.6027118644067797e-05,
"loss": 1.6161,
"step": 6360
},
{
"epoch": 0.16,
"grad_norm": 1.9462825059890747,
"learning_rate": 1.602033898305085e-05,
"loss": 1.4455,
"step": 6370
},
{
"epoch": 0.16,
"grad_norm": 1.9478490352630615,
"learning_rate": 1.60135593220339e-05,
"loss": 1.5183,
"step": 6380
},
{
"epoch": 0.16,
"grad_norm": 2.781726598739624,
"learning_rate": 1.600677966101695e-05,
"loss": 1.3517,
"step": 6390
},
{
"epoch": 0.16,
"grad_norm": 2.242448091506958,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.6257,
"step": 6400
},
{
"epoch": 0.16,
"grad_norm": 1.3541291952133179,
"learning_rate": 1.5993220338983052e-05,
"loss": 1.4108,
"step": 6410
},
{
"epoch": 0.16,
"grad_norm": 2.1203219890594482,
"learning_rate": 1.5986440677966105e-05,
"loss": 1.5363,
"step": 6420
},
{
"epoch": 0.16,
"grad_norm": 2.9840617179870605,
"learning_rate": 1.5979661016949152e-05,
"loss": 1.4282,
"step": 6430
},
{
"epoch": 0.16,
"grad_norm": 3.6846158504486084,
"learning_rate": 1.5972881355932205e-05,
"loss": 1.4535,
"step": 6440
},
{
"epoch": 0.16,
"grad_norm": 1.272042155265808,
"learning_rate": 1.5966101694915255e-05,
"loss": 1.5842,
"step": 6450
},
{
"epoch": 0.16,
"grad_norm": 2.9574079513549805,
"learning_rate": 1.5959322033898304e-05,
"loss": 1.3752,
"step": 6460
},
{
"epoch": 0.16,
"grad_norm": 2.928889751434326,
"learning_rate": 1.5952542372881357e-05,
"loss": 1.4675,
"step": 6470
},
{
"epoch": 0.16,
"grad_norm": 5.292194843292236,
"learning_rate": 1.5945762711864407e-05,
"loss": 1.5172,
"step": 6480
},
{
"epoch": 0.16,
"grad_norm": 3.3911092281341553,
"learning_rate": 1.593898305084746e-05,
"loss": 1.467,
"step": 6490
},
{
"epoch": 0.16,
"grad_norm": 4.30944299697876,
"learning_rate": 1.593220338983051e-05,
"loss": 1.3515,
"step": 6500
},
{
"epoch": 0.16,
"eval_loss": 1.465691089630127,
"eval_runtime": 66.1688,
"eval_samples_per_second": 15.113,
"eval_steps_per_second": 15.113,
"step": 6500
},
{
"epoch": 0.16,
"grad_norm": 2.5098283290863037,
"learning_rate": 1.592542372881356e-05,
"loss": 1.5554,
"step": 6510
},
{
"epoch": 0.16,
"grad_norm": 7.33607292175293,
"learning_rate": 1.5918644067796613e-05,
"loss": 1.6483,
"step": 6520
},
{
"epoch": 0.16,
"grad_norm": 3.0138282775878906,
"learning_rate": 1.5911864406779662e-05,
"loss": 1.5784,
"step": 6530
},
{
"epoch": 0.16,
"grad_norm": 3.527956008911133,
"learning_rate": 1.5905084745762712e-05,
"loss": 1.265,
"step": 6540
},
{
"epoch": 0.16,
"grad_norm": 3.4589829444885254,
"learning_rate": 1.5898305084745765e-05,
"loss": 1.6605,
"step": 6550
},
{
"epoch": 0.16,
"grad_norm": 5.758669853210449,
"learning_rate": 1.5891525423728815e-05,
"loss": 1.2722,
"step": 6560
},
{
"epoch": 0.16,
"grad_norm": 2.0331597328186035,
"learning_rate": 1.5884745762711865e-05,
"loss": 1.5016,
"step": 6570
},
{
"epoch": 0.16,
"grad_norm": 7.730234622955322,
"learning_rate": 1.5877966101694918e-05,
"loss": 1.6401,
"step": 6580
},
{
"epoch": 0.16,
"grad_norm": 1.977071762084961,
"learning_rate": 1.5871186440677967e-05,
"loss": 1.3554,
"step": 6590
},
{
"epoch": 0.17,
"grad_norm": 3.8177216053009033,
"learning_rate": 1.586440677966102e-05,
"loss": 1.6109,
"step": 6600
},
{
"epoch": 0.17,
"grad_norm": 4.232168197631836,
"learning_rate": 1.585762711864407e-05,
"loss": 1.3901,
"step": 6610
},
{
"epoch": 0.17,
"grad_norm": 6.004239082336426,
"learning_rate": 1.585084745762712e-05,
"loss": 1.648,
"step": 6620
},
{
"epoch": 0.17,
"grad_norm": 4.166013240814209,
"learning_rate": 1.584406779661017e-05,
"loss": 1.4462,
"step": 6630
},
{
"epoch": 0.17,
"grad_norm": 2.389451503753662,
"learning_rate": 1.583728813559322e-05,
"loss": 1.5163,
"step": 6640
},
{
"epoch": 0.17,
"grad_norm": 2.042405843734741,
"learning_rate": 1.5830508474576272e-05,
"loss": 1.4931,
"step": 6650
},
{
"epoch": 0.17,
"grad_norm": 2.3263092041015625,
"learning_rate": 1.5823728813559322e-05,
"loss": 1.4758,
"step": 6660
},
{
"epoch": 0.17,
"grad_norm": 5.392455577850342,
"learning_rate": 1.5816949152542372e-05,
"loss": 1.6324,
"step": 6670
},
{
"epoch": 0.17,
"grad_norm": 4.789013385772705,
"learning_rate": 1.5810169491525425e-05,
"loss": 1.5744,
"step": 6680
},
{
"epoch": 0.17,
"grad_norm": 2.284268617630005,
"learning_rate": 1.5803389830508475e-05,
"loss": 1.6875,
"step": 6690
},
{
"epoch": 0.17,
"grad_norm": 2.3472728729248047,
"learning_rate": 1.5796610169491528e-05,
"loss": 1.5855,
"step": 6700
},
{
"epoch": 0.17,
"grad_norm": 4.676074981689453,
"learning_rate": 1.5789830508474577e-05,
"loss": 1.4166,
"step": 6710
},
{
"epoch": 0.17,
"grad_norm": 3.129818916320801,
"learning_rate": 1.5783050847457627e-05,
"loss": 1.3874,
"step": 6720
},
{
"epoch": 0.17,
"grad_norm": 4.993739604949951,
"learning_rate": 1.577627118644068e-05,
"loss": 1.4709,
"step": 6730
},
{
"epoch": 0.17,
"grad_norm": 2.868453025817871,
"learning_rate": 1.576949152542373e-05,
"loss": 1.5804,
"step": 6740
},
{
"epoch": 0.17,
"grad_norm": 1.086580753326416,
"learning_rate": 1.576271186440678e-05,
"loss": 1.3879,
"step": 6750
},
{
"epoch": 0.17,
"grad_norm": 5.881350994110107,
"learning_rate": 1.5755932203389833e-05,
"loss": 1.4923,
"step": 6760
},
{
"epoch": 0.17,
"grad_norm": 3.4320054054260254,
"learning_rate": 1.5749152542372882e-05,
"loss": 1.401,
"step": 6770
},
{
"epoch": 0.17,
"grad_norm": 2.2193899154663086,
"learning_rate": 1.5742372881355936e-05,
"loss": 1.5497,
"step": 6780
},
{
"epoch": 0.17,
"grad_norm": 3.470163583755493,
"learning_rate": 1.5735593220338985e-05,
"loss": 1.4824,
"step": 6790
},
{
"epoch": 0.17,
"grad_norm": 3.5179107189178467,
"learning_rate": 1.5728813559322035e-05,
"loss": 1.5401,
"step": 6800
},
{
"epoch": 0.17,
"grad_norm": 4.4173359870910645,
"learning_rate": 1.5722033898305088e-05,
"loss": 1.3075,
"step": 6810
},
{
"epoch": 0.17,
"grad_norm": 4.17536735534668,
"learning_rate": 1.5715254237288138e-05,
"loss": 1.5135,
"step": 6820
},
{
"epoch": 0.17,
"grad_norm": 3.278881311416626,
"learning_rate": 1.5708474576271187e-05,
"loss": 1.5745,
"step": 6830
},
{
"epoch": 0.17,
"grad_norm": 2.718890905380249,
"learning_rate": 1.5701694915254237e-05,
"loss": 1.5397,
"step": 6840
},
{
"epoch": 0.17,
"grad_norm": 1.8117082118988037,
"learning_rate": 1.5694915254237287e-05,
"loss": 1.5904,
"step": 6850
},
{
"epoch": 0.17,
"grad_norm": 1.4601502418518066,
"learning_rate": 1.568813559322034e-05,
"loss": 1.5485,
"step": 6860
},
{
"epoch": 0.17,
"grad_norm": 2.0269432067871094,
"learning_rate": 1.568135593220339e-05,
"loss": 1.4487,
"step": 6870
},
{
"epoch": 0.17,
"grad_norm": 3.047973394393921,
"learning_rate": 1.5674576271186443e-05,
"loss": 1.2999,
"step": 6880
},
{
"epoch": 0.17,
"grad_norm": 2.7541468143463135,
"learning_rate": 1.5667796610169492e-05,
"loss": 1.4651,
"step": 6890
},
{
"epoch": 0.17,
"grad_norm": 5.5456767082214355,
"learning_rate": 1.5661016949152542e-05,
"loss": 1.4433,
"step": 6900
},
{
"epoch": 0.17,
"grad_norm": 3.4916276931762695,
"learning_rate": 1.5654237288135595e-05,
"loss": 1.5764,
"step": 6910
},
{
"epoch": 0.17,
"grad_norm": 2.6278750896453857,
"learning_rate": 1.5647457627118645e-05,
"loss": 1.3208,
"step": 6920
},
{
"epoch": 0.17,
"grad_norm": 2.9469902515411377,
"learning_rate": 1.5640677966101695e-05,
"loss": 1.6433,
"step": 6930
},
{
"epoch": 0.17,
"grad_norm": 5.235908031463623,
"learning_rate": 1.5633898305084748e-05,
"loss": 1.2608,
"step": 6940
},
{
"epoch": 0.17,
"grad_norm": 1.5849263668060303,
"learning_rate": 1.5627118644067798e-05,
"loss": 1.5088,
"step": 6950
},
{
"epoch": 0.17,
"grad_norm": 3.6014277935028076,
"learning_rate": 1.562033898305085e-05,
"loss": 1.3976,
"step": 6960
},
{
"epoch": 0.17,
"grad_norm": 3.220320701599121,
"learning_rate": 1.56135593220339e-05,
"loss": 1.5011,
"step": 6970
},
{
"epoch": 0.17,
"grad_norm": 3.434382915496826,
"learning_rate": 1.560677966101695e-05,
"loss": 1.4733,
"step": 6980
},
{
"epoch": 0.17,
"grad_norm": 4.612408638000488,
"learning_rate": 1.5600000000000003e-05,
"loss": 1.4618,
"step": 6990
},
{
"epoch": 0.17,
"grad_norm": 2.9301650524139404,
"learning_rate": 1.5593220338983053e-05,
"loss": 1.4165,
"step": 7000
},
{
"epoch": 0.17,
"eval_loss": 1.469406247138977,
"eval_runtime": 66.2443,
"eval_samples_per_second": 15.096,
"eval_steps_per_second": 15.096,
"step": 7000
},
{
"epoch": 0.18,
"grad_norm": 5.865874767303467,
"learning_rate": 1.5586440677966103e-05,
"loss": 1.3156,
"step": 7010
},
{
"epoch": 0.18,
"grad_norm": 5.411038875579834,
"learning_rate": 1.5579661016949156e-05,
"loss": 1.4303,
"step": 7020
},
{
"epoch": 0.18,
"grad_norm": 2.869673728942871,
"learning_rate": 1.5572881355932205e-05,
"loss": 1.5055,
"step": 7030
},
{
"epoch": 0.18,
"grad_norm": 1.0629090070724487,
"learning_rate": 1.5566101694915255e-05,
"loss": 1.6094,
"step": 7040
},
{
"epoch": 0.18,
"grad_norm": 2.271601676940918,
"learning_rate": 1.5559322033898305e-05,
"loss": 1.4648,
"step": 7050
},
{
"epoch": 0.18,
"grad_norm": 1.5404294729232788,
"learning_rate": 1.5552542372881358e-05,
"loss": 1.3793,
"step": 7060
},
{
"epoch": 0.18,
"grad_norm": 3.9599192142486572,
"learning_rate": 1.5545762711864408e-05,
"loss": 1.3563,
"step": 7070
},
{
"epoch": 0.18,
"grad_norm": 6.35668420791626,
"learning_rate": 1.5538983050847457e-05,
"loss": 1.4637,
"step": 7080
},
{
"epoch": 0.18,
"grad_norm": 7.315426349639893,
"learning_rate": 1.553220338983051e-05,
"loss": 1.3284,
"step": 7090
},
{
"epoch": 0.18,
"grad_norm": 1.8330402374267578,
"learning_rate": 1.552542372881356e-05,
"loss": 1.3762,
"step": 7100
},
{
"epoch": 0.18,
"grad_norm": 3.072326183319092,
"learning_rate": 1.551864406779661e-05,
"loss": 1.4194,
"step": 7110
},
{
"epoch": 0.18,
"grad_norm": 1.952779769897461,
"learning_rate": 1.5511864406779663e-05,
"loss": 1.4886,
"step": 7120
},
{
"epoch": 0.18,
"grad_norm": 4.332074165344238,
"learning_rate": 1.5505084745762713e-05,
"loss": 1.4912,
"step": 7130
},
{
"epoch": 0.18,
"grad_norm": 3.962344169616699,
"learning_rate": 1.5498305084745762e-05,
"loss": 1.5502,
"step": 7140
},
{
"epoch": 0.18,
"grad_norm": 7.930359363555908,
"learning_rate": 1.5491525423728815e-05,
"loss": 1.471,
"step": 7150
},
{
"epoch": 0.18,
"grad_norm": 4.925573825836182,
"learning_rate": 1.5484745762711865e-05,
"loss": 1.3155,
"step": 7160
},
{
"epoch": 0.18,
"grad_norm": 3.2943711280822754,
"learning_rate": 1.5477966101694918e-05,
"loss": 1.3094,
"step": 7170
},
{
"epoch": 0.18,
"grad_norm": 1.9433196783065796,
"learning_rate": 1.5471186440677968e-05,
"loss": 1.3453,
"step": 7180
},
{
"epoch": 0.18,
"grad_norm": 1.8384184837341309,
"learning_rate": 1.5464406779661018e-05,
"loss": 1.3788,
"step": 7190
},
{
"epoch": 0.18,
"grad_norm": 2.92781138420105,
"learning_rate": 1.545762711864407e-05,
"loss": 1.5704,
"step": 7200
},
{
"epoch": 0.18,
"grad_norm": 1.0172346830368042,
"learning_rate": 1.545084745762712e-05,
"loss": 1.5442,
"step": 7210
},
{
"epoch": 0.18,
"grad_norm": 1.0915580987930298,
"learning_rate": 1.544406779661017e-05,
"loss": 1.5226,
"step": 7220
},
{
"epoch": 0.18,
"grad_norm": 9.170607566833496,
"learning_rate": 1.5437288135593223e-05,
"loss": 1.5066,
"step": 7230
},
{
"epoch": 0.18,
"grad_norm": 4.84716796875,
"learning_rate": 1.5430508474576273e-05,
"loss": 1.4508,
"step": 7240
},
{
"epoch": 0.18,
"grad_norm": 5.949992656707764,
"learning_rate": 1.5423728813559326e-05,
"loss": 1.3393,
"step": 7250
},
{
"epoch": 0.18,
"grad_norm": 5.912881851196289,
"learning_rate": 1.5416949152542372e-05,
"loss": 1.5918,
"step": 7260
},
{
"epoch": 0.18,
"grad_norm": 1.8128445148468018,
"learning_rate": 1.5410169491525425e-05,
"loss": 1.5794,
"step": 7270
},
{
"epoch": 0.18,
"grad_norm": 3.9023091793060303,
"learning_rate": 1.5403389830508475e-05,
"loss": 1.345,
"step": 7280
},
{
"epoch": 0.18,
"grad_norm": 2.263958692550659,
"learning_rate": 1.5396610169491525e-05,
"loss": 1.3651,
"step": 7290
},
{
"epoch": 0.18,
"grad_norm": 8.103110313415527,
"learning_rate": 1.5389830508474578e-05,
"loss": 1.4784,
"step": 7300
},
{
"epoch": 0.18,
"grad_norm": 4.5859456062316895,
"learning_rate": 1.5383050847457628e-05,
"loss": 1.4087,
"step": 7310
},
{
"epoch": 0.18,
"grad_norm": 5.6322503089904785,
"learning_rate": 1.5376271186440677e-05,
"loss": 1.518,
"step": 7320
},
{
"epoch": 0.18,
"grad_norm": 2.4485771656036377,
"learning_rate": 1.536949152542373e-05,
"loss": 1.3471,
"step": 7330
},
{
"epoch": 0.18,
"grad_norm": 2.0619120597839355,
"learning_rate": 1.536271186440678e-05,
"loss": 1.2563,
"step": 7340
},
{
"epoch": 0.18,
"grad_norm": 4.037415027618408,
"learning_rate": 1.5355932203389833e-05,
"loss": 1.4414,
"step": 7350
},
{
"epoch": 0.18,
"grad_norm": 2.4679503440856934,
"learning_rate": 1.5349152542372883e-05,
"loss": 1.4021,
"step": 7360
},
{
"epoch": 0.18,
"grad_norm": 1.4774163961410522,
"learning_rate": 1.5342372881355933e-05,
"loss": 1.4331,
"step": 7370
},
{
"epoch": 0.18,
"grad_norm": 4.21893310546875,
"learning_rate": 1.5335593220338986e-05,
"loss": 1.4847,
"step": 7380
},
{
"epoch": 0.18,
"grad_norm": 4.261248588562012,
"learning_rate": 1.5328813559322035e-05,
"loss": 1.4098,
"step": 7390
},
{
"epoch": 0.18,
"grad_norm": 2.419386625289917,
"learning_rate": 1.5322033898305085e-05,
"loss": 1.4635,
"step": 7400
},
{
"epoch": 0.19,
"grad_norm": 3.3119351863861084,
"learning_rate": 1.5315254237288138e-05,
"loss": 1.3862,
"step": 7410
},
{
"epoch": 0.19,
"grad_norm": 4.350612163543701,
"learning_rate": 1.5308474576271188e-05,
"loss": 1.5094,
"step": 7420
},
{
"epoch": 0.19,
"grad_norm": 3.089937925338745,
"learning_rate": 1.530169491525424e-05,
"loss": 1.6249,
"step": 7430
},
{
"epoch": 0.19,
"grad_norm": 6.980156898498535,
"learning_rate": 1.529491525423729e-05,
"loss": 1.6102,
"step": 7440
},
{
"epoch": 0.19,
"grad_norm": 3.881622552871704,
"learning_rate": 1.528813559322034e-05,
"loss": 1.5149,
"step": 7450
},
{
"epoch": 0.19,
"grad_norm": 2.8510727882385254,
"learning_rate": 1.528135593220339e-05,
"loss": 1.4542,
"step": 7460
},
{
"epoch": 0.19,
"grad_norm": 4.4826130867004395,
"learning_rate": 1.527457627118644e-05,
"loss": 1.5085,
"step": 7470
},
{
"epoch": 0.19,
"grad_norm": 4.991846084594727,
"learning_rate": 1.5267796610169493e-05,
"loss": 1.4761,
"step": 7480
},
{
"epoch": 0.19,
"grad_norm": 4.136972904205322,
"learning_rate": 1.5261016949152543e-05,
"loss": 1.4851,
"step": 7490
},
{
"epoch": 0.19,
"grad_norm": 3.58878231048584,
"learning_rate": 1.5254237288135594e-05,
"loss": 1.4117,
"step": 7500
},
{
"epoch": 0.19,
"eval_loss": 1.437897801399231,
"eval_runtime": 66.1821,
"eval_samples_per_second": 15.11,
"eval_steps_per_second": 15.11,
"step": 7500
},
{
"epoch": 0.19,
"grad_norm": 2.952547311782837,
"learning_rate": 1.5247457627118645e-05,
"loss": 1.4895,
"step": 7510
},
{
"epoch": 0.19,
"grad_norm": 1.617139458656311,
"learning_rate": 1.5240677966101695e-05,
"loss": 1.3878,
"step": 7520
},
{
"epoch": 0.19,
"grad_norm": 1.6708911657333374,
"learning_rate": 1.5233898305084747e-05,
"loss": 1.5612,
"step": 7530
},
{
"epoch": 0.19,
"grad_norm": 1.824523687362671,
"learning_rate": 1.5227118644067798e-05,
"loss": 1.407,
"step": 7540
},
{
"epoch": 0.19,
"grad_norm": 3.4340052604675293,
"learning_rate": 1.522033898305085e-05,
"loss": 1.2622,
"step": 7550
},
{
"epoch": 0.19,
"grad_norm": 3.8234424591064453,
"learning_rate": 1.5213559322033899e-05,
"loss": 1.5187,
"step": 7560
},
{
"epoch": 0.19,
"grad_norm": 2.398066520690918,
"learning_rate": 1.520677966101695e-05,
"loss": 1.4119,
"step": 7570
},
{
"epoch": 0.19,
"grad_norm": 4.461039066314697,
"learning_rate": 1.5200000000000002e-05,
"loss": 1.6286,
"step": 7580
},
{
"epoch": 0.19,
"grad_norm": 8.626133918762207,
"learning_rate": 1.5193220338983052e-05,
"loss": 1.4452,
"step": 7590
},
{
"epoch": 0.19,
"grad_norm": 4.230071067810059,
"learning_rate": 1.5186440677966103e-05,
"loss": 1.3512,
"step": 7600
},
{
"epoch": 0.19,
"grad_norm": 4.9356279373168945,
"learning_rate": 1.5179661016949154e-05,
"loss": 1.5673,
"step": 7610
},
{
"epoch": 0.19,
"grad_norm": 2.832441806793213,
"learning_rate": 1.5172881355932206e-05,
"loss": 1.3301,
"step": 7620
},
{
"epoch": 0.19,
"grad_norm": 3.501631259918213,
"learning_rate": 1.5166101694915255e-05,
"loss": 1.4446,
"step": 7630
},
{
"epoch": 0.19,
"grad_norm": 1.15791916847229,
"learning_rate": 1.5159322033898307e-05,
"loss": 1.4241,
"step": 7640
},
{
"epoch": 0.19,
"grad_norm": 2.7517549991607666,
"learning_rate": 1.5152542372881358e-05,
"loss": 1.5962,
"step": 7650
},
{
"epoch": 0.19,
"grad_norm": 2.9269521236419678,
"learning_rate": 1.514576271186441e-05,
"loss": 1.3783,
"step": 7660
},
{
"epoch": 0.19,
"grad_norm": 3.538501501083374,
"learning_rate": 1.5138983050847458e-05,
"loss": 1.6107,
"step": 7670
},
{
"epoch": 0.19,
"grad_norm": 9.009359359741211,
"learning_rate": 1.5132203389830509e-05,
"loss": 1.5968,
"step": 7680
},
{
"epoch": 0.19,
"grad_norm": 2.63154673576355,
"learning_rate": 1.512542372881356e-05,
"loss": 1.5102,
"step": 7690
},
{
"epoch": 0.19,
"grad_norm": 3.5580034255981445,
"learning_rate": 1.511864406779661e-05,
"loss": 1.2809,
"step": 7700
},
{
"epoch": 0.19,
"grad_norm": 5.356471061706543,
"learning_rate": 1.5111864406779662e-05,
"loss": 1.214,
"step": 7710
},
{
"epoch": 0.19,
"grad_norm": 4.988667964935303,
"learning_rate": 1.5105084745762713e-05,
"loss": 1.4744,
"step": 7720
},
{
"epoch": 0.19,
"grad_norm": 3.1577963829040527,
"learning_rate": 1.5098305084745763e-05,
"loss": 1.5062,
"step": 7730
},
{
"epoch": 0.19,
"grad_norm": 4.623036861419678,
"learning_rate": 1.5091525423728814e-05,
"loss": 1.5103,
"step": 7740
},
{
"epoch": 0.19,
"grad_norm": 2.3117690086364746,
"learning_rate": 1.5084745762711865e-05,
"loss": 1.4344,
"step": 7750
},
{
"epoch": 0.19,
"grad_norm": 2.289649248123169,
"learning_rate": 1.5077966101694917e-05,
"loss": 1.5636,
"step": 7760
},
{
"epoch": 0.19,
"grad_norm": 1.5120543241500854,
"learning_rate": 1.5071186440677967e-05,
"loss": 1.4689,
"step": 7770
},
{
"epoch": 0.19,
"grad_norm": 2.479252576828003,
"learning_rate": 1.5064406779661018e-05,
"loss": 1.3047,
"step": 7780
},
{
"epoch": 0.19,
"grad_norm": 4.039693832397461,
"learning_rate": 1.505762711864407e-05,
"loss": 1.3197,
"step": 7790
},
{
"epoch": 0.2,
"grad_norm": 3.129855155944824,
"learning_rate": 1.505084745762712e-05,
"loss": 1.4407,
"step": 7800
},
{
"epoch": 0.2,
"grad_norm": 4.973848819732666,
"learning_rate": 1.504406779661017e-05,
"loss": 1.3066,
"step": 7810
},
{
"epoch": 0.2,
"grad_norm": 1.673112392425537,
"learning_rate": 1.5037288135593222e-05,
"loss": 1.4593,
"step": 7820
},
{
"epoch": 0.2,
"grad_norm": 2.6259243488311768,
"learning_rate": 1.5030508474576273e-05,
"loss": 1.4993,
"step": 7830
},
{
"epoch": 0.2,
"grad_norm": 1.8319427967071533,
"learning_rate": 1.5023728813559325e-05,
"loss": 1.444,
"step": 7840
},
{
"epoch": 0.2,
"grad_norm": 3.4944241046905518,
"learning_rate": 1.5016949152542374e-05,
"loss": 1.2785,
"step": 7850
},
{
"epoch": 0.2,
"grad_norm": 2.0709774494171143,
"learning_rate": 1.5010169491525426e-05,
"loss": 1.4925,
"step": 7860
},
{
"epoch": 0.2,
"grad_norm": 1.6483750343322754,
"learning_rate": 1.5003389830508477e-05,
"loss": 1.5283,
"step": 7870
},
{
"epoch": 0.2,
"grad_norm": 4.786270618438721,
"learning_rate": 1.4996610169491525e-05,
"loss": 1.4946,
"step": 7880
},
{
"epoch": 0.2,
"grad_norm": 1.536661148071289,
"learning_rate": 1.4989830508474577e-05,
"loss": 1.4044,
"step": 7890
},
{
"epoch": 0.2,
"grad_norm": 2.717747926712036,
"learning_rate": 1.4983050847457628e-05,
"loss": 1.4569,
"step": 7900
},
{
"epoch": 0.2,
"grad_norm": 2.2604639530181885,
"learning_rate": 1.4976271186440678e-05,
"loss": 1.5154,
"step": 7910
},
{
"epoch": 0.2,
"grad_norm": 3.6678876876831055,
"learning_rate": 1.4969491525423729e-05,
"loss": 1.1938,
"step": 7920
},
{
"epoch": 0.2,
"grad_norm": 3.3134288787841797,
"learning_rate": 1.496271186440678e-05,
"loss": 1.4676,
"step": 7930
},
{
"epoch": 0.2,
"grad_norm": 1.8791242837905884,
"learning_rate": 1.4955932203389832e-05,
"loss": 1.5698,
"step": 7940
},
{
"epoch": 0.2,
"grad_norm": 3.279029369354248,
"learning_rate": 1.4949152542372882e-05,
"loss": 1.4374,
"step": 7950
},
{
"epoch": 0.2,
"grad_norm": 6.518031597137451,
"learning_rate": 1.4942372881355933e-05,
"loss": 1.3021,
"step": 7960
},
{
"epoch": 0.2,
"grad_norm": 5.85239315032959,
"learning_rate": 1.4935593220338984e-05,
"loss": 1.479,
"step": 7970
},
{
"epoch": 0.2,
"grad_norm": 5.397909641265869,
"learning_rate": 1.4928813559322036e-05,
"loss": 1.4118,
"step": 7980
},
{
"epoch": 0.2,
"grad_norm": 3.4075229167938232,
"learning_rate": 1.4922033898305086e-05,
"loss": 1.4959,
"step": 7990
},
{
"epoch": 0.2,
"grad_norm": 6.57758092880249,
"learning_rate": 1.4915254237288137e-05,
"loss": 1.6252,
"step": 8000
},
{
"epoch": 0.2,
"eval_loss": 1.4186667203903198,
"eval_runtime": 66.1276,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 8000
},
{
"epoch": 0.2,
"grad_norm": 1.395254373550415,
"learning_rate": 1.4908474576271188e-05,
"loss": 1.4234,
"step": 8010
},
{
"epoch": 0.2,
"grad_norm": 1.7249040603637695,
"learning_rate": 1.490169491525424e-05,
"loss": 1.3279,
"step": 8020
},
{
"epoch": 0.2,
"grad_norm": 6.21826171875,
"learning_rate": 1.489491525423729e-05,
"loss": 1.3013,
"step": 8030
},
{
"epoch": 0.2,
"grad_norm": 2.4089813232421875,
"learning_rate": 1.4888135593220341e-05,
"loss": 1.4544,
"step": 8040
},
{
"epoch": 0.2,
"grad_norm": 1.597097635269165,
"learning_rate": 1.4881355932203392e-05,
"loss": 1.4793,
"step": 8050
},
{
"epoch": 0.2,
"grad_norm": 1.9910588264465332,
"learning_rate": 1.4874576271186442e-05,
"loss": 1.4182,
"step": 8060
},
{
"epoch": 0.2,
"grad_norm": 4.10093355178833,
"learning_rate": 1.4867796610169493e-05,
"loss": 1.1596,
"step": 8070
},
{
"epoch": 0.2,
"grad_norm": 3.8215675354003906,
"learning_rate": 1.4861016949152545e-05,
"loss": 1.5739,
"step": 8080
},
{
"epoch": 0.2,
"grad_norm": 3.1352415084838867,
"learning_rate": 1.4854237288135593e-05,
"loss": 1.3123,
"step": 8090
},
{
"epoch": 0.2,
"grad_norm": 10.560978889465332,
"learning_rate": 1.4847457627118644e-05,
"loss": 1.5507,
"step": 8100
},
{
"epoch": 0.2,
"grad_norm": 1.808532953262329,
"learning_rate": 1.4840677966101696e-05,
"loss": 1.5942,
"step": 8110
},
{
"epoch": 0.2,
"grad_norm": 4.674659252166748,
"learning_rate": 1.4833898305084747e-05,
"loss": 1.5669,
"step": 8120
},
{
"epoch": 0.2,
"grad_norm": 3.7762465476989746,
"learning_rate": 1.4827118644067797e-05,
"loss": 1.4565,
"step": 8130
},
{
"epoch": 0.2,
"grad_norm": 8.079657554626465,
"learning_rate": 1.4820338983050848e-05,
"loss": 1.4097,
"step": 8140
},
{
"epoch": 0.2,
"grad_norm": 1.739395260810852,
"learning_rate": 1.48135593220339e-05,
"loss": 1.5126,
"step": 8150
},
{
"epoch": 0.2,
"grad_norm": 12.929269790649414,
"learning_rate": 1.4806779661016951e-05,
"loss": 1.4822,
"step": 8160
},
{
"epoch": 0.2,
"grad_norm": 2.2452526092529297,
"learning_rate": 1.48e-05,
"loss": 1.3672,
"step": 8170
},
{
"epoch": 0.2,
"grad_norm": 5.2609333992004395,
"learning_rate": 1.4793220338983052e-05,
"loss": 1.44,
"step": 8180
},
{
"epoch": 0.2,
"grad_norm": 3.3144960403442383,
"learning_rate": 1.4786440677966103e-05,
"loss": 1.4122,
"step": 8190
},
{
"epoch": 0.2,
"grad_norm": 3.837510824203491,
"learning_rate": 1.4779661016949153e-05,
"loss": 1.5105,
"step": 8200
},
{
"epoch": 0.21,
"grad_norm": 2.3328332901000977,
"learning_rate": 1.4772881355932205e-05,
"loss": 1.4721,
"step": 8210
},
{
"epoch": 0.21,
"grad_norm": 3.8686797618865967,
"learning_rate": 1.4766101694915256e-05,
"loss": 1.3236,
"step": 8220
},
{
"epoch": 0.21,
"grad_norm": 3.545306444168091,
"learning_rate": 1.4759322033898307e-05,
"loss": 1.3628,
"step": 8230
},
{
"epoch": 0.21,
"grad_norm": 5.3495965003967285,
"learning_rate": 1.4752542372881357e-05,
"loss": 1.4093,
"step": 8240
},
{
"epoch": 0.21,
"grad_norm": 2.040356159210205,
"learning_rate": 1.4745762711864408e-05,
"loss": 1.4116,
"step": 8250
},
{
"epoch": 0.21,
"grad_norm": 3.531248092651367,
"learning_rate": 1.473898305084746e-05,
"loss": 1.4185,
"step": 8260
},
{
"epoch": 0.21,
"grad_norm": 1.453710675239563,
"learning_rate": 1.4732203389830511e-05,
"loss": 1.3446,
"step": 8270
},
{
"epoch": 0.21,
"grad_norm": 5.032666206359863,
"learning_rate": 1.4725423728813561e-05,
"loss": 1.521,
"step": 8280
},
{
"epoch": 0.21,
"grad_norm": 4.052346229553223,
"learning_rate": 1.4718644067796612e-05,
"loss": 1.5021,
"step": 8290
},
{
"epoch": 0.21,
"grad_norm": 3.378077268600464,
"learning_rate": 1.4711864406779662e-05,
"loss": 1.3995,
"step": 8300
},
{
"epoch": 0.21,
"grad_norm": 3.423922061920166,
"learning_rate": 1.4705084745762712e-05,
"loss": 1.4133,
"step": 8310
},
{
"epoch": 0.21,
"grad_norm": 7.027323246002197,
"learning_rate": 1.4698305084745763e-05,
"loss": 1.2491,
"step": 8320
},
{
"epoch": 0.21,
"grad_norm": 3.2872166633605957,
"learning_rate": 1.4691525423728815e-05,
"loss": 1.1965,
"step": 8330
},
{
"epoch": 0.21,
"grad_norm": 3.879119396209717,
"learning_rate": 1.4684745762711864e-05,
"loss": 1.3579,
"step": 8340
},
{
"epoch": 0.21,
"grad_norm": 4.697524070739746,
"learning_rate": 1.4677966101694916e-05,
"loss": 1.4266,
"step": 8350
},
{
"epoch": 0.21,
"grad_norm": 5.742332458496094,
"learning_rate": 1.4671186440677967e-05,
"loss": 1.4115,
"step": 8360
},
{
"epoch": 0.21,
"grad_norm": 5.360010623931885,
"learning_rate": 1.4664406779661018e-05,
"loss": 1.443,
"step": 8370
},
{
"epoch": 0.21,
"grad_norm": 5.725327968597412,
"learning_rate": 1.4657627118644068e-05,
"loss": 1.3006,
"step": 8380
},
{
"epoch": 0.21,
"grad_norm": 3.76000714302063,
"learning_rate": 1.465084745762712e-05,
"loss": 1.317,
"step": 8390
},
{
"epoch": 0.21,
"grad_norm": 2.6450018882751465,
"learning_rate": 1.4644067796610171e-05,
"loss": 1.5148,
"step": 8400
},
{
"epoch": 0.21,
"grad_norm": 4.761094093322754,
"learning_rate": 1.4637288135593222e-05,
"loss": 1.3873,
"step": 8410
},
{
"epoch": 0.21,
"grad_norm": 3.0644989013671875,
"learning_rate": 1.4630508474576272e-05,
"loss": 1.4294,
"step": 8420
},
{
"epoch": 0.21,
"grad_norm": 2.3098878860473633,
"learning_rate": 1.4623728813559323e-05,
"loss": 1.4443,
"step": 8430
},
{
"epoch": 0.21,
"grad_norm": 4.367359638214111,
"learning_rate": 1.4616949152542375e-05,
"loss": 1.5441,
"step": 8440
},
{
"epoch": 0.21,
"grad_norm": 9.395272254943848,
"learning_rate": 1.4610169491525426e-05,
"loss": 1.4793,
"step": 8450
},
{
"epoch": 0.21,
"grad_norm": 2.6706058979034424,
"learning_rate": 1.4603389830508476e-05,
"loss": 1.3537,
"step": 8460
},
{
"epoch": 0.21,
"grad_norm": 2.5797505378723145,
"learning_rate": 1.4596610169491527e-05,
"loss": 1.3109,
"step": 8470
},
{
"epoch": 0.21,
"grad_norm": 2.9615049362182617,
"learning_rate": 1.4589830508474579e-05,
"loss": 1.4649,
"step": 8480
},
{
"epoch": 0.21,
"grad_norm": 2.650679349899292,
"learning_rate": 1.458305084745763e-05,
"loss": 1.4196,
"step": 8490
},
{
"epoch": 0.21,
"grad_norm": 1.7347412109375,
"learning_rate": 1.4576271186440678e-05,
"loss": 1.4922,
"step": 8500
},
{
"epoch": 0.21,
"eval_loss": 1.4525456428527832,
"eval_runtime": 66.1301,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 8500
},
{
"epoch": 0.21,
"grad_norm": 4.196395397186279,
"learning_rate": 1.456949152542373e-05,
"loss": 1.4091,
"step": 8510
},
{
"epoch": 0.21,
"grad_norm": 1.4905701875686646,
"learning_rate": 1.456271186440678e-05,
"loss": 1.3764,
"step": 8520
},
{
"epoch": 0.21,
"grad_norm": 6.2871174812316895,
"learning_rate": 1.455593220338983e-05,
"loss": 1.5042,
"step": 8530
},
{
"epoch": 0.21,
"grad_norm": 1.7789188623428345,
"learning_rate": 1.4549152542372882e-05,
"loss": 1.4442,
"step": 8540
},
{
"epoch": 0.21,
"grad_norm": 2.0368621349334717,
"learning_rate": 1.4542372881355933e-05,
"loss": 1.3332,
"step": 8550
},
{
"epoch": 0.21,
"grad_norm": 2.3782999515533447,
"learning_rate": 1.4535593220338983e-05,
"loss": 1.4489,
"step": 8560
},
{
"epoch": 0.21,
"grad_norm": 1.9039316177368164,
"learning_rate": 1.4528813559322035e-05,
"loss": 1.4103,
"step": 8570
},
{
"epoch": 0.21,
"grad_norm": 8.607722282409668,
"learning_rate": 1.4522033898305086e-05,
"loss": 1.494,
"step": 8580
},
{
"epoch": 0.21,
"grad_norm": 4.126497268676758,
"learning_rate": 1.4515254237288137e-05,
"loss": 1.5005,
"step": 8590
},
{
"epoch": 0.21,
"grad_norm": 2.387799024581909,
"learning_rate": 1.4508474576271187e-05,
"loss": 1.5427,
"step": 8600
},
{
"epoch": 0.22,
"grad_norm": 2.6460745334625244,
"learning_rate": 1.4501694915254239e-05,
"loss": 1.3386,
"step": 8610
},
{
"epoch": 0.22,
"grad_norm": 13.728014945983887,
"learning_rate": 1.449491525423729e-05,
"loss": 1.3242,
"step": 8620
},
{
"epoch": 0.22,
"grad_norm": 2.1503047943115234,
"learning_rate": 1.4488135593220341e-05,
"loss": 1.5502,
"step": 8630
},
{
"epoch": 0.22,
"grad_norm": 2.9898569583892822,
"learning_rate": 1.4481355932203391e-05,
"loss": 1.5844,
"step": 8640
},
{
"epoch": 0.22,
"grad_norm": 2.0317842960357666,
"learning_rate": 1.4474576271186442e-05,
"loss": 1.4781,
"step": 8650
},
{
"epoch": 0.22,
"grad_norm": 3.0708272457122803,
"learning_rate": 1.4467796610169494e-05,
"loss": 1.4521,
"step": 8660
},
{
"epoch": 0.22,
"grad_norm": 5.950216770172119,
"learning_rate": 1.4461016949152544e-05,
"loss": 1.4619,
"step": 8670
},
{
"epoch": 0.22,
"grad_norm": 3.4892992973327637,
"learning_rate": 1.4454237288135595e-05,
"loss": 1.5312,
"step": 8680
},
{
"epoch": 0.22,
"grad_norm": 2.799809694290161,
"learning_rate": 1.4447457627118646e-05,
"loss": 1.6236,
"step": 8690
},
{
"epoch": 0.22,
"grad_norm": 2.8113033771514893,
"learning_rate": 1.4440677966101698e-05,
"loss": 1.4817,
"step": 8700
},
{
"epoch": 0.22,
"grad_norm": 4.948983192443848,
"learning_rate": 1.4433898305084746e-05,
"loss": 1.4583,
"step": 8710
},
{
"epoch": 0.22,
"grad_norm": 4.663936138153076,
"learning_rate": 1.4427118644067797e-05,
"loss": 1.3879,
"step": 8720
},
{
"epoch": 0.22,
"grad_norm": 5.0930256843566895,
"learning_rate": 1.4420338983050849e-05,
"loss": 1.4138,
"step": 8730
},
{
"epoch": 0.22,
"grad_norm": 2.463430404663086,
"learning_rate": 1.4413559322033898e-05,
"loss": 1.4856,
"step": 8740
},
{
"epoch": 0.22,
"grad_norm": 3.769231081008911,
"learning_rate": 1.440677966101695e-05,
"loss": 1.5743,
"step": 8750
},
{
"epoch": 0.22,
"grad_norm": 2.731410503387451,
"learning_rate": 1.4400000000000001e-05,
"loss": 1.6308,
"step": 8760
},
{
"epoch": 0.22,
"grad_norm": 0.9978777766227722,
"learning_rate": 1.4393220338983052e-05,
"loss": 1.5564,
"step": 8770
},
{
"epoch": 0.22,
"grad_norm": 2.8928000926971436,
"learning_rate": 1.4386440677966102e-05,
"loss": 1.4164,
"step": 8780
},
{
"epoch": 0.22,
"grad_norm": 11.18360424041748,
"learning_rate": 1.4379661016949154e-05,
"loss": 1.4095,
"step": 8790
},
{
"epoch": 0.22,
"grad_norm": 2.7960269451141357,
"learning_rate": 1.4372881355932205e-05,
"loss": 1.5618,
"step": 8800
},
{
"epoch": 0.22,
"grad_norm": 3.2135298252105713,
"learning_rate": 1.4366101694915255e-05,
"loss": 1.4324,
"step": 8810
},
{
"epoch": 0.22,
"grad_norm": 1.8829147815704346,
"learning_rate": 1.4359322033898306e-05,
"loss": 1.5242,
"step": 8820
},
{
"epoch": 0.22,
"grad_norm": 3.833895206451416,
"learning_rate": 1.4352542372881357e-05,
"loss": 1.2926,
"step": 8830
},
{
"epoch": 0.22,
"grad_norm": 5.0034356117248535,
"learning_rate": 1.4345762711864409e-05,
"loss": 1.3666,
"step": 8840
},
{
"epoch": 0.22,
"grad_norm": 2.923778533935547,
"learning_rate": 1.4338983050847459e-05,
"loss": 1.4768,
"step": 8850
},
{
"epoch": 0.22,
"grad_norm": 2.227283477783203,
"learning_rate": 1.433220338983051e-05,
"loss": 1.5738,
"step": 8860
},
{
"epoch": 0.22,
"grad_norm": 5.208776473999023,
"learning_rate": 1.4325423728813561e-05,
"loss": 1.5767,
"step": 8870
},
{
"epoch": 0.22,
"grad_norm": 1.79727041721344,
"learning_rate": 1.4318644067796613e-05,
"loss": 1.3856,
"step": 8880
},
{
"epoch": 0.22,
"grad_norm": 1.7950516939163208,
"learning_rate": 1.4311864406779662e-05,
"loss": 1.4558,
"step": 8890
},
{
"epoch": 0.22,
"grad_norm": 3.7515742778778076,
"learning_rate": 1.4305084745762714e-05,
"loss": 1.4925,
"step": 8900
},
{
"epoch": 0.22,
"grad_norm": 3.0850913524627686,
"learning_rate": 1.4298305084745765e-05,
"loss": 1.5348,
"step": 8910
},
{
"epoch": 0.22,
"grad_norm": 3.571545124053955,
"learning_rate": 1.4291525423728813e-05,
"loss": 1.3513,
"step": 8920
},
{
"epoch": 0.22,
"grad_norm": 3.581817150115967,
"learning_rate": 1.4284745762711865e-05,
"loss": 1.2864,
"step": 8930
},
{
"epoch": 0.22,
"grad_norm": 1.8698399066925049,
"learning_rate": 1.4277966101694916e-05,
"loss": 1.3102,
"step": 8940
},
{
"epoch": 0.22,
"grad_norm": 3.0762484073638916,
"learning_rate": 1.4271186440677966e-05,
"loss": 1.6684,
"step": 8950
},
{
"epoch": 0.22,
"grad_norm": 9.45728588104248,
"learning_rate": 1.4264406779661017e-05,
"loss": 1.4243,
"step": 8960
},
{
"epoch": 0.22,
"grad_norm": 2.8703598976135254,
"learning_rate": 1.4257627118644069e-05,
"loss": 1.6217,
"step": 8970
},
{
"epoch": 0.22,
"grad_norm": 1.5414369106292725,
"learning_rate": 1.425084745762712e-05,
"loss": 1.4434,
"step": 8980
},
{
"epoch": 0.22,
"grad_norm": 2.332796335220337,
"learning_rate": 1.424406779661017e-05,
"loss": 1.3841,
"step": 8990
},
{
"epoch": 0.23,
"grad_norm": 5.087715148925781,
"learning_rate": 1.4237288135593221e-05,
"loss": 1.3598,
"step": 9000
},
{
"epoch": 0.23,
"eval_loss": 1.4213091135025024,
"eval_runtime": 66.1473,
"eval_samples_per_second": 15.118,
"eval_steps_per_second": 15.118,
"step": 9000
},
{
"epoch": 0.23,
"grad_norm": 3.272977590560913,
"learning_rate": 1.4230508474576273e-05,
"loss": 1.4555,
"step": 9010
},
{
"epoch": 0.23,
"grad_norm": 3.6894586086273193,
"learning_rate": 1.4223728813559324e-05,
"loss": 1.3464,
"step": 9020
},
{
"epoch": 0.23,
"grad_norm": 1.8790464401245117,
"learning_rate": 1.4216949152542374e-05,
"loss": 1.6049,
"step": 9030
},
{
"epoch": 0.23,
"grad_norm": 2.313295364379883,
"learning_rate": 1.4210169491525425e-05,
"loss": 1.538,
"step": 9040
},
{
"epoch": 0.23,
"grad_norm": 3.532170295715332,
"learning_rate": 1.4203389830508476e-05,
"loss": 1.4518,
"step": 9050
},
{
"epoch": 0.23,
"grad_norm": 3.459240674972534,
"learning_rate": 1.4196610169491528e-05,
"loss": 1.4913,
"step": 9060
},
{
"epoch": 0.23,
"grad_norm": 2.919579029083252,
"learning_rate": 1.4189830508474578e-05,
"loss": 1.1391,
"step": 9070
},
{
"epoch": 0.23,
"grad_norm": 9.67332935333252,
"learning_rate": 1.4183050847457629e-05,
"loss": 1.4643,
"step": 9080
},
{
"epoch": 0.23,
"grad_norm": 1.2171639204025269,
"learning_rate": 1.417627118644068e-05,
"loss": 1.5016,
"step": 9090
},
{
"epoch": 0.23,
"grad_norm": 9.06068229675293,
"learning_rate": 1.416949152542373e-05,
"loss": 1.5658,
"step": 9100
},
{
"epoch": 0.23,
"grad_norm": 4.348093032836914,
"learning_rate": 1.4162711864406781e-05,
"loss": 1.4176,
"step": 9110
},
{
"epoch": 0.23,
"grad_norm": 4.279529094696045,
"learning_rate": 1.4155932203389833e-05,
"loss": 1.4751,
"step": 9120
},
{
"epoch": 0.23,
"grad_norm": 3.2655725479125977,
"learning_rate": 1.414915254237288e-05,
"loss": 1.4425,
"step": 9130
},
{
"epoch": 0.23,
"grad_norm": 3.8446030616760254,
"learning_rate": 1.4142372881355932e-05,
"loss": 1.4249,
"step": 9140
},
{
"epoch": 0.23,
"grad_norm": 3.5366909503936768,
"learning_rate": 1.4135593220338984e-05,
"loss": 1.3664,
"step": 9150
},
{
"epoch": 0.23,
"grad_norm": 2.4776804447174072,
"learning_rate": 1.4128813559322035e-05,
"loss": 1.4624,
"step": 9160
},
{
"epoch": 0.23,
"grad_norm": 2.9319727420806885,
"learning_rate": 1.4122033898305085e-05,
"loss": 1.2617,
"step": 9170
},
{
"epoch": 0.23,
"grad_norm": 4.934535503387451,
"learning_rate": 1.4115254237288136e-05,
"loss": 1.3883,
"step": 9180
},
{
"epoch": 0.23,
"grad_norm": 5.425727367401123,
"learning_rate": 1.4108474576271188e-05,
"loss": 1.4821,
"step": 9190
},
{
"epoch": 0.23,
"grad_norm": 2.792351007461548,
"learning_rate": 1.4101694915254239e-05,
"loss": 1.5057,
"step": 9200
},
{
"epoch": 0.23,
"grad_norm": 2.8711283206939697,
"learning_rate": 1.4094915254237289e-05,
"loss": 1.3026,
"step": 9210
},
{
"epoch": 0.23,
"grad_norm": 2.8754985332489014,
"learning_rate": 1.408813559322034e-05,
"loss": 1.5327,
"step": 9220
},
{
"epoch": 0.23,
"grad_norm": 4.520450592041016,
"learning_rate": 1.4081355932203391e-05,
"loss": 1.4834,
"step": 9230
},
{
"epoch": 0.23,
"grad_norm": 4.829522609710693,
"learning_rate": 1.4074576271186441e-05,
"loss": 1.4513,
"step": 9240
},
{
"epoch": 0.23,
"grad_norm": 3.405491352081299,
"learning_rate": 1.4067796610169493e-05,
"loss": 1.3591,
"step": 9250
},
{
"epoch": 0.23,
"grad_norm": 3.4185872077941895,
"learning_rate": 1.4061016949152544e-05,
"loss": 1.4836,
"step": 9260
},
{
"epoch": 0.23,
"grad_norm": 2.9153568744659424,
"learning_rate": 1.4054237288135595e-05,
"loss": 1.3154,
"step": 9270
},
{
"epoch": 0.23,
"grad_norm": 5.7017412185668945,
"learning_rate": 1.4047457627118645e-05,
"loss": 1.5036,
"step": 9280
},
{
"epoch": 0.23,
"grad_norm": 2.359771728515625,
"learning_rate": 1.4040677966101696e-05,
"loss": 1.3116,
"step": 9290
},
{
"epoch": 0.23,
"grad_norm": 2.206207036972046,
"learning_rate": 1.4033898305084748e-05,
"loss": 1.4501,
"step": 9300
},
{
"epoch": 0.23,
"grad_norm": 5.006134986877441,
"learning_rate": 1.40271186440678e-05,
"loss": 1.3749,
"step": 9310
},
{
"epoch": 0.23,
"grad_norm": 5.707823753356934,
"learning_rate": 1.4020338983050849e-05,
"loss": 1.5003,
"step": 9320
},
{
"epoch": 0.23,
"grad_norm": 2.2034099102020264,
"learning_rate": 1.4013559322033899e-05,
"loss": 1.398,
"step": 9330
},
{
"epoch": 0.23,
"grad_norm": 4.742650985717773,
"learning_rate": 1.400677966101695e-05,
"loss": 1.4648,
"step": 9340
},
{
"epoch": 0.23,
"grad_norm": 2.050645589828491,
"learning_rate": 1.4e-05,
"loss": 1.388,
"step": 9350
},
{
"epoch": 0.23,
"grad_norm": 5.325618267059326,
"learning_rate": 1.3993220338983051e-05,
"loss": 1.5215,
"step": 9360
},
{
"epoch": 0.23,
"grad_norm": 1.011096715927124,
"learning_rate": 1.3986440677966103e-05,
"loss": 1.3718,
"step": 9370
},
{
"epoch": 0.23,
"grad_norm": 3.435424566268921,
"learning_rate": 1.3979661016949152e-05,
"loss": 1.607,
"step": 9380
},
{
"epoch": 0.23,
"grad_norm": 2.3544552326202393,
"learning_rate": 1.3972881355932204e-05,
"loss": 1.487,
"step": 9390
},
{
"epoch": 0.23,
"grad_norm": 3.571235418319702,
"learning_rate": 1.3966101694915255e-05,
"loss": 1.3305,
"step": 9400
},
{
"epoch": 0.24,
"grad_norm": 2.248509407043457,
"learning_rate": 1.3959322033898306e-05,
"loss": 1.3421,
"step": 9410
},
{
"epoch": 0.24,
"grad_norm": 5.0531005859375,
"learning_rate": 1.3952542372881356e-05,
"loss": 1.5146,
"step": 9420
},
{
"epoch": 0.24,
"grad_norm": 6.318211078643799,
"learning_rate": 1.3945762711864408e-05,
"loss": 1.4586,
"step": 9430
},
{
"epoch": 0.24,
"grad_norm": 5.778545379638672,
"learning_rate": 1.3938983050847459e-05,
"loss": 1.3418,
"step": 9440
},
{
"epoch": 0.24,
"grad_norm": 4.6284871101379395,
"learning_rate": 1.393220338983051e-05,
"loss": 1.6445,
"step": 9450
},
{
"epoch": 0.24,
"grad_norm": 5.105873107910156,
"learning_rate": 1.392542372881356e-05,
"loss": 1.4662,
"step": 9460
},
{
"epoch": 0.24,
"grad_norm": 1.4448604583740234,
"learning_rate": 1.3918644067796612e-05,
"loss": 1.4413,
"step": 9470
},
{
"epoch": 0.24,
"grad_norm": 1.9221776723861694,
"learning_rate": 1.3911864406779663e-05,
"loss": 1.5862,
"step": 9480
},
{
"epoch": 0.24,
"grad_norm": 5.959461688995361,
"learning_rate": 1.3905084745762714e-05,
"loss": 1.3307,
"step": 9490
},
{
"epoch": 0.24,
"grad_norm": 2.975703477859497,
"learning_rate": 1.3898305084745764e-05,
"loss": 1.3601,
"step": 9500
},
{
"epoch": 0.24,
"eval_loss": 1.438934087753296,
"eval_runtime": 66.1288,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 9500
},
{
"epoch": 0.24,
"grad_norm": 1.9357355833053589,
"learning_rate": 1.3891525423728815e-05,
"loss": 1.6061,
"step": 9510
},
{
"epoch": 0.24,
"grad_norm": 5.015805244445801,
"learning_rate": 1.3884745762711867e-05,
"loss": 1.4235,
"step": 9520
},
{
"epoch": 0.24,
"grad_norm": 3.4826226234436035,
"learning_rate": 1.3877966101694918e-05,
"loss": 1.474,
"step": 9530
},
{
"epoch": 0.24,
"grad_norm": 4.293178558349609,
"learning_rate": 1.3871186440677966e-05,
"loss": 1.5052,
"step": 9540
},
{
"epoch": 0.24,
"grad_norm": 3.036905527114868,
"learning_rate": 1.3864406779661018e-05,
"loss": 1.419,
"step": 9550
},
{
"epoch": 0.24,
"grad_norm": 4.020019054412842,
"learning_rate": 1.3857627118644067e-05,
"loss": 1.4257,
"step": 9560
},
{
"epoch": 0.24,
"grad_norm": 7.403942584991455,
"learning_rate": 1.3850847457627119e-05,
"loss": 1.3108,
"step": 9570
},
{
"epoch": 0.24,
"grad_norm": 2.8201069831848145,
"learning_rate": 1.384406779661017e-05,
"loss": 1.3367,
"step": 9580
},
{
"epoch": 0.24,
"grad_norm": 1.7385600805282593,
"learning_rate": 1.3837288135593222e-05,
"loss": 1.511,
"step": 9590
},
{
"epoch": 0.24,
"grad_norm": 1.3381562232971191,
"learning_rate": 1.3830508474576271e-05,
"loss": 1.304,
"step": 9600
},
{
"epoch": 0.24,
"grad_norm": 2.895104169845581,
"learning_rate": 1.3823728813559323e-05,
"loss": 1.4794,
"step": 9610
},
{
"epoch": 0.24,
"grad_norm": 5.847405910491943,
"learning_rate": 1.3816949152542374e-05,
"loss": 1.5482,
"step": 9620
},
{
"epoch": 0.24,
"grad_norm": 1.4990681409835815,
"learning_rate": 1.3810169491525425e-05,
"loss": 1.2669,
"step": 9630
},
{
"epoch": 0.24,
"grad_norm": 4.587515830993652,
"learning_rate": 1.3803389830508475e-05,
"loss": 1.3339,
"step": 9640
},
{
"epoch": 0.24,
"grad_norm": 4.533742904663086,
"learning_rate": 1.3796610169491527e-05,
"loss": 1.3709,
"step": 9650
},
{
"epoch": 0.24,
"grad_norm": 6.8022003173828125,
"learning_rate": 1.3789830508474578e-05,
"loss": 1.2791,
"step": 9660
},
{
"epoch": 0.24,
"grad_norm": 7.4975175857543945,
"learning_rate": 1.378305084745763e-05,
"loss": 1.5221,
"step": 9670
},
{
"epoch": 0.24,
"grad_norm": 3.8370859622955322,
"learning_rate": 1.3776271186440679e-05,
"loss": 1.295,
"step": 9680
},
{
"epoch": 0.24,
"grad_norm": 3.027163505554199,
"learning_rate": 1.376949152542373e-05,
"loss": 1.4657,
"step": 9690
},
{
"epoch": 0.24,
"grad_norm": 7.874743938446045,
"learning_rate": 1.3762711864406782e-05,
"loss": 1.4806,
"step": 9700
},
{
"epoch": 0.24,
"grad_norm": 3.1617636680603027,
"learning_rate": 1.3755932203389832e-05,
"loss": 1.4772,
"step": 9710
},
{
"epoch": 0.24,
"grad_norm": 7.0615458488464355,
"learning_rate": 1.3749152542372883e-05,
"loss": 1.4963,
"step": 9720
},
{
"epoch": 0.24,
"grad_norm": 4.889975070953369,
"learning_rate": 1.3742372881355934e-05,
"loss": 1.4434,
"step": 9730
},
{
"epoch": 0.24,
"grad_norm": 3.209303855895996,
"learning_rate": 1.3735593220338986e-05,
"loss": 1.5107,
"step": 9740
},
{
"epoch": 0.24,
"grad_norm": 3.6406736373901367,
"learning_rate": 1.3728813559322034e-05,
"loss": 1.3425,
"step": 9750
},
{
"epoch": 0.24,
"grad_norm": 1.520709753036499,
"learning_rate": 1.3722033898305085e-05,
"loss": 1.4677,
"step": 9760
},
{
"epoch": 0.24,
"grad_norm": 3.9511709213256836,
"learning_rate": 1.3715254237288137e-05,
"loss": 1.5378,
"step": 9770
},
{
"epoch": 0.24,
"grad_norm": 3.744352102279663,
"learning_rate": 1.3708474576271186e-05,
"loss": 1.4105,
"step": 9780
},
{
"epoch": 0.24,
"grad_norm": 3.820054292678833,
"learning_rate": 1.3701694915254238e-05,
"loss": 1.3589,
"step": 9790
},
{
"epoch": 0.24,
"grad_norm": 4.339268684387207,
"learning_rate": 1.3694915254237289e-05,
"loss": 1.5092,
"step": 9800
},
{
"epoch": 0.25,
"grad_norm": 2.449904680252075,
"learning_rate": 1.368813559322034e-05,
"loss": 1.4878,
"step": 9810
},
{
"epoch": 0.25,
"grad_norm": 9.548687934875488,
"learning_rate": 1.368135593220339e-05,
"loss": 1.4862,
"step": 9820
},
{
"epoch": 0.25,
"grad_norm": 4.738724708557129,
"learning_rate": 1.3674576271186442e-05,
"loss": 1.4042,
"step": 9830
},
{
"epoch": 0.25,
"grad_norm": 4.652593612670898,
"learning_rate": 1.3667796610169493e-05,
"loss": 1.3038,
"step": 9840
},
{
"epoch": 0.25,
"grad_norm": 4.120163917541504,
"learning_rate": 1.3661016949152543e-05,
"loss": 1.4582,
"step": 9850
},
{
"epoch": 0.25,
"grad_norm": 4.297613143920898,
"learning_rate": 1.3654237288135594e-05,
"loss": 1.4914,
"step": 9860
},
{
"epoch": 0.25,
"grad_norm": 3.7870216369628906,
"learning_rate": 1.3647457627118646e-05,
"loss": 1.3886,
"step": 9870
},
{
"epoch": 0.25,
"grad_norm": 4.052813529968262,
"learning_rate": 1.3640677966101697e-05,
"loss": 1.5646,
"step": 9880
},
{
"epoch": 0.25,
"grad_norm": 4.547369003295898,
"learning_rate": 1.3633898305084747e-05,
"loss": 1.3154,
"step": 9890
},
{
"epoch": 0.25,
"grad_norm": 1.9027762413024902,
"learning_rate": 1.3627118644067798e-05,
"loss": 1.4038,
"step": 9900
},
{
"epoch": 0.25,
"grad_norm": 3.208552360534668,
"learning_rate": 1.362033898305085e-05,
"loss": 1.3637,
"step": 9910
},
{
"epoch": 0.25,
"grad_norm": 13.44774055480957,
"learning_rate": 1.36135593220339e-05,
"loss": 1.2825,
"step": 9920
},
{
"epoch": 0.25,
"grad_norm": 8.423622131347656,
"learning_rate": 1.360677966101695e-05,
"loss": 1.4533,
"step": 9930
},
{
"epoch": 0.25,
"grad_norm": 4.321129322052002,
"learning_rate": 1.3600000000000002e-05,
"loss": 1.3952,
"step": 9940
},
{
"epoch": 0.25,
"grad_norm": 4.656356334686279,
"learning_rate": 1.3593220338983053e-05,
"loss": 1.4925,
"step": 9950
},
{
"epoch": 0.25,
"grad_norm": 3.8355157375335693,
"learning_rate": 1.3586440677966101e-05,
"loss": 1.4598,
"step": 9960
},
{
"epoch": 0.25,
"grad_norm": 3.602102041244507,
"learning_rate": 1.3579661016949153e-05,
"loss": 1.4722,
"step": 9970
},
{
"epoch": 0.25,
"grad_norm": 1.8679078817367554,
"learning_rate": 1.3572881355932204e-05,
"loss": 1.5094,
"step": 9980
},
{
"epoch": 0.25,
"grad_norm": 3.2152957916259766,
"learning_rate": 1.3566101694915254e-05,
"loss": 1.4489,
"step": 9990
},
{
"epoch": 0.25,
"grad_norm": 2.101644515991211,
"learning_rate": 1.3559322033898305e-05,
"loss": 1.5278,
"step": 10000
},
{
"epoch": 0.25,
"eval_loss": 1.419925332069397,
"eval_runtime": 66.1325,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 10000
},
{
"epoch": 0.25,
"grad_norm": 6.951940536499023,
"learning_rate": 1.3552542372881357e-05,
"loss": 1.4138,
"step": 10010
},
{
"epoch": 0.25,
"grad_norm": 1.5260531902313232,
"learning_rate": 1.3545762711864408e-05,
"loss": 1.3812,
"step": 10020
},
{
"epoch": 0.25,
"grad_norm": 2.1888010501861572,
"learning_rate": 1.3538983050847458e-05,
"loss": 1.5596,
"step": 10030
},
{
"epoch": 0.25,
"grad_norm": 1.2334835529327393,
"learning_rate": 1.353220338983051e-05,
"loss": 1.5287,
"step": 10040
},
{
"epoch": 0.25,
"grad_norm": 4.4969353675842285,
"learning_rate": 1.352542372881356e-05,
"loss": 1.4268,
"step": 10050
},
{
"epoch": 0.25,
"grad_norm": 4.795516490936279,
"learning_rate": 1.3518644067796612e-05,
"loss": 1.2992,
"step": 10060
},
{
"epoch": 0.25,
"grad_norm": 2.478320837020874,
"learning_rate": 1.3511864406779662e-05,
"loss": 1.4974,
"step": 10070
},
{
"epoch": 0.25,
"grad_norm": 1.9508737325668335,
"learning_rate": 1.3505084745762713e-05,
"loss": 1.3097,
"step": 10080
},
{
"epoch": 0.25,
"grad_norm": 3.163217306137085,
"learning_rate": 1.3498305084745764e-05,
"loss": 1.4782,
"step": 10090
},
{
"epoch": 0.25,
"grad_norm": 5.9888787269592285,
"learning_rate": 1.3491525423728816e-05,
"loss": 1.3798,
"step": 10100
},
{
"epoch": 0.25,
"grad_norm": 4.777866840362549,
"learning_rate": 1.3484745762711866e-05,
"loss": 1.4042,
"step": 10110
},
{
"epoch": 0.25,
"grad_norm": 2.4334771633148193,
"learning_rate": 1.3477966101694917e-05,
"loss": 1.4185,
"step": 10120
},
{
"epoch": 0.25,
"grad_norm": 2.8577849864959717,
"learning_rate": 1.3471186440677968e-05,
"loss": 1.4438,
"step": 10130
},
{
"epoch": 0.25,
"grad_norm": 2.9416913986206055,
"learning_rate": 1.346440677966102e-05,
"loss": 1.4679,
"step": 10140
},
{
"epoch": 0.25,
"grad_norm": 5.749608516693115,
"learning_rate": 1.345762711864407e-05,
"loss": 1.3858,
"step": 10150
},
{
"epoch": 0.25,
"grad_norm": 3.1441190242767334,
"learning_rate": 1.3450847457627121e-05,
"loss": 1.3954,
"step": 10160
},
{
"epoch": 0.25,
"grad_norm": 4.9486188888549805,
"learning_rate": 1.3444067796610169e-05,
"loss": 1.4373,
"step": 10170
},
{
"epoch": 0.25,
"grad_norm": 4.6098785400390625,
"learning_rate": 1.343728813559322e-05,
"loss": 1.3843,
"step": 10180
},
{
"epoch": 0.25,
"grad_norm": 2.3991878032684326,
"learning_rate": 1.3430508474576272e-05,
"loss": 1.3708,
"step": 10190
},
{
"epoch": 0.26,
"grad_norm": 4.673692226409912,
"learning_rate": 1.3423728813559323e-05,
"loss": 1.5219,
"step": 10200
},
{
"epoch": 0.26,
"grad_norm": 1.292708158493042,
"learning_rate": 1.3416949152542373e-05,
"loss": 1.4646,
"step": 10210
},
{
"epoch": 0.26,
"grad_norm": 6.8573479652404785,
"learning_rate": 1.3410169491525424e-05,
"loss": 1.5663,
"step": 10220
},
{
"epoch": 0.26,
"grad_norm": 1.6200875043869019,
"learning_rate": 1.3403389830508476e-05,
"loss": 1.4902,
"step": 10230
},
{
"epoch": 0.26,
"grad_norm": 5.667778491973877,
"learning_rate": 1.3396610169491527e-05,
"loss": 1.4259,
"step": 10240
},
{
"epoch": 0.26,
"grad_norm": 6.129855155944824,
"learning_rate": 1.3389830508474577e-05,
"loss": 1.43,
"step": 10250
},
{
"epoch": 0.26,
"grad_norm": 2.988898992538452,
"learning_rate": 1.3383050847457628e-05,
"loss": 1.3867,
"step": 10260
},
{
"epoch": 0.26,
"grad_norm": 2.862208366394043,
"learning_rate": 1.337627118644068e-05,
"loss": 1.4641,
"step": 10270
},
{
"epoch": 0.26,
"grad_norm": 4.67335319519043,
"learning_rate": 1.3369491525423731e-05,
"loss": 1.4377,
"step": 10280
},
{
"epoch": 0.26,
"grad_norm": 6.531377792358398,
"learning_rate": 1.336271186440678e-05,
"loss": 1.3493,
"step": 10290
},
{
"epoch": 0.26,
"grad_norm": 2.349855899810791,
"learning_rate": 1.3355932203389832e-05,
"loss": 1.3054,
"step": 10300
},
{
"epoch": 0.26,
"grad_norm": 1.38920259475708,
"learning_rate": 1.3349152542372883e-05,
"loss": 1.4467,
"step": 10310
},
{
"epoch": 0.26,
"grad_norm": 4.019247055053711,
"learning_rate": 1.3342372881355933e-05,
"loss": 1.5132,
"step": 10320
},
{
"epoch": 0.26,
"grad_norm": 2.967440128326416,
"learning_rate": 1.3335593220338985e-05,
"loss": 1.391,
"step": 10330
},
{
"epoch": 0.26,
"grad_norm": 8.294310569763184,
"learning_rate": 1.3328813559322036e-05,
"loss": 1.3113,
"step": 10340
},
{
"epoch": 0.26,
"grad_norm": 5.314875602722168,
"learning_rate": 1.3322033898305087e-05,
"loss": 1.4274,
"step": 10350
},
{
"epoch": 0.26,
"grad_norm": 9.737332344055176,
"learning_rate": 1.3315254237288137e-05,
"loss": 1.4872,
"step": 10360
},
{
"epoch": 0.26,
"grad_norm": 5.628332614898682,
"learning_rate": 1.3308474576271187e-05,
"loss": 1.4917,
"step": 10370
},
{
"epoch": 0.26,
"grad_norm": 2.7575583457946777,
"learning_rate": 1.3301694915254238e-05,
"loss": 1.5113,
"step": 10380
},
{
"epoch": 0.26,
"grad_norm": 6.598272800445557,
"learning_rate": 1.3294915254237288e-05,
"loss": 1.321,
"step": 10390
},
{
"epoch": 0.26,
"grad_norm": 5.086571216583252,
"learning_rate": 1.328813559322034e-05,
"loss": 1.4841,
"step": 10400
},
{
"epoch": 0.26,
"grad_norm": 2.9577059745788574,
"learning_rate": 1.328135593220339e-05,
"loss": 1.3408,
"step": 10410
},
{
"epoch": 0.26,
"grad_norm": 7.587204933166504,
"learning_rate": 1.3274576271186442e-05,
"loss": 1.4564,
"step": 10420
},
{
"epoch": 0.26,
"grad_norm": 3.409858226776123,
"learning_rate": 1.3267796610169492e-05,
"loss": 1.5365,
"step": 10430
},
{
"epoch": 0.26,
"grad_norm": 4.1367387771606445,
"learning_rate": 1.3261016949152543e-05,
"loss": 1.3273,
"step": 10440
},
{
"epoch": 0.26,
"grad_norm": 4.932620048522949,
"learning_rate": 1.3254237288135595e-05,
"loss": 1.5361,
"step": 10450
},
{
"epoch": 0.26,
"grad_norm": 2.9673991203308105,
"learning_rate": 1.3247457627118644e-05,
"loss": 1.5074,
"step": 10460
},
{
"epoch": 0.26,
"grad_norm": 8.657196998596191,
"learning_rate": 1.3240677966101696e-05,
"loss": 1.4046,
"step": 10470
},
{
"epoch": 0.26,
"grad_norm": 0.6172802448272705,
"learning_rate": 1.3233898305084747e-05,
"loss": 1.2062,
"step": 10480
},
{
"epoch": 0.26,
"grad_norm": 2.4086947441101074,
"learning_rate": 1.3227118644067798e-05,
"loss": 1.4763,
"step": 10490
},
{
"epoch": 0.26,
"grad_norm": 2.687913656234741,
"learning_rate": 1.3220338983050848e-05,
"loss": 1.4159,
"step": 10500
},
{
"epoch": 0.26,
"eval_loss": 1.3867793083190918,
"eval_runtime": 66.119,
"eval_samples_per_second": 15.124,
"eval_steps_per_second": 15.124,
"step": 10500
},
{
"epoch": 0.26,
"grad_norm": 3.875162363052368,
"learning_rate": 1.32135593220339e-05,
"loss": 1.6311,
"step": 10510
},
{
"epoch": 0.26,
"grad_norm": 2.257335901260376,
"learning_rate": 1.3206779661016951e-05,
"loss": 1.467,
"step": 10520
},
{
"epoch": 0.26,
"grad_norm": 2.7035036087036133,
"learning_rate": 1.3200000000000002e-05,
"loss": 1.5601,
"step": 10530
},
{
"epoch": 0.26,
"grad_norm": 4.788657188415527,
"learning_rate": 1.3193220338983052e-05,
"loss": 1.4355,
"step": 10540
},
{
"epoch": 0.26,
"grad_norm": 3.4326958656311035,
"learning_rate": 1.3186440677966103e-05,
"loss": 1.4474,
"step": 10550
},
{
"epoch": 0.26,
"grad_norm": 5.0102009773254395,
"learning_rate": 1.3179661016949155e-05,
"loss": 1.579,
"step": 10560
},
{
"epoch": 0.26,
"grad_norm": 2.6414384841918945,
"learning_rate": 1.3172881355932206e-05,
"loss": 1.4042,
"step": 10570
},
{
"epoch": 0.26,
"grad_norm": 2.270298480987549,
"learning_rate": 1.3166101694915254e-05,
"loss": 1.4761,
"step": 10580
},
{
"epoch": 0.26,
"grad_norm": 2.361912250518799,
"learning_rate": 1.3159322033898306e-05,
"loss": 1.4999,
"step": 10590
},
{
"epoch": 0.27,
"grad_norm": 5.356451988220215,
"learning_rate": 1.3152542372881355e-05,
"loss": 1.4594,
"step": 10600
},
{
"epoch": 0.27,
"grad_norm": 5.603447914123535,
"learning_rate": 1.3145762711864407e-05,
"loss": 1.355,
"step": 10610
},
{
"epoch": 0.27,
"grad_norm": 2.6022191047668457,
"learning_rate": 1.3138983050847458e-05,
"loss": 1.3742,
"step": 10620
},
{
"epoch": 0.27,
"grad_norm": 5.697202205657959,
"learning_rate": 1.313220338983051e-05,
"loss": 1.3713,
"step": 10630
},
{
"epoch": 0.27,
"grad_norm": 2.859365224838257,
"learning_rate": 1.312542372881356e-05,
"loss": 1.5086,
"step": 10640
},
{
"epoch": 0.27,
"grad_norm": 7.860188007354736,
"learning_rate": 1.311864406779661e-05,
"loss": 1.2122,
"step": 10650
},
{
"epoch": 0.27,
"grad_norm": 2.77958607673645,
"learning_rate": 1.3111864406779662e-05,
"loss": 1.4479,
"step": 10660
},
{
"epoch": 0.27,
"grad_norm": 9.004295349121094,
"learning_rate": 1.3105084745762714e-05,
"loss": 1.5347,
"step": 10670
},
{
"epoch": 0.27,
"grad_norm": 4.261473178863525,
"learning_rate": 1.3098305084745763e-05,
"loss": 1.4104,
"step": 10680
},
{
"epoch": 0.27,
"grad_norm": 1.7853983640670776,
"learning_rate": 1.3091525423728815e-05,
"loss": 1.3679,
"step": 10690
},
{
"epoch": 0.27,
"grad_norm": 4.3893609046936035,
"learning_rate": 1.3084745762711866e-05,
"loss": 1.4985,
"step": 10700
},
{
"epoch": 0.27,
"grad_norm": 1.7915986776351929,
"learning_rate": 1.3077966101694917e-05,
"loss": 1.4847,
"step": 10710
},
{
"epoch": 0.27,
"grad_norm": 5.252231121063232,
"learning_rate": 1.3071186440677967e-05,
"loss": 1.4013,
"step": 10720
},
{
"epoch": 0.27,
"grad_norm": 2.2784807682037354,
"learning_rate": 1.3064406779661019e-05,
"loss": 1.395,
"step": 10730
},
{
"epoch": 0.27,
"grad_norm": 4.382660388946533,
"learning_rate": 1.305762711864407e-05,
"loss": 1.3082,
"step": 10740
},
{
"epoch": 0.27,
"grad_norm": 3.5009450912475586,
"learning_rate": 1.305084745762712e-05,
"loss": 1.355,
"step": 10750
},
{
"epoch": 0.27,
"grad_norm": 6.310500621795654,
"learning_rate": 1.3044067796610171e-05,
"loss": 1.4164,
"step": 10760
},
{
"epoch": 0.27,
"grad_norm": 2.8609213829040527,
"learning_rate": 1.3037288135593222e-05,
"loss": 1.4316,
"step": 10770
},
{
"epoch": 0.27,
"grad_norm": 5.6872382164001465,
"learning_rate": 1.3030508474576274e-05,
"loss": 1.4071,
"step": 10780
},
{
"epoch": 0.27,
"grad_norm": 4.75257682800293,
"learning_rate": 1.3023728813559322e-05,
"loss": 1.4906,
"step": 10790
},
{
"epoch": 0.27,
"grad_norm": 3.286771297454834,
"learning_rate": 1.3016949152542373e-05,
"loss": 1.3927,
"step": 10800
},
{
"epoch": 0.27,
"grad_norm": 5.782283782958984,
"learning_rate": 1.3010169491525425e-05,
"loss": 1.5204,
"step": 10810
},
{
"epoch": 0.27,
"grad_norm": 2.9265480041503906,
"learning_rate": 1.3003389830508474e-05,
"loss": 1.5996,
"step": 10820
},
{
"epoch": 0.27,
"grad_norm": 7.570978164672852,
"learning_rate": 1.2996610169491526e-05,
"loss": 1.3926,
"step": 10830
},
{
"epoch": 0.27,
"grad_norm": 2.7784712314605713,
"learning_rate": 1.2989830508474577e-05,
"loss": 1.4601,
"step": 10840
},
{
"epoch": 0.27,
"grad_norm": 2.448601007461548,
"learning_rate": 1.2983050847457629e-05,
"loss": 1.4095,
"step": 10850
},
{
"epoch": 0.27,
"grad_norm": 3.784883975982666,
"learning_rate": 1.2976271186440678e-05,
"loss": 1.3626,
"step": 10860
},
{
"epoch": 0.27,
"grad_norm": 5.855194091796875,
"learning_rate": 1.296949152542373e-05,
"loss": 1.3263,
"step": 10870
},
{
"epoch": 0.27,
"grad_norm": 3.4989304542541504,
"learning_rate": 1.2962711864406781e-05,
"loss": 1.4125,
"step": 10880
},
{
"epoch": 0.27,
"grad_norm": 4.406559467315674,
"learning_rate": 1.295593220338983e-05,
"loss": 1.561,
"step": 10890
},
{
"epoch": 0.27,
"grad_norm": 3.153087615966797,
"learning_rate": 1.2949152542372882e-05,
"loss": 1.4574,
"step": 10900
},
{
"epoch": 0.27,
"grad_norm": 3.5177876949310303,
"learning_rate": 1.2942372881355934e-05,
"loss": 1.3867,
"step": 10910
},
{
"epoch": 0.27,
"grad_norm": 4.136317729949951,
"learning_rate": 1.2935593220338985e-05,
"loss": 1.4718,
"step": 10920
},
{
"epoch": 0.27,
"grad_norm": 2.291229009628296,
"learning_rate": 1.2928813559322035e-05,
"loss": 1.3122,
"step": 10930
},
{
"epoch": 0.27,
"grad_norm": 3.3874454498291016,
"learning_rate": 1.2922033898305086e-05,
"loss": 1.4073,
"step": 10940
},
{
"epoch": 0.27,
"grad_norm": 4.799066066741943,
"learning_rate": 1.2915254237288137e-05,
"loss": 1.4412,
"step": 10950
},
{
"epoch": 0.27,
"grad_norm": 6.790704250335693,
"learning_rate": 1.2908474576271189e-05,
"loss": 1.5664,
"step": 10960
},
{
"epoch": 0.27,
"grad_norm": 3.8187286853790283,
"learning_rate": 1.2901694915254239e-05,
"loss": 1.2302,
"step": 10970
},
{
"epoch": 0.27,
"grad_norm": 2.162944793701172,
"learning_rate": 1.289491525423729e-05,
"loss": 1.4602,
"step": 10980
},
{
"epoch": 0.27,
"grad_norm": 7.523599147796631,
"learning_rate": 1.2888135593220341e-05,
"loss": 1.4643,
"step": 10990
},
{
"epoch": 0.28,
"grad_norm": 3.7828078269958496,
"learning_rate": 1.288135593220339e-05,
"loss": 1.6134,
"step": 11000
},
{
"epoch": 0.28,
"eval_loss": 1.4437503814697266,
"eval_runtime": 66.1284,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 11000
},
{
"epoch": 0.28,
"grad_norm": 10.092121124267578,
"learning_rate": 1.287457627118644e-05,
"loss": 1.5293,
"step": 11010
},
{
"epoch": 0.28,
"grad_norm": 7.2347235679626465,
"learning_rate": 1.2867796610169492e-05,
"loss": 1.4884,
"step": 11020
},
{
"epoch": 0.28,
"grad_norm": 2.231647491455078,
"learning_rate": 1.2861016949152542e-05,
"loss": 1.5299,
"step": 11030
},
{
"epoch": 0.28,
"grad_norm": 6.09018611907959,
"learning_rate": 1.2854237288135593e-05,
"loss": 1.588,
"step": 11040
},
{
"epoch": 0.28,
"grad_norm": 5.75846529006958,
"learning_rate": 1.2847457627118645e-05,
"loss": 1.4481,
"step": 11050
},
{
"epoch": 0.28,
"grad_norm": 2.08042311668396,
"learning_rate": 1.2840677966101696e-05,
"loss": 1.4066,
"step": 11060
},
{
"epoch": 0.28,
"grad_norm": 4.944487571716309,
"learning_rate": 1.2833898305084746e-05,
"loss": 1.3442,
"step": 11070
},
{
"epoch": 0.28,
"grad_norm": 2.4912710189819336,
"learning_rate": 1.2827118644067797e-05,
"loss": 1.6345,
"step": 11080
},
{
"epoch": 0.28,
"grad_norm": 2.6630759239196777,
"learning_rate": 1.2820338983050849e-05,
"loss": 1.4073,
"step": 11090
},
{
"epoch": 0.28,
"grad_norm": 1.5061525106430054,
"learning_rate": 1.28135593220339e-05,
"loss": 1.5894,
"step": 11100
},
{
"epoch": 0.28,
"grad_norm": 2.7690367698669434,
"learning_rate": 1.280677966101695e-05,
"loss": 1.3617,
"step": 11110
},
{
"epoch": 0.28,
"grad_norm": 6.3855462074279785,
"learning_rate": 1.2800000000000001e-05,
"loss": 1.3119,
"step": 11120
},
{
"epoch": 0.28,
"grad_norm": 7.427648544311523,
"learning_rate": 1.2793220338983053e-05,
"loss": 1.5297,
"step": 11130
},
{
"epoch": 0.28,
"grad_norm": 8.82369613647461,
"learning_rate": 1.2786440677966104e-05,
"loss": 1.3032,
"step": 11140
},
{
"epoch": 0.28,
"grad_norm": 3.6501011848449707,
"learning_rate": 1.2779661016949154e-05,
"loss": 1.3527,
"step": 11150
},
{
"epoch": 0.28,
"grad_norm": 2.652341604232788,
"learning_rate": 1.2772881355932205e-05,
"loss": 1.2817,
"step": 11160
},
{
"epoch": 0.28,
"grad_norm": 5.405377388000488,
"learning_rate": 1.2766101694915256e-05,
"loss": 1.2666,
"step": 11170
},
{
"epoch": 0.28,
"grad_norm": 7.799898147583008,
"learning_rate": 1.2759322033898308e-05,
"loss": 1.5301,
"step": 11180
},
{
"epoch": 0.28,
"grad_norm": 3.052325487136841,
"learning_rate": 1.2752542372881358e-05,
"loss": 1.2761,
"step": 11190
},
{
"epoch": 0.28,
"grad_norm": 5.282183647155762,
"learning_rate": 1.2745762711864407e-05,
"loss": 1.4701,
"step": 11200
},
{
"epoch": 0.28,
"grad_norm": 3.8688693046569824,
"learning_rate": 1.2738983050847457e-05,
"loss": 1.4549,
"step": 11210
},
{
"epoch": 0.28,
"grad_norm": 5.308023929595947,
"learning_rate": 1.2732203389830508e-05,
"loss": 1.3581,
"step": 11220
},
{
"epoch": 0.28,
"grad_norm": 5.7319841384887695,
"learning_rate": 1.272542372881356e-05,
"loss": 1.5457,
"step": 11230
},
{
"epoch": 0.28,
"grad_norm": 9.156388282775879,
"learning_rate": 1.2718644067796611e-05,
"loss": 1.5201,
"step": 11240
},
{
"epoch": 0.28,
"grad_norm": 5.315608024597168,
"learning_rate": 1.2711864406779661e-05,
"loss": 1.3505,
"step": 11250
},
{
"epoch": 0.28,
"grad_norm": 1.7720659971237183,
"learning_rate": 1.2705084745762712e-05,
"loss": 1.3938,
"step": 11260
},
{
"epoch": 0.28,
"grad_norm": 4.514403343200684,
"learning_rate": 1.2698305084745764e-05,
"loss": 1.6121,
"step": 11270
},
{
"epoch": 0.28,
"grad_norm": 6.911466121673584,
"learning_rate": 1.2691525423728815e-05,
"loss": 1.3458,
"step": 11280
},
{
"epoch": 0.28,
"grad_norm": 4.472634792327881,
"learning_rate": 1.2684745762711865e-05,
"loss": 1.2857,
"step": 11290
},
{
"epoch": 0.28,
"grad_norm": 2.4667749404907227,
"learning_rate": 1.2677966101694916e-05,
"loss": 1.3835,
"step": 11300
},
{
"epoch": 0.28,
"grad_norm": 4.501352787017822,
"learning_rate": 1.2671186440677968e-05,
"loss": 1.3502,
"step": 11310
},
{
"epoch": 0.28,
"grad_norm": 4.203758239746094,
"learning_rate": 1.2664406779661019e-05,
"loss": 1.412,
"step": 11320
},
{
"epoch": 0.28,
"grad_norm": 4.501011848449707,
"learning_rate": 1.2657627118644069e-05,
"loss": 1.4239,
"step": 11330
},
{
"epoch": 0.28,
"grad_norm": 6.272605895996094,
"learning_rate": 1.265084745762712e-05,
"loss": 1.4452,
"step": 11340
},
{
"epoch": 0.28,
"grad_norm": 1.9776923656463623,
"learning_rate": 1.2644067796610171e-05,
"loss": 1.3449,
"step": 11350
},
{
"epoch": 0.28,
"grad_norm": 2.2955100536346436,
"learning_rate": 1.2637288135593221e-05,
"loss": 1.4088,
"step": 11360
},
{
"epoch": 0.28,
"grad_norm": 5.005813121795654,
"learning_rate": 1.2630508474576273e-05,
"loss": 1.4663,
"step": 11370
},
{
"epoch": 0.28,
"grad_norm": 3.1890101432800293,
"learning_rate": 1.2623728813559324e-05,
"loss": 1.4736,
"step": 11380
},
{
"epoch": 0.28,
"grad_norm": 5.545890808105469,
"learning_rate": 1.2616949152542375e-05,
"loss": 1.3913,
"step": 11390
},
{
"epoch": 0.28,
"grad_norm": 3.0610640048980713,
"learning_rate": 1.2610169491525425e-05,
"loss": 1.1965,
"step": 11400
},
{
"epoch": 0.29,
"grad_norm": 5.919715881347656,
"learning_rate": 1.2603389830508475e-05,
"loss": 1.6285,
"step": 11410
},
{
"epoch": 0.29,
"grad_norm": 5.279333591461182,
"learning_rate": 1.2596610169491526e-05,
"loss": 1.4144,
"step": 11420
},
{
"epoch": 0.29,
"grad_norm": 6.2425737380981445,
"learning_rate": 1.2589830508474576e-05,
"loss": 1.4376,
"step": 11430
},
{
"epoch": 0.29,
"grad_norm": 1.8463287353515625,
"learning_rate": 1.2583050847457627e-05,
"loss": 1.4008,
"step": 11440
},
{
"epoch": 0.29,
"grad_norm": 6.911423206329346,
"learning_rate": 1.2576271186440679e-05,
"loss": 1.4218,
"step": 11450
},
{
"epoch": 0.29,
"grad_norm": 1.8190709352493286,
"learning_rate": 1.256949152542373e-05,
"loss": 1.3105,
"step": 11460
},
{
"epoch": 0.29,
"grad_norm": 2.9692275524139404,
"learning_rate": 1.256271186440678e-05,
"loss": 1.4071,
"step": 11470
},
{
"epoch": 0.29,
"grad_norm": 2.9249064922332764,
"learning_rate": 1.2555932203389831e-05,
"loss": 1.4017,
"step": 11480
},
{
"epoch": 0.29,
"grad_norm": 4.724937915802002,
"learning_rate": 1.2549152542372883e-05,
"loss": 1.3826,
"step": 11490
},
{
"epoch": 0.29,
"grad_norm": 2.75592303276062,
"learning_rate": 1.2542372881355932e-05,
"loss": 1.4793,
"step": 11500
},
{
"epoch": 0.29,
"eval_loss": 1.4384660720825195,
"eval_runtime": 66.1691,
"eval_samples_per_second": 15.113,
"eval_steps_per_second": 15.113,
"step": 11500
},
{
"epoch": 0.29,
"grad_norm": 4.604795455932617,
"learning_rate": 1.2535593220338984e-05,
"loss": 1.5059,
"step": 11510
},
{
"epoch": 0.29,
"grad_norm": 2.278486490249634,
"learning_rate": 1.2528813559322035e-05,
"loss": 1.307,
"step": 11520
},
{
"epoch": 0.29,
"grad_norm": 3.228501796722412,
"learning_rate": 1.2522033898305087e-05,
"loss": 1.4515,
"step": 11530
},
{
"epoch": 0.29,
"grad_norm": 5.253673076629639,
"learning_rate": 1.2515254237288136e-05,
"loss": 1.3283,
"step": 11540
},
{
"epoch": 0.29,
"grad_norm": 5.198097229003906,
"learning_rate": 1.2508474576271188e-05,
"loss": 1.25,
"step": 11550
},
{
"epoch": 0.29,
"grad_norm": 5.645752429962158,
"learning_rate": 1.2501694915254239e-05,
"loss": 1.555,
"step": 11560
},
{
"epoch": 0.29,
"grad_norm": 5.3141913414001465,
"learning_rate": 1.249491525423729e-05,
"loss": 1.5135,
"step": 11570
},
{
"epoch": 0.29,
"grad_norm": 1.6161258220672607,
"learning_rate": 1.248813559322034e-05,
"loss": 1.3809,
"step": 11580
},
{
"epoch": 0.29,
"grad_norm": 3.477424383163452,
"learning_rate": 1.2481355932203392e-05,
"loss": 1.3237,
"step": 11590
},
{
"epoch": 0.29,
"grad_norm": 3.588980197906494,
"learning_rate": 1.2474576271186443e-05,
"loss": 1.3171,
"step": 11600
},
{
"epoch": 0.29,
"grad_norm": 2.2803046703338623,
"learning_rate": 1.2467796610169494e-05,
"loss": 1.3816,
"step": 11610
},
{
"epoch": 0.29,
"grad_norm": 1.7084169387817383,
"learning_rate": 1.2461016949152542e-05,
"loss": 1.3663,
"step": 11620
},
{
"epoch": 0.29,
"grad_norm": 2.5784099102020264,
"learning_rate": 1.2454237288135594e-05,
"loss": 1.4399,
"step": 11630
},
{
"epoch": 0.29,
"grad_norm": 4.369024276733398,
"learning_rate": 1.2447457627118643e-05,
"loss": 1.3491,
"step": 11640
},
{
"epoch": 0.29,
"grad_norm": 6.239041328430176,
"learning_rate": 1.2440677966101695e-05,
"loss": 1.2562,
"step": 11650
},
{
"epoch": 0.29,
"grad_norm": 4.257971286773682,
"learning_rate": 1.2433898305084746e-05,
"loss": 1.4829,
"step": 11660
},
{
"epoch": 0.29,
"grad_norm": 4.7997660636901855,
"learning_rate": 1.2427118644067798e-05,
"loss": 1.3841,
"step": 11670
},
{
"epoch": 0.29,
"grad_norm": 3.7475717067718506,
"learning_rate": 1.2420338983050847e-05,
"loss": 1.3625,
"step": 11680
},
{
"epoch": 0.29,
"grad_norm": 2.6195523738861084,
"learning_rate": 1.2413559322033899e-05,
"loss": 1.4386,
"step": 11690
},
{
"epoch": 0.29,
"grad_norm": 1.9052702188491821,
"learning_rate": 1.240677966101695e-05,
"loss": 1.3974,
"step": 11700
},
{
"epoch": 0.29,
"grad_norm": 3.2266929149627686,
"learning_rate": 1.2400000000000002e-05,
"loss": 1.4356,
"step": 11710
},
{
"epoch": 0.29,
"grad_norm": 4.4783806800842285,
"learning_rate": 1.2393220338983051e-05,
"loss": 1.5246,
"step": 11720
},
{
"epoch": 0.29,
"grad_norm": 5.959903717041016,
"learning_rate": 1.2386440677966103e-05,
"loss": 1.2639,
"step": 11730
},
{
"epoch": 0.29,
"grad_norm": 3.8165271282196045,
"learning_rate": 1.2379661016949154e-05,
"loss": 1.2957,
"step": 11740
},
{
"epoch": 0.29,
"grad_norm": 6.7595367431640625,
"learning_rate": 1.2372881355932205e-05,
"loss": 1.4522,
"step": 11750
},
{
"epoch": 0.29,
"grad_norm": 2.844536066055298,
"learning_rate": 1.2366101694915255e-05,
"loss": 1.5718,
"step": 11760
},
{
"epoch": 0.29,
"grad_norm": 7.719458103179932,
"learning_rate": 1.2359322033898307e-05,
"loss": 1.579,
"step": 11770
},
{
"epoch": 0.29,
"grad_norm": 1.1778481006622314,
"learning_rate": 1.2352542372881358e-05,
"loss": 1.5497,
"step": 11780
},
{
"epoch": 0.29,
"grad_norm": 4.286680221557617,
"learning_rate": 1.234576271186441e-05,
"loss": 1.4507,
"step": 11790
},
{
"epoch": 0.29,
"grad_norm": 1.9755306243896484,
"learning_rate": 1.2338983050847459e-05,
"loss": 1.2822,
"step": 11800
},
{
"epoch": 0.3,
"grad_norm": 3.383787155151367,
"learning_rate": 1.233220338983051e-05,
"loss": 1.3459,
"step": 11810
},
{
"epoch": 0.3,
"grad_norm": 2.444471836090088,
"learning_rate": 1.2325423728813562e-05,
"loss": 1.3843,
"step": 11820
},
{
"epoch": 0.3,
"grad_norm": 6.240593433380127,
"learning_rate": 1.231864406779661e-05,
"loss": 1.696,
"step": 11830
},
{
"epoch": 0.3,
"grad_norm": 2.954618215560913,
"learning_rate": 1.2311864406779661e-05,
"loss": 1.3513,
"step": 11840
},
{
"epoch": 0.3,
"grad_norm": 4.6317243576049805,
"learning_rate": 1.2305084745762713e-05,
"loss": 1.4252,
"step": 11850
},
{
"epoch": 0.3,
"grad_norm": 7.941683769226074,
"learning_rate": 1.2298305084745762e-05,
"loss": 1.3129,
"step": 11860
},
{
"epoch": 0.3,
"grad_norm": 4.002432823181152,
"learning_rate": 1.2291525423728814e-05,
"loss": 1.5798,
"step": 11870
},
{
"epoch": 0.3,
"grad_norm": 2.3357415199279785,
"learning_rate": 1.2284745762711865e-05,
"loss": 1.2337,
"step": 11880
},
{
"epoch": 0.3,
"grad_norm": 2.709670305252075,
"learning_rate": 1.2277966101694917e-05,
"loss": 1.3422,
"step": 11890
},
{
"epoch": 0.3,
"grad_norm": 4.940807342529297,
"learning_rate": 1.2271186440677966e-05,
"loss": 1.4152,
"step": 11900
},
{
"epoch": 0.3,
"grad_norm": 6.757214546203613,
"learning_rate": 1.2264406779661018e-05,
"loss": 1.6552,
"step": 11910
},
{
"epoch": 0.3,
"grad_norm": 15.510869026184082,
"learning_rate": 1.2257627118644069e-05,
"loss": 1.3236,
"step": 11920
},
{
"epoch": 0.3,
"grad_norm": 1.592451810836792,
"learning_rate": 1.225084745762712e-05,
"loss": 1.4875,
"step": 11930
},
{
"epoch": 0.3,
"grad_norm": 5.799335479736328,
"learning_rate": 1.224406779661017e-05,
"loss": 1.5547,
"step": 11940
},
{
"epoch": 0.3,
"grad_norm": 2.2848143577575684,
"learning_rate": 1.2237288135593222e-05,
"loss": 1.5282,
"step": 11950
},
{
"epoch": 0.3,
"grad_norm": 1.5773121118545532,
"learning_rate": 1.2230508474576273e-05,
"loss": 1.3649,
"step": 11960
},
{
"epoch": 0.3,
"grad_norm": 9.566110610961914,
"learning_rate": 1.2223728813559323e-05,
"loss": 1.4326,
"step": 11970
},
{
"epoch": 0.3,
"grad_norm": 2.2648942470550537,
"learning_rate": 1.2216949152542374e-05,
"loss": 1.2719,
"step": 11980
},
{
"epoch": 0.3,
"grad_norm": 8.525247573852539,
"learning_rate": 1.2210169491525426e-05,
"loss": 1.3143,
"step": 11990
},
{
"epoch": 0.3,
"grad_norm": 2.6508049964904785,
"learning_rate": 1.2203389830508477e-05,
"loss": 1.5295,
"step": 12000
},
{
"epoch": 0.3,
"eval_loss": 1.422916293144226,
"eval_runtime": 66.1433,
"eval_samples_per_second": 15.119,
"eval_steps_per_second": 15.119,
"step": 12000
},
{
"epoch": 0.3,
"grad_norm": 6.6923394203186035,
"learning_rate": 1.2196610169491527e-05,
"loss": 1.4182,
"step": 12010
},
{
"epoch": 0.3,
"grad_norm": 4.860045433044434,
"learning_rate": 1.2189830508474578e-05,
"loss": 1.3847,
"step": 12020
},
{
"epoch": 0.3,
"grad_norm": 5.437830924987793,
"learning_rate": 1.2183050847457628e-05,
"loss": 1.3516,
"step": 12030
},
{
"epoch": 0.3,
"grad_norm": 2.578477382659912,
"learning_rate": 1.2176271186440677e-05,
"loss": 1.4079,
"step": 12040
},
{
"epoch": 0.3,
"grad_norm": 6.06549072265625,
"learning_rate": 1.2169491525423729e-05,
"loss": 1.3579,
"step": 12050
},
{
"epoch": 0.3,
"grad_norm": 7.3074750900268555,
"learning_rate": 1.216271186440678e-05,
"loss": 1.3398,
"step": 12060
},
{
"epoch": 0.3,
"grad_norm": 6.550406455993652,
"learning_rate": 1.2155932203389832e-05,
"loss": 1.4268,
"step": 12070
},
{
"epoch": 0.3,
"grad_norm": 8.258136749267578,
"learning_rate": 1.2149152542372881e-05,
"loss": 1.4757,
"step": 12080
},
{
"epoch": 0.3,
"grad_norm": 3.204159736633301,
"learning_rate": 1.2142372881355933e-05,
"loss": 1.4967,
"step": 12090
},
{
"epoch": 0.3,
"grad_norm": 6.911435127258301,
"learning_rate": 1.2135593220338984e-05,
"loss": 1.3255,
"step": 12100
},
{
"epoch": 0.3,
"grad_norm": 7.780226230621338,
"learning_rate": 1.2128813559322034e-05,
"loss": 1.3092,
"step": 12110
},
{
"epoch": 0.3,
"grad_norm": 3.6478915214538574,
"learning_rate": 1.2122033898305085e-05,
"loss": 1.5233,
"step": 12120
},
{
"epoch": 0.3,
"grad_norm": 2.810391664505005,
"learning_rate": 1.2115254237288137e-05,
"loss": 1.3415,
"step": 12130
},
{
"epoch": 0.3,
"grad_norm": 2.897092580795288,
"learning_rate": 1.2108474576271188e-05,
"loss": 1.3495,
"step": 12140
},
{
"epoch": 0.3,
"grad_norm": 7.986344814300537,
"learning_rate": 1.2101694915254238e-05,
"loss": 1.2973,
"step": 12150
},
{
"epoch": 0.3,
"grad_norm": 5.259188652038574,
"learning_rate": 1.209491525423729e-05,
"loss": 1.5173,
"step": 12160
},
{
"epoch": 0.3,
"grad_norm": 3.554582357406616,
"learning_rate": 1.208813559322034e-05,
"loss": 1.4427,
"step": 12170
},
{
"epoch": 0.3,
"grad_norm": 1.8010913133621216,
"learning_rate": 1.2081355932203392e-05,
"loss": 1.2691,
"step": 12180
},
{
"epoch": 0.3,
"grad_norm": 7.936720371246338,
"learning_rate": 1.2074576271186442e-05,
"loss": 1.2707,
"step": 12190
},
{
"epoch": 0.3,
"grad_norm": 3.37536883354187,
"learning_rate": 1.2067796610169493e-05,
"loss": 1.3896,
"step": 12200
},
{
"epoch": 0.31,
"grad_norm": 3.7712466716766357,
"learning_rate": 1.2061016949152544e-05,
"loss": 1.6235,
"step": 12210
},
{
"epoch": 0.31,
"grad_norm": 6.127082824707031,
"learning_rate": 1.2054237288135596e-05,
"loss": 1.465,
"step": 12220
},
{
"epoch": 0.31,
"grad_norm": 2.607111930847168,
"learning_rate": 1.2047457627118646e-05,
"loss": 1.2618,
"step": 12230
},
{
"epoch": 0.31,
"grad_norm": 5.540430545806885,
"learning_rate": 1.2040677966101695e-05,
"loss": 1.3243,
"step": 12240
},
{
"epoch": 0.31,
"grad_norm": 2.288236141204834,
"learning_rate": 1.2033898305084745e-05,
"loss": 1.4512,
"step": 12250
},
{
"epoch": 0.31,
"grad_norm": 4.430274963378906,
"learning_rate": 1.2027118644067796e-05,
"loss": 1.3736,
"step": 12260
},
{
"epoch": 0.31,
"grad_norm": 5.795346260070801,
"learning_rate": 1.2020338983050848e-05,
"loss": 1.4164,
"step": 12270
},
{
"epoch": 0.31,
"grad_norm": 1.6566057205200195,
"learning_rate": 1.20135593220339e-05,
"loss": 1.4127,
"step": 12280
},
{
"epoch": 0.31,
"grad_norm": 2.7761037349700928,
"learning_rate": 1.2006779661016949e-05,
"loss": 1.4318,
"step": 12290
},
{
"epoch": 0.31,
"grad_norm": 3.363727569580078,
"learning_rate": 1.2e-05,
"loss": 1.4556,
"step": 12300
},
{
"epoch": 0.31,
"grad_norm": 5.918773174285889,
"learning_rate": 1.1993220338983052e-05,
"loss": 1.4288,
"step": 12310
},
{
"epoch": 0.31,
"grad_norm": 3.683154821395874,
"learning_rate": 1.1986440677966103e-05,
"loss": 1.3806,
"step": 12320
},
{
"epoch": 0.31,
"grad_norm": 10.262763977050781,
"learning_rate": 1.1979661016949153e-05,
"loss": 1.4731,
"step": 12330
},
{
"epoch": 0.31,
"grad_norm": 2.709179162979126,
"learning_rate": 1.1972881355932204e-05,
"loss": 1.43,
"step": 12340
},
{
"epoch": 0.31,
"grad_norm": 6.632396221160889,
"learning_rate": 1.1966101694915256e-05,
"loss": 1.5524,
"step": 12350
},
{
"epoch": 0.31,
"grad_norm": 2.5457146167755127,
"learning_rate": 1.1959322033898307e-05,
"loss": 1.4103,
"step": 12360
},
{
"epoch": 0.31,
"grad_norm": 3.7321617603302,
"learning_rate": 1.1952542372881357e-05,
"loss": 1.7357,
"step": 12370
},
{
"epoch": 0.31,
"grad_norm": 13.385096549987793,
"learning_rate": 1.1945762711864408e-05,
"loss": 1.5424,
"step": 12380
},
{
"epoch": 0.31,
"grad_norm": 4.128885746002197,
"learning_rate": 1.193898305084746e-05,
"loss": 1.4083,
"step": 12390
},
{
"epoch": 0.31,
"grad_norm": 3.696180820465088,
"learning_rate": 1.1932203389830511e-05,
"loss": 1.4562,
"step": 12400
},
{
"epoch": 0.31,
"grad_norm": 2.035158157348633,
"learning_rate": 1.192542372881356e-05,
"loss": 1.4134,
"step": 12410
},
{
"epoch": 0.31,
"grad_norm": 5.3175787925720215,
"learning_rate": 1.1918644067796612e-05,
"loss": 1.4388,
"step": 12420
},
{
"epoch": 0.31,
"grad_norm": 4.401218891143799,
"learning_rate": 1.1911864406779663e-05,
"loss": 1.3129,
"step": 12430
},
{
"epoch": 0.31,
"grad_norm": 2.708665132522583,
"learning_rate": 1.1905084745762713e-05,
"loss": 1.3967,
"step": 12440
},
{
"epoch": 0.31,
"grad_norm": 4.167813301086426,
"learning_rate": 1.1898305084745763e-05,
"loss": 1.5785,
"step": 12450
},
{
"epoch": 0.31,
"grad_norm": 6.84459114074707,
"learning_rate": 1.1891525423728814e-05,
"loss": 1.4729,
"step": 12460
},
{
"epoch": 0.31,
"grad_norm": 7.7822651863098145,
"learning_rate": 1.1884745762711864e-05,
"loss": 1.3682,
"step": 12470
},
{
"epoch": 0.31,
"grad_norm": 14.827698707580566,
"learning_rate": 1.1877966101694915e-05,
"loss": 1.316,
"step": 12480
},
{
"epoch": 0.31,
"grad_norm": 7.276782512664795,
"learning_rate": 1.1871186440677967e-05,
"loss": 1.4088,
"step": 12490
},
{
"epoch": 0.31,
"grad_norm": 2.0169296264648438,
"learning_rate": 1.1864406779661018e-05,
"loss": 1.287,
"step": 12500
},
{
"epoch": 0.31,
"eval_loss": 1.3866758346557617,
"eval_runtime": 66.1437,
"eval_samples_per_second": 15.119,
"eval_steps_per_second": 15.119,
"step": 12500
},
{
"epoch": 0.31,
"grad_norm": 3.162349224090576,
"learning_rate": 1.1857627118644068e-05,
"loss": 1.6078,
"step": 12510
},
{
"epoch": 0.31,
"grad_norm": 2.835212469100952,
"learning_rate": 1.185084745762712e-05,
"loss": 1.3934,
"step": 12520
},
{
"epoch": 0.31,
"grad_norm": 3.5523407459259033,
"learning_rate": 1.184406779661017e-05,
"loss": 1.4201,
"step": 12530
},
{
"epoch": 0.31,
"grad_norm": 2.8521957397460938,
"learning_rate": 1.183728813559322e-05,
"loss": 1.1765,
"step": 12540
},
{
"epoch": 0.31,
"grad_norm": 6.991311073303223,
"learning_rate": 1.1830508474576272e-05,
"loss": 1.2703,
"step": 12550
},
{
"epoch": 0.31,
"grad_norm": 9.624125480651855,
"learning_rate": 1.1823728813559323e-05,
"loss": 1.3824,
"step": 12560
},
{
"epoch": 0.31,
"grad_norm": 2.1948940753936768,
"learning_rate": 1.1816949152542375e-05,
"loss": 1.4304,
"step": 12570
},
{
"epoch": 0.31,
"grad_norm": 2.5481557846069336,
"learning_rate": 1.1810169491525424e-05,
"loss": 1.3297,
"step": 12580
},
{
"epoch": 0.31,
"grad_norm": 2.6598098278045654,
"learning_rate": 1.1803389830508476e-05,
"loss": 1.1585,
"step": 12590
},
{
"epoch": 0.32,
"grad_norm": 6.169371128082275,
"learning_rate": 1.1796610169491527e-05,
"loss": 1.4042,
"step": 12600
},
{
"epoch": 0.32,
"grad_norm": 2.9693713188171387,
"learning_rate": 1.1789830508474578e-05,
"loss": 1.4801,
"step": 12610
},
{
"epoch": 0.32,
"grad_norm": 5.916517734527588,
"learning_rate": 1.1783050847457628e-05,
"loss": 1.3524,
"step": 12620
},
{
"epoch": 0.32,
"grad_norm": 12.339306831359863,
"learning_rate": 1.177627118644068e-05,
"loss": 1.4822,
"step": 12630
},
{
"epoch": 0.32,
"grad_norm": 5.854163646697998,
"learning_rate": 1.1769491525423731e-05,
"loss": 1.6638,
"step": 12640
},
{
"epoch": 0.32,
"grad_norm": 4.270496368408203,
"learning_rate": 1.1762711864406782e-05,
"loss": 1.3407,
"step": 12650
},
{
"epoch": 0.32,
"grad_norm": 6.106138706207275,
"learning_rate": 1.175593220338983e-05,
"loss": 1.4408,
"step": 12660
},
{
"epoch": 0.32,
"grad_norm": 1.9153695106506348,
"learning_rate": 1.1749152542372882e-05,
"loss": 1.3821,
"step": 12670
},
{
"epoch": 0.32,
"grad_norm": 4.479553699493408,
"learning_rate": 1.1742372881355931e-05,
"loss": 1.3252,
"step": 12680
},
{
"epoch": 0.32,
"grad_norm": 4.480614185333252,
"learning_rate": 1.1735593220338983e-05,
"loss": 1.5234,
"step": 12690
},
{
"epoch": 0.32,
"grad_norm": 5.578995704650879,
"learning_rate": 1.1728813559322034e-05,
"loss": 1.4655,
"step": 12700
},
{
"epoch": 0.32,
"grad_norm": 4.981246471405029,
"learning_rate": 1.1722033898305086e-05,
"loss": 1.3632,
"step": 12710
},
{
"epoch": 0.32,
"grad_norm": 12.358545303344727,
"learning_rate": 1.1715254237288135e-05,
"loss": 1.3998,
"step": 12720
},
{
"epoch": 0.32,
"grad_norm": 14.668237686157227,
"learning_rate": 1.1708474576271187e-05,
"loss": 1.3784,
"step": 12730
},
{
"epoch": 0.32,
"grad_norm": 3.8441002368927,
"learning_rate": 1.1701694915254238e-05,
"loss": 1.482,
"step": 12740
},
{
"epoch": 0.32,
"grad_norm": 4.92882776260376,
"learning_rate": 1.169491525423729e-05,
"loss": 1.4855,
"step": 12750
},
{
"epoch": 0.32,
"grad_norm": 5.74147891998291,
"learning_rate": 1.168813559322034e-05,
"loss": 1.4695,
"step": 12760
},
{
"epoch": 0.32,
"grad_norm": 4.936017990112305,
"learning_rate": 1.168135593220339e-05,
"loss": 1.3325,
"step": 12770
},
{
"epoch": 0.32,
"grad_norm": 4.67604923248291,
"learning_rate": 1.1674576271186442e-05,
"loss": 1.5104,
"step": 12780
},
{
"epoch": 0.32,
"grad_norm": 2.439807891845703,
"learning_rate": 1.1667796610169494e-05,
"loss": 1.3626,
"step": 12790
},
{
"epoch": 0.32,
"grad_norm": 1.6871551275253296,
"learning_rate": 1.1661016949152543e-05,
"loss": 1.5824,
"step": 12800
},
{
"epoch": 0.32,
"grad_norm": 6.6204094886779785,
"learning_rate": 1.1654237288135595e-05,
"loss": 1.2672,
"step": 12810
},
{
"epoch": 0.32,
"grad_norm": 3.2927229404449463,
"learning_rate": 1.1647457627118646e-05,
"loss": 1.3032,
"step": 12820
},
{
"epoch": 0.32,
"grad_norm": 8.059797286987305,
"learning_rate": 1.1640677966101697e-05,
"loss": 1.3087,
"step": 12830
},
{
"epoch": 0.32,
"grad_norm": 3.8133039474487305,
"learning_rate": 1.1633898305084747e-05,
"loss": 1.4655,
"step": 12840
},
{
"epoch": 0.32,
"grad_norm": 4.857828140258789,
"learning_rate": 1.1627118644067799e-05,
"loss": 1.3882,
"step": 12850
},
{
"epoch": 0.32,
"grad_norm": 4.760891914367676,
"learning_rate": 1.162033898305085e-05,
"loss": 1.438,
"step": 12860
},
{
"epoch": 0.32,
"grad_norm": 5.746943473815918,
"learning_rate": 1.1613559322033898e-05,
"loss": 1.29,
"step": 12870
},
{
"epoch": 0.32,
"grad_norm": 10.054022789001465,
"learning_rate": 1.160677966101695e-05,
"loss": 1.4788,
"step": 12880
},
{
"epoch": 0.32,
"grad_norm": 11.264670372009277,
"learning_rate": 1.16e-05,
"loss": 1.4834,
"step": 12890
},
{
"epoch": 0.32,
"grad_norm": 2.883831739425659,
"learning_rate": 1.159322033898305e-05,
"loss": 1.3276,
"step": 12900
},
{
"epoch": 0.32,
"grad_norm": 3.7215306758880615,
"learning_rate": 1.1586440677966102e-05,
"loss": 1.3578,
"step": 12910
},
{
"epoch": 0.32,
"grad_norm": 3.2970972061157227,
"learning_rate": 1.1579661016949153e-05,
"loss": 1.3238,
"step": 12920
},
{
"epoch": 0.32,
"grad_norm": 5.821772575378418,
"learning_rate": 1.1572881355932205e-05,
"loss": 1.3903,
"step": 12930
},
{
"epoch": 0.32,
"grad_norm": 2.4745774269104004,
"learning_rate": 1.1566101694915254e-05,
"loss": 1.166,
"step": 12940
},
{
"epoch": 0.32,
"grad_norm": 2.578946590423584,
"learning_rate": 1.1559322033898306e-05,
"loss": 1.483,
"step": 12950
},
{
"epoch": 0.32,
"grad_norm": 3.525611400604248,
"learning_rate": 1.1552542372881357e-05,
"loss": 1.3907,
"step": 12960
},
{
"epoch": 0.32,
"grad_norm": 4.851688385009766,
"learning_rate": 1.1545762711864409e-05,
"loss": 1.2686,
"step": 12970
},
{
"epoch": 0.32,
"grad_norm": 2.2654104232788086,
"learning_rate": 1.1538983050847458e-05,
"loss": 1.3566,
"step": 12980
},
{
"epoch": 0.32,
"grad_norm": 4.508155345916748,
"learning_rate": 1.153220338983051e-05,
"loss": 1.434,
"step": 12990
},
{
"epoch": 0.33,
"grad_norm": 4.140927791595459,
"learning_rate": 1.1525423728813561e-05,
"loss": 1.4034,
"step": 13000
},
{
"epoch": 0.33,
"eval_loss": 1.4143681526184082,
"eval_runtime": 66.1661,
"eval_samples_per_second": 15.113,
"eval_steps_per_second": 15.113,
"step": 13000
},
{
"epoch": 0.33,
"grad_norm": 4.795019149780273,
"learning_rate": 1.151864406779661e-05,
"loss": 1.617,
"step": 13010
},
{
"epoch": 0.33,
"grad_norm": 4.142047882080078,
"learning_rate": 1.1511864406779662e-05,
"loss": 1.4168,
"step": 13020
},
{
"epoch": 0.33,
"grad_norm": 3.405993700027466,
"learning_rate": 1.1505084745762714e-05,
"loss": 1.4029,
"step": 13030
},
{
"epoch": 0.33,
"grad_norm": 4.590165615081787,
"learning_rate": 1.1498305084745765e-05,
"loss": 1.2981,
"step": 13040
},
{
"epoch": 0.33,
"grad_norm": 1.7291467189788818,
"learning_rate": 1.1491525423728815e-05,
"loss": 1.3752,
"step": 13050
},
{
"epoch": 0.33,
"grad_norm": 2.5406222343444824,
"learning_rate": 1.1484745762711866e-05,
"loss": 1.4635,
"step": 13060
},
{
"epoch": 0.33,
"grad_norm": 5.161932468414307,
"learning_rate": 1.1477966101694916e-05,
"loss": 1.3763,
"step": 13070
},
{
"epoch": 0.33,
"grad_norm": 5.527559280395508,
"learning_rate": 1.1471186440677965e-05,
"loss": 1.3618,
"step": 13080
},
{
"epoch": 0.33,
"grad_norm": 3.843920946121216,
"learning_rate": 1.1464406779661017e-05,
"loss": 1.312,
"step": 13090
},
{
"epoch": 0.33,
"grad_norm": 4.652730464935303,
"learning_rate": 1.1457627118644068e-05,
"loss": 1.3818,
"step": 13100
},
{
"epoch": 0.33,
"grad_norm": 3.841367244720459,
"learning_rate": 1.145084745762712e-05,
"loss": 1.4252,
"step": 13110
},
{
"epoch": 0.33,
"grad_norm": 4.024381637573242,
"learning_rate": 1.144406779661017e-05,
"loss": 1.4014,
"step": 13120
},
{
"epoch": 0.33,
"grad_norm": 3.9260082244873047,
"learning_rate": 1.143728813559322e-05,
"loss": 1.4733,
"step": 13130
},
{
"epoch": 0.33,
"grad_norm": 7.934506893157959,
"learning_rate": 1.1430508474576272e-05,
"loss": 1.6533,
"step": 13140
},
{
"epoch": 0.33,
"grad_norm": 4.169222354888916,
"learning_rate": 1.1423728813559322e-05,
"loss": 1.4004,
"step": 13150
},
{
"epoch": 0.33,
"grad_norm": 3.165480852127075,
"learning_rate": 1.1416949152542373e-05,
"loss": 1.2961,
"step": 13160
},
{
"epoch": 0.33,
"grad_norm": 1.703479528427124,
"learning_rate": 1.1410169491525425e-05,
"loss": 1.5203,
"step": 13170
},
{
"epoch": 0.33,
"grad_norm": 15.553948402404785,
"learning_rate": 1.1403389830508476e-05,
"loss": 1.4719,
"step": 13180
},
{
"epoch": 0.33,
"grad_norm": 3.422497034072876,
"learning_rate": 1.1396610169491526e-05,
"loss": 1.4232,
"step": 13190
},
{
"epoch": 0.33,
"grad_norm": 3.0440897941589355,
"learning_rate": 1.1389830508474577e-05,
"loss": 1.3813,
"step": 13200
},
{
"epoch": 0.33,
"grad_norm": 3.650101661682129,
"learning_rate": 1.1383050847457629e-05,
"loss": 1.4398,
"step": 13210
},
{
"epoch": 0.33,
"grad_norm": 9.34521484375,
"learning_rate": 1.137627118644068e-05,
"loss": 1.4457,
"step": 13220
},
{
"epoch": 0.33,
"grad_norm": 2.5624570846557617,
"learning_rate": 1.136949152542373e-05,
"loss": 1.3008,
"step": 13230
},
{
"epoch": 0.33,
"grad_norm": 2.584078311920166,
"learning_rate": 1.1362711864406781e-05,
"loss": 1.3545,
"step": 13240
},
{
"epoch": 0.33,
"grad_norm": 4.867392539978027,
"learning_rate": 1.1355932203389833e-05,
"loss": 1.4067,
"step": 13250
},
{
"epoch": 0.33,
"grad_norm": 2.185886859893799,
"learning_rate": 1.1349152542372884e-05,
"loss": 1.3976,
"step": 13260
},
{
"epoch": 0.33,
"grad_norm": 4.314910888671875,
"learning_rate": 1.1342372881355934e-05,
"loss": 1.3568,
"step": 13270
},
{
"epoch": 0.33,
"grad_norm": 5.425469875335693,
"learning_rate": 1.1335593220338983e-05,
"loss": 1.5861,
"step": 13280
},
{
"epoch": 0.33,
"grad_norm": 8.086491584777832,
"learning_rate": 1.1328813559322033e-05,
"loss": 1.567,
"step": 13290
},
{
"epoch": 0.33,
"grad_norm": 4.411368370056152,
"learning_rate": 1.1322033898305084e-05,
"loss": 1.2149,
"step": 13300
},
{
"epoch": 0.33,
"grad_norm": 5.792349815368652,
"learning_rate": 1.1315254237288136e-05,
"loss": 1.4701,
"step": 13310
},
{
"epoch": 0.33,
"grad_norm": 2.967857837677002,
"learning_rate": 1.1308474576271187e-05,
"loss": 1.3723,
"step": 13320
},
{
"epoch": 0.33,
"grad_norm": 4.052553176879883,
"learning_rate": 1.1301694915254237e-05,
"loss": 1.3078,
"step": 13330
},
{
"epoch": 0.33,
"grad_norm": 1.5086994171142578,
"learning_rate": 1.1294915254237288e-05,
"loss": 1.4948,
"step": 13340
},
{
"epoch": 0.33,
"grad_norm": 8.732611656188965,
"learning_rate": 1.128813559322034e-05,
"loss": 1.5011,
"step": 13350
},
{
"epoch": 0.33,
"grad_norm": 6.4710235595703125,
"learning_rate": 1.1281355932203391e-05,
"loss": 1.3214,
"step": 13360
},
{
"epoch": 0.33,
"grad_norm": 5.698025226593018,
"learning_rate": 1.1274576271186441e-05,
"loss": 1.4719,
"step": 13370
},
{
"epoch": 0.33,
"grad_norm": 1.9803974628448486,
"learning_rate": 1.1267796610169492e-05,
"loss": 1.4078,
"step": 13380
},
{
"epoch": 0.33,
"grad_norm": 6.019112586975098,
"learning_rate": 1.1261016949152544e-05,
"loss": 1.4646,
"step": 13390
},
{
"epoch": 0.34,
"grad_norm": 5.664279937744141,
"learning_rate": 1.1254237288135595e-05,
"loss": 1.5236,
"step": 13400
},
{
"epoch": 0.34,
"grad_norm": 3.8619132041931152,
"learning_rate": 1.1247457627118645e-05,
"loss": 1.3184,
"step": 13410
},
{
"epoch": 0.34,
"grad_norm": 2.45538330078125,
"learning_rate": 1.1240677966101696e-05,
"loss": 1.3868,
"step": 13420
},
{
"epoch": 0.34,
"grad_norm": 3.7637529373168945,
"learning_rate": 1.1233898305084748e-05,
"loss": 1.4405,
"step": 13430
},
{
"epoch": 0.34,
"grad_norm": 4.774598121643066,
"learning_rate": 1.1227118644067799e-05,
"loss": 1.287,
"step": 13440
},
{
"epoch": 0.34,
"grad_norm": 3.2298827171325684,
"learning_rate": 1.1220338983050849e-05,
"loss": 1.169,
"step": 13450
},
{
"epoch": 0.34,
"grad_norm": 1.9264500141143799,
"learning_rate": 1.12135593220339e-05,
"loss": 1.4071,
"step": 13460
},
{
"epoch": 0.34,
"grad_norm": 3.1234660148620605,
"learning_rate": 1.1206779661016951e-05,
"loss": 1.3734,
"step": 13470
},
{
"epoch": 0.34,
"grad_norm": 1.4275059700012207,
"learning_rate": 1.1200000000000001e-05,
"loss": 1.4451,
"step": 13480
},
{
"epoch": 0.34,
"grad_norm": 4.883017063140869,
"learning_rate": 1.1193220338983051e-05,
"loss": 1.4083,
"step": 13490
},
{
"epoch": 0.34,
"grad_norm": 7.2406325340271,
"learning_rate": 1.1186440677966102e-05,
"loss": 1.209,
"step": 13500
},
{
"epoch": 0.34,
"eval_loss": 1.4201830625534058,
"eval_runtime": 66.1321,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 13500
},
{
"epoch": 0.34,
"grad_norm": 5.777285099029541,
"learning_rate": 1.1179661016949152e-05,
"loss": 1.3445,
"step": 13510
},
{
"epoch": 0.34,
"grad_norm": 3.1429507732391357,
"learning_rate": 1.1172881355932203e-05,
"loss": 1.4894,
"step": 13520
},
{
"epoch": 0.34,
"grad_norm": 10.670025825500488,
"learning_rate": 1.1166101694915255e-05,
"loss": 1.4,
"step": 13530
},
{
"epoch": 0.34,
"grad_norm": 3.0807442665100098,
"learning_rate": 1.1159322033898306e-05,
"loss": 1.3409,
"step": 13540
},
{
"epoch": 0.34,
"grad_norm": 7.570891857147217,
"learning_rate": 1.1152542372881356e-05,
"loss": 1.3676,
"step": 13550
},
{
"epoch": 0.34,
"grad_norm": 11.09575080871582,
"learning_rate": 1.1145762711864407e-05,
"loss": 1.402,
"step": 13560
},
{
"epoch": 0.34,
"grad_norm": 6.450993537902832,
"learning_rate": 1.1138983050847459e-05,
"loss": 1.4072,
"step": 13570
},
{
"epoch": 0.34,
"grad_norm": 6.089435577392578,
"learning_rate": 1.113220338983051e-05,
"loss": 1.3705,
"step": 13580
},
{
"epoch": 0.34,
"grad_norm": 4.839201927185059,
"learning_rate": 1.112542372881356e-05,
"loss": 1.5674,
"step": 13590
},
{
"epoch": 0.34,
"grad_norm": 12.40379810333252,
"learning_rate": 1.1118644067796611e-05,
"loss": 1.2697,
"step": 13600
},
{
"epoch": 0.34,
"grad_norm": 5.095344066619873,
"learning_rate": 1.1111864406779663e-05,
"loss": 1.1917,
"step": 13610
},
{
"epoch": 0.34,
"grad_norm": 3.252272367477417,
"learning_rate": 1.1105084745762712e-05,
"loss": 1.3542,
"step": 13620
},
{
"epoch": 0.34,
"grad_norm": 8.330405235290527,
"learning_rate": 1.1098305084745764e-05,
"loss": 1.4464,
"step": 13630
},
{
"epoch": 0.34,
"grad_norm": 1.9642021656036377,
"learning_rate": 1.1091525423728815e-05,
"loss": 1.2904,
"step": 13640
},
{
"epoch": 0.34,
"grad_norm": 3.2120866775512695,
"learning_rate": 1.1084745762711867e-05,
"loss": 1.3887,
"step": 13650
},
{
"epoch": 0.34,
"grad_norm": 3.582340717315674,
"learning_rate": 1.1077966101694916e-05,
"loss": 1.5277,
"step": 13660
},
{
"epoch": 0.34,
"grad_norm": 3.2663116455078125,
"learning_rate": 1.1071186440677968e-05,
"loss": 1.2393,
"step": 13670
},
{
"epoch": 0.34,
"grad_norm": 7.755127906799316,
"learning_rate": 1.1064406779661019e-05,
"loss": 1.3327,
"step": 13680
},
{
"epoch": 0.34,
"grad_norm": 4.777354717254639,
"learning_rate": 1.105762711864407e-05,
"loss": 1.4229,
"step": 13690
},
{
"epoch": 0.34,
"grad_norm": 4.864866733551025,
"learning_rate": 1.1050847457627118e-05,
"loss": 1.5831,
"step": 13700
},
{
"epoch": 0.34,
"grad_norm": 3.381706953048706,
"learning_rate": 1.104406779661017e-05,
"loss": 1.5776,
"step": 13710
},
{
"epoch": 0.34,
"grad_norm": 2.4377899169921875,
"learning_rate": 1.1037288135593221e-05,
"loss": 1.4538,
"step": 13720
},
{
"epoch": 0.34,
"grad_norm": 3.592681646347046,
"learning_rate": 1.1030508474576271e-05,
"loss": 1.5015,
"step": 13730
},
{
"epoch": 0.34,
"grad_norm": 4.690391540527344,
"learning_rate": 1.1023728813559322e-05,
"loss": 1.3766,
"step": 13740
},
{
"epoch": 0.34,
"grad_norm": 6.648055553436279,
"learning_rate": 1.1016949152542374e-05,
"loss": 1.319,
"step": 13750
},
{
"epoch": 0.34,
"grad_norm": 4.713019847869873,
"learning_rate": 1.1010169491525423e-05,
"loss": 1.4213,
"step": 13760
},
{
"epoch": 0.34,
"grad_norm": 4.177305698394775,
"learning_rate": 1.1003389830508475e-05,
"loss": 1.5414,
"step": 13770
},
{
"epoch": 0.34,
"grad_norm": 4.832136154174805,
"learning_rate": 1.0996610169491526e-05,
"loss": 1.457,
"step": 13780
},
{
"epoch": 0.34,
"grad_norm": 5.470792770385742,
"learning_rate": 1.0989830508474578e-05,
"loss": 1.3565,
"step": 13790
},
{
"epoch": 0.34,
"grad_norm": 2.7103376388549805,
"learning_rate": 1.0983050847457627e-05,
"loss": 1.0967,
"step": 13800
},
{
"epoch": 0.35,
"grad_norm": 5.0832390785217285,
"learning_rate": 1.0976271186440679e-05,
"loss": 1.5281,
"step": 13810
},
{
"epoch": 0.35,
"grad_norm": 9.628556251525879,
"learning_rate": 1.096949152542373e-05,
"loss": 1.3223,
"step": 13820
},
{
"epoch": 0.35,
"grad_norm": 5.501499176025391,
"learning_rate": 1.0962711864406782e-05,
"loss": 1.452,
"step": 13830
},
{
"epoch": 0.35,
"grad_norm": 1.4253193140029907,
"learning_rate": 1.0955932203389831e-05,
"loss": 1.5329,
"step": 13840
},
{
"epoch": 0.35,
"grad_norm": 3.109076976776123,
"learning_rate": 1.0949152542372883e-05,
"loss": 1.2804,
"step": 13850
},
{
"epoch": 0.35,
"grad_norm": 7.186499118804932,
"learning_rate": 1.0942372881355934e-05,
"loss": 1.3599,
"step": 13860
},
{
"epoch": 0.35,
"grad_norm": 3.0720372200012207,
"learning_rate": 1.0935593220338985e-05,
"loss": 1.4686,
"step": 13870
},
{
"epoch": 0.35,
"grad_norm": 3.5827410221099854,
"learning_rate": 1.0928813559322035e-05,
"loss": 1.3139,
"step": 13880
},
{
"epoch": 0.35,
"grad_norm": 12.06527328491211,
"learning_rate": 1.0922033898305087e-05,
"loss": 1.4268,
"step": 13890
},
{
"epoch": 0.35,
"grad_norm": 4.608367919921875,
"learning_rate": 1.0915254237288135e-05,
"loss": 1.2591,
"step": 13900
},
{
"epoch": 0.35,
"grad_norm": 13.63802433013916,
"learning_rate": 1.0908474576271186e-05,
"loss": 1.2512,
"step": 13910
},
{
"epoch": 0.35,
"grad_norm": 2.2884724140167236,
"learning_rate": 1.0901694915254237e-05,
"loss": 1.5108,
"step": 13920
},
{
"epoch": 0.35,
"grad_norm": 4.693191051483154,
"learning_rate": 1.0894915254237289e-05,
"loss": 1.1383,
"step": 13930
},
{
"epoch": 0.35,
"grad_norm": 4.830557823181152,
"learning_rate": 1.0888135593220339e-05,
"loss": 1.3608,
"step": 13940
},
{
"epoch": 0.35,
"grad_norm": 5.044919490814209,
"learning_rate": 1.088135593220339e-05,
"loss": 1.2956,
"step": 13950
},
{
"epoch": 0.35,
"grad_norm": 2.327402114868164,
"learning_rate": 1.0874576271186441e-05,
"loss": 1.4006,
"step": 13960
},
{
"epoch": 0.35,
"grad_norm": 8.76153564453125,
"learning_rate": 1.0867796610169493e-05,
"loss": 1.4544,
"step": 13970
},
{
"epoch": 0.35,
"grad_norm": 3.762763023376465,
"learning_rate": 1.0861016949152542e-05,
"loss": 1.4545,
"step": 13980
},
{
"epoch": 0.35,
"grad_norm": 4.206634521484375,
"learning_rate": 1.0854237288135594e-05,
"loss": 1.3256,
"step": 13990
},
{
"epoch": 0.35,
"grad_norm": 3.6637351512908936,
"learning_rate": 1.0847457627118645e-05,
"loss": 1.5003,
"step": 14000
},
{
"epoch": 0.35,
"eval_loss": 1.414922833442688,
"eval_runtime": 66.1224,
"eval_samples_per_second": 15.123,
"eval_steps_per_second": 15.123,
"step": 14000
},
{
"epoch": 0.35,
"grad_norm": 3.5580198764801025,
"learning_rate": 1.0840677966101697e-05,
"loss": 1.2528,
"step": 14010
},
{
"epoch": 0.35,
"grad_norm": 4.473653316497803,
"learning_rate": 1.0833898305084746e-05,
"loss": 1.4678,
"step": 14020
},
{
"epoch": 0.35,
"grad_norm": 2.7036588191986084,
"learning_rate": 1.0827118644067798e-05,
"loss": 1.5327,
"step": 14030
},
{
"epoch": 0.35,
"grad_norm": 2.9595816135406494,
"learning_rate": 1.0820338983050849e-05,
"loss": 1.3218,
"step": 14040
},
{
"epoch": 0.35,
"grad_norm": 2.984138011932373,
"learning_rate": 1.08135593220339e-05,
"loss": 1.3408,
"step": 14050
},
{
"epoch": 0.35,
"grad_norm": 3.022242307662964,
"learning_rate": 1.080677966101695e-05,
"loss": 1.5133,
"step": 14060
},
{
"epoch": 0.35,
"grad_norm": 4.826704978942871,
"learning_rate": 1.0800000000000002e-05,
"loss": 1.3917,
"step": 14070
},
{
"epoch": 0.35,
"grad_norm": 7.675862789154053,
"learning_rate": 1.0793220338983053e-05,
"loss": 1.2337,
"step": 14080
},
{
"epoch": 0.35,
"grad_norm": 4.611223220825195,
"learning_rate": 1.0786440677966103e-05,
"loss": 1.3582,
"step": 14090
},
{
"epoch": 0.35,
"grad_norm": 2.353292942047119,
"learning_rate": 1.0779661016949154e-05,
"loss": 1.4237,
"step": 14100
},
{
"epoch": 0.35,
"grad_norm": 4.331914901733398,
"learning_rate": 1.0772881355932204e-05,
"loss": 1.4808,
"step": 14110
},
{
"epoch": 0.35,
"grad_norm": 2.759838819503784,
"learning_rate": 1.0766101694915254e-05,
"loss": 1.4499,
"step": 14120
},
{
"epoch": 0.35,
"grad_norm": 4.08477783203125,
"learning_rate": 1.0759322033898305e-05,
"loss": 1.3125,
"step": 14130
},
{
"epoch": 0.35,
"grad_norm": 3.8619682788848877,
"learning_rate": 1.0752542372881356e-05,
"loss": 1.2508,
"step": 14140
},
{
"epoch": 0.35,
"grad_norm": 5.26367712020874,
"learning_rate": 1.0745762711864408e-05,
"loss": 1.4067,
"step": 14150
},
{
"epoch": 0.35,
"grad_norm": 8.280834197998047,
"learning_rate": 1.0738983050847457e-05,
"loss": 1.381,
"step": 14160
},
{
"epoch": 0.35,
"grad_norm": 1.3759233951568604,
"learning_rate": 1.0732203389830509e-05,
"loss": 1.4211,
"step": 14170
},
{
"epoch": 0.35,
"grad_norm": 1.8453333377838135,
"learning_rate": 1.072542372881356e-05,
"loss": 1.4187,
"step": 14180
},
{
"epoch": 0.35,
"grad_norm": 3.5439627170562744,
"learning_rate": 1.0718644067796612e-05,
"loss": 1.2863,
"step": 14190
},
{
"epoch": 0.35,
"grad_norm": 5.693528652191162,
"learning_rate": 1.0711864406779661e-05,
"loss": 1.3819,
"step": 14200
},
{
"epoch": 0.36,
"grad_norm": 6.000442028045654,
"learning_rate": 1.0705084745762713e-05,
"loss": 1.4,
"step": 14210
},
{
"epoch": 0.36,
"grad_norm": 4.452291011810303,
"learning_rate": 1.0698305084745764e-05,
"loss": 1.3242,
"step": 14220
},
{
"epoch": 0.36,
"grad_norm": 8.150368690490723,
"learning_rate": 1.0691525423728814e-05,
"loss": 1.4536,
"step": 14230
},
{
"epoch": 0.36,
"grad_norm": 6.411572456359863,
"learning_rate": 1.0684745762711865e-05,
"loss": 1.2176,
"step": 14240
},
{
"epoch": 0.36,
"grad_norm": 6.748562335968018,
"learning_rate": 1.0677966101694917e-05,
"loss": 1.3727,
"step": 14250
},
{
"epoch": 0.36,
"grad_norm": 5.915468215942383,
"learning_rate": 1.0671186440677968e-05,
"loss": 1.2339,
"step": 14260
},
{
"epoch": 0.36,
"grad_norm": 2.7658674716949463,
"learning_rate": 1.0664406779661018e-05,
"loss": 1.3729,
"step": 14270
},
{
"epoch": 0.36,
"grad_norm": 2.117175579071045,
"learning_rate": 1.065762711864407e-05,
"loss": 1.4233,
"step": 14280
},
{
"epoch": 0.36,
"grad_norm": 3.734428882598877,
"learning_rate": 1.065084745762712e-05,
"loss": 1.4449,
"step": 14290
},
{
"epoch": 0.36,
"grad_norm": 3.934438467025757,
"learning_rate": 1.0644067796610172e-05,
"loss": 1.3337,
"step": 14300
},
{
"epoch": 0.36,
"grad_norm": 10.599371910095215,
"learning_rate": 1.0637288135593222e-05,
"loss": 1.3862,
"step": 14310
},
{
"epoch": 0.36,
"grad_norm": 3.4198312759399414,
"learning_rate": 1.0630508474576271e-05,
"loss": 1.5868,
"step": 14320
},
{
"epoch": 0.36,
"grad_norm": 1.7814559936523438,
"learning_rate": 1.0623728813559323e-05,
"loss": 1.3958,
"step": 14330
},
{
"epoch": 0.36,
"grad_norm": 10.756269454956055,
"learning_rate": 1.0616949152542373e-05,
"loss": 1.516,
"step": 14340
},
{
"epoch": 0.36,
"grad_norm": 1.8235437870025635,
"learning_rate": 1.0610169491525424e-05,
"loss": 1.54,
"step": 14350
},
{
"epoch": 0.36,
"grad_norm": 7.628040790557861,
"learning_rate": 1.0603389830508475e-05,
"loss": 1.2579,
"step": 14360
},
{
"epoch": 0.36,
"grad_norm": 3.0332067012786865,
"learning_rate": 1.0596610169491525e-05,
"loss": 1.3637,
"step": 14370
},
{
"epoch": 0.36,
"grad_norm": 3.866598606109619,
"learning_rate": 1.0589830508474576e-05,
"loss": 1.5127,
"step": 14380
},
{
"epoch": 0.36,
"grad_norm": 3.0996134281158447,
"learning_rate": 1.0583050847457628e-05,
"loss": 1.538,
"step": 14390
},
{
"epoch": 0.36,
"grad_norm": 3.3183581829071045,
"learning_rate": 1.057627118644068e-05,
"loss": 1.4018,
"step": 14400
},
{
"epoch": 0.36,
"grad_norm": 5.61666202545166,
"learning_rate": 1.0569491525423729e-05,
"loss": 1.374,
"step": 14410
},
{
"epoch": 0.36,
"grad_norm": 4.340878963470459,
"learning_rate": 1.056271186440678e-05,
"loss": 1.4605,
"step": 14420
},
{
"epoch": 0.36,
"grad_norm": 3.6541993618011475,
"learning_rate": 1.0555932203389832e-05,
"loss": 1.4579,
"step": 14430
},
{
"epoch": 0.36,
"grad_norm": 6.3943772315979,
"learning_rate": 1.0549152542372883e-05,
"loss": 1.6246,
"step": 14440
},
{
"epoch": 0.36,
"grad_norm": 3.543633460998535,
"learning_rate": 1.0542372881355933e-05,
"loss": 1.5073,
"step": 14450
},
{
"epoch": 0.36,
"grad_norm": 2.0119078159332275,
"learning_rate": 1.0535593220338984e-05,
"loss": 1.367,
"step": 14460
},
{
"epoch": 0.36,
"grad_norm": 14.83475112915039,
"learning_rate": 1.0528813559322036e-05,
"loss": 1.3893,
"step": 14470
},
{
"epoch": 0.36,
"grad_norm": 2.9194772243499756,
"learning_rate": 1.0522033898305087e-05,
"loss": 1.3436,
"step": 14480
},
{
"epoch": 0.36,
"grad_norm": 1.9060604572296143,
"learning_rate": 1.0515254237288137e-05,
"loss": 1.2047,
"step": 14490
},
{
"epoch": 0.36,
"grad_norm": 2.6039791107177734,
"learning_rate": 1.0508474576271188e-05,
"loss": 1.5192,
"step": 14500
},
{
"epoch": 0.36,
"eval_loss": 1.3242599964141846,
"eval_runtime": 66.116,
"eval_samples_per_second": 15.125,
"eval_steps_per_second": 15.125,
"step": 14500
},
{
"epoch": 0.36,
"grad_norm": 6.422837257385254,
"learning_rate": 1.050169491525424e-05,
"loss": 1.4687,
"step": 14510
},
{
"epoch": 0.36,
"grad_norm": 4.683182716369629,
"learning_rate": 1.049491525423729e-05,
"loss": 1.4816,
"step": 14520
},
{
"epoch": 0.36,
"grad_norm": 3.1711997985839844,
"learning_rate": 1.0488135593220339e-05,
"loss": 1.354,
"step": 14530
},
{
"epoch": 0.36,
"grad_norm": 2.8648619651794434,
"learning_rate": 1.048135593220339e-05,
"loss": 1.3518,
"step": 14540
},
{
"epoch": 0.36,
"grad_norm": 10.453191757202148,
"learning_rate": 1.047457627118644e-05,
"loss": 1.3256,
"step": 14550
},
{
"epoch": 0.36,
"grad_norm": 4.701107978820801,
"learning_rate": 1.0467796610169491e-05,
"loss": 1.3005,
"step": 14560
},
{
"epoch": 0.36,
"grad_norm": 3.5893495082855225,
"learning_rate": 1.0461016949152543e-05,
"loss": 1.3151,
"step": 14570
},
{
"epoch": 0.36,
"grad_norm": 12.855472564697266,
"learning_rate": 1.0454237288135594e-05,
"loss": 1.5216,
"step": 14580
},
{
"epoch": 0.36,
"grad_norm": 3.115908145904541,
"learning_rate": 1.0447457627118644e-05,
"loss": 1.6356,
"step": 14590
},
{
"epoch": 0.36,
"grad_norm": 5.046034812927246,
"learning_rate": 1.0440677966101695e-05,
"loss": 1.4374,
"step": 14600
},
{
"epoch": 0.37,
"grad_norm": 8.317269325256348,
"learning_rate": 1.0433898305084747e-05,
"loss": 1.3663,
"step": 14610
},
{
"epoch": 0.37,
"grad_norm": 9.950586318969727,
"learning_rate": 1.0427118644067798e-05,
"loss": 1.5752,
"step": 14620
},
{
"epoch": 0.37,
"grad_norm": 8.241649627685547,
"learning_rate": 1.0420338983050848e-05,
"loss": 1.3704,
"step": 14630
},
{
"epoch": 0.37,
"grad_norm": 5.070446491241455,
"learning_rate": 1.04135593220339e-05,
"loss": 1.4483,
"step": 14640
},
{
"epoch": 0.37,
"grad_norm": 1.575085997581482,
"learning_rate": 1.040677966101695e-05,
"loss": 1.5115,
"step": 14650
},
{
"epoch": 0.37,
"grad_norm": 8.428694725036621,
"learning_rate": 1.04e-05,
"loss": 1.3522,
"step": 14660
},
{
"epoch": 0.37,
"grad_norm": 7.704261302947998,
"learning_rate": 1.0393220338983052e-05,
"loss": 1.2121,
"step": 14670
},
{
"epoch": 0.37,
"grad_norm": 3.4666318893432617,
"learning_rate": 1.0386440677966103e-05,
"loss": 1.4718,
"step": 14680
},
{
"epoch": 0.37,
"grad_norm": 8.076580047607422,
"learning_rate": 1.0379661016949155e-05,
"loss": 1.4,
"step": 14690
},
{
"epoch": 0.37,
"grad_norm": 3.1513173580169678,
"learning_rate": 1.0372881355932204e-05,
"loss": 1.3593,
"step": 14700
},
{
"epoch": 0.37,
"grad_norm": 4.961329460144043,
"learning_rate": 1.0366101694915256e-05,
"loss": 1.4394,
"step": 14710
},
{
"epoch": 0.37,
"grad_norm": 7.442839622497559,
"learning_rate": 1.0359322033898307e-05,
"loss": 1.3876,
"step": 14720
},
{
"epoch": 0.37,
"grad_norm": 6.982285499572754,
"learning_rate": 1.0352542372881358e-05,
"loss": 1.455,
"step": 14730
},
{
"epoch": 0.37,
"grad_norm": 10.58169174194336,
"learning_rate": 1.0345762711864406e-05,
"loss": 1.362,
"step": 14740
},
{
"epoch": 0.37,
"grad_norm": 3.1506564617156982,
"learning_rate": 1.0338983050847458e-05,
"loss": 1.3879,
"step": 14750
},
{
"epoch": 0.37,
"grad_norm": 4.143970012664795,
"learning_rate": 1.033220338983051e-05,
"loss": 1.4343,
"step": 14760
},
{
"epoch": 0.37,
"grad_norm": 2.493852376937866,
"learning_rate": 1.0325423728813559e-05,
"loss": 1.4638,
"step": 14770
},
{
"epoch": 0.37,
"grad_norm": 4.69954252243042,
"learning_rate": 1.031864406779661e-05,
"loss": 1.3695,
"step": 14780
},
{
"epoch": 0.37,
"grad_norm": 2.7337687015533447,
"learning_rate": 1.0311864406779662e-05,
"loss": 1.4077,
"step": 14790
},
{
"epoch": 0.37,
"grad_norm": 4.879547119140625,
"learning_rate": 1.0305084745762712e-05,
"loss": 1.3673,
"step": 14800
},
{
"epoch": 0.37,
"grad_norm": 3.318307638168335,
"learning_rate": 1.0298305084745763e-05,
"loss": 1.4526,
"step": 14810
},
{
"epoch": 0.37,
"grad_norm": 1.659396767616272,
"learning_rate": 1.0291525423728814e-05,
"loss": 1.4729,
"step": 14820
},
{
"epoch": 0.37,
"grad_norm": 2.797215223312378,
"learning_rate": 1.0284745762711866e-05,
"loss": 1.5173,
"step": 14830
},
{
"epoch": 0.37,
"grad_norm": 10.495044708251953,
"learning_rate": 1.0277966101694915e-05,
"loss": 1.4145,
"step": 14840
},
{
"epoch": 0.37,
"grad_norm": 8.500311851501465,
"learning_rate": 1.0271186440677967e-05,
"loss": 1.3707,
"step": 14850
},
{
"epoch": 0.37,
"grad_norm": 3.057346820831299,
"learning_rate": 1.0264406779661018e-05,
"loss": 1.2915,
"step": 14860
},
{
"epoch": 0.37,
"grad_norm": 3.8174734115600586,
"learning_rate": 1.025762711864407e-05,
"loss": 1.2654,
"step": 14870
},
{
"epoch": 0.37,
"grad_norm": 2.6704373359680176,
"learning_rate": 1.025084745762712e-05,
"loss": 1.178,
"step": 14880
},
{
"epoch": 0.37,
"grad_norm": 1.770277500152588,
"learning_rate": 1.024406779661017e-05,
"loss": 1.338,
"step": 14890
},
{
"epoch": 0.37,
"grad_norm": 2.0929298400878906,
"learning_rate": 1.0237288135593222e-05,
"loss": 1.3959,
"step": 14900
},
{
"epoch": 0.37,
"grad_norm": 13.965825080871582,
"learning_rate": 1.0230508474576274e-05,
"loss": 1.3289,
"step": 14910
},
{
"epoch": 0.37,
"grad_norm": 5.192085266113281,
"learning_rate": 1.0223728813559323e-05,
"loss": 1.3114,
"step": 14920
},
{
"epoch": 0.37,
"grad_norm": 9.4222412109375,
"learning_rate": 1.0216949152542375e-05,
"loss": 1.4214,
"step": 14930
},
{
"epoch": 0.37,
"grad_norm": 3.792501926422119,
"learning_rate": 1.0210169491525423e-05,
"loss": 1.4965,
"step": 14940
},
{
"epoch": 0.37,
"grad_norm": 2.663039207458496,
"learning_rate": 1.0203389830508474e-05,
"loss": 1.2647,
"step": 14950
},
{
"epoch": 0.37,
"grad_norm": 5.2181806564331055,
"learning_rate": 1.0196610169491525e-05,
"loss": 1.4376,
"step": 14960
},
{
"epoch": 0.37,
"grad_norm": 2.7282190322875977,
"learning_rate": 1.0189830508474577e-05,
"loss": 1.3529,
"step": 14970
},
{
"epoch": 0.37,
"grad_norm": 3.326599597930908,
"learning_rate": 1.0183050847457627e-05,
"loss": 1.4733,
"step": 14980
},
{
"epoch": 0.37,
"grad_norm": 1.9269278049468994,
"learning_rate": 1.0176271186440678e-05,
"loss": 1.4141,
"step": 14990
},
{
"epoch": 0.38,
"grad_norm": 10.291763305664062,
"learning_rate": 1.016949152542373e-05,
"loss": 1.3266,
"step": 15000
},
{
"epoch": 0.38,
"eval_loss": 1.3898714780807495,
"eval_runtime": 66.1314,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 15000
},
{
"epoch": 0.38,
"grad_norm": 3.327223777770996,
"learning_rate": 1.016271186440678e-05,
"loss": 1.4933,
"step": 15010
},
{
"epoch": 0.38,
"grad_norm": 5.010586738586426,
"learning_rate": 1.015593220338983e-05,
"loss": 1.2914,
"step": 15020
},
{
"epoch": 0.38,
"grad_norm": 4.52363395690918,
"learning_rate": 1.0149152542372882e-05,
"loss": 1.3007,
"step": 15030
},
{
"epoch": 0.38,
"grad_norm": 4.301648139953613,
"learning_rate": 1.0142372881355933e-05,
"loss": 1.3652,
"step": 15040
},
{
"epoch": 0.38,
"grad_norm": 8.124075889587402,
"learning_rate": 1.0135593220338985e-05,
"loss": 1.3957,
"step": 15050
},
{
"epoch": 0.38,
"grad_norm": 4.588447093963623,
"learning_rate": 1.0128813559322034e-05,
"loss": 1.1597,
"step": 15060
},
{
"epoch": 0.38,
"grad_norm": 3.6226577758789062,
"learning_rate": 1.0122033898305086e-05,
"loss": 1.5604,
"step": 15070
},
{
"epoch": 0.38,
"grad_norm": 7.039438724517822,
"learning_rate": 1.0115254237288137e-05,
"loss": 1.2913,
"step": 15080
},
{
"epoch": 0.38,
"grad_norm": 1.9853217601776123,
"learning_rate": 1.0108474576271189e-05,
"loss": 1.2087,
"step": 15090
},
{
"epoch": 0.38,
"grad_norm": 6.037110805511475,
"learning_rate": 1.0101694915254238e-05,
"loss": 1.3358,
"step": 15100
},
{
"epoch": 0.38,
"grad_norm": 1.2350672483444214,
"learning_rate": 1.009491525423729e-05,
"loss": 1.4253,
"step": 15110
},
{
"epoch": 0.38,
"grad_norm": 4.601665019989014,
"learning_rate": 1.0088135593220341e-05,
"loss": 1.3471,
"step": 15120
},
{
"epoch": 0.38,
"grad_norm": 7.822099208831787,
"learning_rate": 1.008135593220339e-05,
"loss": 1.4255,
"step": 15130
},
{
"epoch": 0.38,
"grad_norm": 2.1710550785064697,
"learning_rate": 1.0074576271186442e-05,
"loss": 1.4419,
"step": 15140
},
{
"epoch": 0.38,
"grad_norm": 6.896235942840576,
"learning_rate": 1.0067796610169492e-05,
"loss": 1.3166,
"step": 15150
},
{
"epoch": 0.38,
"grad_norm": 2.7912566661834717,
"learning_rate": 1.0061016949152542e-05,
"loss": 1.3577,
"step": 15160
},
{
"epoch": 0.38,
"grad_norm": 2.8911972045898438,
"learning_rate": 1.0054237288135593e-05,
"loss": 1.3415,
"step": 15170
},
{
"epoch": 0.38,
"grad_norm": 1.8997740745544434,
"learning_rate": 1.0047457627118644e-05,
"loss": 1.3933,
"step": 15180
},
{
"epoch": 0.38,
"grad_norm": 5.039108753204346,
"learning_rate": 1.0040677966101696e-05,
"loss": 1.3395,
"step": 15190
},
{
"epoch": 0.38,
"grad_norm": 5.744147300720215,
"learning_rate": 1.0033898305084746e-05,
"loss": 1.4215,
"step": 15200
},
{
"epoch": 0.38,
"grad_norm": 18.610450744628906,
"learning_rate": 1.0027118644067797e-05,
"loss": 1.4639,
"step": 15210
},
{
"epoch": 0.38,
"grad_norm": 4.890042304992676,
"learning_rate": 1.0020338983050848e-05,
"loss": 1.34,
"step": 15220
},
{
"epoch": 0.38,
"grad_norm": 2.5182347297668457,
"learning_rate": 1.00135593220339e-05,
"loss": 1.5222,
"step": 15230
},
{
"epoch": 0.38,
"grad_norm": 2.647080183029175,
"learning_rate": 1.000677966101695e-05,
"loss": 1.3716,
"step": 15240
},
{
"epoch": 0.38,
"grad_norm": 7.10032320022583,
"learning_rate": 1e-05,
"loss": 1.4433,
"step": 15250
},
{
"epoch": 0.38,
"grad_norm": 10.124835968017578,
"learning_rate": 9.993220338983052e-06,
"loss": 1.402,
"step": 15260
},
{
"epoch": 0.38,
"grad_norm": 13.186834335327148,
"learning_rate": 9.986440677966102e-06,
"loss": 1.4211,
"step": 15270
},
{
"epoch": 0.38,
"grad_norm": 3.1523687839508057,
"learning_rate": 9.979661016949153e-06,
"loss": 1.3585,
"step": 15280
},
{
"epoch": 0.38,
"grad_norm": 5.048830032348633,
"learning_rate": 9.972881355932205e-06,
"loss": 1.4079,
"step": 15290
},
{
"epoch": 0.38,
"grad_norm": 9.165743827819824,
"learning_rate": 9.966101694915256e-06,
"loss": 1.4847,
"step": 15300
},
{
"epoch": 0.38,
"grad_norm": 10.89134693145752,
"learning_rate": 9.959322033898306e-06,
"loss": 1.4922,
"step": 15310
},
{
"epoch": 0.38,
"grad_norm": 2.4734840393066406,
"learning_rate": 9.952542372881356e-06,
"loss": 1.496,
"step": 15320
},
{
"epoch": 0.38,
"grad_norm": 2.6290500164031982,
"learning_rate": 9.945762711864407e-06,
"loss": 1.4658,
"step": 15330
},
{
"epoch": 0.38,
"grad_norm": 2.3270649909973145,
"learning_rate": 9.938983050847458e-06,
"loss": 1.3203,
"step": 15340
},
{
"epoch": 0.38,
"grad_norm": 3.256084442138672,
"learning_rate": 9.93220338983051e-06,
"loss": 1.3317,
"step": 15350
},
{
"epoch": 0.38,
"grad_norm": 7.33984899520874,
"learning_rate": 9.92542372881356e-06,
"loss": 1.2692,
"step": 15360
},
{
"epoch": 0.38,
"grad_norm": 5.8543877601623535,
"learning_rate": 9.918644067796611e-06,
"loss": 1.4074,
"step": 15370
},
{
"epoch": 0.38,
"grad_norm": 3.914440870285034,
"learning_rate": 9.911864406779662e-06,
"loss": 1.4998,
"step": 15380
},
{
"epoch": 0.38,
"grad_norm": 4.911335468292236,
"learning_rate": 9.905084745762714e-06,
"loss": 1.2676,
"step": 15390
},
{
"epoch": 0.39,
"grad_norm": 2.9057512283325195,
"learning_rate": 9.898305084745763e-06,
"loss": 1.4411,
"step": 15400
},
{
"epoch": 0.39,
"grad_norm": 6.507194519042969,
"learning_rate": 9.891525423728813e-06,
"loss": 1.3075,
"step": 15410
},
{
"epoch": 0.39,
"grad_norm": 2.6351258754730225,
"learning_rate": 9.884745762711864e-06,
"loss": 1.368,
"step": 15420
},
{
"epoch": 0.39,
"grad_norm": 10.542738914489746,
"learning_rate": 9.877966101694916e-06,
"loss": 1.3388,
"step": 15430
},
{
"epoch": 0.39,
"grad_norm": 5.773244857788086,
"learning_rate": 9.871186440677967e-06,
"loss": 1.3478,
"step": 15440
},
{
"epoch": 0.39,
"grad_norm": 3.678858757019043,
"learning_rate": 9.864406779661017e-06,
"loss": 1.3866,
"step": 15450
},
{
"epoch": 0.39,
"grad_norm": 6.312896251678467,
"learning_rate": 9.857627118644068e-06,
"loss": 1.418,
"step": 15460
},
{
"epoch": 0.39,
"grad_norm": 6.026124954223633,
"learning_rate": 9.85084745762712e-06,
"loss": 1.584,
"step": 15470
},
{
"epoch": 0.39,
"grad_norm": 3.8132076263427734,
"learning_rate": 9.844067796610171e-06,
"loss": 1.4433,
"step": 15480
},
{
"epoch": 0.39,
"grad_norm": 3.9010612964630127,
"learning_rate": 9.837288135593221e-06,
"loss": 1.3962,
"step": 15490
},
{
"epoch": 0.39,
"grad_norm": 2.6610724925994873,
"learning_rate": 9.830508474576272e-06,
"loss": 1.3506,
"step": 15500
},
{
"epoch": 0.39,
"eval_loss": 1.4206969738006592,
"eval_runtime": 66.1225,
"eval_samples_per_second": 15.123,
"eval_steps_per_second": 15.123,
"step": 15500
},
{
"epoch": 0.39,
"grad_norm": 10.500436782836914,
"learning_rate": 9.823728813559322e-06,
"loss": 1.4373,
"step": 15510
},
{
"epoch": 0.39,
"grad_norm": 4.49819278717041,
"learning_rate": 9.816949152542373e-06,
"loss": 1.3353,
"step": 15520
},
{
"epoch": 0.39,
"grad_norm": 4.517664909362793,
"learning_rate": 9.810169491525425e-06,
"loss": 1.4062,
"step": 15530
},
{
"epoch": 0.39,
"grad_norm": 11.253808975219727,
"learning_rate": 9.803389830508474e-06,
"loss": 1.4507,
"step": 15540
},
{
"epoch": 0.39,
"grad_norm": 2.5719611644744873,
"learning_rate": 9.796610169491526e-06,
"loss": 1.2784,
"step": 15550
},
{
"epoch": 0.39,
"grad_norm": 2.9650001525878906,
"learning_rate": 9.789830508474577e-06,
"loss": 1.3021,
"step": 15560
},
{
"epoch": 0.39,
"grad_norm": 4.1970319747924805,
"learning_rate": 9.783050847457629e-06,
"loss": 1.4252,
"step": 15570
},
{
"epoch": 0.39,
"grad_norm": 1.926850438117981,
"learning_rate": 9.776271186440678e-06,
"loss": 1.2455,
"step": 15580
},
{
"epoch": 0.39,
"grad_norm": 8.753373146057129,
"learning_rate": 9.76949152542373e-06,
"loss": 1.1841,
"step": 15590
},
{
"epoch": 0.39,
"grad_norm": 11.411351203918457,
"learning_rate": 9.762711864406781e-06,
"loss": 1.391,
"step": 15600
},
{
"epoch": 0.39,
"grad_norm": 9.503700256347656,
"learning_rate": 9.755932203389833e-06,
"loss": 1.4071,
"step": 15610
},
{
"epoch": 0.39,
"grad_norm": 3.8297741413116455,
"learning_rate": 9.749152542372882e-06,
"loss": 1.4042,
"step": 15620
},
{
"epoch": 0.39,
"grad_norm": 7.153141498565674,
"learning_rate": 9.742372881355932e-06,
"loss": 1.3474,
"step": 15630
},
{
"epoch": 0.39,
"grad_norm": 2.806121826171875,
"learning_rate": 9.735593220338983e-06,
"loss": 1.4126,
"step": 15640
},
{
"epoch": 0.39,
"grad_norm": 7.874058246612549,
"learning_rate": 9.728813559322035e-06,
"loss": 1.5157,
"step": 15650
},
{
"epoch": 0.39,
"grad_norm": 3.6346802711486816,
"learning_rate": 9.722033898305086e-06,
"loss": 1.3617,
"step": 15660
},
{
"epoch": 0.39,
"grad_norm": 19.185550689697266,
"learning_rate": 9.715254237288136e-06,
"loss": 1.4801,
"step": 15670
},
{
"epoch": 0.39,
"grad_norm": 15.9682035446167,
"learning_rate": 9.708474576271187e-06,
"loss": 1.1923,
"step": 15680
},
{
"epoch": 0.39,
"grad_norm": 2.7673888206481934,
"learning_rate": 9.701694915254239e-06,
"loss": 1.5568,
"step": 15690
},
{
"epoch": 0.39,
"grad_norm": 7.633556842803955,
"learning_rate": 9.69491525423729e-06,
"loss": 1.4391,
"step": 15700
},
{
"epoch": 0.39,
"grad_norm": 11.961231231689453,
"learning_rate": 9.68813559322034e-06,
"loss": 1.4212,
"step": 15710
},
{
"epoch": 0.39,
"grad_norm": 2.426793098449707,
"learning_rate": 9.68135593220339e-06,
"loss": 1.1991,
"step": 15720
},
{
"epoch": 0.39,
"grad_norm": 7.8566107749938965,
"learning_rate": 9.674576271186441e-06,
"loss": 1.2991,
"step": 15730
},
{
"epoch": 0.39,
"grad_norm": 10.158808708190918,
"learning_rate": 9.667796610169492e-06,
"loss": 1.4826,
"step": 15740
},
{
"epoch": 0.39,
"grad_norm": 6.697884559631348,
"learning_rate": 9.661016949152544e-06,
"loss": 1.4448,
"step": 15750
},
{
"epoch": 0.39,
"grad_norm": 3.4867446422576904,
"learning_rate": 9.654237288135593e-06,
"loss": 1.4839,
"step": 15760
},
{
"epoch": 0.39,
"grad_norm": 1.6158361434936523,
"learning_rate": 9.647457627118645e-06,
"loss": 1.2674,
"step": 15770
},
{
"epoch": 0.39,
"grad_norm": 3.913695812225342,
"learning_rate": 9.640677966101696e-06,
"loss": 1.3351,
"step": 15780
},
{
"epoch": 0.39,
"grad_norm": 1.9007806777954102,
"learning_rate": 9.633898305084746e-06,
"loss": 1.3478,
"step": 15790
},
{
"epoch": 0.4,
"grad_norm": 2.1645798683166504,
"learning_rate": 9.627118644067797e-06,
"loss": 1.3018,
"step": 15800
},
{
"epoch": 0.4,
"grad_norm": 3.865454912185669,
"learning_rate": 9.620338983050849e-06,
"loss": 1.5398,
"step": 15810
},
{
"epoch": 0.4,
"grad_norm": 4.120809555053711,
"learning_rate": 9.6135593220339e-06,
"loss": 1.4546,
"step": 15820
},
{
"epoch": 0.4,
"grad_norm": 5.410675525665283,
"learning_rate": 9.60677966101695e-06,
"loss": 1.4648,
"step": 15830
},
{
"epoch": 0.4,
"grad_norm": 8.03393268585205,
"learning_rate": 9.600000000000001e-06,
"loss": 1.2891,
"step": 15840
},
{
"epoch": 0.4,
"grad_norm": 7.1069655418396,
"learning_rate": 9.593220338983051e-06,
"loss": 1.314,
"step": 15850
},
{
"epoch": 0.4,
"grad_norm": 6.312065601348877,
"learning_rate": 9.586440677966102e-06,
"loss": 1.2608,
"step": 15860
},
{
"epoch": 0.4,
"grad_norm": 3.499418258666992,
"learning_rate": 9.579661016949154e-06,
"loss": 1.3816,
"step": 15870
},
{
"epoch": 0.4,
"grad_norm": 5.211281776428223,
"learning_rate": 9.572881355932203e-06,
"loss": 1.1021,
"step": 15880
},
{
"epoch": 0.4,
"grad_norm": 4.372194290161133,
"learning_rate": 9.566101694915255e-06,
"loss": 1.4179,
"step": 15890
},
{
"epoch": 0.4,
"grad_norm": 5.44555139541626,
"learning_rate": 9.559322033898306e-06,
"loss": 1.4848,
"step": 15900
},
{
"epoch": 0.4,
"grad_norm": 3.483454942703247,
"learning_rate": 9.552542372881358e-06,
"loss": 1.2985,
"step": 15910
},
{
"epoch": 0.4,
"grad_norm": 3.8721625804901123,
"learning_rate": 9.545762711864407e-06,
"loss": 1.4731,
"step": 15920
},
{
"epoch": 0.4,
"grad_norm": 4.317465305328369,
"learning_rate": 9.538983050847457e-06,
"loss": 1.501,
"step": 15930
},
{
"epoch": 0.4,
"grad_norm": 7.075793743133545,
"learning_rate": 9.532203389830508e-06,
"loss": 1.5138,
"step": 15940
},
{
"epoch": 0.4,
"grad_norm": 5.606925964355469,
"learning_rate": 9.52542372881356e-06,
"loss": 1.3771,
"step": 15950
},
{
"epoch": 0.4,
"grad_norm": 2.937833547592163,
"learning_rate": 9.518644067796611e-06,
"loss": 1.4764,
"step": 15960
},
{
"epoch": 0.4,
"grad_norm": 4.39876651763916,
"learning_rate": 9.511864406779661e-06,
"loss": 1.4022,
"step": 15970
},
{
"epoch": 0.4,
"grad_norm": 6.455230712890625,
"learning_rate": 9.505084745762712e-06,
"loss": 1.3569,
"step": 15980
},
{
"epoch": 0.4,
"grad_norm": 4.89456033706665,
"learning_rate": 9.498305084745764e-06,
"loss": 1.3426,
"step": 15990
},
{
"epoch": 0.4,
"grad_norm": 5.696287631988525,
"learning_rate": 9.491525423728815e-06,
"loss": 1.5381,
"step": 16000
},
{
"epoch": 0.4,
"eval_loss": 1.3579893112182617,
"eval_runtime": 66.2433,
"eval_samples_per_second": 15.096,
"eval_steps_per_second": 15.096,
"step": 16000
},
{
"epoch": 0.4,
"grad_norm": 6.06011962890625,
"learning_rate": 9.484745762711865e-06,
"loss": 1.4422,
"step": 16010
},
{
"epoch": 0.4,
"grad_norm": 0.9439699053764343,
"learning_rate": 9.477966101694916e-06,
"loss": 1.3661,
"step": 16020
},
{
"epoch": 0.4,
"grad_norm": 2.9158008098602295,
"learning_rate": 9.471186440677966e-06,
"loss": 1.1349,
"step": 16030
},
{
"epoch": 0.4,
"grad_norm": 3.3317673206329346,
"learning_rate": 9.464406779661017e-06,
"loss": 1.4161,
"step": 16040
},
{
"epoch": 0.4,
"grad_norm": 2.866821050643921,
"learning_rate": 9.457627118644069e-06,
"loss": 1.2952,
"step": 16050
},
{
"epoch": 0.4,
"grad_norm": 8.00590991973877,
"learning_rate": 9.450847457627119e-06,
"loss": 1.3493,
"step": 16060
},
{
"epoch": 0.4,
"grad_norm": 2.1362667083740234,
"learning_rate": 9.44406779661017e-06,
"loss": 1.3385,
"step": 16070
},
{
"epoch": 0.4,
"grad_norm": 4.567724227905273,
"learning_rate": 9.437288135593221e-06,
"loss": 1.2988,
"step": 16080
},
{
"epoch": 0.4,
"grad_norm": 3.382605791091919,
"learning_rate": 9.430508474576273e-06,
"loss": 1.2131,
"step": 16090
},
{
"epoch": 0.4,
"grad_norm": 13.581380844116211,
"learning_rate": 9.423728813559322e-06,
"loss": 1.3962,
"step": 16100
},
{
"epoch": 0.4,
"grad_norm": 6.365314483642578,
"learning_rate": 9.416949152542374e-06,
"loss": 1.4928,
"step": 16110
},
{
"epoch": 0.4,
"grad_norm": 5.413196563720703,
"learning_rate": 9.410169491525425e-06,
"loss": 1.2728,
"step": 16120
},
{
"epoch": 0.4,
"grad_norm": 3.6787874698638916,
"learning_rate": 9.403389830508477e-06,
"loss": 1.1831,
"step": 16130
},
{
"epoch": 0.4,
"grad_norm": 2.8850882053375244,
"learning_rate": 9.396610169491526e-06,
"loss": 1.3025,
"step": 16140
},
{
"epoch": 0.4,
"grad_norm": 3.6025898456573486,
"learning_rate": 9.389830508474576e-06,
"loss": 1.3553,
"step": 16150
},
{
"epoch": 0.4,
"grad_norm": 3.00011944770813,
"learning_rate": 9.383050847457627e-06,
"loss": 1.5127,
"step": 16160
},
{
"epoch": 0.4,
"grad_norm": 4.33295202255249,
"learning_rate": 9.376271186440679e-06,
"loss": 1.5001,
"step": 16170
},
{
"epoch": 0.4,
"grad_norm": 3.6982882022857666,
"learning_rate": 9.36949152542373e-06,
"loss": 1.4729,
"step": 16180
},
{
"epoch": 0.4,
"grad_norm": 7.049408435821533,
"learning_rate": 9.36271186440678e-06,
"loss": 1.276,
"step": 16190
},
{
"epoch": 0.41,
"grad_norm": 1.6569970846176147,
"learning_rate": 9.355932203389831e-06,
"loss": 1.3321,
"step": 16200
},
{
"epoch": 0.41,
"grad_norm": 2.4531731605529785,
"learning_rate": 9.349152542372883e-06,
"loss": 1.435,
"step": 16210
},
{
"epoch": 0.41,
"grad_norm": 2.776325225830078,
"learning_rate": 9.342372881355934e-06,
"loss": 1.2586,
"step": 16220
},
{
"epoch": 0.41,
"grad_norm": 2.805823564529419,
"learning_rate": 9.335593220338984e-06,
"loss": 1.3393,
"step": 16230
},
{
"epoch": 0.41,
"grad_norm": 2.3886725902557373,
"learning_rate": 9.328813559322034e-06,
"loss": 1.4241,
"step": 16240
},
{
"epoch": 0.41,
"grad_norm": 10.983514785766602,
"learning_rate": 9.322033898305085e-06,
"loss": 1.5032,
"step": 16250
},
{
"epoch": 0.41,
"grad_norm": 2.922609329223633,
"learning_rate": 9.315254237288136e-06,
"loss": 1.2289,
"step": 16260
},
{
"epoch": 0.41,
"grad_norm": 3.623088836669922,
"learning_rate": 9.308474576271188e-06,
"loss": 1.2854,
"step": 16270
},
{
"epoch": 0.41,
"grad_norm": 3.6580920219421387,
"learning_rate": 9.301694915254237e-06,
"loss": 1.3371,
"step": 16280
},
{
"epoch": 0.41,
"grad_norm": 3.010573625564575,
"learning_rate": 9.294915254237289e-06,
"loss": 1.465,
"step": 16290
},
{
"epoch": 0.41,
"grad_norm": 3.1502161026000977,
"learning_rate": 9.28813559322034e-06,
"loss": 1.2245,
"step": 16300
},
{
"epoch": 0.41,
"grad_norm": 2.7607476711273193,
"learning_rate": 9.28135593220339e-06,
"loss": 1.4668,
"step": 16310
},
{
"epoch": 0.41,
"grad_norm": 6.022114276885986,
"learning_rate": 9.274576271186441e-06,
"loss": 1.1924,
"step": 16320
},
{
"epoch": 0.41,
"grad_norm": 5.824821949005127,
"learning_rate": 9.267796610169493e-06,
"loss": 1.3773,
"step": 16330
},
{
"epoch": 0.41,
"grad_norm": 3.7949423789978027,
"learning_rate": 9.261016949152544e-06,
"loss": 1.3334,
"step": 16340
},
{
"epoch": 0.41,
"grad_norm": 1.7965031862258911,
"learning_rate": 9.254237288135594e-06,
"loss": 1.5022,
"step": 16350
},
{
"epoch": 0.41,
"grad_norm": 4.245846271514893,
"learning_rate": 9.247457627118645e-06,
"loss": 1.2312,
"step": 16360
},
{
"epoch": 0.41,
"grad_norm": 7.283750057220459,
"learning_rate": 9.240677966101695e-06,
"loss": 1.2448,
"step": 16370
},
{
"epoch": 0.41,
"grad_norm": 3.51558780670166,
"learning_rate": 9.233898305084746e-06,
"loss": 1.4823,
"step": 16380
},
{
"epoch": 0.41,
"grad_norm": 7.842286586761475,
"learning_rate": 9.227118644067798e-06,
"loss": 1.3931,
"step": 16390
},
{
"epoch": 0.41,
"grad_norm": 2.6128389835357666,
"learning_rate": 9.220338983050847e-06,
"loss": 1.4359,
"step": 16400
},
{
"epoch": 0.41,
"grad_norm": 2.5427112579345703,
"learning_rate": 9.213559322033899e-06,
"loss": 1.4364,
"step": 16410
},
{
"epoch": 0.41,
"grad_norm": 3.0635411739349365,
"learning_rate": 9.20677966101695e-06,
"loss": 1.4804,
"step": 16420
},
{
"epoch": 0.41,
"grad_norm": 4.00955057144165,
"learning_rate": 9.200000000000002e-06,
"loss": 1.1819,
"step": 16430
},
{
"epoch": 0.41,
"grad_norm": 2.0021607875823975,
"learning_rate": 9.193220338983051e-06,
"loss": 1.312,
"step": 16440
},
{
"epoch": 0.41,
"grad_norm": 6.373915672302246,
"learning_rate": 9.186440677966101e-06,
"loss": 1.5638,
"step": 16450
},
{
"epoch": 0.41,
"grad_norm": 7.486032962799072,
"learning_rate": 9.179661016949153e-06,
"loss": 1.5156,
"step": 16460
},
{
"epoch": 0.41,
"grad_norm": 7.894711971282959,
"learning_rate": 9.172881355932204e-06,
"loss": 1.3042,
"step": 16470
},
{
"epoch": 0.41,
"grad_norm": 4.769121170043945,
"learning_rate": 9.166101694915255e-06,
"loss": 1.5031,
"step": 16480
},
{
"epoch": 0.41,
"grad_norm": 5.979970455169678,
"learning_rate": 9.159322033898305e-06,
"loss": 1.2542,
"step": 16490
},
{
"epoch": 0.41,
"grad_norm": 6.4072651863098145,
"learning_rate": 9.152542372881356e-06,
"loss": 1.319,
"step": 16500
},
{
"epoch": 0.41,
"eval_loss": 1.3969231843948364,
"eval_runtime": 66.1684,
"eval_samples_per_second": 15.113,
"eval_steps_per_second": 15.113,
"step": 16500
},
{
"epoch": 0.41,
"grad_norm": 6.439510822296143,
"learning_rate": 9.145762711864408e-06,
"loss": 1.6106,
"step": 16510
},
{
"epoch": 0.41,
"grad_norm": 3.3735623359680176,
"learning_rate": 9.13898305084746e-06,
"loss": 1.3179,
"step": 16520
},
{
"epoch": 0.41,
"grad_norm": 4.679867267608643,
"learning_rate": 9.132203389830509e-06,
"loss": 1.465,
"step": 16530
},
{
"epoch": 0.41,
"grad_norm": 3.878396987915039,
"learning_rate": 9.12542372881356e-06,
"loss": 1.2999,
"step": 16540
},
{
"epoch": 0.41,
"grad_norm": 5.397853374481201,
"learning_rate": 9.11864406779661e-06,
"loss": 1.3575,
"step": 16550
},
{
"epoch": 0.41,
"grad_norm": 3.2240700721740723,
"learning_rate": 9.111864406779661e-06,
"loss": 1.5086,
"step": 16560
},
{
"epoch": 0.41,
"grad_norm": 4.239155292510986,
"learning_rate": 9.105084745762713e-06,
"loss": 1.3695,
"step": 16570
},
{
"epoch": 0.41,
"grad_norm": 9.791718482971191,
"learning_rate": 9.098305084745763e-06,
"loss": 1.4928,
"step": 16580
},
{
"epoch": 0.41,
"grad_norm": 4.839406967163086,
"learning_rate": 9.091525423728814e-06,
"loss": 1.4183,
"step": 16590
},
{
"epoch": 0.41,
"grad_norm": 3.728440046310425,
"learning_rate": 9.084745762711865e-06,
"loss": 1.2819,
"step": 16600
},
{
"epoch": 0.42,
"grad_norm": 2.5298233032226562,
"learning_rate": 9.077966101694917e-06,
"loss": 1.3922,
"step": 16610
},
{
"epoch": 0.42,
"grad_norm": 4.599859714508057,
"learning_rate": 9.071186440677966e-06,
"loss": 1.4627,
"step": 16620
},
{
"epoch": 0.42,
"grad_norm": 2.5592544078826904,
"learning_rate": 9.064406779661018e-06,
"loss": 1.544,
"step": 16630
},
{
"epoch": 0.42,
"grad_norm": 8.281981468200684,
"learning_rate": 9.05762711864407e-06,
"loss": 1.4084,
"step": 16640
},
{
"epoch": 0.42,
"grad_norm": 5.964704990386963,
"learning_rate": 9.05084745762712e-06,
"loss": 1.4186,
"step": 16650
},
{
"epoch": 0.42,
"grad_norm": 4.577818393707275,
"learning_rate": 9.04406779661017e-06,
"loss": 1.2691,
"step": 16660
},
{
"epoch": 0.42,
"grad_norm": 3.4889256954193115,
"learning_rate": 9.03728813559322e-06,
"loss": 1.3494,
"step": 16670
},
{
"epoch": 0.42,
"grad_norm": 4.458433628082275,
"learning_rate": 9.030508474576271e-06,
"loss": 1.4639,
"step": 16680
},
{
"epoch": 0.42,
"grad_norm": 5.813331127166748,
"learning_rate": 9.023728813559323e-06,
"loss": 1.2206,
"step": 16690
},
{
"epoch": 0.42,
"grad_norm": 2.583115339279175,
"learning_rate": 9.016949152542374e-06,
"loss": 1.4299,
"step": 16700
},
{
"epoch": 0.42,
"grad_norm": 7.3467206954956055,
"learning_rate": 9.010169491525424e-06,
"loss": 1.3001,
"step": 16710
},
{
"epoch": 0.42,
"grad_norm": 3.896238088607788,
"learning_rate": 9.003389830508475e-06,
"loss": 1.3541,
"step": 16720
},
{
"epoch": 0.42,
"grad_norm": 3.967496156692505,
"learning_rate": 8.996610169491527e-06,
"loss": 1.3315,
"step": 16730
},
{
"epoch": 0.42,
"grad_norm": 2.9316656589508057,
"learning_rate": 8.989830508474578e-06,
"loss": 1.3935,
"step": 16740
},
{
"epoch": 0.42,
"grad_norm": 1.619555950164795,
"learning_rate": 8.983050847457628e-06,
"loss": 1.2879,
"step": 16750
},
{
"epoch": 0.42,
"grad_norm": 3.6497581005096436,
"learning_rate": 8.976271186440678e-06,
"loss": 1.5195,
"step": 16760
},
{
"epoch": 0.42,
"grad_norm": 5.33650016784668,
"learning_rate": 8.969491525423729e-06,
"loss": 1.4568,
"step": 16770
},
{
"epoch": 0.42,
"grad_norm": 7.370872974395752,
"learning_rate": 8.96271186440678e-06,
"loss": 1.5122,
"step": 16780
},
{
"epoch": 0.42,
"grad_norm": 6.295271873474121,
"learning_rate": 8.955932203389832e-06,
"loss": 1.177,
"step": 16790
},
{
"epoch": 0.42,
"grad_norm": 7.165903091430664,
"learning_rate": 8.949152542372881e-06,
"loss": 1.3597,
"step": 16800
},
{
"epoch": 0.42,
"grad_norm": 4.152050018310547,
"learning_rate": 8.942372881355933e-06,
"loss": 1.3837,
"step": 16810
},
{
"epoch": 0.42,
"grad_norm": 2.2240993976593018,
"learning_rate": 8.935593220338984e-06,
"loss": 1.3974,
"step": 16820
},
{
"epoch": 0.42,
"grad_norm": 2.854851245880127,
"learning_rate": 8.928813559322036e-06,
"loss": 1.4478,
"step": 16830
},
{
"epoch": 0.42,
"grad_norm": 4.584380149841309,
"learning_rate": 8.922033898305085e-06,
"loss": 1.4808,
"step": 16840
},
{
"epoch": 0.42,
"grad_norm": 5.496336460113525,
"learning_rate": 8.915254237288137e-06,
"loss": 1.4587,
"step": 16850
},
{
"epoch": 0.42,
"grad_norm": 9.56225872039795,
"learning_rate": 8.908474576271188e-06,
"loss": 1.2726,
"step": 16860
},
{
"epoch": 0.42,
"grad_norm": 3.6302268505096436,
"learning_rate": 8.901694915254238e-06,
"loss": 1.4753,
"step": 16870
},
{
"epoch": 0.42,
"grad_norm": 9.815473556518555,
"learning_rate": 8.89491525423729e-06,
"loss": 1.3163,
"step": 16880
},
{
"epoch": 0.42,
"grad_norm": 6.138949871063232,
"learning_rate": 8.888135593220339e-06,
"loss": 1.3637,
"step": 16890
},
{
"epoch": 0.42,
"grad_norm": 5.0360212326049805,
"learning_rate": 8.88135593220339e-06,
"loss": 1.5961,
"step": 16900
},
{
"epoch": 0.42,
"grad_norm": 11.983102798461914,
"learning_rate": 8.874576271186442e-06,
"loss": 1.3359,
"step": 16910
},
{
"epoch": 0.42,
"grad_norm": 3.8704030513763428,
"learning_rate": 8.867796610169492e-06,
"loss": 1.2624,
"step": 16920
},
{
"epoch": 0.42,
"grad_norm": 8.61357593536377,
"learning_rate": 8.861016949152543e-06,
"loss": 1.4325,
"step": 16930
},
{
"epoch": 0.42,
"grad_norm": 1.4079731702804565,
"learning_rate": 8.854237288135594e-06,
"loss": 1.3984,
"step": 16940
},
{
"epoch": 0.42,
"grad_norm": 4.425031661987305,
"learning_rate": 8.847457627118646e-06,
"loss": 1.3202,
"step": 16950
},
{
"epoch": 0.42,
"grad_norm": 7.565074920654297,
"learning_rate": 8.840677966101695e-06,
"loss": 1.4561,
"step": 16960
},
{
"epoch": 0.42,
"grad_norm": 4.513836860656738,
"learning_rate": 8.833898305084747e-06,
"loss": 1.52,
"step": 16970
},
{
"epoch": 0.42,
"grad_norm": 7.830351829528809,
"learning_rate": 8.827118644067797e-06,
"loss": 1.4321,
"step": 16980
},
{
"epoch": 0.42,
"grad_norm": 4.269097328186035,
"learning_rate": 8.820338983050848e-06,
"loss": 1.4247,
"step": 16990
},
{
"epoch": 0.42,
"grad_norm": 3.2125535011291504,
"learning_rate": 8.8135593220339e-06,
"loss": 1.1334,
"step": 17000
},
{
"epoch": 0.42,
"eval_loss": 1.3314954042434692,
"eval_runtime": 66.1036,
"eval_samples_per_second": 15.128,
"eval_steps_per_second": 15.128,
"step": 17000
},
{
"epoch": 0.43,
"grad_norm": 3.5944836139678955,
"learning_rate": 8.806779661016949e-06,
"loss": 1.2363,
"step": 17010
},
{
"epoch": 0.43,
"grad_norm": 3.680687189102173,
"learning_rate": 8.8e-06,
"loss": 1.2968,
"step": 17020
},
{
"epoch": 0.43,
"grad_norm": 9.756001472473145,
"learning_rate": 8.793220338983052e-06,
"loss": 1.378,
"step": 17030
},
{
"epoch": 0.43,
"grad_norm": 3.8217949867248535,
"learning_rate": 8.786440677966103e-06,
"loss": 1.1933,
"step": 17040
},
{
"epoch": 0.43,
"grad_norm": 4.635326862335205,
"learning_rate": 8.779661016949153e-06,
"loss": 1.4635,
"step": 17050
},
{
"epoch": 0.43,
"grad_norm": 12.349688529968262,
"learning_rate": 8.772881355932204e-06,
"loss": 1.1233,
"step": 17060
},
{
"epoch": 0.43,
"grad_norm": 2.0350751876831055,
"learning_rate": 8.766101694915254e-06,
"loss": 1.5517,
"step": 17070
},
{
"epoch": 0.43,
"grad_norm": 2.533076524734497,
"learning_rate": 8.759322033898305e-06,
"loss": 1.2468,
"step": 17080
},
{
"epoch": 0.43,
"grad_norm": 5.4851813316345215,
"learning_rate": 8.752542372881357e-06,
"loss": 1.3057,
"step": 17090
},
{
"epoch": 0.43,
"grad_norm": 5.9718499183654785,
"learning_rate": 8.745762711864407e-06,
"loss": 1.3685,
"step": 17100
},
{
"epoch": 0.43,
"grad_norm": 6.759610176086426,
"learning_rate": 8.738983050847458e-06,
"loss": 1.2429,
"step": 17110
},
{
"epoch": 0.43,
"grad_norm": 10.515771865844727,
"learning_rate": 8.73220338983051e-06,
"loss": 1.1885,
"step": 17120
},
{
"epoch": 0.43,
"grad_norm": 5.439294815063477,
"learning_rate": 8.72542372881356e-06,
"loss": 1.4804,
"step": 17130
},
{
"epoch": 0.43,
"grad_norm": 3.2483952045440674,
"learning_rate": 8.71864406779661e-06,
"loss": 1.2879,
"step": 17140
},
{
"epoch": 0.43,
"grad_norm": 12.16537857055664,
"learning_rate": 8.711864406779662e-06,
"loss": 1.365,
"step": 17150
},
{
"epoch": 0.43,
"grad_norm": 7.272519588470459,
"learning_rate": 8.705084745762713e-06,
"loss": 1.2356,
"step": 17160
},
{
"epoch": 0.43,
"grad_norm": 5.10579252243042,
"learning_rate": 8.698305084745765e-06,
"loss": 1.2475,
"step": 17170
},
{
"epoch": 0.43,
"grad_norm": 2.060873031616211,
"learning_rate": 8.691525423728814e-06,
"loss": 1.2728,
"step": 17180
},
{
"epoch": 0.43,
"grad_norm": 4.275463581085205,
"learning_rate": 8.684745762711864e-06,
"loss": 1.3372,
"step": 17190
},
{
"epoch": 0.43,
"grad_norm": 7.369024753570557,
"learning_rate": 8.677966101694915e-06,
"loss": 1.3319,
"step": 17200
},
{
"epoch": 0.43,
"grad_norm": 4.748632431030273,
"learning_rate": 8.671186440677967e-06,
"loss": 1.3954,
"step": 17210
},
{
"epoch": 0.43,
"grad_norm": 6.593578815460205,
"learning_rate": 8.664406779661018e-06,
"loss": 1.218,
"step": 17220
},
{
"epoch": 0.43,
"grad_norm": 9.88740348815918,
"learning_rate": 8.657627118644068e-06,
"loss": 1.3813,
"step": 17230
},
{
"epoch": 0.43,
"grad_norm": 5.244362831115723,
"learning_rate": 8.65084745762712e-06,
"loss": 1.3307,
"step": 17240
},
{
"epoch": 0.43,
"grad_norm": 6.388607025146484,
"learning_rate": 8.64406779661017e-06,
"loss": 1.3975,
"step": 17250
},
{
"epoch": 0.43,
"grad_norm": 5.940032005310059,
"learning_rate": 8.637288135593222e-06,
"loss": 1.452,
"step": 17260
},
{
"epoch": 0.43,
"grad_norm": 1.8261200189590454,
"learning_rate": 8.630508474576272e-06,
"loss": 1.3114,
"step": 17270
},
{
"epoch": 0.43,
"grad_norm": 4.878362655639648,
"learning_rate": 8.623728813559322e-06,
"loss": 1.3152,
"step": 17280
},
{
"epoch": 0.43,
"grad_norm": 5.558933258056641,
"learning_rate": 8.616949152542373e-06,
"loss": 1.5567,
"step": 17290
},
{
"epoch": 0.43,
"grad_norm": 2.346237897872925,
"learning_rate": 8.610169491525424e-06,
"loss": 1.4462,
"step": 17300
},
{
"epoch": 0.43,
"grad_norm": 5.712830066680908,
"learning_rate": 8.603389830508476e-06,
"loss": 1.2658,
"step": 17310
},
{
"epoch": 0.43,
"grad_norm": 6.277383804321289,
"learning_rate": 8.596610169491526e-06,
"loss": 1.3562,
"step": 17320
},
{
"epoch": 0.43,
"grad_norm": 2.8792171478271484,
"learning_rate": 8.589830508474577e-06,
"loss": 1.2409,
"step": 17330
},
{
"epoch": 0.43,
"grad_norm": 4.000193119049072,
"learning_rate": 8.583050847457628e-06,
"loss": 1.3503,
"step": 17340
},
{
"epoch": 0.43,
"grad_norm": 6.255504131317139,
"learning_rate": 8.57627118644068e-06,
"loss": 1.3411,
"step": 17350
},
{
"epoch": 0.43,
"grad_norm": 5.366970539093018,
"learning_rate": 8.56949152542373e-06,
"loss": 1.2918,
"step": 17360
},
{
"epoch": 0.43,
"grad_norm": 1.9505482912063599,
"learning_rate": 8.56271186440678e-06,
"loss": 1.3139,
"step": 17370
},
{
"epoch": 0.43,
"grad_norm": 8.022522926330566,
"learning_rate": 8.55593220338983e-06,
"loss": 1.515,
"step": 17380
},
{
"epoch": 0.43,
"grad_norm": 5.483621120452881,
"learning_rate": 8.549152542372882e-06,
"loss": 1.4464,
"step": 17390
},
{
"epoch": 0.43,
"grad_norm": 5.349776268005371,
"learning_rate": 8.542372881355933e-06,
"loss": 1.5864,
"step": 17400
},
{
"epoch": 0.44,
"grad_norm": 3.4136993885040283,
"learning_rate": 8.535593220338983e-06,
"loss": 1.2964,
"step": 17410
},
{
"epoch": 0.44,
"grad_norm": 4.270326614379883,
"learning_rate": 8.528813559322034e-06,
"loss": 1.3736,
"step": 17420
},
{
"epoch": 0.44,
"grad_norm": 2.886993169784546,
"learning_rate": 8.522033898305086e-06,
"loss": 1.3513,
"step": 17430
},
{
"epoch": 0.44,
"grad_norm": 4.42819881439209,
"learning_rate": 8.515254237288136e-06,
"loss": 1.4226,
"step": 17440
},
{
"epoch": 0.44,
"grad_norm": 5.844710350036621,
"learning_rate": 8.508474576271187e-06,
"loss": 1.3135,
"step": 17450
},
{
"epoch": 0.44,
"grad_norm": 2.4220869541168213,
"learning_rate": 8.501694915254238e-06,
"loss": 1.165,
"step": 17460
},
{
"epoch": 0.44,
"grad_norm": 4.976444244384766,
"learning_rate": 8.49491525423729e-06,
"loss": 1.1964,
"step": 17470
},
{
"epoch": 0.44,
"grad_norm": 12.55490779876709,
"learning_rate": 8.48813559322034e-06,
"loss": 1.5891,
"step": 17480
},
{
"epoch": 0.44,
"grad_norm": 4.102578163146973,
"learning_rate": 8.481355932203391e-06,
"loss": 1.445,
"step": 17490
},
{
"epoch": 0.44,
"grad_norm": 2.6767678260803223,
"learning_rate": 8.47457627118644e-06,
"loss": 1.536,
"step": 17500
},
{
"epoch": 0.44,
"eval_loss": 1.388473391532898,
"eval_runtime": 66.17,
"eval_samples_per_second": 15.113,
"eval_steps_per_second": 15.113,
"step": 17500
},
{
"epoch": 0.44,
"grad_norm": 5.189449310302734,
"learning_rate": 8.467796610169492e-06,
"loss": 1.352,
"step": 17510
},
{
"epoch": 0.44,
"grad_norm": 2.765326499938965,
"learning_rate": 8.461016949152543e-06,
"loss": 1.5275,
"step": 17520
},
{
"epoch": 0.44,
"grad_norm": 3.60880446434021,
"learning_rate": 8.454237288135593e-06,
"loss": 1.2793,
"step": 17530
},
{
"epoch": 0.44,
"grad_norm": 3.072174310684204,
"learning_rate": 8.447457627118644e-06,
"loss": 1.2746,
"step": 17540
},
{
"epoch": 0.44,
"grad_norm": 3.3222594261169434,
"learning_rate": 8.440677966101696e-06,
"loss": 1.3359,
"step": 17550
},
{
"epoch": 0.44,
"grad_norm": 1.3298100233078003,
"learning_rate": 8.433898305084747e-06,
"loss": 1.3826,
"step": 17560
},
{
"epoch": 0.44,
"grad_norm": 4.06503438949585,
"learning_rate": 8.427118644067797e-06,
"loss": 1.3339,
"step": 17570
},
{
"epoch": 0.44,
"grad_norm": 6.150786399841309,
"learning_rate": 8.420338983050848e-06,
"loss": 1.3414,
"step": 17580
},
{
"epoch": 0.44,
"grad_norm": 9.045748710632324,
"learning_rate": 8.413559322033898e-06,
"loss": 1.3183,
"step": 17590
},
{
"epoch": 0.44,
"grad_norm": 4.415981769561768,
"learning_rate": 8.40677966101695e-06,
"loss": 1.3323,
"step": 17600
},
{
"epoch": 0.44,
"grad_norm": 8.40713882446289,
"learning_rate": 8.400000000000001e-06,
"loss": 1.2519,
"step": 17610
},
{
"epoch": 0.44,
"grad_norm": 6.841925144195557,
"learning_rate": 8.39322033898305e-06,
"loss": 1.4799,
"step": 17620
},
{
"epoch": 0.44,
"grad_norm": 10.993582725524902,
"learning_rate": 8.386440677966102e-06,
"loss": 1.3495,
"step": 17630
},
{
"epoch": 0.44,
"grad_norm": 2.115006685256958,
"learning_rate": 8.379661016949153e-06,
"loss": 1.3902,
"step": 17640
},
{
"epoch": 0.44,
"grad_norm": 3.485889434814453,
"learning_rate": 8.372881355932205e-06,
"loss": 1.4462,
"step": 17650
},
{
"epoch": 0.44,
"grad_norm": 5.725930690765381,
"learning_rate": 8.366101694915255e-06,
"loss": 1.4603,
"step": 17660
},
{
"epoch": 0.44,
"grad_norm": 6.5530548095703125,
"learning_rate": 8.359322033898306e-06,
"loss": 1.4954,
"step": 17670
},
{
"epoch": 0.44,
"grad_norm": 6.006740570068359,
"learning_rate": 8.352542372881357e-06,
"loss": 1.3556,
"step": 17680
},
{
"epoch": 0.44,
"grad_norm": 5.5967936515808105,
"learning_rate": 8.345762711864409e-06,
"loss": 1.0719,
"step": 17690
},
{
"epoch": 0.44,
"grad_norm": 5.70643949508667,
"learning_rate": 8.338983050847458e-06,
"loss": 1.2968,
"step": 17700
},
{
"epoch": 0.44,
"grad_norm": 3.280465841293335,
"learning_rate": 8.332203389830508e-06,
"loss": 1.2669,
"step": 17710
},
{
"epoch": 0.44,
"grad_norm": 3.098782777786255,
"learning_rate": 8.32542372881356e-06,
"loss": 1.4704,
"step": 17720
},
{
"epoch": 0.44,
"grad_norm": 2.825495958328247,
"learning_rate": 8.318644067796611e-06,
"loss": 1.2165,
"step": 17730
},
{
"epoch": 0.44,
"grad_norm": 3.8040249347686768,
"learning_rate": 8.311864406779662e-06,
"loss": 1.3623,
"step": 17740
},
{
"epoch": 0.44,
"grad_norm": 3.4438130855560303,
"learning_rate": 8.305084745762712e-06,
"loss": 1.4255,
"step": 17750
},
{
"epoch": 0.44,
"grad_norm": 2.187945604324341,
"learning_rate": 8.298305084745763e-06,
"loss": 1.2181,
"step": 17760
},
{
"epoch": 0.44,
"grad_norm": 2.0644514560699463,
"learning_rate": 8.291525423728815e-06,
"loss": 1.4669,
"step": 17770
},
{
"epoch": 0.44,
"grad_norm": 4.314215660095215,
"learning_rate": 8.284745762711866e-06,
"loss": 1.3776,
"step": 17780
},
{
"epoch": 0.44,
"grad_norm": 2.948415517807007,
"learning_rate": 8.277966101694916e-06,
"loss": 1.3198,
"step": 17790
},
{
"epoch": 0.45,
"grad_norm": 7.77257776260376,
"learning_rate": 8.271186440677966e-06,
"loss": 1.3826,
"step": 17800
},
{
"epoch": 0.45,
"grad_norm": 2.2085230350494385,
"learning_rate": 8.264406779661017e-06,
"loss": 1.3443,
"step": 17810
},
{
"epoch": 0.45,
"grad_norm": 2.8318214416503906,
"learning_rate": 8.257627118644068e-06,
"loss": 1.3126,
"step": 17820
},
{
"epoch": 0.45,
"grad_norm": 4.544606685638428,
"learning_rate": 8.25084745762712e-06,
"loss": 1.2955,
"step": 17830
},
{
"epoch": 0.45,
"grad_norm": 5.595729351043701,
"learning_rate": 8.24406779661017e-06,
"loss": 1.3814,
"step": 17840
},
{
"epoch": 0.45,
"grad_norm": 6.201912879943848,
"learning_rate": 8.237288135593221e-06,
"loss": 1.4243,
"step": 17850
},
{
"epoch": 0.45,
"grad_norm": 7.697713851928711,
"learning_rate": 8.230508474576272e-06,
"loss": 1.326,
"step": 17860
},
{
"epoch": 0.45,
"grad_norm": 3.7588555812835693,
"learning_rate": 8.223728813559324e-06,
"loss": 1.1265,
"step": 17870
},
{
"epoch": 0.45,
"grad_norm": 8.472000122070312,
"learning_rate": 8.216949152542373e-06,
"loss": 1.2678,
"step": 17880
},
{
"epoch": 0.45,
"grad_norm": 4.802706241607666,
"learning_rate": 8.210169491525425e-06,
"loss": 1.5116,
"step": 17890
},
{
"epoch": 0.45,
"grad_norm": 5.357900142669678,
"learning_rate": 8.203389830508475e-06,
"loss": 1.4435,
"step": 17900
},
{
"epoch": 0.45,
"grad_norm": 3.5331151485443115,
"learning_rate": 8.196610169491526e-06,
"loss": 1.3838,
"step": 17910
},
{
"epoch": 0.45,
"grad_norm": 2.591485023498535,
"learning_rate": 8.189830508474577e-06,
"loss": 1.3595,
"step": 17920
},
{
"epoch": 0.45,
"grad_norm": 9.1065034866333,
"learning_rate": 8.183050847457627e-06,
"loss": 1.4109,
"step": 17930
},
{
"epoch": 0.45,
"grad_norm": 3.2992680072784424,
"learning_rate": 8.176271186440678e-06,
"loss": 1.3554,
"step": 17940
},
{
"epoch": 0.45,
"grad_norm": 10.191650390625,
"learning_rate": 8.16949152542373e-06,
"loss": 1.3561,
"step": 17950
},
{
"epoch": 0.45,
"grad_norm": 5.769218444824219,
"learning_rate": 8.162711864406781e-06,
"loss": 1.4044,
"step": 17960
},
{
"epoch": 0.45,
"grad_norm": 47.62531661987305,
"learning_rate": 8.155932203389831e-06,
"loss": 1.6211,
"step": 17970
},
{
"epoch": 0.45,
"grad_norm": 3.514390230178833,
"learning_rate": 8.149152542372882e-06,
"loss": 1.288,
"step": 17980
},
{
"epoch": 0.45,
"grad_norm": 3.9963278770446777,
"learning_rate": 8.142372881355934e-06,
"loss": 1.3625,
"step": 17990
},
{
"epoch": 0.45,
"grad_norm": 4.2144856452941895,
"learning_rate": 8.135593220338983e-06,
"loss": 1.3154,
"step": 18000
},
{
"epoch": 0.45,
"eval_loss": 1.3663108348846436,
"eval_runtime": 66.1644,
"eval_samples_per_second": 15.114,
"eval_steps_per_second": 15.114,
"step": 18000
},
{
"epoch": 0.45,
"grad_norm": 3.047025680541992,
"learning_rate": 8.128813559322035e-06,
"loss": 1.3131,
"step": 18010
},
{
"epoch": 0.45,
"grad_norm": 5.798911094665527,
"learning_rate": 8.122033898305085e-06,
"loss": 1.4371,
"step": 18020
},
{
"epoch": 0.45,
"grad_norm": 3.8710880279541016,
"learning_rate": 8.115254237288136e-06,
"loss": 1.5147,
"step": 18030
},
{
"epoch": 0.45,
"grad_norm": 6.445448398590088,
"learning_rate": 8.108474576271187e-06,
"loss": 1.3332,
"step": 18040
},
{
"epoch": 0.45,
"grad_norm": 5.0877275466918945,
"learning_rate": 8.101694915254237e-06,
"loss": 1.5076,
"step": 18050
},
{
"epoch": 0.45,
"grad_norm": 6.0062737464904785,
"learning_rate": 8.094915254237289e-06,
"loss": 1.3257,
"step": 18060
},
{
"epoch": 0.45,
"grad_norm": 3.2079477310180664,
"learning_rate": 8.08813559322034e-06,
"loss": 1.2162,
"step": 18070
},
{
"epoch": 0.45,
"grad_norm": 2.974025249481201,
"learning_rate": 8.081355932203391e-06,
"loss": 1.3433,
"step": 18080
},
{
"epoch": 0.45,
"grad_norm": 11.321368217468262,
"learning_rate": 8.074576271186441e-06,
"loss": 1.4845,
"step": 18090
},
{
"epoch": 0.45,
"grad_norm": 3.768314838409424,
"learning_rate": 8.067796610169492e-06,
"loss": 1.3288,
"step": 18100
},
{
"epoch": 0.45,
"grad_norm": 3.9936201572418213,
"learning_rate": 8.061016949152542e-06,
"loss": 1.2507,
"step": 18110
},
{
"epoch": 0.45,
"grad_norm": 3.69484543800354,
"learning_rate": 8.054237288135594e-06,
"loss": 1.2615,
"step": 18120
},
{
"epoch": 0.45,
"grad_norm": 6.352110862731934,
"learning_rate": 8.047457627118645e-06,
"loss": 1.2888,
"step": 18130
},
{
"epoch": 0.45,
"grad_norm": 8.259808540344238,
"learning_rate": 8.040677966101695e-06,
"loss": 1.3588,
"step": 18140
},
{
"epoch": 0.45,
"grad_norm": 8.203502655029297,
"learning_rate": 8.033898305084746e-06,
"loss": 1.2716,
"step": 18150
},
{
"epoch": 0.45,
"grad_norm": 3.8740904331207275,
"learning_rate": 8.027118644067797e-06,
"loss": 1.4191,
"step": 18160
},
{
"epoch": 0.45,
"grad_norm": 3.9104349613189697,
"learning_rate": 8.020338983050849e-06,
"loss": 1.1126,
"step": 18170
},
{
"epoch": 0.45,
"grad_norm": 4.973033428192139,
"learning_rate": 8.013559322033899e-06,
"loss": 1.4034,
"step": 18180
},
{
"epoch": 0.45,
"grad_norm": 5.2367329597473145,
"learning_rate": 8.00677966101695e-06,
"loss": 1.4094,
"step": 18190
},
{
"epoch": 0.46,
"grad_norm": 4.842703819274902,
"learning_rate": 8.000000000000001e-06,
"loss": 1.4423,
"step": 18200
},
{
"epoch": 0.46,
"grad_norm": 6.39046573638916,
"learning_rate": 7.993220338983053e-06,
"loss": 1.3491,
"step": 18210
},
{
"epoch": 0.46,
"grad_norm": 3.424907922744751,
"learning_rate": 7.986440677966102e-06,
"loss": 1.268,
"step": 18220
},
{
"epoch": 0.46,
"grad_norm": 4.363860607147217,
"learning_rate": 7.979661016949152e-06,
"loss": 1.2921,
"step": 18230
},
{
"epoch": 0.46,
"grad_norm": 10.35586929321289,
"learning_rate": 7.972881355932204e-06,
"loss": 1.4456,
"step": 18240
},
{
"epoch": 0.46,
"grad_norm": 11.094548225402832,
"learning_rate": 7.966101694915255e-06,
"loss": 1.5682,
"step": 18250
},
{
"epoch": 0.46,
"grad_norm": 10.801241874694824,
"learning_rate": 7.959322033898306e-06,
"loss": 1.3475,
"step": 18260
},
{
"epoch": 0.46,
"grad_norm": 3.893005132675171,
"learning_rate": 7.952542372881356e-06,
"loss": 1.3858,
"step": 18270
},
{
"epoch": 0.46,
"grad_norm": 8.01429557800293,
"learning_rate": 7.945762711864407e-06,
"loss": 1.3695,
"step": 18280
},
{
"epoch": 0.46,
"grad_norm": 9.507207870483398,
"learning_rate": 7.938983050847459e-06,
"loss": 1.2442,
"step": 18290
},
{
"epoch": 0.46,
"grad_norm": 4.206308364868164,
"learning_rate": 7.93220338983051e-06,
"loss": 1.241,
"step": 18300
},
{
"epoch": 0.46,
"grad_norm": 7.3320794105529785,
"learning_rate": 7.92542372881356e-06,
"loss": 1.3338,
"step": 18310
},
{
"epoch": 0.46,
"grad_norm": 4.259428977966309,
"learning_rate": 7.91864406779661e-06,
"loss": 1.3902,
"step": 18320
},
{
"epoch": 0.46,
"grad_norm": 15.748085975646973,
"learning_rate": 7.911864406779661e-06,
"loss": 1.282,
"step": 18330
},
{
"epoch": 0.46,
"grad_norm": 7.123409748077393,
"learning_rate": 7.905084745762712e-06,
"loss": 1.3311,
"step": 18340
},
{
"epoch": 0.46,
"grad_norm": 7.022697925567627,
"learning_rate": 7.898305084745764e-06,
"loss": 1.292,
"step": 18350
},
{
"epoch": 0.46,
"grad_norm": 3.3042855262756348,
"learning_rate": 7.891525423728814e-06,
"loss": 1.3587,
"step": 18360
},
{
"epoch": 0.46,
"grad_norm": 3.5053913593292236,
"learning_rate": 7.884745762711865e-06,
"loss": 1.5188,
"step": 18370
},
{
"epoch": 0.46,
"grad_norm": 2.6854681968688965,
"learning_rate": 7.877966101694916e-06,
"loss": 1.3633,
"step": 18380
},
{
"epoch": 0.46,
"grad_norm": 3.8879082202911377,
"learning_rate": 7.871186440677968e-06,
"loss": 1.5491,
"step": 18390
},
{
"epoch": 0.46,
"grad_norm": 9.43989086151123,
"learning_rate": 7.864406779661017e-06,
"loss": 1.2686,
"step": 18400
},
{
"epoch": 0.46,
"grad_norm": 4.684770107269287,
"learning_rate": 7.857627118644069e-06,
"loss": 1.2463,
"step": 18410
},
{
"epoch": 0.46,
"grad_norm": 2.699023485183716,
"learning_rate": 7.850847457627119e-06,
"loss": 1.5216,
"step": 18420
},
{
"epoch": 0.46,
"grad_norm": 3.1254053115844727,
"learning_rate": 7.84406779661017e-06,
"loss": 1.4231,
"step": 18430
},
{
"epoch": 0.46,
"grad_norm": 3.4634742736816406,
"learning_rate": 7.837288135593221e-06,
"loss": 1.4174,
"step": 18440
},
{
"epoch": 0.46,
"grad_norm": 2.7554078102111816,
"learning_rate": 7.830508474576271e-06,
"loss": 1.2646,
"step": 18450
},
{
"epoch": 0.46,
"grad_norm": 7.782464504241943,
"learning_rate": 7.823728813559322e-06,
"loss": 1.273,
"step": 18460
},
{
"epoch": 0.46,
"grad_norm": 12.717724800109863,
"learning_rate": 7.816949152542374e-06,
"loss": 1.3536,
"step": 18470
},
{
"epoch": 0.46,
"grad_norm": 4.644845008850098,
"learning_rate": 7.810169491525425e-06,
"loss": 1.567,
"step": 18480
},
{
"epoch": 0.46,
"grad_norm": 2.032013177871704,
"learning_rate": 7.803389830508475e-06,
"loss": 1.4428,
"step": 18490
},
{
"epoch": 0.46,
"grad_norm": 8.91115951538086,
"learning_rate": 7.796610169491526e-06,
"loss": 1.4071,
"step": 18500
},
{
"epoch": 0.46,
"eval_loss": 1.3208402395248413,
"eval_runtime": 66.1773,
"eval_samples_per_second": 15.111,
"eval_steps_per_second": 15.111,
"step": 18500
},
{
"epoch": 0.46,
"grad_norm": 4.1798248291015625,
"learning_rate": 7.789830508474578e-06,
"loss": 1.3752,
"step": 18510
},
{
"epoch": 0.46,
"grad_norm": 3.6975016593933105,
"learning_rate": 7.783050847457628e-06,
"loss": 1.3685,
"step": 18520
},
{
"epoch": 0.46,
"grad_norm": 4.421247959136963,
"learning_rate": 7.776271186440679e-06,
"loss": 1.5264,
"step": 18530
},
{
"epoch": 0.46,
"grad_norm": 8.739845275878906,
"learning_rate": 7.769491525423729e-06,
"loss": 1.4609,
"step": 18540
},
{
"epoch": 0.46,
"grad_norm": 3.287424325942993,
"learning_rate": 7.76271186440678e-06,
"loss": 1.1154,
"step": 18550
},
{
"epoch": 0.46,
"grad_norm": 3.1192266941070557,
"learning_rate": 7.755932203389831e-06,
"loss": 1.2787,
"step": 18560
},
{
"epoch": 0.46,
"grad_norm": 8.368815422058105,
"learning_rate": 7.749152542372881e-06,
"loss": 1.3339,
"step": 18570
},
{
"epoch": 0.46,
"grad_norm": 6.241825580596924,
"learning_rate": 7.742372881355933e-06,
"loss": 1.2293,
"step": 18580
},
{
"epoch": 0.46,
"grad_norm": 9.267044067382812,
"learning_rate": 7.735593220338984e-06,
"loss": 1.4354,
"step": 18590
},
{
"epoch": 0.47,
"grad_norm": 5.569955348968506,
"learning_rate": 7.728813559322035e-06,
"loss": 1.3941,
"step": 18600
},
{
"epoch": 0.47,
"grad_norm": 11.273942947387695,
"learning_rate": 7.722033898305085e-06,
"loss": 1.1127,
"step": 18610
},
{
"epoch": 0.47,
"grad_norm": 3.7952206134796143,
"learning_rate": 7.715254237288136e-06,
"loss": 1.3437,
"step": 18620
},
{
"epoch": 0.47,
"grad_norm": 2.9084484577178955,
"learning_rate": 7.708474576271186e-06,
"loss": 1.3646,
"step": 18630
},
{
"epoch": 0.47,
"grad_norm": 4.15964937210083,
"learning_rate": 7.701694915254238e-06,
"loss": 1.4666,
"step": 18640
},
{
"epoch": 0.47,
"grad_norm": 4.938930988311768,
"learning_rate": 7.694915254237289e-06,
"loss": 1.4065,
"step": 18650
},
{
"epoch": 0.47,
"grad_norm": 8.298666954040527,
"learning_rate": 7.688135593220339e-06,
"loss": 1.354,
"step": 18660
},
{
"epoch": 0.47,
"grad_norm": 3.1837921142578125,
"learning_rate": 7.68135593220339e-06,
"loss": 1.3879,
"step": 18670
},
{
"epoch": 0.47,
"grad_norm": 2.978053331375122,
"learning_rate": 7.674576271186441e-06,
"loss": 1.3765,
"step": 18680
},
{
"epoch": 0.47,
"grad_norm": 5.5602827072143555,
"learning_rate": 7.667796610169493e-06,
"loss": 1.2135,
"step": 18690
},
{
"epoch": 0.47,
"grad_norm": 5.481113910675049,
"learning_rate": 7.661016949152543e-06,
"loss": 1.3841,
"step": 18700
},
{
"epoch": 0.47,
"grad_norm": 9.419681549072266,
"learning_rate": 7.654237288135594e-06,
"loss": 1.2349,
"step": 18710
},
{
"epoch": 0.47,
"grad_norm": 4.955466270446777,
"learning_rate": 7.647457627118645e-06,
"loss": 1.4208,
"step": 18720
},
{
"epoch": 0.47,
"grad_norm": 5.841220855712891,
"learning_rate": 7.640677966101695e-06,
"loss": 1.4526,
"step": 18730
},
{
"epoch": 0.47,
"grad_norm": 3.1412012577056885,
"learning_rate": 7.633898305084746e-06,
"loss": 1.4606,
"step": 18740
},
{
"epoch": 0.47,
"grad_norm": 4.624940395355225,
"learning_rate": 7.627118644067797e-06,
"loss": 1.3875,
"step": 18750
},
{
"epoch": 0.47,
"grad_norm": 9.828381538391113,
"learning_rate": 7.6203389830508476e-06,
"loss": 1.4079,
"step": 18760
},
{
"epoch": 0.47,
"grad_norm": 5.299017429351807,
"learning_rate": 7.613559322033899e-06,
"loss": 1.278,
"step": 18770
},
{
"epoch": 0.47,
"grad_norm": 6.449117183685303,
"learning_rate": 7.6067796610169495e-06,
"loss": 1.4141,
"step": 18780
},
{
"epoch": 0.47,
"grad_norm": 6.961145401000977,
"learning_rate": 7.600000000000001e-06,
"loss": 1.2828,
"step": 18790
},
{
"epoch": 0.47,
"grad_norm": 8.588834762573242,
"learning_rate": 7.5932203389830515e-06,
"loss": 1.3544,
"step": 18800
},
{
"epoch": 0.47,
"grad_norm": 2.5683794021606445,
"learning_rate": 7.586440677966103e-06,
"loss": 1.4127,
"step": 18810
},
{
"epoch": 0.47,
"grad_norm": 2.101924180984497,
"learning_rate": 7.5796610169491534e-06,
"loss": 1.229,
"step": 18820
},
{
"epoch": 0.47,
"grad_norm": 5.389444351196289,
"learning_rate": 7.572881355932205e-06,
"loss": 1.3112,
"step": 18830
},
{
"epoch": 0.47,
"grad_norm": 3.154507875442505,
"learning_rate": 7.5661016949152545e-06,
"loss": 1.2603,
"step": 18840
},
{
"epoch": 0.47,
"grad_norm": 4.938506126403809,
"learning_rate": 7.559322033898305e-06,
"loss": 1.5423,
"step": 18850
},
{
"epoch": 0.47,
"grad_norm": 5.838810443878174,
"learning_rate": 7.5525423728813565e-06,
"loss": 1.0651,
"step": 18860
},
{
"epoch": 0.47,
"grad_norm": 7.946393013000488,
"learning_rate": 7.545762711864407e-06,
"loss": 1.4706,
"step": 18870
},
{
"epoch": 0.47,
"grad_norm": 12.6395902633667,
"learning_rate": 7.5389830508474584e-06,
"loss": 1.4036,
"step": 18880
},
{
"epoch": 0.47,
"grad_norm": 4.769916534423828,
"learning_rate": 7.532203389830509e-06,
"loss": 1.33,
"step": 18890
},
{
"epoch": 0.47,
"grad_norm": 5.928445816040039,
"learning_rate": 7.52542372881356e-06,
"loss": 1.4844,
"step": 18900
},
{
"epoch": 0.47,
"grad_norm": 4.121998310089111,
"learning_rate": 7.518644067796611e-06,
"loss": 1.2672,
"step": 18910
},
{
"epoch": 0.47,
"grad_norm": 1.256914496421814,
"learning_rate": 7.511864406779662e-06,
"loss": 1.337,
"step": 18920
},
{
"epoch": 0.47,
"grad_norm": 9.951703071594238,
"learning_rate": 7.505084745762713e-06,
"loss": 1.3359,
"step": 18930
},
{
"epoch": 0.47,
"grad_norm": 5.373715400695801,
"learning_rate": 7.498305084745763e-06,
"loss": 1.2437,
"step": 18940
},
{
"epoch": 0.47,
"grad_norm": 7.120504379272461,
"learning_rate": 7.491525423728814e-06,
"loss": 1.1883,
"step": 18950
},
{
"epoch": 0.47,
"grad_norm": 7.661159992218018,
"learning_rate": 7.4847457627118646e-06,
"loss": 1.2648,
"step": 18960
},
{
"epoch": 0.47,
"grad_norm": 5.859286308288574,
"learning_rate": 7.477966101694916e-06,
"loss": 1.3106,
"step": 18970
},
{
"epoch": 0.47,
"grad_norm": 7.276169300079346,
"learning_rate": 7.4711864406779665e-06,
"loss": 1.393,
"step": 18980
},
{
"epoch": 0.47,
"grad_norm": 6.31447172164917,
"learning_rate": 7.464406779661018e-06,
"loss": 1.4051,
"step": 18990
},
{
"epoch": 0.47,
"grad_norm": 8.36728572845459,
"learning_rate": 7.4576271186440685e-06,
"loss": 1.4003,
"step": 19000
},
{
"epoch": 0.47,
"eval_loss": 1.347457766532898,
"eval_runtime": 66.1461,
"eval_samples_per_second": 15.118,
"eval_steps_per_second": 15.118,
"step": 19000
},
{
"epoch": 0.48,
"grad_norm": 3.911505937576294,
"learning_rate": 7.45084745762712e-06,
"loss": 1.2646,
"step": 19010
},
{
"epoch": 0.48,
"grad_norm": 3.6765291690826416,
"learning_rate": 7.4440677966101704e-06,
"loss": 1.3827,
"step": 19020
},
{
"epoch": 0.48,
"grad_norm": 3.899599313735962,
"learning_rate": 7.437288135593221e-06,
"loss": 1.4144,
"step": 19030
},
{
"epoch": 0.48,
"grad_norm": 6.436791896820068,
"learning_rate": 7.430508474576272e-06,
"loss": 1.2087,
"step": 19040
},
{
"epoch": 0.48,
"grad_norm": 4.624211311340332,
"learning_rate": 7.423728813559322e-06,
"loss": 1.3824,
"step": 19050
},
{
"epoch": 0.48,
"grad_norm": 6.657593727111816,
"learning_rate": 7.4169491525423735e-06,
"loss": 1.3856,
"step": 19060
},
{
"epoch": 0.48,
"grad_norm": 7.132912635803223,
"learning_rate": 7.410169491525424e-06,
"loss": 1.4319,
"step": 19070
},
{
"epoch": 0.48,
"grad_norm": 2.8681843280792236,
"learning_rate": 7.4033898305084754e-06,
"loss": 1.419,
"step": 19080
},
{
"epoch": 0.48,
"grad_norm": 5.819919109344482,
"learning_rate": 7.396610169491526e-06,
"loss": 1.3988,
"step": 19090
},
{
"epoch": 0.48,
"grad_norm": 6.482515811920166,
"learning_rate": 7.3898305084745766e-06,
"loss": 1.3366,
"step": 19100
},
{
"epoch": 0.48,
"grad_norm": 10.341208457946777,
"learning_rate": 7.383050847457628e-06,
"loss": 1.3502,
"step": 19110
},
{
"epoch": 0.48,
"grad_norm": 2.820133686065674,
"learning_rate": 7.3762711864406785e-06,
"loss": 1.415,
"step": 19120
},
{
"epoch": 0.48,
"grad_norm": 7.112204551696777,
"learning_rate": 7.36949152542373e-06,
"loss": 1.2609,
"step": 19130
},
{
"epoch": 0.48,
"grad_norm": 6.172835826873779,
"learning_rate": 7.3627118644067805e-06,
"loss": 1.2456,
"step": 19140
},
{
"epoch": 0.48,
"grad_norm": 2.9154605865478516,
"learning_rate": 7.355932203389831e-06,
"loss": 1.3332,
"step": 19150
},
{
"epoch": 0.48,
"grad_norm": 6.8542256355285645,
"learning_rate": 7.3491525423728816e-06,
"loss": 1.2728,
"step": 19160
},
{
"epoch": 0.48,
"grad_norm": 7.636801719665527,
"learning_rate": 7.342372881355932e-06,
"loss": 1.3744,
"step": 19170
},
{
"epoch": 0.48,
"grad_norm": 7.556679725646973,
"learning_rate": 7.3355932203389835e-06,
"loss": 1.446,
"step": 19180
},
{
"epoch": 0.48,
"grad_norm": 22.831199645996094,
"learning_rate": 7.328813559322034e-06,
"loss": 1.358,
"step": 19190
},
{
"epoch": 0.48,
"grad_norm": 10.184800148010254,
"learning_rate": 7.3220338983050855e-06,
"loss": 1.2379,
"step": 19200
},
{
"epoch": 0.48,
"grad_norm": 1.9062511920928955,
"learning_rate": 7.315254237288136e-06,
"loss": 1.3724,
"step": 19210
},
{
"epoch": 0.48,
"grad_norm": 9.429797172546387,
"learning_rate": 7.3084745762711874e-06,
"loss": 1.33,
"step": 19220
},
{
"epoch": 0.48,
"grad_norm": 3.129505157470703,
"learning_rate": 7.301694915254238e-06,
"loss": 1.1795,
"step": 19230
},
{
"epoch": 0.48,
"grad_norm": 6.839028835296631,
"learning_rate": 7.294915254237289e-06,
"loss": 1.2983,
"step": 19240
},
{
"epoch": 0.48,
"grad_norm": 15.552626609802246,
"learning_rate": 7.288135593220339e-06,
"loss": 1.4867,
"step": 19250
},
{
"epoch": 0.48,
"grad_norm": 3.4032084941864014,
"learning_rate": 7.28135593220339e-06,
"loss": 1.4256,
"step": 19260
},
{
"epoch": 0.48,
"grad_norm": 6.437047481536865,
"learning_rate": 7.274576271186441e-06,
"loss": 1.3215,
"step": 19270
},
{
"epoch": 0.48,
"grad_norm": 5.592834949493408,
"learning_rate": 7.267796610169492e-06,
"loss": 1.3466,
"step": 19280
},
{
"epoch": 0.48,
"grad_norm": 10.394824028015137,
"learning_rate": 7.261016949152543e-06,
"loss": 1.3806,
"step": 19290
},
{
"epoch": 0.48,
"grad_norm": 6.921483039855957,
"learning_rate": 7.2542372881355936e-06,
"loss": 1.5624,
"step": 19300
},
{
"epoch": 0.48,
"grad_norm": 3.078634262084961,
"learning_rate": 7.247457627118645e-06,
"loss": 1.1216,
"step": 19310
},
{
"epoch": 0.48,
"grad_norm": 6.003988742828369,
"learning_rate": 7.2406779661016955e-06,
"loss": 1.2759,
"step": 19320
},
{
"epoch": 0.48,
"grad_norm": 2.419410467147827,
"learning_rate": 7.233898305084747e-06,
"loss": 1.2067,
"step": 19330
},
{
"epoch": 0.48,
"grad_norm": 3.014220952987671,
"learning_rate": 7.2271186440677975e-06,
"loss": 1.1819,
"step": 19340
},
{
"epoch": 0.48,
"grad_norm": 5.271297454833984,
"learning_rate": 7.220338983050849e-06,
"loss": 1.4742,
"step": 19350
},
{
"epoch": 0.48,
"grad_norm": 2.456838607788086,
"learning_rate": 7.2135593220338986e-06,
"loss": 1.4656,
"step": 19360
},
{
"epoch": 0.48,
"grad_norm": 5.350526332855225,
"learning_rate": 7.206779661016949e-06,
"loss": 1.1238,
"step": 19370
},
{
"epoch": 0.48,
"grad_norm": 5.42751932144165,
"learning_rate": 7.2000000000000005e-06,
"loss": 1.4704,
"step": 19380
},
{
"epoch": 0.48,
"grad_norm": 3.97116756439209,
"learning_rate": 7.193220338983051e-06,
"loss": 1.4455,
"step": 19390
},
{
"epoch": 0.48,
"grad_norm": 5.157166481018066,
"learning_rate": 7.1864406779661025e-06,
"loss": 1.3725,
"step": 19400
},
{
"epoch": 0.49,
"grad_norm": 4.2230072021484375,
"learning_rate": 7.179661016949153e-06,
"loss": 1.3768,
"step": 19410
},
{
"epoch": 0.49,
"grad_norm": 5.513181686401367,
"learning_rate": 7.1728813559322044e-06,
"loss": 1.3018,
"step": 19420
},
{
"epoch": 0.49,
"grad_norm": 7.353570461273193,
"learning_rate": 7.166101694915255e-06,
"loss": 1.513,
"step": 19430
},
{
"epoch": 0.49,
"grad_norm": 6.263894081115723,
"learning_rate": 7.159322033898306e-06,
"loss": 1.4521,
"step": 19440
},
{
"epoch": 0.49,
"grad_norm": 13.368189811706543,
"learning_rate": 7.152542372881357e-06,
"loss": 1.3158,
"step": 19450
},
{
"epoch": 0.49,
"grad_norm": 4.644199371337891,
"learning_rate": 7.145762711864407e-06,
"loss": 1.4184,
"step": 19460
},
{
"epoch": 0.49,
"grad_norm": 5.206189155578613,
"learning_rate": 7.138983050847458e-06,
"loss": 1.3848,
"step": 19470
},
{
"epoch": 0.49,
"grad_norm": 7.3628153800964355,
"learning_rate": 7.132203389830509e-06,
"loss": 1.2512,
"step": 19480
},
{
"epoch": 0.49,
"grad_norm": 7.519322395324707,
"learning_rate": 7.12542372881356e-06,
"loss": 1.261,
"step": 19490
},
{
"epoch": 0.49,
"grad_norm": 1.5350415706634521,
"learning_rate": 7.1186440677966106e-06,
"loss": 1.2517,
"step": 19500
},
{
"epoch": 0.49,
"eval_loss": 1.3194345235824585,
"eval_runtime": 66.1305,
"eval_samples_per_second": 15.122,
"eval_steps_per_second": 15.122,
"step": 19500
},
{
"epoch": 0.49,
"grad_norm": 3.0648765563964844,
"learning_rate": 7.111864406779662e-06,
"loss": 1.398,
"step": 19510
},
{
"epoch": 0.49,
"grad_norm": 5.124095439910889,
"learning_rate": 7.1050847457627125e-06,
"loss": 1.404,
"step": 19520
},
{
"epoch": 0.49,
"grad_norm": 2.8961544036865234,
"learning_rate": 7.098305084745764e-06,
"loss": 1.2651,
"step": 19530
},
{
"epoch": 0.49,
"grad_norm": 3.0641605854034424,
"learning_rate": 7.0915254237288145e-06,
"loss": 1.3587,
"step": 19540
},
{
"epoch": 0.49,
"grad_norm": 3.618454933166504,
"learning_rate": 7.084745762711865e-06,
"loss": 1.3533,
"step": 19550
},
{
"epoch": 0.49,
"grad_norm": 3.3083536624908447,
"learning_rate": 7.077966101694916e-06,
"loss": 1.2859,
"step": 19560
},
{
"epoch": 0.49,
"grad_norm": 1.260048508644104,
"learning_rate": 7.071186440677966e-06,
"loss": 1.2224,
"step": 19570
},
{
"epoch": 0.49,
"grad_norm": 5.410586833953857,
"learning_rate": 7.0644067796610175e-06,
"loss": 1.2916,
"step": 19580
},
{
"epoch": 0.49,
"grad_norm": 1.9295905828475952,
"learning_rate": 7.057627118644068e-06,
"loss": 1.3004,
"step": 19590
},
{
"epoch": 0.49,
"grad_norm": 10.48406982421875,
"learning_rate": 7.0508474576271195e-06,
"loss": 1.3105,
"step": 19600
},
{
"epoch": 0.49,
"grad_norm": 4.99778413772583,
"learning_rate": 7.04406779661017e-06,
"loss": 1.2481,
"step": 19610
},
{
"epoch": 0.49,
"grad_norm": 5.476469993591309,
"learning_rate": 7.037288135593221e-06,
"loss": 1.345,
"step": 19620
},
{
"epoch": 0.49,
"grad_norm": 7.250665664672852,
"learning_rate": 7.030508474576272e-06,
"loss": 1.2993,
"step": 19630
},
{
"epoch": 0.49,
"grad_norm": 7.802820682525635,
"learning_rate": 7.0237288135593225e-06,
"loss": 1.3106,
"step": 19640
},
{
"epoch": 0.49,
"grad_norm": 2.5077409744262695,
"learning_rate": 7.016949152542374e-06,
"loss": 1.1298,
"step": 19650
},
{
"epoch": 0.49,
"grad_norm": 2.3012547492980957,
"learning_rate": 7.0101694915254245e-06,
"loss": 1.2944,
"step": 19660
},
{
"epoch": 0.49,
"grad_norm": 10.731738090515137,
"learning_rate": 7.003389830508475e-06,
"loss": 1.2545,
"step": 19670
},
{
"epoch": 0.49,
"grad_norm": 11.851224899291992,
"learning_rate": 6.996610169491526e-06,
"loss": 1.3364,
"step": 19680
},
{
"epoch": 0.49,
"grad_norm": 2.060750722885132,
"learning_rate": 6.989830508474576e-06,
"loss": 1.2909,
"step": 19690
},
{
"epoch": 0.49,
"grad_norm": 8.549054145812988,
"learning_rate": 6.9830508474576275e-06,
"loss": 1.405,
"step": 19700
},
{
"epoch": 0.49,
"grad_norm": 5.37472677230835,
"learning_rate": 6.976271186440678e-06,
"loss": 1.5615,
"step": 19710
},
{
"epoch": 0.49,
"grad_norm": 4.753462314605713,
"learning_rate": 6.9694915254237295e-06,
"loss": 1.362,
"step": 19720
},
{
"epoch": 0.49,
"grad_norm": 6.330374717712402,
"learning_rate": 6.96271186440678e-06,
"loss": 1.3563,
"step": 19730
},
{
"epoch": 0.49,
"grad_norm": 6.962733268737793,
"learning_rate": 6.9559322033898315e-06,
"loss": 1.3272,
"step": 19740
},
{
"epoch": 0.49,
"grad_norm": 13.372142791748047,
"learning_rate": 6.949152542372882e-06,
"loss": 1.4422,
"step": 19750
},
{
"epoch": 0.49,
"grad_norm": 4.561896800994873,
"learning_rate": 6.942372881355933e-06,
"loss": 1.3691,
"step": 19760
},
{
"epoch": 0.49,
"grad_norm": 2.2547922134399414,
"learning_rate": 6.935593220338983e-06,
"loss": 1.4905,
"step": 19770
},
{
"epoch": 0.49,
"grad_norm": 4.15011739730835,
"learning_rate": 6.928813559322034e-06,
"loss": 1.3899,
"step": 19780
},
{
"epoch": 0.49,
"grad_norm": 11.278037071228027,
"learning_rate": 6.922033898305085e-06,
"loss": 1.0803,
"step": 19790
},
{
"epoch": 0.49,
"grad_norm": 9.100043296813965,
"learning_rate": 6.915254237288136e-06,
"loss": 1.4593,
"step": 19800
},
{
"epoch": 0.5,
"grad_norm": 10.643383026123047,
"learning_rate": 6.908474576271187e-06,
"loss": 1.3571,
"step": 19810
},
{
"epoch": 0.5,
"grad_norm": 11.314017295837402,
"learning_rate": 6.901694915254238e-06,
"loss": 1.2376,
"step": 19820
},
{
"epoch": 0.5,
"grad_norm": 3.855220317840576,
"learning_rate": 6.894915254237289e-06,
"loss": 1.1095,
"step": 19830
},
{
"epoch": 0.5,
"grad_norm": 5.896849155426025,
"learning_rate": 6.8881355932203395e-06,
"loss": 1.2078,
"step": 19840
},
{
"epoch": 0.5,
"grad_norm": 2.1142632961273193,
"learning_rate": 6.881355932203391e-06,
"loss": 1.3017,
"step": 19850
},
{
"epoch": 0.5,
"grad_norm": 7.116094589233398,
"learning_rate": 6.8745762711864415e-06,
"loss": 1.217,
"step": 19860
},
{
"epoch": 0.5,
"grad_norm": 3.392282724380493,
"learning_rate": 6.867796610169493e-06,
"loss": 1.4027,
"step": 19870
},
{
"epoch": 0.5,
"grad_norm": 5.760110855102539,
"learning_rate": 6.861016949152543e-06,
"loss": 1.2888,
"step": 19880
},
{
"epoch": 0.5,
"grad_norm": 2.312904119491577,
"learning_rate": 6.854237288135593e-06,
"loss": 1.39,
"step": 19890
},
{
"epoch": 0.5,
"grad_norm": 9.748230934143066,
"learning_rate": 6.8474576271186445e-06,
"loss": 1.408,
"step": 19900
},
{
"epoch": 0.5,
"grad_norm": 8.029982566833496,
"learning_rate": 6.840677966101695e-06,
"loss": 1.2631,
"step": 19910
},
{
"epoch": 0.5,
"grad_norm": 5.408463954925537,
"learning_rate": 6.8338983050847465e-06,
"loss": 1.3545,
"step": 19920
},
{
"epoch": 0.5,
"grad_norm": 4.480403900146484,
"learning_rate": 6.827118644067797e-06,
"loss": 1.4549,
"step": 19930
},
{
"epoch": 0.5,
"grad_norm": 8.981225967407227,
"learning_rate": 6.8203389830508485e-06,
"loss": 1.3352,
"step": 19940
},
{
"epoch": 0.5,
"grad_norm": 1.9335066080093384,
"learning_rate": 6.813559322033899e-06,
"loss": 1.2825,
"step": 19950
},
{
"epoch": 0.5,
"grad_norm": 7.6206464767456055,
"learning_rate": 6.80677966101695e-06,
"loss": 1.4976,
"step": 19960
},
{
"epoch": 0.5,
"grad_norm": 14.687816619873047,
"learning_rate": 6.800000000000001e-06,
"loss": 1.4687,
"step": 19970
},
{
"epoch": 0.5,
"grad_norm": 9.034219741821289,
"learning_rate": 6.793220338983051e-06,
"loss": 1.4715,
"step": 19980
},
{
"epoch": 0.5,
"grad_norm": 8.120539665222168,
"learning_rate": 6.786440677966102e-06,
"loss": 1.4331,
"step": 19990
},
{
"epoch": 0.5,
"grad_norm": 6.594362258911133,
"learning_rate": 6.779661016949153e-06,
"loss": 1.4516,
"step": 20000
},
{
"epoch": 0.5,
"eval_loss": 1.348677158355713,
"eval_runtime": 66.1188,
"eval_samples_per_second": 15.124,
"eval_steps_per_second": 15.124,
"step": 20000
},
{
"epoch": 0.5,
"grad_norm": 2.159421920776367,
"learning_rate": 6.772881355932204e-06,
"loss": 1.4193,
"step": 20010
},
{
"epoch": 0.5,
"grad_norm": 7.60541296005249,
"learning_rate": 6.766101694915255e-06,
"loss": 1.4747,
"step": 20020
},
{
"epoch": 0.5,
"grad_norm": 2.0991897583007812,
"learning_rate": 6.759322033898306e-06,
"loss": 1.2775,
"step": 20030
},
{
"epoch": 0.5,
"grad_norm": 7.756378173828125,
"learning_rate": 6.7525423728813565e-06,
"loss": 1.343,
"step": 20040
},
{
"epoch": 0.5,
"grad_norm": 7.062932014465332,
"learning_rate": 6.745762711864408e-06,
"loss": 1.5168,
"step": 20050
},
{
"epoch": 0.5,
"grad_norm": 6.689007759094238,
"learning_rate": 6.7389830508474585e-06,
"loss": 1.2894,
"step": 20060
},
{
"epoch": 0.5,
"grad_norm": 5.31237268447876,
"learning_rate": 6.73220338983051e-06,
"loss": 1.3382,
"step": 20070
},
{
"epoch": 0.5,
"grad_norm": 3.6598782539367676,
"learning_rate": 6.7254237288135604e-06,
"loss": 1.3839,
"step": 20080
},
{
"epoch": 0.5,
"grad_norm": 3.3091630935668945,
"learning_rate": 6.71864406779661e-06,
"loss": 1.3654,
"step": 20090
},
{
"epoch": 0.5,
"grad_norm": 2.1399312019348145,
"learning_rate": 6.7118644067796615e-06,
"loss": 1.5429,
"step": 20100
},
{
"epoch": 0.5,
"grad_norm": 4.380695819854736,
"learning_rate": 6.705084745762712e-06,
"loss": 1.4629,
"step": 20110
},
{
"epoch": 0.5,
"grad_norm": 9.330241203308105,
"learning_rate": 6.6983050847457635e-06,
"loss": 1.3547,
"step": 20120
},
{
"epoch": 0.5,
"grad_norm": 5.803046226501465,
"learning_rate": 6.691525423728814e-06,
"loss": 1.3534,
"step": 20130
},
{
"epoch": 0.5,
"grad_norm": 4.967430591583252,
"learning_rate": 6.6847457627118655e-06,
"loss": 1.4965,
"step": 20140
},
{
"epoch": 0.5,
"grad_norm": 8.131784439086914,
"learning_rate": 6.677966101694916e-06,
"loss": 1.4012,
"step": 20150
},
{
"epoch": 0.5,
"grad_norm": 9.220160484313965,
"learning_rate": 6.6711864406779666e-06,
"loss": 1.3637,
"step": 20160
},
{
"epoch": 0.5,
"grad_norm": 6.833899974822998,
"learning_rate": 6.664406779661018e-06,
"loss": 1.3748,
"step": 20170
},
{
"epoch": 0.5,
"grad_norm": 13.869956970214844,
"learning_rate": 6.6576271186440685e-06,
"loss": 1.4474,
"step": 20180
},
{
"epoch": 0.5,
"grad_norm": 9.462739944458008,
"learning_rate": 6.650847457627119e-06,
"loss": 1.3893,
"step": 20190
},
{
"epoch": 0.51,
"grad_norm": 2.321714162826538,
"learning_rate": 6.64406779661017e-06,
"loss": 1.3836,
"step": 20200
},
{
"epoch": 0.51,
"grad_norm": 11.65925407409668,
"learning_rate": 6.637288135593221e-06,
"loss": 1.493,
"step": 20210
},
{
"epoch": 0.51,
"grad_norm": 6.808838367462158,
"learning_rate": 6.6305084745762716e-06,
"loss": 1.3632,
"step": 20220
},
{
"epoch": 0.51,
"grad_norm": 4.534874439239502,
"learning_rate": 6.623728813559322e-06,
"loss": 1.4303,
"step": 20230
},
{
"epoch": 0.51,
"grad_norm": 1.9837802648544312,
"learning_rate": 6.6169491525423735e-06,
"loss": 1.446,
"step": 20240
},
{
"epoch": 0.51,
"grad_norm": 7.779874324798584,
"learning_rate": 6.610169491525424e-06,
"loss": 1.4566,
"step": 20250
},
{
"epoch": 0.51,
"grad_norm": 5.729377746582031,
"learning_rate": 6.6033898305084755e-06,
"loss": 1.401,
"step": 20260
},
{
"epoch": 0.51,
"grad_norm": 12.633749008178711,
"learning_rate": 6.596610169491526e-06,
"loss": 1.2969,
"step": 20270
},
{
"epoch": 0.51,
"grad_norm": 2.7961552143096924,
"learning_rate": 6.5898305084745774e-06,
"loss": 1.4151,
"step": 20280
},
{
"epoch": 0.51,
"grad_norm": 8.461606979370117,
"learning_rate": 6.583050847457627e-06,
"loss": 1.2665,
"step": 20290
},
{
"epoch": 0.51,
"grad_norm": 10.325592041015625,
"learning_rate": 6.576271186440678e-06,
"loss": 1.3666,
"step": 20300
},
{
"epoch": 0.51,
"grad_norm": 3.0818538665771484,
"learning_rate": 6.569491525423729e-06,
"loss": 1.4307,
"step": 20310
},
{
"epoch": 0.51,
"grad_norm": 7.449318885803223,
"learning_rate": 6.56271186440678e-06,
"loss": 1.4585,
"step": 20320
},
{
"epoch": 0.51,
"grad_norm": 4.6587042808532715,
"learning_rate": 6.555932203389831e-06,
"loss": 1.4053,
"step": 20330
},
{
"epoch": 0.51,
"grad_norm": 5.837299346923828,
"learning_rate": 6.549152542372882e-06,
"loss": 1.3456,
"step": 20340
},
{
"epoch": 0.51,
"grad_norm": 7.345305442810059,
"learning_rate": 6.542372881355933e-06,
"loss": 1.3618,
"step": 20350
},
{
"epoch": 0.51,
"grad_norm": 5.679592609405518,
"learning_rate": 6.5355932203389836e-06,
"loss": 1.3875,
"step": 20360
},
{
"epoch": 0.51,
"grad_norm": 1.062429428100586,
"learning_rate": 6.528813559322035e-06,
"loss": 1.3083,
"step": 20370
},
{
"epoch": 0.51,
"grad_norm": 5.2157769203186035,
"learning_rate": 6.5220338983050855e-06,
"loss": 1.3582,
"step": 20380
},
{
"epoch": 0.51,
"grad_norm": 4.890625476837158,
"learning_rate": 6.515254237288137e-06,
"loss": 1.3184,
"step": 20390
},
{
"epoch": 0.51,
"grad_norm": 1.587312936782837,
"learning_rate": 6.508474576271187e-06,
"loss": 1.2191,
"step": 20400
},
{
"epoch": 0.51,
"grad_norm": 4.013046741485596,
"learning_rate": 6.501694915254237e-06,
"loss": 1.3331,
"step": 20410
},
{
"epoch": 0.51,
"grad_norm": 4.076744556427002,
"learning_rate": 6.4949152542372886e-06,
"loss": 1.4351,
"step": 20420
},
{
"epoch": 0.51,
"grad_norm": 6.493119239807129,
"learning_rate": 6.488135593220339e-06,
"loss": 1.3519,
"step": 20430
},
{
"epoch": 0.51,
"grad_norm": 6.127922058105469,
"learning_rate": 6.4813559322033905e-06,
"loss": 1.4255,
"step": 20440
},
{
"epoch": 0.51,
"grad_norm": 4.424916744232178,
"learning_rate": 6.474576271186441e-06,
"loss": 1.44,
"step": 20450
},
{
"epoch": 0.51,
"grad_norm": 9.894845008850098,
"learning_rate": 6.4677966101694925e-06,
"loss": 1.2922,
"step": 20460
},
{
"epoch": 0.51,
"grad_norm": 5.174190521240234,
"learning_rate": 6.461016949152543e-06,
"loss": 1.439,
"step": 20470
},
{
"epoch": 0.51,
"grad_norm": 6.313706398010254,
"learning_rate": 6.4542372881355944e-06,
"loss": 1.3229,
"step": 20480
},
{
"epoch": 0.51,
"grad_norm": 3.240434169769287,
"learning_rate": 6.447457627118645e-06,
"loss": 1.3911,
"step": 20490
},
{
"epoch": 0.51,
"grad_norm": 7.466948986053467,
"learning_rate": 6.440677966101695e-06,
"loss": 1.4467,
"step": 20500
},
{
"epoch": 0.51,
"eval_loss": 1.3339221477508545,
"eval_runtime": 66.152,
"eval_samples_per_second": 15.117,
"eval_steps_per_second": 15.117,
"step": 20500
},
{
"epoch": 0.51,
"grad_norm": 32.95411682128906,
"learning_rate": 6.433898305084746e-06,
"loss": 1.2821,
"step": 20510
},
{
"epoch": 0.51,
"grad_norm": 11.001514434814453,
"learning_rate": 6.427118644067797e-06,
"loss": 1.3523,
"step": 20520
},
{
"epoch": 0.51,
"grad_norm": 1.6923043727874756,
"learning_rate": 6.420338983050848e-06,
"loss": 1.3424,
"step": 20530
},
{
"epoch": 0.51,
"grad_norm": 6.267253398895264,
"learning_rate": 6.413559322033899e-06,
"loss": 1.2163,
"step": 20540
},
{
"epoch": 0.51,
"grad_norm": 15.797646522521973,
"learning_rate": 6.40677966101695e-06,
"loss": 1.2104,
"step": 20550
},
{
"epoch": 0.51,
"grad_norm": 3.1398627758026123,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.2264,
"step": 20560
},
{
"epoch": 0.51,
"grad_norm": 1.7001848220825195,
"learning_rate": 6.393220338983052e-06,
"loss": 1.345,
"step": 20570
},
{
"epoch": 0.51,
"grad_norm": 6.551231861114502,
"learning_rate": 6.3864406779661025e-06,
"loss": 1.295,
"step": 20580
},
{
"epoch": 0.51,
"grad_norm": 2.501774787902832,
"learning_rate": 6.379661016949154e-06,
"loss": 1.2727,
"step": 20590
},
{
"epoch": 0.52,
"grad_norm": 5.2971906661987305,
"learning_rate": 6.372881355932204e-06,
"loss": 1.1617,
"step": 20600
},
{
"epoch": 0.52,
"grad_norm": 8.737128257751465,
"learning_rate": 6.366101694915254e-06,
"loss": 1.3792,
"step": 20610
},
{
"epoch": 0.52,
"grad_norm": 1.8445500135421753,
"learning_rate": 6.3593220338983056e-06,
"loss": 1.2818,
"step": 20620
},
{
"epoch": 0.52,
"grad_norm": 6.586302280426025,
"learning_rate": 6.352542372881356e-06,
"loss": 1.471,
"step": 20630
},
{
"epoch": 0.52,
"grad_norm": 2.3581125736236572,
"learning_rate": 6.3457627118644075e-06,
"loss": 1.26,
"step": 20640
},
{
"epoch": 0.52,
"grad_norm": 3.006883382797241,
"learning_rate": 6.338983050847458e-06,
"loss": 1.4312,
"step": 20650
},
{
"epoch": 0.52,
"grad_norm": 1.8936930894851685,
"learning_rate": 6.3322033898305095e-06,
"loss": 1.3705,
"step": 20660
},
{
"epoch": 0.52,
"grad_norm": 5.22953462600708,
"learning_rate": 6.32542372881356e-06,
"loss": 1.3252,
"step": 20670
},
{
"epoch": 0.52,
"grad_norm": 7.669064044952393,
"learning_rate": 6.318644067796611e-06,
"loss": 1.4673,
"step": 20680
},
{
"epoch": 0.52,
"grad_norm": 3.7508702278137207,
"learning_rate": 6.311864406779662e-06,
"loss": 1.3465,
"step": 20690
},
{
"epoch": 0.52,
"grad_norm": 7.959824085235596,
"learning_rate": 6.3050847457627125e-06,
"loss": 1.4148,
"step": 20700
},
{
"epoch": 0.52,
"grad_norm": 8.888174057006836,
"learning_rate": 6.298305084745763e-06,
"loss": 1.4874,
"step": 20710
},
{
"epoch": 0.52,
"grad_norm": 11.385110855102539,
"learning_rate": 6.291525423728814e-06,
"loss": 1.3519,
"step": 20720
},
{
"epoch": 0.52,
"grad_norm": 11.636566162109375,
"learning_rate": 6.284745762711865e-06,
"loss": 1.3237,
"step": 20730
},
{
"epoch": 0.52,
"grad_norm": 2.6630842685699463,
"learning_rate": 6.277966101694916e-06,
"loss": 1.5615,
"step": 20740
},
{
"epoch": 0.52,
"grad_norm": 15.339405059814453,
"learning_rate": 6.271186440677966e-06,
"loss": 1.0971,
"step": 20750
},
{
"epoch": 0.52,
"grad_norm": 5.97685432434082,
"learning_rate": 6.2644067796610176e-06,
"loss": 1.2582,
"step": 20760
},
{
"epoch": 0.52,
"grad_norm": 5.218130111694336,
"learning_rate": 6.257627118644068e-06,
"loss": 1.3272,
"step": 20770
},
{
"epoch": 0.52,
"grad_norm": 8.943811416625977,
"learning_rate": 6.2508474576271195e-06,
"loss": 1.4001,
"step": 20780
},
{
"epoch": 0.52,
"grad_norm": 3.288783550262451,
"learning_rate": 6.24406779661017e-06,
"loss": 1.3729,
"step": 20790
},
{
"epoch": 0.52,
"grad_norm": 2.7321832180023193,
"learning_rate": 6.2372881355932215e-06,
"loss": 1.4365,
"step": 20800
},
{
"epoch": 0.52,
"grad_norm": 7.656303405761719,
"learning_rate": 6.230508474576271e-06,
"loss": 1.3646,
"step": 20810
},
{
"epoch": 0.52,
"grad_norm": 3.2849278450012207,
"learning_rate": 6.223728813559322e-06,
"loss": 1.0755,
"step": 20820
},
{
"epoch": 0.52,
"grad_norm": 3.6145408153533936,
"learning_rate": 6.216949152542373e-06,
"loss": 1.2836,
"step": 20830
},
{
"epoch": 0.52,
"grad_norm": 6.668073654174805,
"learning_rate": 6.210169491525424e-06,
"loss": 1.3706,
"step": 20840
},
{
"epoch": 0.52,
"grad_norm": 10.408007621765137,
"learning_rate": 6.203389830508475e-06,
"loss": 1.5154,
"step": 20850
},
{
"epoch": 0.52,
"grad_norm": 2.708711862564087,
"learning_rate": 6.196610169491526e-06,
"loss": 1.3841,
"step": 20860
},
{
"epoch": 0.52,
"grad_norm": 1.7991631031036377,
"learning_rate": 6.189830508474577e-06,
"loss": 1.261,
"step": 20870
},
{
"epoch": 0.52,
"grad_norm": 6.492557048797607,
"learning_rate": 6.183050847457628e-06,
"loss": 1.469,
"step": 20880
},
{
"epoch": 0.52,
"grad_norm": 3.5670902729034424,
"learning_rate": 6.176271186440679e-06,
"loss": 1.4484,
"step": 20890
},
{
"epoch": 0.52,
"grad_norm": 3.4511330127716064,
"learning_rate": 6.1694915254237295e-06,
"loss": 1.3566,
"step": 20900
},
{
"epoch": 0.52,
"grad_norm": 7.958609580993652,
"learning_rate": 6.162711864406781e-06,
"loss": 1.3272,
"step": 20910
},
{
"epoch": 0.52,
"grad_norm": 6.594333648681641,
"learning_rate": 6.155932203389831e-06,
"loss": 1.3926,
"step": 20920
},
{
"epoch": 0.52,
"grad_norm": 7.329288005828857,
"learning_rate": 6.149152542372881e-06,
"loss": 1.4948,
"step": 20930
},
{
"epoch": 0.52,
"grad_norm": 10.120712280273438,
"learning_rate": 6.142372881355933e-06,
"loss": 1.5394,
"step": 20940
},
{
"epoch": 0.52,
"grad_norm": 4.3365983963012695,
"learning_rate": 6.135593220338983e-06,
"loss": 1.2156,
"step": 20950
},
{
"epoch": 0.52,
"grad_norm": 2.0522942543029785,
"learning_rate": 6.1288135593220346e-06,
"loss": 1.3206,
"step": 20960
},
{
"epoch": 0.52,
"grad_norm": 5.730597496032715,
"learning_rate": 6.122033898305085e-06,
"loss": 1.4396,
"step": 20970
},
{
"epoch": 0.52,
"grad_norm": 13.125938415527344,
"learning_rate": 6.1152542372881365e-06,
"loss": 1.1969,
"step": 20980
},
{
"epoch": 0.52,
"grad_norm": 3.8410260677337646,
"learning_rate": 6.108474576271187e-06,
"loss": 1.3842,
"step": 20990
},
{
"epoch": 0.53,
"grad_norm": 3.415696144104004,
"learning_rate": 6.1016949152542385e-06,
"loss": 1.3296,
"step": 21000
},
{
"epoch": 0.53,
"eval_loss": 1.344160556793213,
"eval_runtime": 66.1512,
"eval_samples_per_second": 15.117,
"eval_steps_per_second": 15.117,
"step": 21000
},
{
"epoch": 0.53,
"grad_norm": 9.95438003540039,
"learning_rate": 6.094915254237289e-06,
"loss": 1.3206,
"step": 21010
},
{
"epoch": 0.53,
"grad_norm": 10.410721778869629,
"learning_rate": 6.088135593220339e-06,
"loss": 1.4014,
"step": 21020
},
{
"epoch": 0.53,
"grad_norm": 9.159972190856934,
"learning_rate": 6.08135593220339e-06,
"loss": 1.3827,
"step": 21030
},
{
"epoch": 0.53,
"grad_norm": 4.856491565704346,
"learning_rate": 6.074576271186441e-06,
"loss": 1.4561,
"step": 21040
},
{
"epoch": 0.53,
"grad_norm": 5.863302707672119,
"learning_rate": 6.067796610169492e-06,
"loss": 1.3225,
"step": 21050
},
{
"epoch": 0.53,
"grad_norm": 2.968809127807617,
"learning_rate": 6.061016949152543e-06,
"loss": 1.3713,
"step": 21060
},
{
"epoch": 0.53,
"grad_norm": 4.19352388381958,
"learning_rate": 6.054237288135594e-06,
"loss": 1.3894,
"step": 21070
},
{
"epoch": 0.53,
"grad_norm": 4.841989040374756,
"learning_rate": 6.047457627118645e-06,
"loss": 1.3609,
"step": 21080
},
{
"epoch": 0.53,
"grad_norm": 8.693398475646973,
"learning_rate": 6.040677966101696e-06,
"loss": 1.4118,
"step": 21090
},
{
"epoch": 0.53,
"grad_norm": 10.902780532836914,
"learning_rate": 6.0338983050847465e-06,
"loss": 1.4529,
"step": 21100
},
{
"epoch": 0.53,
"grad_norm": 5.1729607582092285,
"learning_rate": 6.027118644067798e-06,
"loss": 1.3302,
"step": 21110
},
{
"epoch": 0.53,
"grad_norm": 1.9906094074249268,
"learning_rate": 6.020338983050848e-06,
"loss": 1.2294,
"step": 21120
},
{
"epoch": 0.53,
"grad_norm": 3.285928964614868,
"learning_rate": 6.013559322033898e-06,
"loss": 1.5006,
"step": 21130
},
{
"epoch": 0.53,
"grad_norm": 13.098603248596191,
"learning_rate": 6.00677966101695e-06,
"loss": 1.425,
"step": 21140
},
{
"epoch": 0.53,
"grad_norm": 4.391754150390625,
"learning_rate": 6e-06,
"loss": 1.2668,
"step": 21150
},
{
"epoch": 0.53,
"grad_norm": 8.409893035888672,
"learning_rate": 5.9932203389830516e-06,
"loss": 1.2061,
"step": 21160
},
{
"epoch": 0.53,
"grad_norm": 10.568397521972656,
"learning_rate": 5.986440677966102e-06,
"loss": 1.4191,
"step": 21170
},
{
"epoch": 0.53,
"grad_norm": 7.371358394622803,
"learning_rate": 5.9796610169491535e-06,
"loss": 1.2942,
"step": 21180
},
{
"epoch": 0.53,
"grad_norm": 2.6429450511932373,
"learning_rate": 5.972881355932204e-06,
"loss": 1.4216,
"step": 21190
},
{
"epoch": 0.53,
"grad_norm": 5.85234260559082,
"learning_rate": 5.9661016949152555e-06,
"loss": 1.3541,
"step": 21200
},
{
"epoch": 0.53,
"grad_norm": 3.2903425693511963,
"learning_rate": 5.959322033898306e-06,
"loss": 1.2969,
"step": 21210
},
{
"epoch": 0.53,
"grad_norm": 3.9652106761932373,
"learning_rate": 5.9525423728813566e-06,
"loss": 1.2853,
"step": 21220
},
{
"epoch": 0.53,
"grad_norm": 3.8703153133392334,
"learning_rate": 5.945762711864407e-06,
"loss": 1.3094,
"step": 21230
},
{
"epoch": 0.53,
"grad_norm": 8.018983840942383,
"learning_rate": 5.938983050847458e-06,
"loss": 1.4798,
"step": 21240
},
{
"epoch": 0.53,
"grad_norm": 2.752399206161499,
"learning_rate": 5.932203389830509e-06,
"loss": 1.4141,
"step": 21250
},
{
"epoch": 0.53,
"grad_norm": 5.574487686157227,
"learning_rate": 5.92542372881356e-06,
"loss": 1.2121,
"step": 21260
},
{
"epoch": 0.53,
"grad_norm": 10.352456092834473,
"learning_rate": 5.91864406779661e-06,
"loss": 1.3264,
"step": 21270
},
{
"epoch": 0.53,
"grad_norm": 4.182732582092285,
"learning_rate": 5.911864406779662e-06,
"loss": 1.4334,
"step": 21280
},
{
"epoch": 0.53,
"grad_norm": 5.310421943664551,
"learning_rate": 5.905084745762712e-06,
"loss": 1.5641,
"step": 21290
},
{
"epoch": 0.53,
"grad_norm": 5.489622592926025,
"learning_rate": 5.8983050847457635e-06,
"loss": 1.4699,
"step": 21300
},
{
"epoch": 0.53,
"grad_norm": 6.644534111022949,
"learning_rate": 5.891525423728814e-06,
"loss": 1.3273,
"step": 21310
},
{
"epoch": 0.53,
"grad_norm": 13.480459213256836,
"learning_rate": 5.8847457627118655e-06,
"loss": 1.2957,
"step": 21320
},
{
"epoch": 0.53,
"grad_norm": 9.34183120727539,
"learning_rate": 5.877966101694915e-06,
"loss": 1.4145,
"step": 21330
},
{
"epoch": 0.53,
"grad_norm": 4.792973518371582,
"learning_rate": 5.871186440677966e-06,
"loss": 1.4912,
"step": 21340
},
{
"epoch": 0.53,
"grad_norm": 4.098564147949219,
"learning_rate": 5.864406779661017e-06,
"loss": 1.1895,
"step": 21350
},
{
"epoch": 0.53,
"grad_norm": 7.23917293548584,
"learning_rate": 5.857627118644068e-06,
"loss": 1.3875,
"step": 21360
},
{
"epoch": 0.53,
"grad_norm": 4.86613130569458,
"learning_rate": 5.850847457627119e-06,
"loss": 1.4827,
"step": 21370
},
{
"epoch": 0.53,
"grad_norm": 7.629755973815918,
"learning_rate": 5.84406779661017e-06,
"loss": 1.2189,
"step": 21380
},
{
"epoch": 0.53,
"grad_norm": 3.80531907081604,
"learning_rate": 5.837288135593221e-06,
"loss": 1.4064,
"step": 21390
},
{
"epoch": 0.54,
"grad_norm": 3.432089328765869,
"learning_rate": 5.830508474576272e-06,
"loss": 1.1929,
"step": 21400
},
{
"epoch": 0.54,
"grad_norm": 9.766077995300293,
"learning_rate": 5.823728813559323e-06,
"loss": 1.4525,
"step": 21410
},
{
"epoch": 0.54,
"grad_norm": 4.745760440826416,
"learning_rate": 5.8169491525423736e-06,
"loss": 1.4924,
"step": 21420
},
{
"epoch": 0.54,
"grad_norm": 5.188168525695801,
"learning_rate": 5.810169491525425e-06,
"loss": 1.27,
"step": 21430
},
{
"epoch": 0.54,
"grad_norm": 4.576213359832764,
"learning_rate": 5.803389830508475e-06,
"loss": 1.3165,
"step": 21440
},
{
"epoch": 0.54,
"grad_norm": 10.540860176086426,
"learning_rate": 5.796610169491525e-06,
"loss": 1.2652,
"step": 21450
},
{
"epoch": 0.54,
"grad_norm": 4.210390567779541,
"learning_rate": 5.789830508474577e-06,
"loss": 1.4372,
"step": 21460
},
{
"epoch": 0.54,
"grad_norm": 8.733638763427734,
"learning_rate": 5.783050847457627e-06,
"loss": 1.2691,
"step": 21470
},
{
"epoch": 0.54,
"grad_norm": 4.997326374053955,
"learning_rate": 5.776271186440679e-06,
"loss": 1.1454,
"step": 21480
},
{
"epoch": 0.54,
"grad_norm": 10.108692169189453,
"learning_rate": 5.769491525423729e-06,
"loss": 1.2982,
"step": 21490
},
{
"epoch": 0.54,
"grad_norm": 13.393025398254395,
"learning_rate": 5.7627118644067805e-06,
"loss": 1.4995,
"step": 21500
},
{
"epoch": 0.54,
"eval_loss": 1.3565526008605957,
"eval_runtime": 66.1537,
"eval_samples_per_second": 15.116,
"eval_steps_per_second": 15.116,
"step": 21500
},
{
"epoch": 0.54,
"grad_norm": 5.675364017486572,
"learning_rate": 5.755932203389831e-06,
"loss": 1.2256,
"step": 21510
},
{
"epoch": 0.54,
"grad_norm": 2.069751262664795,
"learning_rate": 5.7491525423728825e-06,
"loss": 1.2355,
"step": 21520
},
{
"epoch": 0.54,
"grad_norm": 4.536093711853027,
"learning_rate": 5.742372881355933e-06,
"loss": 1.1957,
"step": 21530
},
{
"epoch": 0.54,
"grad_norm": 2.5274765491485596,
"learning_rate": 5.735593220338983e-06,
"loss": 1.1843,
"step": 21540
},
{
"epoch": 0.54,
"grad_norm": 4.418458461761475,
"learning_rate": 5.728813559322034e-06,
"loss": 1.1994,
"step": 21550
},
{
"epoch": 0.54,
"grad_norm": 13.488496780395508,
"learning_rate": 5.722033898305085e-06,
"loss": 1.4173,
"step": 21560
},
{
"epoch": 0.54,
"grad_norm": 5.223592758178711,
"learning_rate": 5.715254237288136e-06,
"loss": 1.3032,
"step": 21570
},
{
"epoch": 0.54,
"grad_norm": 5.894464492797852,
"learning_rate": 5.708474576271187e-06,
"loss": 1.3027,
"step": 21580
},
{
"epoch": 0.54,
"grad_norm": 6.945793151855469,
"learning_rate": 5.701694915254238e-06,
"loss": 1.3102,
"step": 21590
},
{
"epoch": 0.54,
"grad_norm": 7.386875629425049,
"learning_rate": 5.694915254237289e-06,
"loss": 1.3643,
"step": 21600
},
{
"epoch": 0.54,
"grad_norm": 3.5999162197113037,
"learning_rate": 5.68813559322034e-06,
"loss": 1.2456,
"step": 21610
},
{
"epoch": 0.54,
"grad_norm": 5.8258490562438965,
"learning_rate": 5.6813559322033906e-06,
"loss": 1.1922,
"step": 21620
},
{
"epoch": 0.54,
"grad_norm": 10.920169830322266,
"learning_rate": 5.674576271186442e-06,
"loss": 1.5312,
"step": 21630
},
{
"epoch": 0.54,
"grad_norm": 3.898834705352783,
"learning_rate": 5.667796610169492e-06,
"loss": 1.5012,
"step": 21640
},
{
"epoch": 0.54,
"grad_norm": 6.2130866050720215,
"learning_rate": 5.661016949152542e-06,
"loss": 1.543,
"step": 21650
},
{
"epoch": 0.54,
"grad_norm": 3.604144811630249,
"learning_rate": 5.654237288135594e-06,
"loss": 1.4586,
"step": 21660
},
{
"epoch": 0.54,
"grad_norm": 4.859696388244629,
"learning_rate": 5.647457627118644e-06,
"loss": 1.3056,
"step": 21670
},
{
"epoch": 0.54,
"grad_norm": 2.9596614837646484,
"learning_rate": 5.640677966101696e-06,
"loss": 1.3888,
"step": 21680
},
{
"epoch": 0.54,
"grad_norm": 5.489665985107422,
"learning_rate": 5.633898305084746e-06,
"loss": 1.2491,
"step": 21690
},
{
"epoch": 0.54,
"grad_norm": 5.618114948272705,
"learning_rate": 5.6271186440677975e-06,
"loss": 1.2746,
"step": 21700
},
{
"epoch": 0.54,
"grad_norm": 2.145024061203003,
"learning_rate": 5.620338983050848e-06,
"loss": 1.4036,
"step": 21710
},
{
"epoch": 0.54,
"grad_norm": 2.4870400428771973,
"learning_rate": 5.6135593220338995e-06,
"loss": 1.4355,
"step": 21720
},
{
"epoch": 0.54,
"grad_norm": 2.576144218444824,
"learning_rate": 5.60677966101695e-06,
"loss": 1.2988,
"step": 21730
},
{
"epoch": 0.54,
"grad_norm": 5.971595764160156,
"learning_rate": 5.600000000000001e-06,
"loss": 1.3201,
"step": 21740
},
{
"epoch": 0.54,
"grad_norm": 7.581085205078125,
"learning_rate": 5.593220338983051e-06,
"loss": 1.3358,
"step": 21750
},
{
"epoch": 0.54,
"grad_norm": 4.148537635803223,
"learning_rate": 5.586440677966102e-06,
"loss": 1.2464,
"step": 21760
},
{
"epoch": 0.54,
"grad_norm": 6.613537788391113,
"learning_rate": 5.579661016949153e-06,
"loss": 1.3156,
"step": 21770
},
{
"epoch": 0.54,
"grad_norm": 10.526129722595215,
"learning_rate": 5.572881355932204e-06,
"loss": 1.2483,
"step": 21780
},
{
"epoch": 0.54,
"grad_norm": 7.221047401428223,
"learning_rate": 5.566101694915255e-06,
"loss": 1.5878,
"step": 21790
},
{
"epoch": 0.55,
"grad_norm": 6.365529537200928,
"learning_rate": 5.559322033898306e-06,
"loss": 1.4117,
"step": 21800
},
{
"epoch": 0.55,
"grad_norm": 3.8305916786193848,
"learning_rate": 5.552542372881356e-06,
"loss": 1.1894,
"step": 21810
},
{
"epoch": 0.55,
"grad_norm": 3.672477960586548,
"learning_rate": 5.5457627118644076e-06,
"loss": 1.2199,
"step": 21820
},
{
"epoch": 0.55,
"grad_norm": 2.586512565612793,
"learning_rate": 5.538983050847458e-06,
"loss": 1.392,
"step": 21830
},
{
"epoch": 0.55,
"grad_norm": 4.2184624671936035,
"learning_rate": 5.5322033898305095e-06,
"loss": 1.3008,
"step": 21840
},
{
"epoch": 0.55,
"grad_norm": 7.834671974182129,
"learning_rate": 5.525423728813559e-06,
"loss": 1.1794,
"step": 21850
},
{
"epoch": 0.55,
"grad_norm": 3.5877692699432373,
"learning_rate": 5.518644067796611e-06,
"loss": 1.4339,
"step": 21860
},
{
"epoch": 0.55,
"grad_norm": 3.0174179077148438,
"learning_rate": 5.511864406779661e-06,
"loss": 1.2704,
"step": 21870
},
{
"epoch": 0.55,
"grad_norm": 2.9889588356018066,
"learning_rate": 5.505084745762712e-06,
"loss": 1.3588,
"step": 21880
},
{
"epoch": 0.55,
"grad_norm": 10.810959815979004,
"learning_rate": 5.498305084745763e-06,
"loss": 1.3539,
"step": 21890
},
{
"epoch": 0.55,
"grad_norm": 5.771850109100342,
"learning_rate": 5.491525423728814e-06,
"loss": 1.4278,
"step": 21900
},
{
"epoch": 0.55,
"grad_norm": 4.13969612121582,
"learning_rate": 5.484745762711865e-06,
"loss": 1.2798,
"step": 21910
},
{
"epoch": 0.55,
"grad_norm": 15.295929908752441,
"learning_rate": 5.477966101694916e-06,
"loss": 1.5213,
"step": 21920
},
{
"epoch": 0.55,
"grad_norm": 6.445948600769043,
"learning_rate": 5.471186440677967e-06,
"loss": 1.3498,
"step": 21930
},
{
"epoch": 0.55,
"grad_norm": 9.28097152709961,
"learning_rate": 5.464406779661018e-06,
"loss": 1.3046,
"step": 21940
},
{
"epoch": 0.55,
"grad_norm": 6.094447135925293,
"learning_rate": 5.457627118644067e-06,
"loss": 1.4336,
"step": 21950
},
{
"epoch": 0.55,
"grad_norm": 9.818504333496094,
"learning_rate": 5.450847457627119e-06,
"loss": 1.516,
"step": 21960
},
{
"epoch": 0.55,
"grad_norm": 11.956009864807129,
"learning_rate": 5.444067796610169e-06,
"loss": 1.3435,
"step": 21970
},
{
"epoch": 0.55,
"grad_norm": 7.544681072235107,
"learning_rate": 5.437288135593221e-06,
"loss": 1.4625,
"step": 21980
},
{
"epoch": 0.55,
"grad_norm": 3.394897222518921,
"learning_rate": 5.430508474576271e-06,
"loss": 1.4565,
"step": 21990
},
{
"epoch": 0.55,
"grad_norm": 5.722468852996826,
"learning_rate": 5.423728813559323e-06,
"loss": 1.2133,
"step": 22000
},
{
"epoch": 0.55,
"eval_loss": 1.3403723239898682,
"eval_runtime": 66.1471,
"eval_samples_per_second": 15.118,
"eval_steps_per_second": 15.118,
"step": 22000
},
{
"epoch": 0.55,
"grad_norm": 7.046730995178223,
"learning_rate": 5.416949152542373e-06,
"loss": 1.1353,
"step": 22010
},
{
"epoch": 0.55,
"grad_norm": 7.013365268707275,
"learning_rate": 5.4101694915254246e-06,
"loss": 1.4333,
"step": 22020
},
{
"epoch": 0.55,
"grad_norm": 3.9469892978668213,
"learning_rate": 5.403389830508475e-06,
"loss": 1.3601,
"step": 22030
},
{
"epoch": 0.55,
"grad_norm": 3.740983486175537,
"learning_rate": 5.3966101694915265e-06,
"loss": 1.0232,
"step": 22040
},
{
"epoch": 0.55,
"grad_norm": 4.604060649871826,
"learning_rate": 5.389830508474577e-06,
"loss": 1.4011,
"step": 22050
},
{
"epoch": 0.55,
"grad_norm": 0.5601249933242798,
"learning_rate": 5.383050847457627e-06,
"loss": 1.2068,
"step": 22060
},
{
"epoch": 0.55,
"grad_norm": 2.7781484127044678,
"learning_rate": 5.376271186440678e-06,
"loss": 1.3976,
"step": 22070
},
{
"epoch": 0.55,
"grad_norm": 10.417901992797852,
"learning_rate": 5.369491525423729e-06,
"loss": 1.2444,
"step": 22080
},
{
"epoch": 0.55,
"grad_norm": 8.263280868530273,
"learning_rate": 5.36271186440678e-06,
"loss": 1.0937,
"step": 22090
},
{
"epoch": 0.55,
"grad_norm": 6.128343105316162,
"learning_rate": 5.355932203389831e-06,
"loss": 1.2891,
"step": 22100
},
{
"epoch": 0.55,
"grad_norm": 4.099800109863281,
"learning_rate": 5.349152542372882e-06,
"loss": 1.117,
"step": 22110
},
{
"epoch": 0.55,
"grad_norm": 7.756937026977539,
"learning_rate": 5.342372881355933e-06,
"loss": 1.3112,
"step": 22120
},
{
"epoch": 0.55,
"grad_norm": 5.372160911560059,
"learning_rate": 5.335593220338984e-06,
"loss": 1.5276,
"step": 22130
},
{
"epoch": 0.55,
"grad_norm": 5.017634391784668,
"learning_rate": 5.328813559322035e-06,
"loss": 1.2825,
"step": 22140
},
{
"epoch": 0.55,
"grad_norm": 5.629271030426025,
"learning_rate": 5.322033898305086e-06,
"loss": 1.2428,
"step": 22150
},
{
"epoch": 0.55,
"grad_norm": 6.938544273376465,
"learning_rate": 5.315254237288136e-06,
"loss": 1.4563,
"step": 22160
},
{
"epoch": 0.55,
"grad_norm": 13.804441452026367,
"learning_rate": 5.308474576271186e-06,
"loss": 1.332,
"step": 22170
},
{
"epoch": 0.55,
"grad_norm": 10.347596168518066,
"learning_rate": 5.301694915254238e-06,
"loss": 1.3879,
"step": 22180
},
{
"epoch": 0.55,
"grad_norm": 2.613632917404175,
"learning_rate": 5.294915254237288e-06,
"loss": 1.3298,
"step": 22190
},
{
"epoch": 0.56,
"grad_norm": 14.637787818908691,
"learning_rate": 5.28813559322034e-06,
"loss": 1.3301,
"step": 22200
},
{
"epoch": 0.56,
"grad_norm": 2.5796003341674805,
"learning_rate": 5.28135593220339e-06,
"loss": 1.41,
"step": 22210
},
{
"epoch": 0.56,
"grad_norm": 5.326439380645752,
"learning_rate": 5.2745762711864416e-06,
"loss": 1.2483,
"step": 22220
},
{
"epoch": 0.56,
"grad_norm": 8.928110122680664,
"learning_rate": 5.267796610169492e-06,
"loss": 1.2566,
"step": 22230
},
{
"epoch": 0.56,
"grad_norm": 9.285192489624023,
"learning_rate": 5.2610169491525435e-06,
"loss": 1.2982,
"step": 22240
},
{
"epoch": 0.56,
"grad_norm": 2.101649045944214,
"learning_rate": 5.254237288135594e-06,
"loss": 1.3181,
"step": 22250
},
{
"epoch": 0.56,
"grad_norm": 5.57994270324707,
"learning_rate": 5.247457627118645e-06,
"loss": 1.3745,
"step": 22260
},
{
"epoch": 0.56,
"grad_norm": 12.0460844039917,
"learning_rate": 5.240677966101695e-06,
"loss": 1.202,
"step": 22270
},
{
"epoch": 0.56,
"grad_norm": 5.367649078369141,
"learning_rate": 5.233898305084746e-06,
"loss": 1.2969,
"step": 22280
},
{
"epoch": 0.56,
"grad_norm": 4.784932613372803,
"learning_rate": 5.227118644067797e-06,
"loss": 1.363,
"step": 22290
},
{
"epoch": 0.56,
"grad_norm": 8.103744506835938,
"learning_rate": 5.220338983050848e-06,
"loss": 1.4144,
"step": 22300
},
{
"epoch": 0.56,
"grad_norm": 5.561450958251953,
"learning_rate": 5.213559322033899e-06,
"loss": 1.3018,
"step": 22310
},
{
"epoch": 0.56,
"grad_norm": 8.294450759887695,
"learning_rate": 5.20677966101695e-06,
"loss": 1.3509,
"step": 22320
},
{
"epoch": 0.56,
"grad_norm": 8.498016357421875,
"learning_rate": 5.2e-06,
"loss": 1.4932,
"step": 22330
},
{
"epoch": 0.56,
"grad_norm": 3.3640925884246826,
"learning_rate": 5.193220338983052e-06,
"loss": 1.4413,
"step": 22340
},
{
"epoch": 0.56,
"grad_norm": 3.236924886703491,
"learning_rate": 5.186440677966102e-06,
"loss": 1.2535,
"step": 22350
},
{
"epoch": 0.56,
"grad_norm": 6.234226703643799,
"learning_rate": 5.1796610169491535e-06,
"loss": 1.1816,
"step": 22360
},
{
"epoch": 0.56,
"grad_norm": 6.9503703117370605,
"learning_rate": 5.172881355932203e-06,
"loss": 1.2022,
"step": 22370
},
{
"epoch": 0.56,
"grad_norm": 11.745012283325195,
"learning_rate": 5.166101694915255e-06,
"loss": 1.543,
"step": 22380
},
{
"epoch": 0.56,
"grad_norm": 8.977341651916504,
"learning_rate": 5.159322033898305e-06,
"loss": 1.4185,
"step": 22390
},
{
"epoch": 0.56,
"grad_norm": 2.515448570251465,
"learning_rate": 5.152542372881356e-06,
"loss": 1.2665,
"step": 22400
},
{
"epoch": 0.56,
"grad_norm": 9.47354507446289,
"learning_rate": 5.145762711864407e-06,
"loss": 1.3086,
"step": 22410
},
{
"epoch": 0.56,
"grad_norm": 6.550108432769775,
"learning_rate": 5.138983050847458e-06,
"loss": 1.3129,
"step": 22420
},
{
"epoch": 0.56,
"grad_norm": 3.860886335372925,
"learning_rate": 5.132203389830509e-06,
"loss": 1.4046,
"step": 22430
},
{
"epoch": 0.56,
"grad_norm": 8.062068939208984,
"learning_rate": 5.12542372881356e-06,
"loss": 1.207,
"step": 22440
},
{
"epoch": 0.56,
"grad_norm": 5.823556423187256,
"learning_rate": 5.118644067796611e-06,
"loss": 1.2547,
"step": 22450
},
{
"epoch": 0.56,
"grad_norm": 7.168029308319092,
"learning_rate": 5.111864406779662e-06,
"loss": 1.2772,
"step": 22460
},
{
"epoch": 0.56,
"grad_norm": 9.224081039428711,
"learning_rate": 5.105084745762711e-06,
"loss": 1.3066,
"step": 22470
},
{
"epoch": 0.56,
"grad_norm": 3.878537178039551,
"learning_rate": 5.098305084745763e-06,
"loss": 1.2413,
"step": 22480
},
{
"epoch": 0.56,
"grad_norm": 5.821982383728027,
"learning_rate": 5.091525423728813e-06,
"loss": 1.3609,
"step": 22490
},
{
"epoch": 0.56,
"grad_norm": 9.057456970214844,
"learning_rate": 5.084745762711865e-06,
"loss": 1.2255,
"step": 22500
},
{
"epoch": 0.56,
"eval_loss": 1.3439137935638428,
"eval_runtime": 66.2088,
"eval_samples_per_second": 15.104,
"eval_steps_per_second": 15.104,
"step": 22500
},
{
"epoch": 0.56,
"grad_norm": 3.232191562652588,
"learning_rate": 5.077966101694915e-06,
"loss": 1.3644,
"step": 22510
},
{
"epoch": 0.56,
"grad_norm": 5.3384528160095215,
"learning_rate": 5.071186440677967e-06,
"loss": 1.3093,
"step": 22520
},
{
"epoch": 0.56,
"grad_norm": 6.911621570587158,
"learning_rate": 5.064406779661017e-06,
"loss": 1.2571,
"step": 22530
},
{
"epoch": 0.56,
"grad_norm": 6.84785270690918,
"learning_rate": 5.057627118644069e-06,
"loss": 1.324,
"step": 22540
},
{
"epoch": 0.56,
"grad_norm": 4.634193420410156,
"learning_rate": 5.050847457627119e-06,
"loss": 1.4672,
"step": 22550
},
{
"epoch": 0.56,
"grad_norm": 3.4189460277557373,
"learning_rate": 5.0440677966101705e-06,
"loss": 1.325,
"step": 22560
},
{
"epoch": 0.56,
"grad_norm": 7.2140793800354,
"learning_rate": 5.037288135593221e-06,
"loss": 1.3923,
"step": 22570
},
{
"epoch": 0.56,
"grad_norm": 3.916853189468384,
"learning_rate": 5.030508474576271e-06,
"loss": 1.3431,
"step": 22580
},
{
"epoch": 0.56,
"grad_norm": 6.434349536895752,
"learning_rate": 5.023728813559322e-06,
"loss": 1.4615,
"step": 22590
},
{
"epoch": 0.56,
"grad_norm": 6.851098537445068,
"learning_rate": 5.016949152542373e-06,
"loss": 1.3013,
"step": 22600
},
{
"epoch": 0.57,
"grad_norm": 7.541562080383301,
"learning_rate": 5.010169491525424e-06,
"loss": 1.3696,
"step": 22610
},
{
"epoch": 0.57,
"grad_norm": 4.100308895111084,
"learning_rate": 5.003389830508475e-06,
"loss": 1.3802,
"step": 22620
},
{
"epoch": 0.57,
"grad_norm": 12.694953918457031,
"learning_rate": 4.996610169491526e-06,
"loss": 1.2328,
"step": 22630
},
{
"epoch": 0.57,
"grad_norm": 12.468879699707031,
"learning_rate": 4.989830508474577e-06,
"loss": 1.3995,
"step": 22640
},
{
"epoch": 0.57,
"grad_norm": 3.142810821533203,
"learning_rate": 4.983050847457628e-06,
"loss": 1.2403,
"step": 22650
},
{
"epoch": 0.57,
"grad_norm": 7.71440315246582,
"learning_rate": 4.976271186440678e-06,
"loss": 1.2902,
"step": 22660
},
{
"epoch": 0.57,
"grad_norm": 4.5355634689331055,
"learning_rate": 4.969491525423729e-06,
"loss": 1.3243,
"step": 22670
},
{
"epoch": 0.57,
"grad_norm": 1.5578103065490723,
"learning_rate": 4.96271186440678e-06,
"loss": 1.2796,
"step": 22680
},
{
"epoch": 0.57,
"grad_norm": 4.997007846832275,
"learning_rate": 4.955932203389831e-06,
"loss": 1.1288,
"step": 22690
},
{
"epoch": 0.57,
"grad_norm": 4.240728855133057,
"learning_rate": 4.949152542372882e-06,
"loss": 1.408,
"step": 22700
},
{
"epoch": 0.57,
"grad_norm": 7.517406940460205,
"learning_rate": 4.942372881355932e-06,
"loss": 1.3879,
"step": 22710
},
{
"epoch": 0.57,
"grad_norm": 1.1581978797912598,
"learning_rate": 4.935593220338984e-06,
"loss": 1.3547,
"step": 22720
},
{
"epoch": 0.57,
"grad_norm": 5.073269367218018,
"learning_rate": 4.928813559322034e-06,
"loss": 1.3664,
"step": 22730
},
{
"epoch": 0.57,
"grad_norm": 3.7317769527435303,
"learning_rate": 4.922033898305086e-06,
"loss": 1.3942,
"step": 22740
},
{
"epoch": 0.57,
"grad_norm": 4.843672275543213,
"learning_rate": 4.915254237288136e-06,
"loss": 1.3723,
"step": 22750
},
{
"epoch": 0.57,
"grad_norm": 7.751224040985107,
"learning_rate": 4.908474576271187e-06,
"loss": 1.2057,
"step": 22760
},
{
"epoch": 0.57,
"grad_norm": 2.4519495964050293,
"learning_rate": 4.901694915254237e-06,
"loss": 1.3836,
"step": 22770
},
{
"epoch": 0.57,
"grad_norm": 8.7233304977417,
"learning_rate": 4.894915254237289e-06,
"loss": 1.3842,
"step": 22780
},
{
"epoch": 0.57,
"grad_norm": 2.717367172241211,
"learning_rate": 4.888135593220339e-06,
"loss": 1.4519,
"step": 22790
},
{
"epoch": 0.57,
"grad_norm": 4.797736167907715,
"learning_rate": 4.881355932203391e-06,
"loss": 1.1785,
"step": 22800
},
{
"epoch": 0.57,
"grad_norm": 11.28987979888916,
"learning_rate": 4.874576271186441e-06,
"loss": 1.3058,
"step": 22810
},
{
"epoch": 0.57,
"grad_norm": 5.097863674163818,
"learning_rate": 4.867796610169492e-06,
"loss": 1.232,
"step": 22820
},
{
"epoch": 0.57,
"grad_norm": 5.479716777801514,
"learning_rate": 4.861016949152543e-06,
"loss": 1.4677,
"step": 22830
},
{
"epoch": 0.57,
"grad_norm": 11.921456336975098,
"learning_rate": 4.854237288135594e-06,
"loss": 1.5442,
"step": 22840
},
{
"epoch": 0.57,
"grad_norm": 9.934676170349121,
"learning_rate": 4.847457627118645e-06,
"loss": 1.4147,
"step": 22850
},
{
"epoch": 0.57,
"grad_norm": 3.3140487670898438,
"learning_rate": 4.840677966101695e-06,
"loss": 1.2146,
"step": 22860
},
{
"epoch": 0.57,
"grad_norm": 2.1303317546844482,
"learning_rate": 4.833898305084746e-06,
"loss": 1.3021,
"step": 22870
},
{
"epoch": 0.57,
"grad_norm": 3.943474769592285,
"learning_rate": 4.827118644067797e-06,
"loss": 1.2028,
"step": 22880
},
{
"epoch": 0.57,
"grad_norm": 5.444009780883789,
"learning_rate": 4.820338983050848e-06,
"loss": 1.3873,
"step": 22890
},
{
"epoch": 0.57,
"grad_norm": 5.230558395385742,
"learning_rate": 4.813559322033899e-06,
"loss": 1.2907,
"step": 22900
},
{
"epoch": 0.57,
"grad_norm": 2.652158498764038,
"learning_rate": 4.80677966101695e-06,
"loss": 1.27,
"step": 22910
},
{
"epoch": 0.57,
"grad_norm": 5.423113822937012,
"learning_rate": 4.800000000000001e-06,
"loss": 1.3488,
"step": 22920
},
{
"epoch": 0.57,
"grad_norm": 6.731577396392822,
"learning_rate": 4.793220338983051e-06,
"loss": 1.3411,
"step": 22930
},
{
"epoch": 0.57,
"grad_norm": 6.142993927001953,
"learning_rate": 4.786440677966102e-06,
"loss": 1.4032,
"step": 22940
},
{
"epoch": 0.57,
"grad_norm": 7.845600605010986,
"learning_rate": 4.779661016949153e-06,
"loss": 1.3567,
"step": 22950
},
{
"epoch": 0.57,
"grad_norm": 3.123938798904419,
"learning_rate": 4.772881355932204e-06,
"loss": 1.2106,
"step": 22960
},
{
"epoch": 0.57,
"grad_norm": 4.684544086456299,
"learning_rate": 4.766101694915254e-06,
"loss": 1.3407,
"step": 22970
},
{
"epoch": 0.57,
"grad_norm": 11.232462882995605,
"learning_rate": 4.759322033898306e-06,
"loss": 1.136,
"step": 22980
},
{
"epoch": 0.57,
"grad_norm": 3.3728113174438477,
"learning_rate": 4.752542372881356e-06,
"loss": 1.199,
"step": 22990
},
{
"epoch": 0.57,
"grad_norm": 3.5851376056671143,
"learning_rate": 4.745762711864408e-06,
"loss": 1.463,
"step": 23000
},
{
"epoch": 0.57,
"eval_loss": 1.3161565065383911,
"eval_runtime": 66.103,
"eval_samples_per_second": 15.128,
"eval_steps_per_second": 15.128,
"step": 23000
},
{
"epoch": 0.58,
"grad_norm": 5.720829010009766,
"learning_rate": 4.738983050847458e-06,
"loss": 1.4393,
"step": 23010
},
{
"epoch": 0.58,
"grad_norm": 11.663517951965332,
"learning_rate": 4.732203389830509e-06,
"loss": 1.3158,
"step": 23020
},
{
"epoch": 0.58,
"grad_norm": 10.158949851989746,
"learning_rate": 4.725423728813559e-06,
"loss": 1.4086,
"step": 23030
},
{
"epoch": 0.58,
"grad_norm": 6.925542831420898,
"learning_rate": 4.718644067796611e-06,
"loss": 1.3655,
"step": 23040
},
{
"epoch": 0.58,
"grad_norm": 1.4193698167800903,
"learning_rate": 4.711864406779661e-06,
"loss": 1.2348,
"step": 23050
},
{
"epoch": 0.58,
"grad_norm": 6.884500980377197,
"learning_rate": 4.705084745762713e-06,
"loss": 1.4053,
"step": 23060
},
{
"epoch": 0.58,
"grad_norm": 4.412232875823975,
"learning_rate": 4.698305084745763e-06,
"loss": 1.2476,
"step": 23070
},
{
"epoch": 0.58,
"grad_norm": 2.403428077697754,
"learning_rate": 4.691525423728814e-06,
"loss": 1.4168,
"step": 23080
},
{
"epoch": 0.58,
"grad_norm": 7.022388458251953,
"learning_rate": 4.684745762711865e-06,
"loss": 1.359,
"step": 23090
},
{
"epoch": 0.58,
"grad_norm": 5.728121280670166,
"learning_rate": 4.677966101694916e-06,
"loss": 1.3916,
"step": 23100
},
{
"epoch": 0.58,
"grad_norm": 14.995932579040527,
"learning_rate": 4.671186440677967e-06,
"loss": 1.4541,
"step": 23110
},
{
"epoch": 0.58,
"grad_norm": 12.448729515075684,
"learning_rate": 4.664406779661017e-06,
"loss": 1.2989,
"step": 23120
},
{
"epoch": 0.58,
"grad_norm": 2.5807507038116455,
"learning_rate": 4.657627118644068e-06,
"loss": 1.4881,
"step": 23130
},
{
"epoch": 0.58,
"grad_norm": 4.670041561126709,
"learning_rate": 4.650847457627119e-06,
"loss": 1.4595,
"step": 23140
},
{
"epoch": 0.58,
"grad_norm": 3.8450145721435547,
"learning_rate": 4.64406779661017e-06,
"loss": 1.245,
"step": 23150
},
{
"epoch": 0.58,
"grad_norm": 15.546969413757324,
"learning_rate": 4.637288135593221e-06,
"loss": 1.2199,
"step": 23160
},
{
"epoch": 0.58,
"grad_norm": 8.563859939575195,
"learning_rate": 4.630508474576272e-06,
"loss": 1.3752,
"step": 23170
},
{
"epoch": 0.58,
"grad_norm": 8.742653846740723,
"learning_rate": 4.623728813559323e-06,
"loss": 1.2802,
"step": 23180
},
{
"epoch": 0.58,
"grad_norm": 6.253279685974121,
"learning_rate": 4.616949152542373e-06,
"loss": 1.4559,
"step": 23190
},
{
"epoch": 0.58,
"grad_norm": 5.3615827560424805,
"learning_rate": 4.610169491525424e-06,
"loss": 1.1649,
"step": 23200
},
{
"epoch": 0.58,
"grad_norm": 9.165109634399414,
"learning_rate": 4.603389830508475e-06,
"loss": 1.4435,
"step": 23210
},
{
"epoch": 0.58,
"grad_norm": 6.625391483306885,
"learning_rate": 4.596610169491526e-06,
"loss": 1.2911,
"step": 23220
},
{
"epoch": 0.58,
"grad_norm": 3.898466110229492,
"learning_rate": 4.589830508474576e-06,
"loss": 1.2653,
"step": 23230
},
{
"epoch": 0.58,
"grad_norm": 1.5483791828155518,
"learning_rate": 4.583050847457628e-06,
"loss": 1.1665,
"step": 23240
},
{
"epoch": 0.58,
"grad_norm": 4.248947620391846,
"learning_rate": 4.576271186440678e-06,
"loss": 1.2275,
"step": 23250
},
{
"epoch": 0.58,
"grad_norm": 3.522451639175415,
"learning_rate": 4.56949152542373e-06,
"loss": 1.3427,
"step": 23260
},
{
"epoch": 0.58,
"grad_norm": 5.79518461227417,
"learning_rate": 4.56271186440678e-06,
"loss": 1.2545,
"step": 23270
},
{
"epoch": 0.58,
"grad_norm": 7.211407661437988,
"learning_rate": 4.555932203389831e-06,
"loss": 1.3256,
"step": 23280
},
{
"epoch": 0.58,
"grad_norm": 2.218186855316162,
"learning_rate": 4.549152542372881e-06,
"loss": 1.327,
"step": 23290
},
{
"epoch": 0.58,
"grad_norm": 3.0725247859954834,
"learning_rate": 4.542372881355933e-06,
"loss": 1.1152,
"step": 23300
},
{
"epoch": 0.58,
"grad_norm": 10.184927940368652,
"learning_rate": 4.535593220338983e-06,
"loss": 1.3153,
"step": 23310
},
{
"epoch": 0.58,
"grad_norm": 6.575405597686768,
"learning_rate": 4.528813559322035e-06,
"loss": 1.2879,
"step": 23320
},
{
"epoch": 0.58,
"grad_norm": 5.2145094871521,
"learning_rate": 4.522033898305085e-06,
"loss": 1.2836,
"step": 23330
},
{
"epoch": 0.58,
"grad_norm": 16.586687088012695,
"learning_rate": 4.515254237288136e-06,
"loss": 1.245,
"step": 23340
},
{
"epoch": 0.58,
"grad_norm": 5.264496803283691,
"learning_rate": 4.508474576271187e-06,
"loss": 1.1731,
"step": 23350
},
{
"epoch": 0.58,
"grad_norm": 5.267670631408691,
"learning_rate": 4.501694915254238e-06,
"loss": 1.3662,
"step": 23360
},
{
"epoch": 0.58,
"grad_norm": 7.335075855255127,
"learning_rate": 4.494915254237289e-06,
"loss": 1.3261,
"step": 23370
},
{
"epoch": 0.58,
"grad_norm": 7.068376064300537,
"learning_rate": 4.488135593220339e-06,
"loss": 1.2805,
"step": 23380
},
{
"epoch": 0.58,
"grad_norm": 15.14303207397461,
"learning_rate": 4.48135593220339e-06,
"loss": 1.3605,
"step": 23390
},
{
"epoch": 0.58,
"grad_norm": 12.552229881286621,
"learning_rate": 4.474576271186441e-06,
"loss": 1.4804,
"step": 23400
},
{
"epoch": 0.59,
"grad_norm": 6.760104656219482,
"learning_rate": 4.467796610169492e-06,
"loss": 1.2604,
"step": 23410
},
{
"epoch": 0.59,
"grad_norm": 6.444414138793945,
"learning_rate": 4.461016949152543e-06,
"loss": 1.3677,
"step": 23420
},
{
"epoch": 0.59,
"grad_norm": 2.223396062850952,
"learning_rate": 4.454237288135594e-06,
"loss": 1.4585,
"step": 23430
},
{
"epoch": 0.59,
"grad_norm": 3.0469980239868164,
"learning_rate": 4.447457627118645e-06,
"loss": 1.2125,
"step": 23440
},
{
"epoch": 0.59,
"grad_norm": 9.140281677246094,
"learning_rate": 4.440677966101695e-06,
"loss": 1.2847,
"step": 23450
},
{
"epoch": 0.59,
"grad_norm": 10.596829414367676,
"learning_rate": 4.433898305084746e-06,
"loss": 1.1919,
"step": 23460
},
{
"epoch": 0.59,
"grad_norm": 6.769688129425049,
"learning_rate": 4.427118644067797e-06,
"loss": 1.3911,
"step": 23470
},
{
"epoch": 0.59,
"grad_norm": 8.3526029586792,
"learning_rate": 4.420338983050848e-06,
"loss": 1.3243,
"step": 23480
},
{
"epoch": 0.59,
"grad_norm": 1.516774296760559,
"learning_rate": 4.413559322033898e-06,
"loss": 1.4328,
"step": 23490
},
{
"epoch": 0.59,
"grad_norm": 5.758790493011475,
"learning_rate": 4.40677966101695e-06,
"loss": 1.1158,
"step": 23500
},
{
"epoch": 0.59,
"eval_loss": 1.3372619152069092,
"eval_runtime": 66.1468,
"eval_samples_per_second": 15.118,
"eval_steps_per_second": 15.118,
"step": 23500
},
{
"epoch": 0.59,
"grad_norm": 10.471700668334961,
"learning_rate": 4.4e-06,
"loss": 1.1029,
"step": 23510
},
{
"epoch": 0.59,
"grad_norm": 6.78934383392334,
"learning_rate": 4.393220338983052e-06,
"loss": 1.2838,
"step": 23520
},
{
"epoch": 0.59,
"grad_norm": 4.890566825866699,
"learning_rate": 4.386440677966102e-06,
"loss": 1.2861,
"step": 23530
},
{
"epoch": 0.59,
"grad_norm": 9.901065826416016,
"learning_rate": 4.379661016949153e-06,
"loss": 1.3043,
"step": 23540
},
{
"epoch": 0.59,
"grad_norm": 3.332019805908203,
"learning_rate": 4.372881355932203e-06,
"loss": 1.5098,
"step": 23550
},
{
"epoch": 0.59,
"grad_norm": 9.1102876663208,
"learning_rate": 4.366101694915255e-06,
"loss": 1.2222,
"step": 23560
},
{
"epoch": 0.59,
"grad_norm": 2.800964832305908,
"learning_rate": 4.359322033898305e-06,
"loss": 1.1916,
"step": 23570
},
{
"epoch": 0.59,
"grad_norm": 4.45274019241333,
"learning_rate": 4.352542372881357e-06,
"loss": 1.496,
"step": 23580
},
{
"epoch": 0.59,
"grad_norm": 7.7979350090026855,
"learning_rate": 4.345762711864407e-06,
"loss": 1.3483,
"step": 23590
},
{
"epoch": 0.59,
"grad_norm": 5.517279148101807,
"learning_rate": 4.338983050847458e-06,
"loss": 1.1602,
"step": 23600
},
{
"epoch": 0.59,
"grad_norm": 8.224603652954102,
"learning_rate": 4.332203389830509e-06,
"loss": 1.2216,
"step": 23610
},
{
"epoch": 0.59,
"grad_norm": 3.9079153537750244,
"learning_rate": 4.32542372881356e-06,
"loss": 1.4094,
"step": 23620
},
{
"epoch": 0.59,
"grad_norm": 7.209962844848633,
"learning_rate": 4.318644067796611e-06,
"loss": 1.4033,
"step": 23630
},
{
"epoch": 0.59,
"grad_norm": 6.915498733520508,
"learning_rate": 4.311864406779661e-06,
"loss": 1.3988,
"step": 23640
},
{
"epoch": 0.59,
"grad_norm": 6.8702778816223145,
"learning_rate": 4.305084745762712e-06,
"loss": 1.3308,
"step": 23650
},
{
"epoch": 0.59,
"grad_norm": 6.673946380615234,
"learning_rate": 4.298305084745763e-06,
"loss": 1.2756,
"step": 23660
},
{
"epoch": 0.59,
"grad_norm": 2.729367971420288,
"learning_rate": 4.291525423728814e-06,
"loss": 1.2702,
"step": 23670
},
{
"epoch": 0.59,
"grad_norm": 4.333055019378662,
"learning_rate": 4.284745762711865e-06,
"loss": 1.3365,
"step": 23680
},
{
"epoch": 0.59,
"grad_norm": 8.36184024810791,
"learning_rate": 4.277966101694915e-06,
"loss": 1.2783,
"step": 23690
},
{
"epoch": 0.59,
"grad_norm": 4.62699031829834,
"learning_rate": 4.271186440677967e-06,
"loss": 1.4385,
"step": 23700
},
{
"epoch": 0.59,
"grad_norm": 3.193026304244995,
"learning_rate": 4.264406779661017e-06,
"loss": 1.4843,
"step": 23710
},
{
"epoch": 0.59,
"grad_norm": 8.289533615112305,
"learning_rate": 4.257627118644068e-06,
"loss": 1.5263,
"step": 23720
},
{
"epoch": 0.59,
"grad_norm": 3.887775182723999,
"learning_rate": 4.250847457627119e-06,
"loss": 1.228,
"step": 23730
},
{
"epoch": 0.59,
"grad_norm": 10.728804588317871,
"learning_rate": 4.24406779661017e-06,
"loss": 1.4097,
"step": 23740
},
{
"epoch": 0.59,
"grad_norm": 5.405580997467041,
"learning_rate": 4.23728813559322e-06,
"loss": 1.227,
"step": 23750
},
{
"epoch": 0.59,
"grad_norm": 2.104985237121582,
"learning_rate": 4.230508474576272e-06,
"loss": 1.3258,
"step": 23760
},
{
"epoch": 0.59,
"grad_norm": 11.678805351257324,
"learning_rate": 4.223728813559322e-06,
"loss": 1.1797,
"step": 23770
},
{
"epoch": 0.59,
"grad_norm": 12.024051666259766,
"learning_rate": 4.216949152542374e-06,
"loss": 1.3278,
"step": 23780
},
{
"epoch": 0.59,
"grad_norm": 12.879485130310059,
"learning_rate": 4.210169491525424e-06,
"loss": 1.2552,
"step": 23790
},
{
"epoch": 0.59,
"grad_norm": 6.001992702484131,
"learning_rate": 4.203389830508475e-06,
"loss": 1.4639,
"step": 23800
},
{
"epoch": 0.6,
"grad_norm": 7.713657855987549,
"learning_rate": 4.196610169491525e-06,
"loss": 1.3542,
"step": 23810
},
{
"epoch": 0.6,
"grad_norm": 25.137435913085938,
"learning_rate": 4.189830508474577e-06,
"loss": 1.2694,
"step": 23820
},
{
"epoch": 0.6,
"grad_norm": 13.080780029296875,
"learning_rate": 4.183050847457627e-06,
"loss": 1.5512,
"step": 23830
},
{
"epoch": 0.6,
"grad_norm": 3.648967981338501,
"learning_rate": 4.176271186440679e-06,
"loss": 1.4919,
"step": 23840
},
{
"epoch": 0.6,
"grad_norm": 2.8366498947143555,
"learning_rate": 4.169491525423729e-06,
"loss": 1.3528,
"step": 23850
},
{
"epoch": 0.6,
"grad_norm": 2.3198916912078857,
"learning_rate": 4.16271186440678e-06,
"loss": 1.4445,
"step": 23860
},
{
"epoch": 0.6,
"grad_norm": 9.170830726623535,
"learning_rate": 4.155932203389831e-06,
"loss": 1.1801,
"step": 23870
},
{
"epoch": 0.6,
"grad_norm": 7.985089302062988,
"learning_rate": 4.149152542372882e-06,
"loss": 1.3195,
"step": 23880
},
{
"epoch": 0.6,
"grad_norm": 10.688753128051758,
"learning_rate": 4.142372881355933e-06,
"loss": 1.4527,
"step": 23890
},
{
"epoch": 0.6,
"grad_norm": 12.181285858154297,
"learning_rate": 4.135593220338983e-06,
"loss": 1.3875,
"step": 23900
},
{
"epoch": 0.6,
"grad_norm": 10.353550910949707,
"learning_rate": 4.128813559322034e-06,
"loss": 1.3937,
"step": 23910
},
{
"epoch": 0.6,
"grad_norm": 3.3962326049804688,
"learning_rate": 4.122033898305085e-06,
"loss": 1.2212,
"step": 23920
},
{
"epoch": 0.6,
"grad_norm": 9.191743850708008,
"learning_rate": 4.115254237288136e-06,
"loss": 1.3884,
"step": 23930
},
{
"epoch": 0.6,
"grad_norm": 8.74504566192627,
"learning_rate": 4.108474576271187e-06,
"loss": 1.369,
"step": 23940
},
{
"epoch": 0.6,
"grad_norm": 15.484914779663086,
"learning_rate": 4.101694915254237e-06,
"loss": 1.3607,
"step": 23950
},
{
"epoch": 0.6,
"grad_norm": 8.069631576538086,
"learning_rate": 4.094915254237289e-06,
"loss": 1.1674,
"step": 23960
},
{
"epoch": 0.6,
"grad_norm": 5.688279151916504,
"learning_rate": 4.088135593220339e-06,
"loss": 1.2652,
"step": 23970
},
{
"epoch": 0.6,
"grad_norm": 2.326960325241089,
"learning_rate": 4.081355932203391e-06,
"loss": 1.2149,
"step": 23980
},
{
"epoch": 0.6,
"grad_norm": 9.749725341796875,
"learning_rate": 4.074576271186441e-06,
"loss": 1.2378,
"step": 23990
},
{
"epoch": 0.6,
"grad_norm": 5.552289962768555,
"learning_rate": 4.067796610169492e-06,
"loss": 1.2699,
"step": 24000
},
{
"epoch": 0.6,
"eval_loss": 1.3411681652069092,
"eval_runtime": 66.1396,
"eval_samples_per_second": 15.12,
"eval_steps_per_second": 15.12,
"step": 24000
},
{
"epoch": 0.6,
"grad_norm": 8.768118858337402,
"learning_rate": 4.061016949152542e-06,
"loss": 1.1869,
"step": 24010
},
{
"epoch": 0.6,
"grad_norm": 1.0193852186203003,
"learning_rate": 4.054237288135594e-06,
"loss": 1.1477,
"step": 24020
},
{
"epoch": 0.6,
"grad_norm": 11.04339599609375,
"learning_rate": 4.047457627118644e-06,
"loss": 1.2492,
"step": 24030
},
{
"epoch": 0.6,
"grad_norm": 2.5347607135772705,
"learning_rate": 4.040677966101696e-06,
"loss": 1.2424,
"step": 24040
},
{
"epoch": 0.6,
"grad_norm": 5.121871471405029,
"learning_rate": 4.033898305084746e-06,
"loss": 1.1773,
"step": 24050
},
{
"epoch": 0.6,
"grad_norm": 8.53433609008789,
"learning_rate": 4.027118644067797e-06,
"loss": 1.4407,
"step": 24060
},
{
"epoch": 0.6,
"grad_norm": 11.311376571655273,
"learning_rate": 4.020338983050847e-06,
"loss": 1.4095,
"step": 24070
},
{
"epoch": 0.6,
"grad_norm": 2.8956375122070312,
"learning_rate": 4.013559322033899e-06,
"loss": 1.3076,
"step": 24080
},
{
"epoch": 0.6,
"grad_norm": 3.6406021118164062,
"learning_rate": 4.006779661016949e-06,
"loss": 1.1801,
"step": 24090
},
{
"epoch": 0.6,
"grad_norm": 4.67333459854126,
"learning_rate": 4.000000000000001e-06,
"loss": 1.3627,
"step": 24100
},
{
"epoch": 0.6,
"grad_norm": 4.243159294128418,
"learning_rate": 3.993220338983051e-06,
"loss": 1.2065,
"step": 24110
},
{
"epoch": 0.6,
"grad_norm": 4.570652484893799,
"learning_rate": 3.986440677966102e-06,
"loss": 1.4576,
"step": 24120
},
{
"epoch": 0.6,
"grad_norm": 10.30574893951416,
"learning_rate": 3.979661016949153e-06,
"loss": 1.3798,
"step": 24130
},
{
"epoch": 0.6,
"grad_norm": 1.7883845567703247,
"learning_rate": 3.972881355932204e-06,
"loss": 1.1474,
"step": 24140
},
{
"epoch": 0.6,
"grad_norm": 2.429614305496216,
"learning_rate": 3.966101694915255e-06,
"loss": 1.3992,
"step": 24150
},
{
"epoch": 0.6,
"grad_norm": 5.791226863861084,
"learning_rate": 3.959322033898305e-06,
"loss": 1.4164,
"step": 24160
},
{
"epoch": 0.6,
"grad_norm": 6.212001800537109,
"learning_rate": 3.952542372881356e-06,
"loss": 1.4647,
"step": 24170
},
{
"epoch": 0.6,
"grad_norm": 4.9569292068481445,
"learning_rate": 3.945762711864407e-06,
"loss": 1.2948,
"step": 24180
},
{
"epoch": 0.6,
"grad_norm": 2.2119970321655273,
"learning_rate": 3.938983050847458e-06,
"loss": 1.3955,
"step": 24190
},
{
"epoch": 0.6,
"grad_norm": 10.280770301818848,
"learning_rate": 3.932203389830509e-06,
"loss": 1.4461,
"step": 24200
},
{
"epoch": 0.61,
"grad_norm": 3.701272487640381,
"learning_rate": 3.925423728813559e-06,
"loss": 1.348,
"step": 24210
},
{
"epoch": 0.61,
"grad_norm": 8.827926635742188,
"learning_rate": 3.918644067796611e-06,
"loss": 1.3398,
"step": 24220
},
{
"epoch": 0.61,
"grad_norm": 6.997286319732666,
"learning_rate": 3.911864406779661e-06,
"loss": 1.4724,
"step": 24230
},
{
"epoch": 0.61,
"grad_norm": 5.5268449783325195,
"learning_rate": 3.905084745762713e-06,
"loss": 1.388,
"step": 24240
},
{
"epoch": 0.61,
"grad_norm": 8.842992782592773,
"learning_rate": 3.898305084745763e-06,
"loss": 1.2382,
"step": 24250
},
{
"epoch": 0.61,
"grad_norm": 11.24975872039795,
"learning_rate": 3.891525423728814e-06,
"loss": 1.2194,
"step": 24260
},
{
"epoch": 0.61,
"grad_norm": 2.875722646713257,
"learning_rate": 3.884745762711864e-06,
"loss": 1.3792,
"step": 24270
},
{
"epoch": 0.61,
"grad_norm": 8.459474563598633,
"learning_rate": 3.877966101694916e-06,
"loss": 1.3021,
"step": 24280
},
{
"epoch": 0.61,
"grad_norm": 3.315873861312866,
"learning_rate": 3.871186440677966e-06,
"loss": 1.2976,
"step": 24290
},
{
"epoch": 0.61,
"grad_norm": 6.280729293823242,
"learning_rate": 3.864406779661018e-06,
"loss": 1.3294,
"step": 24300
},
{
"epoch": 0.61,
"grad_norm": 6.004711627960205,
"learning_rate": 3.857627118644068e-06,
"loss": 1.3178,
"step": 24310
},
{
"epoch": 0.61,
"grad_norm": 8.207845687866211,
"learning_rate": 3.850847457627119e-06,
"loss": 1.424,
"step": 24320
},
{
"epoch": 0.61,
"grad_norm": 8.01065444946289,
"learning_rate": 3.844067796610169e-06,
"loss": 1.2842,
"step": 24330
},
{
"epoch": 0.61,
"grad_norm": 9.126721382141113,
"learning_rate": 3.837288135593221e-06,
"loss": 1.368,
"step": 24340
},
{
"epoch": 0.61,
"grad_norm": 11.590188026428223,
"learning_rate": 3.830508474576271e-06,
"loss": 1.375,
"step": 24350
},
{
"epoch": 0.61,
"grad_norm": 7.325139045715332,
"learning_rate": 3.823728813559323e-06,
"loss": 1.2593,
"step": 24360
},
{
"epoch": 0.61,
"grad_norm": 6.3924760818481445,
"learning_rate": 3.816949152542373e-06,
"loss": 1.3238,
"step": 24370
},
{
"epoch": 0.61,
"grad_norm": 5.093543529510498,
"learning_rate": 3.8101694915254238e-06,
"loss": 1.3398,
"step": 24380
},
{
"epoch": 0.61,
"grad_norm": 4.488302707672119,
"learning_rate": 3.8033898305084748e-06,
"loss": 1.2532,
"step": 24390
},
{
"epoch": 0.61,
"grad_norm": 3.5369062423706055,
"learning_rate": 3.7966101694915257e-06,
"loss": 1.1723,
"step": 24400
},
{
"epoch": 0.61,
"grad_norm": 3.2012510299682617,
"learning_rate": 3.7898305084745767e-06,
"loss": 1.3348,
"step": 24410
},
{
"epoch": 0.61,
"grad_norm": 1.5665017366409302,
"learning_rate": 3.7830508474576273e-06,
"loss": 1.4159,
"step": 24420
},
{
"epoch": 0.61,
"grad_norm": 12.912787437438965,
"learning_rate": 3.7762711864406782e-06,
"loss": 1.2406,
"step": 24430
},
{
"epoch": 0.61,
"grad_norm": 6.572142124176025,
"learning_rate": 3.7694915254237292e-06,
"loss": 1.1544,
"step": 24440
},
{
"epoch": 0.61,
"grad_norm": 4.999161720275879,
"learning_rate": 3.76271186440678e-06,
"loss": 1.1744,
"step": 24450
},
{
"epoch": 0.61,
"grad_norm": 3.322866439819336,
"learning_rate": 3.755932203389831e-06,
"loss": 1.3713,
"step": 24460
},
{
"epoch": 0.61,
"grad_norm": 5.197652816772461,
"learning_rate": 3.7491525423728813e-06,
"loss": 1.1858,
"step": 24470
},
{
"epoch": 0.61,
"grad_norm": 6.7361369132995605,
"learning_rate": 3.7423728813559323e-06,
"loss": 1.2778,
"step": 24480
},
{
"epoch": 0.61,
"grad_norm": 1.1276848316192627,
"learning_rate": 3.7355932203389833e-06,
"loss": 1.3971,
"step": 24490
},
{
"epoch": 0.61,
"grad_norm": 4.62593412399292,
"learning_rate": 3.7288135593220342e-06,
"loss": 1.479,
"step": 24500
},
{
"epoch": 0.61,
"eval_loss": 1.3212531805038452,
"eval_runtime": 66.1318,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 24500
},
{
"epoch": 0.61,
"grad_norm": 7.978855133056641,
"learning_rate": 3.7220338983050852e-06,
"loss": 1.1919,
"step": 24510
},
{
"epoch": 0.61,
"grad_norm": 6.347212314605713,
"learning_rate": 3.715254237288136e-06,
"loss": 1.2493,
"step": 24520
},
{
"epoch": 0.61,
"grad_norm": 6.2206573486328125,
"learning_rate": 3.7084745762711867e-06,
"loss": 1.2258,
"step": 24530
},
{
"epoch": 0.61,
"grad_norm": 2.548797607421875,
"learning_rate": 3.7016949152542377e-06,
"loss": 1.262,
"step": 24540
},
{
"epoch": 0.61,
"grad_norm": 9.5992431640625,
"learning_rate": 3.6949152542372883e-06,
"loss": 1.4298,
"step": 24550
},
{
"epoch": 0.61,
"grad_norm": 2.236175298690796,
"learning_rate": 3.6881355932203393e-06,
"loss": 1.3466,
"step": 24560
},
{
"epoch": 0.61,
"grad_norm": 7.134004592895508,
"learning_rate": 3.6813559322033902e-06,
"loss": 1.2642,
"step": 24570
},
{
"epoch": 0.61,
"grad_norm": 12.453125,
"learning_rate": 3.6745762711864408e-06,
"loss": 1.369,
"step": 24580
},
{
"epoch": 0.61,
"grad_norm": 16.243106842041016,
"learning_rate": 3.6677966101694918e-06,
"loss": 1.3269,
"step": 24590
},
{
"epoch": 0.61,
"grad_norm": 11.406882286071777,
"learning_rate": 3.6610169491525427e-06,
"loss": 1.4339,
"step": 24600
},
{
"epoch": 0.62,
"grad_norm": 6.334946632385254,
"learning_rate": 3.6542372881355937e-06,
"loss": 1.2445,
"step": 24610
},
{
"epoch": 0.62,
"grad_norm": 9.451517105102539,
"learning_rate": 3.6474576271186447e-06,
"loss": 1.1902,
"step": 24620
},
{
"epoch": 0.62,
"grad_norm": 1.9071747064590454,
"learning_rate": 3.640677966101695e-06,
"loss": 1.3273,
"step": 24630
},
{
"epoch": 0.62,
"grad_norm": 4.054659843444824,
"learning_rate": 3.633898305084746e-06,
"loss": 1.2808,
"step": 24640
},
{
"epoch": 0.62,
"grad_norm": 6.314877986907959,
"learning_rate": 3.6271186440677968e-06,
"loss": 1.4351,
"step": 24650
},
{
"epoch": 0.62,
"grad_norm": 6.971933841705322,
"learning_rate": 3.6203389830508478e-06,
"loss": 1.2973,
"step": 24660
},
{
"epoch": 0.62,
"grad_norm": 2.7064402103424072,
"learning_rate": 3.6135593220338987e-06,
"loss": 1.4335,
"step": 24670
},
{
"epoch": 0.62,
"grad_norm": 5.698015213012695,
"learning_rate": 3.6067796610169493e-06,
"loss": 1.3042,
"step": 24680
},
{
"epoch": 0.62,
"grad_norm": 5.647088050842285,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.3832,
"step": 24690
},
{
"epoch": 0.62,
"grad_norm": 10.945414543151855,
"learning_rate": 3.5932203389830512e-06,
"loss": 1.2363,
"step": 24700
},
{
"epoch": 0.62,
"grad_norm": 7.516660213470459,
"learning_rate": 3.5864406779661022e-06,
"loss": 1.2615,
"step": 24710
},
{
"epoch": 0.62,
"grad_norm": 3.6117703914642334,
"learning_rate": 3.579661016949153e-06,
"loss": 1.3533,
"step": 24720
},
{
"epoch": 0.62,
"grad_norm": 10.817008972167969,
"learning_rate": 3.5728813559322033e-06,
"loss": 1.3425,
"step": 24730
},
{
"epoch": 0.62,
"grad_norm": 7.7072858810424805,
"learning_rate": 3.5661016949152543e-06,
"loss": 1.2762,
"step": 24740
},
{
"epoch": 0.62,
"grad_norm": 5.9887495040893555,
"learning_rate": 3.5593220338983053e-06,
"loss": 1.3516,
"step": 24750
},
{
"epoch": 0.62,
"grad_norm": 3.4481067657470703,
"learning_rate": 3.5525423728813563e-06,
"loss": 1.0325,
"step": 24760
},
{
"epoch": 0.62,
"grad_norm": 5.2485551834106445,
"learning_rate": 3.5457627118644072e-06,
"loss": 1.4088,
"step": 24770
},
{
"epoch": 0.62,
"grad_norm": 6.970777988433838,
"learning_rate": 3.538983050847458e-06,
"loss": 1.4962,
"step": 24780
},
{
"epoch": 0.62,
"grad_norm": 6.037806034088135,
"learning_rate": 3.5322033898305088e-06,
"loss": 1.3806,
"step": 24790
},
{
"epoch": 0.62,
"grad_norm": 4.5726518630981445,
"learning_rate": 3.5254237288135597e-06,
"loss": 1.3086,
"step": 24800
},
{
"epoch": 0.62,
"grad_norm": 4.184850692749023,
"learning_rate": 3.5186440677966103e-06,
"loss": 1.4004,
"step": 24810
},
{
"epoch": 0.62,
"grad_norm": 9.181009292602539,
"learning_rate": 3.5118644067796613e-06,
"loss": 1.4348,
"step": 24820
},
{
"epoch": 0.62,
"grad_norm": 5.124319076538086,
"learning_rate": 3.5050847457627122e-06,
"loss": 1.2272,
"step": 24830
},
{
"epoch": 0.62,
"grad_norm": 7.30942440032959,
"learning_rate": 3.498305084745763e-06,
"loss": 1.3819,
"step": 24840
},
{
"epoch": 0.62,
"grad_norm": 7.727287769317627,
"learning_rate": 3.4915254237288138e-06,
"loss": 1.1956,
"step": 24850
},
{
"epoch": 0.62,
"grad_norm": 5.2934112548828125,
"learning_rate": 3.4847457627118648e-06,
"loss": 1.3141,
"step": 24860
},
{
"epoch": 0.62,
"grad_norm": 7.7370500564575195,
"learning_rate": 3.4779661016949157e-06,
"loss": 1.2418,
"step": 24870
},
{
"epoch": 0.62,
"grad_norm": 5.379338264465332,
"learning_rate": 3.4711864406779667e-06,
"loss": 1.435,
"step": 24880
},
{
"epoch": 0.62,
"grad_norm": 5.592279434204102,
"learning_rate": 3.464406779661017e-06,
"loss": 1.324,
"step": 24890
},
{
"epoch": 0.62,
"grad_norm": 4.178751468658447,
"learning_rate": 3.457627118644068e-06,
"loss": 1.4095,
"step": 24900
},
{
"epoch": 0.62,
"grad_norm": 2.7852957248687744,
"learning_rate": 3.450847457627119e-06,
"loss": 1.1834,
"step": 24910
},
{
"epoch": 0.62,
"grad_norm": 12.001542091369629,
"learning_rate": 3.4440677966101698e-06,
"loss": 1.3294,
"step": 24920
},
{
"epoch": 0.62,
"grad_norm": 4.010140895843506,
"learning_rate": 3.4372881355932207e-06,
"loss": 1.4615,
"step": 24930
},
{
"epoch": 0.62,
"grad_norm": 2.7130627632141113,
"learning_rate": 3.4305084745762713e-06,
"loss": 1.4789,
"step": 24940
},
{
"epoch": 0.62,
"grad_norm": 5.209987640380859,
"learning_rate": 3.4237288135593223e-06,
"loss": 1.2389,
"step": 24950
},
{
"epoch": 0.62,
"grad_norm": 4.1047515869140625,
"learning_rate": 3.4169491525423733e-06,
"loss": 1.3449,
"step": 24960
},
{
"epoch": 0.62,
"grad_norm": 19.238649368286133,
"learning_rate": 3.4101694915254242e-06,
"loss": 1.3224,
"step": 24970
},
{
"epoch": 0.62,
"grad_norm": 11.792010307312012,
"learning_rate": 3.403389830508475e-06,
"loss": 1.1061,
"step": 24980
},
{
"epoch": 0.62,
"grad_norm": 10.934020042419434,
"learning_rate": 3.3966101694915253e-06,
"loss": 1.4179,
"step": 24990
},
{
"epoch": 0.62,
"grad_norm": 1.5776662826538086,
"learning_rate": 3.3898305084745763e-06,
"loss": 1.3254,
"step": 25000
},
{
"epoch": 0.62,
"eval_loss": 1.3409814834594727,
"eval_runtime": 66.1173,
"eval_samples_per_second": 15.125,
"eval_steps_per_second": 15.125,
"step": 25000
},
{
"epoch": 0.63,
"grad_norm": 7.775777339935303,
"learning_rate": 3.3830508474576273e-06,
"loss": 1.5704,
"step": 25010
},
{
"epoch": 0.63,
"grad_norm": 6.8910698890686035,
"learning_rate": 3.3762711864406783e-06,
"loss": 1.2296,
"step": 25020
},
{
"epoch": 0.63,
"grad_norm": 3.11647367477417,
"learning_rate": 3.3694915254237292e-06,
"loss": 1.1465,
"step": 25030
},
{
"epoch": 0.63,
"grad_norm": 4.026238441467285,
"learning_rate": 3.3627118644067802e-06,
"loss": 1.4266,
"step": 25040
},
{
"epoch": 0.63,
"grad_norm": 3.3097872734069824,
"learning_rate": 3.3559322033898308e-06,
"loss": 1.4,
"step": 25050
},
{
"epoch": 0.63,
"grad_norm": 3.2518560886383057,
"learning_rate": 3.3491525423728817e-06,
"loss": 1.3535,
"step": 25060
},
{
"epoch": 0.63,
"grad_norm": 4.915974140167236,
"learning_rate": 3.3423728813559327e-06,
"loss": 1.3441,
"step": 25070
},
{
"epoch": 0.63,
"grad_norm": 4.099907398223877,
"learning_rate": 3.3355932203389833e-06,
"loss": 1.4649,
"step": 25080
},
{
"epoch": 0.63,
"grad_norm": 7.00651741027832,
"learning_rate": 3.3288135593220343e-06,
"loss": 1.4445,
"step": 25090
},
{
"epoch": 0.63,
"grad_norm": 2.7813143730163574,
"learning_rate": 3.322033898305085e-06,
"loss": 1.2997,
"step": 25100
},
{
"epoch": 0.63,
"grad_norm": 4.583659648895264,
"learning_rate": 3.3152542372881358e-06,
"loss": 1.3112,
"step": 25110
},
{
"epoch": 0.63,
"grad_norm": 9.5046968460083,
"learning_rate": 3.3084745762711868e-06,
"loss": 1.3488,
"step": 25120
},
{
"epoch": 0.63,
"grad_norm": 4.829354286193848,
"learning_rate": 3.3016949152542377e-06,
"loss": 1.2801,
"step": 25130
},
{
"epoch": 0.63,
"grad_norm": 8.664487838745117,
"learning_rate": 3.2949152542372887e-06,
"loss": 1.3001,
"step": 25140
},
{
"epoch": 0.63,
"grad_norm": 1.5184565782546997,
"learning_rate": 3.288135593220339e-06,
"loss": 1.3178,
"step": 25150
},
{
"epoch": 0.63,
"grad_norm": 2.527219295501709,
"learning_rate": 3.28135593220339e-06,
"loss": 1.1742,
"step": 25160
},
{
"epoch": 0.63,
"grad_norm": 1.9469351768493652,
"learning_rate": 3.274576271186441e-06,
"loss": 1.3338,
"step": 25170
},
{
"epoch": 0.63,
"grad_norm": 9.183582305908203,
"learning_rate": 3.2677966101694918e-06,
"loss": 1.2689,
"step": 25180
},
{
"epoch": 0.63,
"grad_norm": 8.95080280303955,
"learning_rate": 3.2610169491525428e-06,
"loss": 1.4399,
"step": 25190
},
{
"epoch": 0.63,
"grad_norm": 3.5885729789733887,
"learning_rate": 3.2542372881355933e-06,
"loss": 1.2411,
"step": 25200
},
{
"epoch": 0.63,
"grad_norm": 6.78897762298584,
"learning_rate": 3.2474576271186443e-06,
"loss": 1.1516,
"step": 25210
},
{
"epoch": 0.63,
"grad_norm": 16.358980178833008,
"learning_rate": 3.2406779661016953e-06,
"loss": 1.3401,
"step": 25220
},
{
"epoch": 0.63,
"grad_norm": 13.469403266906738,
"learning_rate": 3.2338983050847462e-06,
"loss": 1.1778,
"step": 25230
},
{
"epoch": 0.63,
"grad_norm": 9.235873222351074,
"learning_rate": 3.2271186440677972e-06,
"loss": 1.463,
"step": 25240
},
{
"epoch": 0.63,
"grad_norm": 5.634754657745361,
"learning_rate": 3.2203389830508473e-06,
"loss": 1.4592,
"step": 25250
},
{
"epoch": 0.63,
"grad_norm": 7.916177749633789,
"learning_rate": 3.2135593220338983e-06,
"loss": 1.2284,
"step": 25260
},
{
"epoch": 0.63,
"grad_norm": 3.810818672180176,
"learning_rate": 3.2067796610169493e-06,
"loss": 1.3948,
"step": 25270
},
{
"epoch": 0.63,
"grad_norm": 3.151276111602783,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.3252,
"step": 25280
},
{
"epoch": 0.63,
"grad_norm": 4.654315948486328,
"learning_rate": 3.1932203389830513e-06,
"loss": 1.3577,
"step": 25290
},
{
"epoch": 0.63,
"grad_norm": 7.808940410614014,
"learning_rate": 3.186440677966102e-06,
"loss": 1.419,
"step": 25300
},
{
"epoch": 0.63,
"grad_norm": 6.512801647186279,
"learning_rate": 3.1796610169491528e-06,
"loss": 1.2751,
"step": 25310
},
{
"epoch": 0.63,
"grad_norm": 8.710541725158691,
"learning_rate": 3.1728813559322038e-06,
"loss": 1.2923,
"step": 25320
},
{
"epoch": 0.63,
"grad_norm": 3.160752296447754,
"learning_rate": 3.1661016949152547e-06,
"loss": 1.3357,
"step": 25330
},
{
"epoch": 0.63,
"grad_norm": 2.4420571327209473,
"learning_rate": 3.1593220338983053e-06,
"loss": 1.4122,
"step": 25340
},
{
"epoch": 0.63,
"grad_norm": 12.090134620666504,
"learning_rate": 3.1525423728813563e-06,
"loss": 1.1703,
"step": 25350
},
{
"epoch": 0.63,
"grad_norm": 5.109529972076416,
"learning_rate": 3.145762711864407e-06,
"loss": 1.4391,
"step": 25360
},
{
"epoch": 0.63,
"grad_norm": 3.0033655166625977,
"learning_rate": 3.138983050847458e-06,
"loss": 1.2308,
"step": 25370
},
{
"epoch": 0.63,
"grad_norm": 6.253593921661377,
"learning_rate": 3.1322033898305088e-06,
"loss": 1.1566,
"step": 25380
},
{
"epoch": 0.63,
"grad_norm": 5.59571647644043,
"learning_rate": 3.1254237288135598e-06,
"loss": 1.3234,
"step": 25390
},
{
"epoch": 0.64,
"grad_norm": 15.45341968536377,
"learning_rate": 3.1186440677966107e-06,
"loss": 1.3298,
"step": 25400
},
{
"epoch": 0.64,
"grad_norm": 4.012663841247559,
"learning_rate": 3.111864406779661e-06,
"loss": 1.3515,
"step": 25410
},
{
"epoch": 0.64,
"grad_norm": 11.210050582885742,
"learning_rate": 3.105084745762712e-06,
"loss": 1.5219,
"step": 25420
},
{
"epoch": 0.64,
"grad_norm": 12.784728050231934,
"learning_rate": 3.098305084745763e-06,
"loss": 1.1988,
"step": 25430
},
{
"epoch": 0.64,
"grad_norm": 4.20169734954834,
"learning_rate": 3.091525423728814e-06,
"loss": 1.421,
"step": 25440
},
{
"epoch": 0.64,
"grad_norm": 6.456515312194824,
"learning_rate": 3.0847457627118648e-06,
"loss": 1.1633,
"step": 25450
},
{
"epoch": 0.64,
"grad_norm": 6.611431121826172,
"learning_rate": 3.0779661016949153e-06,
"loss": 1.2853,
"step": 25460
},
{
"epoch": 0.64,
"grad_norm": 4.3997015953063965,
"learning_rate": 3.0711864406779663e-06,
"loss": 1.2609,
"step": 25470
},
{
"epoch": 0.64,
"grad_norm": 12.016313552856445,
"learning_rate": 3.0644067796610173e-06,
"loss": 1.1744,
"step": 25480
},
{
"epoch": 0.64,
"grad_norm": 5.508640289306641,
"learning_rate": 3.0576271186440683e-06,
"loss": 1.2545,
"step": 25490
},
{
"epoch": 0.64,
"grad_norm": 7.7658305168151855,
"learning_rate": 3.0508474576271192e-06,
"loss": 1.132,
"step": 25500
},
{
"epoch": 0.64,
"eval_loss": 1.3264555931091309,
"eval_runtime": 66.1247,
"eval_samples_per_second": 15.123,
"eval_steps_per_second": 15.123,
"step": 25500
},
{
"epoch": 0.64,
"grad_norm": 4.297203063964844,
"learning_rate": 3.0440677966101694e-06,
"loss": 1.0663,
"step": 25510
},
{
"epoch": 0.64,
"grad_norm": 12.412454605102539,
"learning_rate": 3.0372881355932203e-06,
"loss": 1.1782,
"step": 25520
},
{
"epoch": 0.64,
"grad_norm": 10.640459060668945,
"learning_rate": 3.0305084745762713e-06,
"loss": 1.4982,
"step": 25530
},
{
"epoch": 0.64,
"grad_norm": 12.518424987792969,
"learning_rate": 3.0237288135593223e-06,
"loss": 1.4771,
"step": 25540
},
{
"epoch": 0.64,
"grad_norm": 7.445399284362793,
"learning_rate": 3.0169491525423733e-06,
"loss": 1.2959,
"step": 25550
},
{
"epoch": 0.64,
"grad_norm": 3.960470676422119,
"learning_rate": 3.010169491525424e-06,
"loss": 1.4464,
"step": 25560
},
{
"epoch": 0.64,
"grad_norm": 11.259561538696289,
"learning_rate": 3.003389830508475e-06,
"loss": 1.2178,
"step": 25570
},
{
"epoch": 0.64,
"grad_norm": 7.605580806732178,
"learning_rate": 2.9966101694915258e-06,
"loss": 1.3286,
"step": 25580
},
{
"epoch": 0.64,
"grad_norm": 4.82462215423584,
"learning_rate": 2.9898305084745768e-06,
"loss": 1.4727,
"step": 25590
},
{
"epoch": 0.64,
"grad_norm": 4.511000633239746,
"learning_rate": 2.9830508474576277e-06,
"loss": 1.3572,
"step": 25600
},
{
"epoch": 0.64,
"grad_norm": 6.090604305267334,
"learning_rate": 2.9762711864406783e-06,
"loss": 1.4802,
"step": 25610
},
{
"epoch": 0.64,
"grad_norm": 9.687544822692871,
"learning_rate": 2.969491525423729e-06,
"loss": 1.3652,
"step": 25620
},
{
"epoch": 0.64,
"grad_norm": 12.760902404785156,
"learning_rate": 2.96271186440678e-06,
"loss": 1.3634,
"step": 25630
},
{
"epoch": 0.64,
"grad_norm": 6.688235759735107,
"learning_rate": 2.955932203389831e-06,
"loss": 1.209,
"step": 25640
},
{
"epoch": 0.64,
"grad_norm": 5.238013744354248,
"learning_rate": 2.9491525423728818e-06,
"loss": 1.5462,
"step": 25650
},
{
"epoch": 0.64,
"grad_norm": 18.444650650024414,
"learning_rate": 2.9423728813559327e-06,
"loss": 1.2963,
"step": 25660
},
{
"epoch": 0.64,
"grad_norm": 5.625192642211914,
"learning_rate": 2.935593220338983e-06,
"loss": 1.3514,
"step": 25670
},
{
"epoch": 0.64,
"grad_norm": 5.223139762878418,
"learning_rate": 2.928813559322034e-06,
"loss": 1.2339,
"step": 25680
},
{
"epoch": 0.64,
"grad_norm": 4.5006422996521,
"learning_rate": 2.922033898305085e-06,
"loss": 1.4424,
"step": 25690
},
{
"epoch": 0.64,
"grad_norm": 9.127293586730957,
"learning_rate": 2.915254237288136e-06,
"loss": 1.2334,
"step": 25700
},
{
"epoch": 0.64,
"grad_norm": 5.78049373626709,
"learning_rate": 2.9084745762711868e-06,
"loss": 1.301,
"step": 25710
},
{
"epoch": 0.64,
"grad_norm": 5.395204544067383,
"learning_rate": 2.9016949152542373e-06,
"loss": 1.1889,
"step": 25720
},
{
"epoch": 0.64,
"grad_norm": 2.3502519130706787,
"learning_rate": 2.8949152542372883e-06,
"loss": 1.41,
"step": 25730
},
{
"epoch": 0.64,
"grad_norm": 3.3766727447509766,
"learning_rate": 2.8881355932203393e-06,
"loss": 1.2768,
"step": 25740
},
{
"epoch": 0.64,
"grad_norm": 6.30269718170166,
"learning_rate": 2.8813559322033903e-06,
"loss": 1.3432,
"step": 25750
},
{
"epoch": 0.64,
"grad_norm": 7.727334022521973,
"learning_rate": 2.8745762711864412e-06,
"loss": 1.3465,
"step": 25760
},
{
"epoch": 0.64,
"grad_norm": 8.983928680419922,
"learning_rate": 2.8677966101694914e-06,
"loss": 1.1202,
"step": 25770
},
{
"epoch": 0.64,
"grad_norm": 4.910791397094727,
"learning_rate": 2.8610169491525424e-06,
"loss": 1.2038,
"step": 25780
},
{
"epoch": 0.64,
"grad_norm": 12.079176902770996,
"learning_rate": 2.8542372881355933e-06,
"loss": 1.1599,
"step": 25790
},
{
"epoch": 0.65,
"grad_norm": 3.6942214965820312,
"learning_rate": 2.8474576271186443e-06,
"loss": 1.2242,
"step": 25800
},
{
"epoch": 0.65,
"grad_norm": 2.4002459049224854,
"learning_rate": 2.8406779661016953e-06,
"loss": 1.2288,
"step": 25810
},
{
"epoch": 0.65,
"grad_norm": 4.8168840408325195,
"learning_rate": 2.833898305084746e-06,
"loss": 1.4489,
"step": 25820
},
{
"epoch": 0.65,
"grad_norm": 10.90778636932373,
"learning_rate": 2.827118644067797e-06,
"loss": 1.4108,
"step": 25830
},
{
"epoch": 0.65,
"grad_norm": 9.852614402770996,
"learning_rate": 2.820338983050848e-06,
"loss": 1.3685,
"step": 25840
},
{
"epoch": 0.65,
"grad_norm": 8.391303062438965,
"learning_rate": 2.8135593220338988e-06,
"loss": 1.2921,
"step": 25850
},
{
"epoch": 0.65,
"grad_norm": 5.348249912261963,
"learning_rate": 2.8067796610169497e-06,
"loss": 1.3463,
"step": 25860
},
{
"epoch": 0.65,
"grad_norm": 3.3569209575653076,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.2877,
"step": 25870
},
{
"epoch": 0.65,
"grad_norm": 9.246468544006348,
"learning_rate": 2.793220338983051e-06,
"loss": 1.3478,
"step": 25880
},
{
"epoch": 0.65,
"grad_norm": 7.3121771812438965,
"learning_rate": 2.786440677966102e-06,
"loss": 1.3116,
"step": 25890
},
{
"epoch": 0.65,
"grad_norm": 7.324717044830322,
"learning_rate": 2.779661016949153e-06,
"loss": 1.4986,
"step": 25900
},
{
"epoch": 0.65,
"grad_norm": 6.7373738288879395,
"learning_rate": 2.7728813559322038e-06,
"loss": 1.5417,
"step": 25910
},
{
"epoch": 0.65,
"grad_norm": 10.05711555480957,
"learning_rate": 2.7661016949152548e-06,
"loss": 1.3439,
"step": 25920
},
{
"epoch": 0.65,
"grad_norm": 3.843493700027466,
"learning_rate": 2.7593220338983053e-06,
"loss": 1.1626,
"step": 25930
},
{
"epoch": 0.65,
"grad_norm": 9.513094902038574,
"learning_rate": 2.752542372881356e-06,
"loss": 1.3412,
"step": 25940
},
{
"epoch": 0.65,
"grad_norm": 8.064391136169434,
"learning_rate": 2.745762711864407e-06,
"loss": 1.3483,
"step": 25950
},
{
"epoch": 0.65,
"grad_norm": 10.491897583007812,
"learning_rate": 2.738983050847458e-06,
"loss": 1.3214,
"step": 25960
},
{
"epoch": 0.65,
"grad_norm": 3.9402971267700195,
"learning_rate": 2.732203389830509e-06,
"loss": 1.3803,
"step": 25970
},
{
"epoch": 0.65,
"grad_norm": 9.979386329650879,
"learning_rate": 2.7254237288135593e-06,
"loss": 1.1408,
"step": 25980
},
{
"epoch": 0.65,
"grad_norm": 7.461246490478516,
"learning_rate": 2.7186440677966103e-06,
"loss": 1.3098,
"step": 25990
},
{
"epoch": 0.65,
"grad_norm": 4.412223815917969,
"learning_rate": 2.7118644067796613e-06,
"loss": 1.457,
"step": 26000
},
{
"epoch": 0.65,
"eval_loss": 1.3418877124786377,
"eval_runtime": 66.1458,
"eval_samples_per_second": 15.118,
"eval_steps_per_second": 15.118,
"step": 26000
},
{
"epoch": 0.65,
"grad_norm": 4.992754936218262,
"learning_rate": 2.7050847457627123e-06,
"loss": 1.1937,
"step": 26010
},
{
"epoch": 0.65,
"grad_norm": 5.497439861297607,
"learning_rate": 2.6983050847457633e-06,
"loss": 1.3435,
"step": 26020
},
{
"epoch": 0.65,
"grad_norm": 4.283881664276123,
"learning_rate": 2.6915254237288134e-06,
"loss": 1.3997,
"step": 26030
},
{
"epoch": 0.65,
"grad_norm": 6.579436779022217,
"learning_rate": 2.6847457627118644e-06,
"loss": 1.4047,
"step": 26040
},
{
"epoch": 0.65,
"grad_norm": 4.805014133453369,
"learning_rate": 2.6779661016949153e-06,
"loss": 1.4021,
"step": 26050
},
{
"epoch": 0.65,
"grad_norm": 4.568453788757324,
"learning_rate": 2.6711864406779663e-06,
"loss": 1.4761,
"step": 26060
},
{
"epoch": 0.65,
"grad_norm": 5.559253692626953,
"learning_rate": 2.6644067796610173e-06,
"loss": 1.2958,
"step": 26070
},
{
"epoch": 0.65,
"grad_norm": 14.279343605041504,
"learning_rate": 2.657627118644068e-06,
"loss": 1.194,
"step": 26080
},
{
"epoch": 0.65,
"grad_norm": 15.45864486694336,
"learning_rate": 2.650847457627119e-06,
"loss": 1.2134,
"step": 26090
},
{
"epoch": 0.65,
"grad_norm": 17.310068130493164,
"learning_rate": 2.64406779661017e-06,
"loss": 1.2946,
"step": 26100
},
{
"epoch": 0.65,
"grad_norm": 4.31659460067749,
"learning_rate": 2.6372881355932208e-06,
"loss": 1.3783,
"step": 26110
},
{
"epoch": 0.65,
"grad_norm": 14.684484481811523,
"learning_rate": 2.6305084745762718e-06,
"loss": 1.3414,
"step": 26120
},
{
"epoch": 0.65,
"grad_norm": 5.934599876403809,
"learning_rate": 2.6237288135593223e-06,
"loss": 1.3266,
"step": 26130
},
{
"epoch": 0.65,
"grad_norm": 7.535737037658691,
"learning_rate": 2.616949152542373e-06,
"loss": 1.2579,
"step": 26140
},
{
"epoch": 0.65,
"grad_norm": 5.8215556144714355,
"learning_rate": 2.610169491525424e-06,
"loss": 1.304,
"step": 26150
},
{
"epoch": 0.65,
"grad_norm": 5.646538734436035,
"learning_rate": 2.603389830508475e-06,
"loss": 1.2792,
"step": 26160
},
{
"epoch": 0.65,
"grad_norm": 3.834282398223877,
"learning_rate": 2.596610169491526e-06,
"loss": 1.4717,
"step": 26170
},
{
"epoch": 0.65,
"grad_norm": 3.313835620880127,
"learning_rate": 2.5898305084745768e-06,
"loss": 1.3312,
"step": 26180
},
{
"epoch": 0.65,
"grad_norm": 20.453479766845703,
"learning_rate": 2.5830508474576273e-06,
"loss": 1.3784,
"step": 26190
},
{
"epoch": 0.66,
"grad_norm": 1.0731091499328613,
"learning_rate": 2.576271186440678e-06,
"loss": 1.4162,
"step": 26200
},
{
"epoch": 0.66,
"grad_norm": 7.5899577140808105,
"learning_rate": 2.569491525423729e-06,
"loss": 1.2536,
"step": 26210
},
{
"epoch": 0.66,
"grad_norm": 4.527303218841553,
"learning_rate": 2.56271186440678e-06,
"loss": 1.5048,
"step": 26220
},
{
"epoch": 0.66,
"grad_norm": 3.702897787094116,
"learning_rate": 2.555932203389831e-06,
"loss": 1.1666,
"step": 26230
},
{
"epoch": 0.66,
"grad_norm": 3.448979139328003,
"learning_rate": 2.5491525423728814e-06,
"loss": 1.2812,
"step": 26240
},
{
"epoch": 0.66,
"grad_norm": 15.188081741333008,
"learning_rate": 2.5423728813559323e-06,
"loss": 1.3521,
"step": 26250
},
{
"epoch": 0.66,
"grad_norm": 8.675294876098633,
"learning_rate": 2.5355932203389833e-06,
"loss": 1.4845,
"step": 26260
},
{
"epoch": 0.66,
"grad_norm": 5.673226356506348,
"learning_rate": 2.5288135593220343e-06,
"loss": 1.2264,
"step": 26270
},
{
"epoch": 0.66,
"grad_norm": 6.015854358673096,
"learning_rate": 2.5220338983050853e-06,
"loss": 1.3036,
"step": 26280
},
{
"epoch": 0.66,
"grad_norm": 12.492931365966797,
"learning_rate": 2.5152542372881354e-06,
"loss": 1.4434,
"step": 26290
},
{
"epoch": 0.66,
"grad_norm": 5.928922176361084,
"learning_rate": 2.5084745762711864e-06,
"loss": 1.3118,
"step": 26300
},
{
"epoch": 0.66,
"grad_norm": 5.230995178222656,
"learning_rate": 2.5016949152542374e-06,
"loss": 1.3524,
"step": 26310
},
{
"epoch": 0.66,
"grad_norm": 6.7296671867370605,
"learning_rate": 2.4949152542372883e-06,
"loss": 1.3651,
"step": 26320
},
{
"epoch": 0.66,
"grad_norm": 2.378596544265747,
"learning_rate": 2.488135593220339e-06,
"loss": 1.4236,
"step": 26330
},
{
"epoch": 0.66,
"grad_norm": 5.858238697052002,
"learning_rate": 2.48135593220339e-06,
"loss": 1.1553,
"step": 26340
},
{
"epoch": 0.66,
"grad_norm": 3.7590560913085938,
"learning_rate": 2.474576271186441e-06,
"loss": 1.4302,
"step": 26350
},
{
"epoch": 0.66,
"grad_norm": 10.852778434753418,
"learning_rate": 2.467796610169492e-06,
"loss": 1.1943,
"step": 26360
},
{
"epoch": 0.66,
"grad_norm": 11.347558975219727,
"learning_rate": 2.461016949152543e-06,
"loss": 1.2559,
"step": 26370
},
{
"epoch": 0.66,
"grad_norm": 6.081493854522705,
"learning_rate": 2.4542372881355933e-06,
"loss": 1.3532,
"step": 26380
},
{
"epoch": 0.66,
"grad_norm": 8.263628959655762,
"learning_rate": 2.4474576271186443e-06,
"loss": 1.2041,
"step": 26390
},
{
"epoch": 0.66,
"grad_norm": 7.144092082977295,
"learning_rate": 2.4406779661016953e-06,
"loss": 1.2193,
"step": 26400
},
{
"epoch": 0.66,
"grad_norm": 6.578696250915527,
"learning_rate": 2.433898305084746e-06,
"loss": 1.3139,
"step": 26410
},
{
"epoch": 0.66,
"grad_norm": 3.019033193588257,
"learning_rate": 2.427118644067797e-06,
"loss": 1.4642,
"step": 26420
},
{
"epoch": 0.66,
"grad_norm": 5.443304538726807,
"learning_rate": 2.4203389830508474e-06,
"loss": 1.3738,
"step": 26430
},
{
"epoch": 0.66,
"grad_norm": 10.051207542419434,
"learning_rate": 2.4135593220338984e-06,
"loss": 1.4396,
"step": 26440
},
{
"epoch": 0.66,
"grad_norm": 10.604415893554688,
"learning_rate": 2.4067796610169493e-06,
"loss": 1.2971,
"step": 26450
},
{
"epoch": 0.66,
"grad_norm": 9.754619598388672,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.3421,
"step": 26460
},
{
"epoch": 0.66,
"grad_norm": 6.159849166870117,
"learning_rate": 2.393220338983051e-06,
"loss": 1.3461,
"step": 26470
},
{
"epoch": 0.66,
"grad_norm": 1.3210265636444092,
"learning_rate": 2.386440677966102e-06,
"loss": 1.2615,
"step": 26480
},
{
"epoch": 0.66,
"grad_norm": 5.67736291885376,
"learning_rate": 2.379661016949153e-06,
"loss": 1.3296,
"step": 26490
},
{
"epoch": 0.66,
"grad_norm": 4.055379867553711,
"learning_rate": 2.372881355932204e-06,
"loss": 1.4705,
"step": 26500
},
{
"epoch": 0.66,
"eval_loss": 1.2714667320251465,
"eval_runtime": 66.1328,
"eval_samples_per_second": 15.121,
"eval_steps_per_second": 15.121,
"step": 26500
},
{
"epoch": 0.66,
"grad_norm": 7.19535493850708,
"learning_rate": 2.3661016949152544e-06,
"loss": 1.2956,
"step": 26510
},
{
"epoch": 0.66,
"grad_norm": 7.952252388000488,
"learning_rate": 2.3593220338983053e-06,
"loss": 1.2952,
"step": 26520
},
{
"epoch": 0.66,
"grad_norm": 1.3878239393234253,
"learning_rate": 2.3525423728813563e-06,
"loss": 1.3957,
"step": 26530
},
{
"epoch": 0.66,
"grad_norm": 9.453435897827148,
"learning_rate": 2.345762711864407e-06,
"loss": 1.3314,
"step": 26540
},
{
"epoch": 0.66,
"grad_norm": 2.3859434127807617,
"learning_rate": 2.338983050847458e-06,
"loss": 1.3959,
"step": 26550
},
{
"epoch": 0.66,
"grad_norm": 4.052861213684082,
"learning_rate": 2.3322033898305084e-06,
"loss": 1.3097,
"step": 26560
},
{
"epoch": 0.66,
"grad_norm": 14.907791137695312,
"learning_rate": 2.3254237288135594e-06,
"loss": 1.3706,
"step": 26570
},
{
"epoch": 0.66,
"grad_norm": 7.020768642425537,
"learning_rate": 2.3186440677966103e-06,
"loss": 1.3536,
"step": 26580
},
{
"epoch": 0.66,
"grad_norm": 15.225879669189453,
"learning_rate": 2.3118644067796613e-06,
"loss": 1.354,
"step": 26590
},
{
"epoch": 0.67,
"grad_norm": 6.249582767486572,
"learning_rate": 2.305084745762712e-06,
"loss": 1.2721,
"step": 26600
},
{
"epoch": 0.67,
"grad_norm": 4.756400108337402,
"learning_rate": 2.298305084745763e-06,
"loss": 1.3044,
"step": 26610
},
{
"epoch": 0.67,
"grad_norm": 9.848575592041016,
"learning_rate": 2.291525423728814e-06,
"loss": 1.2471,
"step": 26620
},
{
"epoch": 0.67,
"grad_norm": 9.997851371765137,
"learning_rate": 2.284745762711865e-06,
"loss": 1.1673,
"step": 26630
},
{
"epoch": 0.67,
"grad_norm": 11.57490062713623,
"learning_rate": 2.2779661016949154e-06,
"loss": 1.3241,
"step": 26640
},
{
"epoch": 0.67,
"grad_norm": 8.282011032104492,
"learning_rate": 2.2711864406779663e-06,
"loss": 1.2821,
"step": 26650
},
{
"epoch": 0.67,
"grad_norm": 6.8680644035339355,
"learning_rate": 2.2644067796610173e-06,
"loss": 1.2355,
"step": 26660
},
{
"epoch": 0.67,
"grad_norm": 15.327707290649414,
"learning_rate": 2.257627118644068e-06,
"loss": 1.2088,
"step": 26670
},
{
"epoch": 0.67,
"grad_norm": 6.978293418884277,
"learning_rate": 2.250847457627119e-06,
"loss": 1.3372,
"step": 26680
},
{
"epoch": 0.67,
"grad_norm": 10.799981117248535,
"learning_rate": 2.2440677966101694e-06,
"loss": 1.4839,
"step": 26690
},
{
"epoch": 0.67,
"grad_norm": 4.861376762390137,
"learning_rate": 2.2372881355932204e-06,
"loss": 1.3361,
"step": 26700
},
{
"epoch": 0.67,
"grad_norm": 5.07356595993042,
"learning_rate": 2.2305084745762714e-06,
"loss": 1.1581,
"step": 26710
},
{
"epoch": 0.67,
"grad_norm": 16.509151458740234,
"learning_rate": 2.2237288135593223e-06,
"loss": 1.3014,
"step": 26720
},
{
"epoch": 0.67,
"grad_norm": 8.232512474060059,
"learning_rate": 2.216949152542373e-06,
"loss": 1.4513,
"step": 26730
},
{
"epoch": 0.67,
"grad_norm": 4.673596382141113,
"learning_rate": 2.210169491525424e-06,
"loss": 1.3162,
"step": 26740
},
{
"epoch": 0.67,
"grad_norm": 3.565079927444458,
"learning_rate": 2.203389830508475e-06,
"loss": 1.3423,
"step": 26750
},
{
"epoch": 0.67,
"grad_norm": 8.415205001831055,
"learning_rate": 2.196610169491526e-06,
"loss": 1.3341,
"step": 26760
},
{
"epoch": 0.67,
"grad_norm": 13.892396926879883,
"learning_rate": 2.1898305084745764e-06,
"loss": 1.3268,
"step": 26770
},
{
"epoch": 0.67,
"grad_norm": 12.167963981628418,
"learning_rate": 2.1830508474576273e-06,
"loss": 1.3763,
"step": 26780
},
{
"epoch": 0.67,
"grad_norm": 9.574051856994629,
"learning_rate": 2.1762711864406783e-06,
"loss": 1.5376,
"step": 26790
},
{
"epoch": 0.67,
"grad_norm": 12.77506160736084,
"learning_rate": 2.169491525423729e-06,
"loss": 1.2457,
"step": 26800
},
{
"epoch": 0.67,
"grad_norm": 7.312320232391357,
"learning_rate": 2.16271186440678e-06,
"loss": 1.3077,
"step": 26810
},
{
"epoch": 0.67,
"grad_norm": 10.53618335723877,
"learning_rate": 2.1559322033898304e-06,
"loss": 1.3706,
"step": 26820
},
{
"epoch": 0.67,
"grad_norm": 4.399540424346924,
"learning_rate": 2.1491525423728814e-06,
"loss": 1.3237,
"step": 26830
},
{
"epoch": 0.67,
"grad_norm": 8.757078170776367,
"learning_rate": 2.1423728813559324e-06,
"loss": 1.2385,
"step": 26840
},
{
"epoch": 0.67,
"grad_norm": 3.866237163543701,
"learning_rate": 2.1355932203389833e-06,
"loss": 1.4038,
"step": 26850
},
{
"epoch": 0.67,
"grad_norm": 4.188089370727539,
"learning_rate": 2.128813559322034e-06,
"loss": 1.5465,
"step": 26860
},
{
"epoch": 0.67,
"grad_norm": 10.617339134216309,
"learning_rate": 2.122033898305085e-06,
"loss": 1.1873,
"step": 26870
},
{
"epoch": 0.67,
"grad_norm": 3.4754912853240967,
"learning_rate": 2.115254237288136e-06,
"loss": 1.2539,
"step": 26880
},
{
"epoch": 0.67,
"grad_norm": 6.057491302490234,
"learning_rate": 2.108474576271187e-06,
"loss": 1.2034,
"step": 26890
},
{
"epoch": 0.67,
"grad_norm": 9.13399887084961,
"learning_rate": 2.1016949152542374e-06,
"loss": 1.4333,
"step": 26900
},
{
"epoch": 0.67,
"grad_norm": 14.797123908996582,
"learning_rate": 2.0949152542372883e-06,
"loss": 1.1961,
"step": 26910
},
{
"epoch": 0.67,
"grad_norm": 6.3039398193359375,
"learning_rate": 2.0881355932203393e-06,
"loss": 1.0776,
"step": 26920
},
{
"epoch": 0.67,
"grad_norm": 5.700479507446289,
"learning_rate": 2.08135593220339e-06,
"loss": 1.5808,
"step": 26930
},
{
"epoch": 0.67,
"grad_norm": 17.28398895263672,
"learning_rate": 2.074576271186441e-06,
"loss": 1.4131,
"step": 26940
},
{
"epoch": 0.67,
"grad_norm": 5.036632061004639,
"learning_rate": 2.0677966101694914e-06,
"loss": 1.1761,
"step": 26950
},
{
"epoch": 0.67,
"grad_norm": 4.72507381439209,
"learning_rate": 2.0610169491525424e-06,
"loss": 1.3633,
"step": 26960
},
{
"epoch": 0.67,
"grad_norm": 18.702543258666992,
"learning_rate": 2.0542372881355934e-06,
"loss": 1.3922,
"step": 26970
},
{
"epoch": 0.67,
"grad_norm": 11.218758583068848,
"learning_rate": 2.0474576271186443e-06,
"loss": 1.3427,
"step": 26980
},
{
"epoch": 0.67,
"grad_norm": 5.716217517852783,
"learning_rate": 2.0406779661016953e-06,
"loss": 1.2603,
"step": 26990
},
{
"epoch": 0.68,
"grad_norm": 8.085409164428711,
"learning_rate": 2.033898305084746e-06,
"loss": 1.4003,
"step": 27000
},
{
"epoch": 0.68,
"eval_loss": 1.3196759223937988,
"eval_runtime": 66.1008,
"eval_samples_per_second": 15.128,
"eval_steps_per_second": 15.128,
"step": 27000
},
{
"epoch": 0.68,
"grad_norm": 2.983642578125,
"learning_rate": 2.027118644067797e-06,
"loss": 1.2987,
"step": 27010
},
{
"epoch": 0.68,
"grad_norm": 3.0962276458740234,
"learning_rate": 2.020338983050848e-06,
"loss": 1.3189,
"step": 27020
},
{
"epoch": 0.68,
"grad_norm": 4.857142925262451,
"learning_rate": 2.0135593220338984e-06,
"loss": 1.3651,
"step": 27030
},
{
"epoch": 0.68,
"grad_norm": 4.3937482833862305,
"learning_rate": 2.0067796610169494e-06,
"loss": 1.4407,
"step": 27040
},
{
"epoch": 0.68,
"grad_norm": 4.091692924499512,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.2065,
"step": 27050
},
{
"epoch": 0.68,
"grad_norm": 7.725543975830078,
"learning_rate": 1.993220338983051e-06,
"loss": 1.1836,
"step": 27060
},
{
"epoch": 0.68,
"grad_norm": 5.124769687652588,
"learning_rate": 1.986440677966102e-06,
"loss": 1.3262,
"step": 27070
},
{
"epoch": 0.68,
"grad_norm": 5.097384452819824,
"learning_rate": 1.9796610169491524e-06,
"loss": 1.3057,
"step": 27080
},
{
"epoch": 0.68,
"grad_norm": 3.207469940185547,
"learning_rate": 1.9728813559322034e-06,
"loss": 1.241,
"step": 27090
},
{
"epoch": 0.68,
"grad_norm": 8.80706787109375,
"learning_rate": 1.9661016949152544e-06,
"loss": 1.4066,
"step": 27100
},
{
"epoch": 0.68,
"grad_norm": 2.1298069953918457,
"learning_rate": 1.9593220338983053e-06,
"loss": 1.3394,
"step": 27110
},
{
"epoch": 0.68,
"grad_norm": 9.768424034118652,
"learning_rate": 1.9525423728813563e-06,
"loss": 1.4714,
"step": 27120
},
{
"epoch": 0.68,
"grad_norm": 16.757766723632812,
"learning_rate": 1.945762711864407e-06,
"loss": 1.4322,
"step": 27130
},
{
"epoch": 0.68,
"grad_norm": 4.268533229827881,
"learning_rate": 1.938983050847458e-06,
"loss": 1.3988,
"step": 27140
},
{
"epoch": 0.68,
"grad_norm": 7.853205680847168,
"learning_rate": 1.932203389830509e-06,
"loss": 1.2837,
"step": 27150
},
{
"epoch": 0.68,
"grad_norm": 3.1779301166534424,
"learning_rate": 1.9254237288135594e-06,
"loss": 1.3075,
"step": 27160
},
{
"epoch": 0.68,
"grad_norm": 7.716470718383789,
"learning_rate": 1.9186440677966104e-06,
"loss": 1.2207,
"step": 27170
},
{
"epoch": 0.68,
"grad_norm": 4.0963029861450195,
"learning_rate": 1.9118644067796613e-06,
"loss": 1.3241,
"step": 27180
},
{
"epoch": 0.68,
"grad_norm": 10.319551467895508,
"learning_rate": 1.9050847457627119e-06,
"loss": 1.4133,
"step": 27190
},
{
"epoch": 0.68,
"grad_norm": 4.663591384887695,
"learning_rate": 1.8983050847457629e-06,
"loss": 1.2505,
"step": 27200
},
{
"epoch": 0.68,
"grad_norm": 8.912657737731934,
"learning_rate": 1.8915254237288136e-06,
"loss": 1.2057,
"step": 27210
},
{
"epoch": 0.68,
"grad_norm": 11.967384338378906,
"learning_rate": 1.8847457627118646e-06,
"loss": 1.4043,
"step": 27220
},
{
"epoch": 0.68,
"grad_norm": 19.304821014404297,
"learning_rate": 1.8779661016949156e-06,
"loss": 1.3083,
"step": 27230
},
{
"epoch": 0.68,
"grad_norm": 4.1671600341796875,
"learning_rate": 1.8711864406779661e-06,
"loss": 1.415,
"step": 27240
},
{
"epoch": 0.68,
"grad_norm": 3.430443525314331,
"learning_rate": 1.8644067796610171e-06,
"loss": 1.3154,
"step": 27250
},
{
"epoch": 0.68,
"grad_norm": 8.876851081848145,
"learning_rate": 1.857627118644068e-06,
"loss": 1.2701,
"step": 27260
},
{
"epoch": 0.68,
"grad_norm": 12.077353477478027,
"learning_rate": 1.8508474576271189e-06,
"loss": 1.3744,
"step": 27270
},
{
"epoch": 0.68,
"grad_norm": 1.86147940158844,
"learning_rate": 1.8440677966101696e-06,
"loss": 1.4091,
"step": 27280
},
{
"epoch": 0.68,
"grad_norm": 6.873291969299316,
"learning_rate": 1.8372881355932204e-06,
"loss": 1.3488,
"step": 27290
},
{
"epoch": 0.68,
"grad_norm": 5.088196754455566,
"learning_rate": 1.8305084745762714e-06,
"loss": 1.437,
"step": 27300
},
{
"epoch": 0.68,
"grad_norm": 3.1845881938934326,
"learning_rate": 1.8237288135593223e-06,
"loss": 1.3961,
"step": 27310
},
{
"epoch": 0.68,
"grad_norm": 8.228582382202148,
"learning_rate": 1.816949152542373e-06,
"loss": 1.1022,
"step": 27320
},
{
"epoch": 0.68,
"grad_norm": 15.932997703552246,
"learning_rate": 1.8101694915254239e-06,
"loss": 1.3529,
"step": 27330
},
{
"epoch": 0.68,
"grad_norm": 1.3720171451568604,
"learning_rate": 1.8033898305084746e-06,
"loss": 1.2631,
"step": 27340
},
{
"epoch": 0.68,
"grad_norm": 3.995771646499634,
"learning_rate": 1.7966101694915256e-06,
"loss": 1.2678,
"step": 27350
},
{
"epoch": 0.68,
"grad_norm": 5.7671895027160645,
"learning_rate": 1.7898305084745766e-06,
"loss": 1.329,
"step": 27360
},
{
"epoch": 0.68,
"grad_norm": 3.57239031791687,
"learning_rate": 1.7830508474576271e-06,
"loss": 1.2557,
"step": 27370
},
{
"epoch": 0.68,
"grad_norm": 8.127071380615234,
"learning_rate": 1.7762711864406781e-06,
"loss": 1.3917,
"step": 27380
},
{
"epoch": 0.68,
"grad_norm": 11.008809089660645,
"learning_rate": 1.769491525423729e-06,
"loss": 1.4125,
"step": 27390
},
{
"epoch": 0.69,
"grad_norm": 3.305449962615967,
"learning_rate": 1.7627118644067799e-06,
"loss": 1.4367,
"step": 27400
},
{
"epoch": 0.69,
"grad_norm": 4.471165180206299,
"learning_rate": 1.7559322033898306e-06,
"loss": 1.1336,
"step": 27410
},
{
"epoch": 0.69,
"grad_norm": 4.836517810821533,
"learning_rate": 1.7491525423728814e-06,
"loss": 1.37,
"step": 27420
},
{
"epoch": 0.69,
"grad_norm": 6.102312088012695,
"learning_rate": 1.7423728813559324e-06,
"loss": 1.3035,
"step": 27430
},
{
"epoch": 0.69,
"grad_norm": 4.232480049133301,
"learning_rate": 1.7355932203389834e-06,
"loss": 1.2197,
"step": 27440
},
{
"epoch": 0.69,
"grad_norm": 9.582504272460938,
"learning_rate": 1.728813559322034e-06,
"loss": 1.2699,
"step": 27450
},
{
"epoch": 0.69,
"grad_norm": 9.52001667022705,
"learning_rate": 1.7220338983050849e-06,
"loss": 1.356,
"step": 27460
},
{
"epoch": 0.69,
"grad_norm": 13.017762184143066,
"learning_rate": 1.7152542372881356e-06,
"loss": 1.2246,
"step": 27470
},
{
"epoch": 0.69,
"grad_norm": 7.059675693511963,
"learning_rate": 1.7084745762711866e-06,
"loss": 1.3978,
"step": 27480
},
{
"epoch": 0.69,
"grad_norm": 4.05330753326416,
"learning_rate": 1.7016949152542376e-06,
"loss": 1.3678,
"step": 27490
},
{
"epoch": 0.69,
"grad_norm": 8.122435569763184,
"learning_rate": 1.6949152542372882e-06,
"loss": 1.3062,
"step": 27500
},
{
"epoch": 0.69,
"eval_loss": 1.262895941734314,
"eval_runtime": 66.2138,
"eval_samples_per_second": 15.103,
"eval_steps_per_second": 15.103,
"step": 27500
},
{
"epoch": 0.69,
"grad_norm": 5.283616542816162,
"learning_rate": 1.6881355932203391e-06,
"loss": 1.3504,
"step": 27510
},
{
"epoch": 0.69,
"grad_norm": 3.6427159309387207,
"learning_rate": 1.6813559322033901e-06,
"loss": 1.3028,
"step": 27520
},
{
"epoch": 0.69,
"grad_norm": 5.701483249664307,
"learning_rate": 1.6745762711864409e-06,
"loss": 1.4436,
"step": 27530
},
{
"epoch": 0.69,
"grad_norm": 9.295339584350586,
"learning_rate": 1.6677966101694916e-06,
"loss": 1.4106,
"step": 27540
},
{
"epoch": 0.69,
"grad_norm": 3.1163859367370605,
"learning_rate": 1.6610169491525424e-06,
"loss": 1.2698,
"step": 27550
},
{
"epoch": 0.69,
"grad_norm": 6.673079967498779,
"learning_rate": 1.6542372881355934e-06,
"loss": 1.2202,
"step": 27560
},
{
"epoch": 0.69,
"grad_norm": 4.866211414337158,
"learning_rate": 1.6474576271186444e-06,
"loss": 1.2538,
"step": 27570
},
{
"epoch": 0.69,
"grad_norm": 3.319688081741333,
"learning_rate": 1.640677966101695e-06,
"loss": 1.4981,
"step": 27580
},
{
"epoch": 0.69,
"grad_norm": 5.170316219329834,
"learning_rate": 1.6338983050847459e-06,
"loss": 1.3682,
"step": 27590
},
{
"epoch": 0.69,
"grad_norm": 9.018121719360352,
"learning_rate": 1.6271186440677967e-06,
"loss": 1.3149,
"step": 27600
},
{
"epoch": 0.69,
"grad_norm": 6.896349906921387,
"learning_rate": 1.6203389830508476e-06,
"loss": 1.317,
"step": 27610
},
{
"epoch": 0.69,
"grad_norm": 7.63593053817749,
"learning_rate": 1.6135593220338986e-06,
"loss": 1.3238,
"step": 27620
},
{
"epoch": 0.69,
"grad_norm": 8.823158264160156,
"learning_rate": 1.6067796610169492e-06,
"loss": 1.4479,
"step": 27630
},
{
"epoch": 0.69,
"grad_norm": 8.540557861328125,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.3663,
"step": 27640
},
{
"epoch": 0.69,
"grad_norm": 8.03847885131836,
"learning_rate": 1.593220338983051e-06,
"loss": 1.2248,
"step": 27650
},
{
"epoch": 0.69,
"grad_norm": 5.210377216339111,
"learning_rate": 1.5864406779661019e-06,
"loss": 1.3373,
"step": 27660
},
{
"epoch": 0.69,
"grad_norm": 6.994758129119873,
"learning_rate": 1.5796610169491526e-06,
"loss": 1.3166,
"step": 27670
},
{
"epoch": 0.69,
"grad_norm": 7.343669414520264,
"learning_rate": 1.5728813559322034e-06,
"loss": 1.3406,
"step": 27680
},
{
"epoch": 0.69,
"grad_norm": 12.689948081970215,
"learning_rate": 1.5661016949152544e-06,
"loss": 1.2073,
"step": 27690
},
{
"epoch": 0.69,
"grad_norm": 7.018815040588379,
"learning_rate": 1.5593220338983054e-06,
"loss": 1.2738,
"step": 27700
},
{
"epoch": 0.69,
"grad_norm": 5.334643840789795,
"learning_rate": 1.552542372881356e-06,
"loss": 1.2211,
"step": 27710
},
{
"epoch": 0.69,
"grad_norm": 1.1950844526290894,
"learning_rate": 1.545762711864407e-06,
"loss": 1.268,
"step": 27720
},
{
"epoch": 0.69,
"grad_norm": 11.679058074951172,
"learning_rate": 1.5389830508474577e-06,
"loss": 1.1873,
"step": 27730
},
{
"epoch": 0.69,
"grad_norm": 7.686078071594238,
"learning_rate": 1.5322033898305086e-06,
"loss": 1.277,
"step": 27740
},
{
"epoch": 0.69,
"grad_norm": 2.579845905303955,
"learning_rate": 1.5254237288135596e-06,
"loss": 1.3752,
"step": 27750
},
{
"epoch": 0.69,
"grad_norm": 11.227088928222656,
"learning_rate": 1.5186440677966102e-06,
"loss": 1.2593,
"step": 27760
},
{
"epoch": 0.69,
"grad_norm": 6.903045654296875,
"learning_rate": 1.5118644067796611e-06,
"loss": 1.2463,
"step": 27770
},
{
"epoch": 0.69,
"grad_norm": 12.534771919250488,
"learning_rate": 1.505084745762712e-06,
"loss": 1.2949,
"step": 27780
},
{
"epoch": 0.69,
"grad_norm": 4.901615619659424,
"learning_rate": 1.4983050847457629e-06,
"loss": 1.2975,
"step": 27790
},
{
"epoch": 0.69,
"grad_norm": 9.669910430908203,
"learning_rate": 1.4915254237288139e-06,
"loss": 1.3476,
"step": 27800
},
{
"epoch": 0.7,
"grad_norm": 6.982202529907227,
"learning_rate": 1.4847457627118644e-06,
"loss": 1.3839,
"step": 27810
},
{
"epoch": 0.7,
"grad_norm": 9.172724723815918,
"learning_rate": 1.4779661016949154e-06,
"loss": 1.3519,
"step": 27820
},
{
"epoch": 0.7,
"grad_norm": 8.561583518981934,
"learning_rate": 1.4711864406779664e-06,
"loss": 1.3402,
"step": 27830
},
{
"epoch": 0.7,
"grad_norm": 11.40105152130127,
"learning_rate": 1.464406779661017e-06,
"loss": 1.2805,
"step": 27840
},
{
"epoch": 0.7,
"grad_norm": 11.502050399780273,
"learning_rate": 1.457627118644068e-06,
"loss": 1.4351,
"step": 27850
},
{
"epoch": 0.7,
"grad_norm": 3.5157549381256104,
"learning_rate": 1.4508474576271187e-06,
"loss": 1.2948,
"step": 27860
},
{
"epoch": 0.7,
"grad_norm": 8.663883209228516,
"learning_rate": 1.4440677966101696e-06,
"loss": 1.3285,
"step": 27870
},
{
"epoch": 0.7,
"grad_norm": 4.502950191497803,
"learning_rate": 1.4372881355932206e-06,
"loss": 1.4507,
"step": 27880
},
{
"epoch": 0.7,
"grad_norm": 2.098318338394165,
"learning_rate": 1.4305084745762712e-06,
"loss": 1.1753,
"step": 27890
},
{
"epoch": 0.7,
"grad_norm": 2.721998453140259,
"learning_rate": 1.4237288135593222e-06,
"loss": 1.3162,
"step": 27900
},
{
"epoch": 0.7,
"grad_norm": 6.501703262329102,
"learning_rate": 1.416949152542373e-06,
"loss": 1.2442,
"step": 27910
},
{
"epoch": 0.7,
"grad_norm": 4.597460746765137,
"learning_rate": 1.410169491525424e-06,
"loss": 1.2659,
"step": 27920
},
{
"epoch": 0.7,
"grad_norm": 1.399740219116211,
"learning_rate": 1.4033898305084749e-06,
"loss": 1.2234,
"step": 27930
},
{
"epoch": 0.7,
"grad_norm": 1.2826783657073975,
"learning_rate": 1.3966101694915254e-06,
"loss": 1.3775,
"step": 27940
},
{
"epoch": 0.7,
"grad_norm": 9.264411926269531,
"learning_rate": 1.3898305084745764e-06,
"loss": 1.299,
"step": 27950
},
{
"epoch": 0.7,
"grad_norm": 11.41451644897461,
"learning_rate": 1.3830508474576274e-06,
"loss": 1.2851,
"step": 27960
},
{
"epoch": 0.7,
"grad_norm": 4.351644515991211,
"learning_rate": 1.376271186440678e-06,
"loss": 1.1979,
"step": 27970
},
{
"epoch": 0.7,
"grad_norm": 9.026026725769043,
"learning_rate": 1.369491525423729e-06,
"loss": 1.3119,
"step": 27980
},
{
"epoch": 0.7,
"grad_norm": 6.000504970550537,
"learning_rate": 1.3627118644067797e-06,
"loss": 1.2543,
"step": 27990
},
{
"epoch": 0.7,
"grad_norm": 22.265581130981445,
"learning_rate": 1.3559322033898307e-06,
"loss": 1.34,
"step": 28000
},
{
"epoch": 0.7,
"eval_loss": 1.3058114051818848,
"eval_runtime": 66.2214,
"eval_samples_per_second": 15.101,
"eval_steps_per_second": 15.101,
"step": 28000
},
{
"epoch": 0.7,
"grad_norm": 4.122670650482178,
"learning_rate": 1.3491525423728816e-06,
"loss": 1.2058,
"step": 28010
},
{
"epoch": 0.7,
"grad_norm": 3.766960859298706,
"learning_rate": 1.3423728813559322e-06,
"loss": 1.3474,
"step": 28020
},
{
"epoch": 0.7,
"grad_norm": 2.3636386394500732,
"learning_rate": 1.3355932203389832e-06,
"loss": 1.1637,
"step": 28030
},
{
"epoch": 0.7,
"grad_norm": 9.471797943115234,
"learning_rate": 1.328813559322034e-06,
"loss": 1.3953,
"step": 28040
},
{
"epoch": 0.7,
"grad_norm": 8.232218742370605,
"learning_rate": 1.322033898305085e-06,
"loss": 1.4269,
"step": 28050
},
{
"epoch": 0.7,
"grad_norm": 1.1584389209747314,
"learning_rate": 1.3152542372881359e-06,
"loss": 1.3231,
"step": 28060
},
{
"epoch": 0.7,
"grad_norm": 4.911566257476807,
"learning_rate": 1.3084745762711864e-06,
"loss": 1.241,
"step": 28070
},
{
"epoch": 0.7,
"grad_norm": 18.741342544555664,
"learning_rate": 1.3016949152542374e-06,
"loss": 1.3838,
"step": 28080
},
{
"epoch": 0.7,
"grad_norm": 8.622126579284668,
"learning_rate": 1.2949152542372884e-06,
"loss": 1.2792,
"step": 28090
},
{
"epoch": 0.7,
"grad_norm": 4.402095794677734,
"learning_rate": 1.288135593220339e-06,
"loss": 1.2308,
"step": 28100
},
{
"epoch": 0.7,
"grad_norm": 5.238518714904785,
"learning_rate": 1.28135593220339e-06,
"loss": 1.3537,
"step": 28110
},
{
"epoch": 0.7,
"grad_norm": 5.757815837860107,
"learning_rate": 1.2745762711864407e-06,
"loss": 1.3987,
"step": 28120
},
{
"epoch": 0.7,
"grad_norm": 9.999316215515137,
"learning_rate": 1.2677966101694917e-06,
"loss": 1.1397,
"step": 28130
},
{
"epoch": 0.7,
"grad_norm": 6.276950359344482,
"learning_rate": 1.2610169491525426e-06,
"loss": 1.5024,
"step": 28140
},
{
"epoch": 0.7,
"grad_norm": 5.9979634284973145,
"learning_rate": 1.2542372881355932e-06,
"loss": 1.4296,
"step": 28150
},
{
"epoch": 0.7,
"grad_norm": 7.124503135681152,
"learning_rate": 1.2474576271186442e-06,
"loss": 1.279,
"step": 28160
},
{
"epoch": 0.7,
"grad_norm": 10.153301239013672,
"learning_rate": 1.240677966101695e-06,
"loss": 1.3013,
"step": 28170
},
{
"epoch": 0.7,
"grad_norm": 6.391571521759033,
"learning_rate": 1.233898305084746e-06,
"loss": 1.2782,
"step": 28180
},
{
"epoch": 0.7,
"grad_norm": 10.13975715637207,
"learning_rate": 1.2271186440677967e-06,
"loss": 1.4136,
"step": 28190
},
{
"epoch": 0.7,
"grad_norm": 5.052265167236328,
"learning_rate": 1.2203389830508477e-06,
"loss": 1.3246,
"step": 28200
},
{
"epoch": 0.71,
"grad_norm": 6.8834638595581055,
"learning_rate": 1.2135593220338984e-06,
"loss": 1.3436,
"step": 28210
},
{
"epoch": 0.71,
"grad_norm": 10.564448356628418,
"learning_rate": 1.2067796610169492e-06,
"loss": 1.1619,
"step": 28220
},
{
"epoch": 0.71,
"grad_norm": 3.8192801475524902,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.2328,
"step": 28230
},
{
"epoch": 0.71,
"grad_norm": 5.635645866394043,
"learning_rate": 1.193220338983051e-06,
"loss": 1.3258,
"step": 28240
},
{
"epoch": 0.71,
"grad_norm": 6.429792881011963,
"learning_rate": 1.186440677966102e-06,
"loss": 1.3368,
"step": 28250
},
{
"epoch": 0.71,
"grad_norm": 11.116402626037598,
"learning_rate": 1.1796610169491527e-06,
"loss": 1.2527,
"step": 28260
},
{
"epoch": 0.71,
"grad_norm": 2.7472074031829834,
"learning_rate": 1.1728813559322034e-06,
"loss": 1.4633,
"step": 28270
},
{
"epoch": 0.71,
"grad_norm": 4.118687629699707,
"learning_rate": 1.1661016949152542e-06,
"loss": 1.4852,
"step": 28280
},
{
"epoch": 0.71,
"grad_norm": 14.502837181091309,
"learning_rate": 1.1593220338983052e-06,
"loss": 1.1964,
"step": 28290
},
{
"epoch": 0.71,
"grad_norm": 3.8546032905578613,
"learning_rate": 1.152542372881356e-06,
"loss": 1.1832,
"step": 28300
},
{
"epoch": 0.71,
"grad_norm": 2.377305507659912,
"learning_rate": 1.145762711864407e-06,
"loss": 1.3164,
"step": 28310
},
{
"epoch": 0.71,
"grad_norm": 2.172879219055176,
"learning_rate": 1.1389830508474577e-06,
"loss": 1.3163,
"step": 28320
},
{
"epoch": 0.71,
"grad_norm": 4.290719032287598,
"learning_rate": 1.1322033898305087e-06,
"loss": 1.2505,
"step": 28330
},
{
"epoch": 0.71,
"grad_norm": 15.109819412231445,
"learning_rate": 1.1254237288135594e-06,
"loss": 1.2868,
"step": 28340
},
{
"epoch": 0.71,
"grad_norm": 11.07304859161377,
"learning_rate": 1.1186440677966102e-06,
"loss": 1.2118,
"step": 28350
},
{
"epoch": 0.71,
"grad_norm": 14.64116096496582,
"learning_rate": 1.1118644067796612e-06,
"loss": 1.2904,
"step": 28360
},
{
"epoch": 0.71,
"grad_norm": 12.49071216583252,
"learning_rate": 1.105084745762712e-06,
"loss": 1.286,
"step": 28370
},
{
"epoch": 0.71,
"grad_norm": 3.750757932662964,
"learning_rate": 1.098305084745763e-06,
"loss": 1.2723,
"step": 28380
},
{
"epoch": 0.71,
"grad_norm": 11.984790802001953,
"learning_rate": 1.0915254237288137e-06,
"loss": 1.1732,
"step": 28390
},
{
"epoch": 0.71,
"grad_norm": 13.987404823303223,
"learning_rate": 1.0847457627118644e-06,
"loss": 1.2168,
"step": 28400
},
{
"epoch": 0.71,
"grad_norm": 10.837672233581543,
"learning_rate": 1.0779661016949152e-06,
"loss": 1.4036,
"step": 28410
},
{
"epoch": 0.71,
"grad_norm": 12.291699409484863,
"learning_rate": 1.0711864406779662e-06,
"loss": 1.1172,
"step": 28420
},
{
"epoch": 0.71,
"grad_norm": 2.2296297550201416,
"learning_rate": 1.064406779661017e-06,
"loss": 1.3241,
"step": 28430
},
{
"epoch": 0.71,
"grad_norm": 6.434119701385498,
"learning_rate": 1.057627118644068e-06,
"loss": 1.3096,
"step": 28440
},
{
"epoch": 0.71,
"grad_norm": 3.2857985496520996,
"learning_rate": 1.0508474576271187e-06,
"loss": 1.3744,
"step": 28450
},
{
"epoch": 0.71,
"grad_norm": 6.822338104248047,
"learning_rate": 1.0440677966101697e-06,
"loss": 1.3724,
"step": 28460
},
{
"epoch": 0.71,
"grad_norm": 6.319363117218018,
"learning_rate": 1.0372881355932204e-06,
"loss": 1.2181,
"step": 28470
},
{
"epoch": 0.71,
"grad_norm": 3.777921676635742,
"learning_rate": 1.0305084745762712e-06,
"loss": 1.3495,
"step": 28480
},
{
"epoch": 0.71,
"grad_norm": 6.540971755981445,
"learning_rate": 1.0237288135593222e-06,
"loss": 1.4589,
"step": 28490
},
{
"epoch": 0.71,
"grad_norm": 7.9154744148254395,
"learning_rate": 1.016949152542373e-06,
"loss": 1.3015,
"step": 28500
},
{
"epoch": 0.71,
"eval_loss": 1.3162422180175781,
"eval_runtime": 66.2512,
"eval_samples_per_second": 15.094,
"eval_steps_per_second": 15.094,
"step": 28500
},
{
"epoch": 0.71,
"grad_norm": 15.148056983947754,
"learning_rate": 1.010169491525424e-06,
"loss": 1.2363,
"step": 28510
},
{
"epoch": 0.71,
"grad_norm": 10.743448257446289,
"learning_rate": 1.0033898305084747e-06,
"loss": 1.1965,
"step": 28520
},
{
"epoch": 0.71,
"grad_norm": 12.191396713256836,
"learning_rate": 9.966101694915254e-07,
"loss": 1.4387,
"step": 28530
},
{
"epoch": 0.71,
"grad_norm": 4.78171968460083,
"learning_rate": 9.898305084745762e-07,
"loss": 1.4977,
"step": 28540
},
{
"epoch": 0.71,
"grad_norm": 5.24019718170166,
"learning_rate": 9.830508474576272e-07,
"loss": 1.5177,
"step": 28550
},
{
"epoch": 0.71,
"grad_norm": 12.000500679016113,
"learning_rate": 9.762711864406782e-07,
"loss": 1.1637,
"step": 28560
},
{
"epoch": 0.71,
"grad_norm": 10.612434387207031,
"learning_rate": 9.69491525423729e-07,
"loss": 1.334,
"step": 28570
},
{
"epoch": 0.71,
"grad_norm": 6.343203544616699,
"learning_rate": 9.627118644067797e-07,
"loss": 1.3152,
"step": 28580
},
{
"epoch": 0.71,
"grad_norm": 5.986273288726807,
"learning_rate": 9.559322033898307e-07,
"loss": 1.3724,
"step": 28590
},
{
"epoch": 0.71,
"grad_norm": 12.447896003723145,
"learning_rate": 9.491525423728814e-07,
"loss": 1.2404,
"step": 28600
},
{
"epoch": 0.72,
"grad_norm": 18.211698532104492,
"learning_rate": 9.423728813559323e-07,
"loss": 1.3091,
"step": 28610
},
{
"epoch": 0.72,
"grad_norm": 5.227106094360352,
"learning_rate": 9.355932203389831e-07,
"loss": 1.1275,
"step": 28620
},
{
"epoch": 0.72,
"grad_norm": 4.874502182006836,
"learning_rate": 9.28813559322034e-07,
"loss": 1.2959,
"step": 28630
},
{
"epoch": 0.72,
"grad_norm": 7.699239253997803,
"learning_rate": 9.220338983050848e-07,
"loss": 1.3687,
"step": 28640
},
{
"epoch": 0.72,
"grad_norm": 7.712405681610107,
"learning_rate": 9.152542372881357e-07,
"loss": 1.3746,
"step": 28650
},
{
"epoch": 0.72,
"grad_norm": 8.854902267456055,
"learning_rate": 9.084745762711864e-07,
"loss": 1.3399,
"step": 28660
},
{
"epoch": 0.72,
"grad_norm": 5.26395320892334,
"learning_rate": 9.016949152542373e-07,
"loss": 1.2296,
"step": 28670
},
{
"epoch": 0.72,
"grad_norm": 5.966766357421875,
"learning_rate": 8.949152542372883e-07,
"loss": 1.2806,
"step": 28680
},
{
"epoch": 0.72,
"grad_norm": 5.988807201385498,
"learning_rate": 8.881355932203391e-07,
"loss": 1.4652,
"step": 28690
},
{
"epoch": 0.72,
"grad_norm": 1.523200273513794,
"learning_rate": 8.813559322033899e-07,
"loss": 1.4848,
"step": 28700
},
{
"epoch": 0.72,
"grad_norm": 3.4972944259643555,
"learning_rate": 8.745762711864407e-07,
"loss": 1.296,
"step": 28710
},
{
"epoch": 0.72,
"grad_norm": 4.062095642089844,
"learning_rate": 8.677966101694917e-07,
"loss": 1.3716,
"step": 28720
},
{
"epoch": 0.72,
"grad_norm": 1.065338134765625,
"learning_rate": 8.610169491525424e-07,
"loss": 1.2375,
"step": 28730
},
{
"epoch": 0.72,
"grad_norm": 6.1202006340026855,
"learning_rate": 8.542372881355933e-07,
"loss": 1.3413,
"step": 28740
},
{
"epoch": 0.72,
"grad_norm": 17.116790771484375,
"learning_rate": 8.474576271186441e-07,
"loss": 1.2014,
"step": 28750
},
{
"epoch": 0.72,
"grad_norm": 6.310455799102783,
"learning_rate": 8.406779661016951e-07,
"loss": 1.2509,
"step": 28760
},
{
"epoch": 0.72,
"grad_norm": 19.508827209472656,
"learning_rate": 8.338983050847458e-07,
"loss": 1.4035,
"step": 28770
},
{
"epoch": 0.72,
"grad_norm": 10.587761878967285,
"learning_rate": 8.271186440677967e-07,
"loss": 1.3374,
"step": 28780
},
{
"epoch": 0.72,
"grad_norm": 15.715071678161621,
"learning_rate": 8.203389830508475e-07,
"loss": 1.3229,
"step": 28790
},
{
"epoch": 0.72,
"grad_norm": 3.0983753204345703,
"learning_rate": 8.135593220338983e-07,
"loss": 1.4297,
"step": 28800
},
{
"epoch": 0.72,
"grad_norm": 4.541349411010742,
"learning_rate": 8.067796610169493e-07,
"loss": 1.3327,
"step": 28810
},
{
"epoch": 0.72,
"grad_norm": 4.102433204650879,
"learning_rate": 8.000000000000001e-07,
"loss": 1.3362,
"step": 28820
},
{
"epoch": 0.72,
"grad_norm": 6.362680435180664,
"learning_rate": 7.932203389830509e-07,
"loss": 1.2275,
"step": 28830
},
{
"epoch": 0.72,
"grad_norm": 9.411408424377441,
"learning_rate": 7.864406779661017e-07,
"loss": 1.4023,
"step": 28840
},
{
"epoch": 0.72,
"grad_norm": 6.113147258758545,
"learning_rate": 7.796610169491527e-07,
"loss": 1.2319,
"step": 28850
},
{
"epoch": 0.72,
"grad_norm": 6.580341815948486,
"learning_rate": 7.728813559322034e-07,
"loss": 1.3559,
"step": 28860
},
{
"epoch": 0.72,
"grad_norm": 10.778473854064941,
"learning_rate": 7.661016949152543e-07,
"loss": 1.2247,
"step": 28870
},
{
"epoch": 0.72,
"grad_norm": 4.692105770111084,
"learning_rate": 7.593220338983051e-07,
"loss": 1.2254,
"step": 28880
},
{
"epoch": 0.72,
"grad_norm": 8.076458930969238,
"learning_rate": 7.52542372881356e-07,
"loss": 1.3264,
"step": 28890
},
{
"epoch": 0.72,
"grad_norm": 11.618599891662598,
"learning_rate": 7.457627118644069e-07,
"loss": 1.174,
"step": 28900
},
{
"epoch": 0.72,
"grad_norm": 10.995185852050781,
"learning_rate": 7.389830508474577e-07,
"loss": 1.21,
"step": 28910
},
{
"epoch": 0.72,
"grad_norm": 12.011213302612305,
"learning_rate": 7.322033898305085e-07,
"loss": 1.3272,
"step": 28920
},
{
"epoch": 0.72,
"grad_norm": 4.367415428161621,
"learning_rate": 7.254237288135593e-07,
"loss": 1.3997,
"step": 28930
},
{
"epoch": 0.72,
"grad_norm": 11.401775360107422,
"learning_rate": 7.186440677966103e-07,
"loss": 1.3507,
"step": 28940
},
{
"epoch": 0.72,
"grad_norm": 8.856404304504395,
"learning_rate": 7.118644067796611e-07,
"loss": 1.3274,
"step": 28950
},
{
"epoch": 0.72,
"grad_norm": 5.873079776763916,
"learning_rate": 7.05084745762712e-07,
"loss": 1.2987,
"step": 28960
},
{
"epoch": 0.72,
"grad_norm": 7.142796993255615,
"learning_rate": 6.983050847457627e-07,
"loss": 1.2049,
"step": 28970
},
{
"epoch": 0.72,
"grad_norm": 5.9704179763793945,
"learning_rate": 6.915254237288137e-07,
"loss": 1.3481,
"step": 28980
},
{
"epoch": 0.72,
"grad_norm": 10.48417854309082,
"learning_rate": 6.847457627118645e-07,
"loss": 1.3498,
"step": 28990
},
{
"epoch": 0.72,
"grad_norm": 12.131574630737305,
"learning_rate": 6.779661016949153e-07,
"loss": 1.3529,
"step": 29000
},
{
"epoch": 0.72,
"eval_loss": 1.3368879556655884,
"eval_runtime": 66.2441,
"eval_samples_per_second": 15.096,
"eval_steps_per_second": 15.096,
"step": 29000
},
{
"epoch": 0.73,
"grad_norm": 7.354034423828125,
"learning_rate": 6.711864406779661e-07,
"loss": 1.2698,
"step": 29010
},
{
"epoch": 0.73,
"grad_norm": 3.9224977493286133,
"learning_rate": 6.64406779661017e-07,
"loss": 1.3377,
"step": 29020
},
{
"epoch": 0.73,
"grad_norm": 8.08880615234375,
"learning_rate": 6.576271186440679e-07,
"loss": 1.4448,
"step": 29030
},
{
"epoch": 0.73,
"grad_norm": 3.625216245651245,
"learning_rate": 6.508474576271187e-07,
"loss": 1.3561,
"step": 29040
},
{
"epoch": 0.73,
"grad_norm": 13.905680656433105,
"learning_rate": 6.440677966101695e-07,
"loss": 1.2956,
"step": 29050
},
{
"epoch": 0.73,
"grad_norm": 3.773040294647217,
"learning_rate": 6.372881355932203e-07,
"loss": 1.2934,
"step": 29060
},
{
"epoch": 0.73,
"grad_norm": 5.38674259185791,
"learning_rate": 6.305084745762713e-07,
"loss": 1.3364,
"step": 29070
},
{
"epoch": 0.73,
"grad_norm": 4.701632022857666,
"learning_rate": 6.237288135593221e-07,
"loss": 1.1754,
"step": 29080
},
{
"epoch": 0.73,
"grad_norm": 8.243915557861328,
"learning_rate": 6.16949152542373e-07,
"loss": 1.4113,
"step": 29090
},
{
"epoch": 0.73,
"grad_norm": 4.2352213859558105,
"learning_rate": 6.101694915254238e-07,
"loss": 1.4673,
"step": 29100
},
{
"epoch": 0.73,
"grad_norm": 19.758262634277344,
"learning_rate": 6.033898305084746e-07,
"loss": 1.0756,
"step": 29110
},
{
"epoch": 0.73,
"grad_norm": 9.024798393249512,
"learning_rate": 5.966101694915255e-07,
"loss": 1.1835,
"step": 29120
},
{
"epoch": 0.73,
"grad_norm": 2.7573533058166504,
"learning_rate": 5.898305084745763e-07,
"loss": 1.3342,
"step": 29130
},
{
"epoch": 0.73,
"grad_norm": 2.6472461223602295,
"learning_rate": 5.830508474576271e-07,
"loss": 1.478,
"step": 29140
},
{
"epoch": 0.73,
"grad_norm": 8.194090843200684,
"learning_rate": 5.76271186440678e-07,
"loss": 1.2781,
"step": 29150
},
{
"epoch": 0.73,
"grad_norm": 7.336125373840332,
"learning_rate": 5.694915254237288e-07,
"loss": 1.3871,
"step": 29160
},
{
"epoch": 0.73,
"grad_norm": 9.800869941711426,
"learning_rate": 5.627118644067797e-07,
"loss": 1.237,
"step": 29170
},
{
"epoch": 0.73,
"grad_norm": 3.7578229904174805,
"learning_rate": 5.559322033898306e-07,
"loss": 1.3408,
"step": 29180
},
{
"epoch": 0.73,
"grad_norm": 11.04594612121582,
"learning_rate": 5.491525423728815e-07,
"loss": 1.4338,
"step": 29190
},
{
"epoch": 0.73,
"grad_norm": 9.137775421142578,
"learning_rate": 5.423728813559322e-07,
"loss": 1.2292,
"step": 29200
},
{
"epoch": 0.73,
"grad_norm": 6.266888618469238,
"learning_rate": 5.355932203389831e-07,
"loss": 1.3781,
"step": 29210
},
{
"epoch": 0.73,
"grad_norm": 10.63198184967041,
"learning_rate": 5.28813559322034e-07,
"loss": 1.2394,
"step": 29220
},
{
"epoch": 0.73,
"grad_norm": 16.352039337158203,
"learning_rate": 5.220338983050848e-07,
"loss": 1.3784,
"step": 29230
},
{
"epoch": 0.73,
"grad_norm": 9.548080444335938,
"learning_rate": 5.152542372881356e-07,
"loss": 1.3201,
"step": 29240
},
{
"epoch": 0.73,
"grad_norm": 2.56595516204834,
"learning_rate": 5.084745762711865e-07,
"loss": 1.2335,
"step": 29250
},
{
"epoch": 0.73,
"grad_norm": 3.092132568359375,
"learning_rate": 5.016949152542373e-07,
"loss": 1.2999,
"step": 29260
},
{
"epoch": 0.73,
"grad_norm": 12.341950416564941,
"learning_rate": 4.949152542372881e-07,
"loss": 1.2719,
"step": 29270
},
{
"epoch": 0.73,
"grad_norm": 16.05048179626465,
"learning_rate": 4.881355932203391e-07,
"loss": 1.1151,
"step": 29280
},
{
"epoch": 0.73,
"grad_norm": 8.888206481933594,
"learning_rate": 4.813559322033898e-07,
"loss": 1.0549,
"step": 29290
},
{
"epoch": 0.73,
"grad_norm": 16.083812713623047,
"learning_rate": 4.745762711864407e-07,
"loss": 1.2443,
"step": 29300
},
{
"epoch": 0.73,
"grad_norm": 5.3232741355896,
"learning_rate": 4.6779661016949154e-07,
"loss": 1.2953,
"step": 29310
},
{
"epoch": 0.73,
"grad_norm": 15.959102630615234,
"learning_rate": 4.610169491525424e-07,
"loss": 1.282,
"step": 29320
},
{
"epoch": 0.73,
"grad_norm": 8.539166450500488,
"learning_rate": 4.542372881355932e-07,
"loss": 1.2538,
"step": 29330
},
{
"epoch": 0.73,
"grad_norm": 4.907639980316162,
"learning_rate": 4.4745762711864415e-07,
"loss": 1.3666,
"step": 29340
},
{
"epoch": 0.73,
"grad_norm": 2.455517530441284,
"learning_rate": 4.4067796610169497e-07,
"loss": 1.2269,
"step": 29350
},
{
"epoch": 0.73,
"grad_norm": 9.032910346984863,
"learning_rate": 4.3389830508474584e-07,
"loss": 1.2813,
"step": 29360
},
{
"epoch": 0.73,
"grad_norm": 2.8847286701202393,
"learning_rate": 4.2711864406779666e-07,
"loss": 1.3336,
"step": 29370
},
{
"epoch": 0.73,
"grad_norm": 11.284420013427734,
"learning_rate": 4.2033898305084753e-07,
"loss": 1.286,
"step": 29380
},
{
"epoch": 0.73,
"grad_norm": 11.412368774414062,
"learning_rate": 4.1355932203389835e-07,
"loss": 1.3317,
"step": 29390
},
{
"epoch": 0.73,
"grad_norm": 4.6027302742004395,
"learning_rate": 4.0677966101694916e-07,
"loss": 1.3263,
"step": 29400
},
{
"epoch": 0.74,
"grad_norm": 2.39072322845459,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.3174,
"step": 29410
},
{
"epoch": 0.74,
"grad_norm": 2.7101926803588867,
"learning_rate": 3.9322033898305085e-07,
"loss": 1.2673,
"step": 29420
},
{
"epoch": 0.74,
"grad_norm": 10.152178764343262,
"learning_rate": 3.864406779661017e-07,
"loss": 1.3241,
"step": 29430
},
{
"epoch": 0.74,
"grad_norm": 4.905571937561035,
"learning_rate": 3.7966101694915254e-07,
"loss": 1.2261,
"step": 29440
},
{
"epoch": 0.74,
"grad_norm": 10.35471248626709,
"learning_rate": 3.7288135593220347e-07,
"loss": 1.2083,
"step": 29450
},
{
"epoch": 0.74,
"grad_norm": 11.7713623046875,
"learning_rate": 3.6610169491525423e-07,
"loss": 1.2647,
"step": 29460
},
{
"epoch": 0.74,
"grad_norm": 3.181910276412964,
"learning_rate": 3.5932203389830516e-07,
"loss": 1.4678,
"step": 29470
},
{
"epoch": 0.74,
"grad_norm": 6.756598472595215,
"learning_rate": 3.52542372881356e-07,
"loss": 1.273,
"step": 29480
},
{
"epoch": 0.74,
"grad_norm": 12.660133361816406,
"learning_rate": 3.4576271186440684e-07,
"loss": 1.3497,
"step": 29490
},
{
"epoch": 0.74,
"grad_norm": 11.418837547302246,
"learning_rate": 3.3898305084745766e-07,
"loss": 1.3025,
"step": 29500
},
{
"epoch": 0.74,
"eval_loss": 1.2983591556549072,
"eval_runtime": 66.2327,
"eval_samples_per_second": 15.098,
"eval_steps_per_second": 15.098,
"step": 29500
},
{
"epoch": 0.74,
"grad_norm": 5.526192665100098,
"learning_rate": 3.322033898305085e-07,
"loss": 1.1364,
"step": 29510
},
{
"epoch": 0.74,
"grad_norm": 4.229588031768799,
"learning_rate": 3.2542372881355935e-07,
"loss": 1.3351,
"step": 29520
},
{
"epoch": 0.74,
"grad_norm": 3.8703114986419678,
"learning_rate": 3.1864406779661017e-07,
"loss": 1.547,
"step": 29530
},
{
"epoch": 0.74,
"grad_norm": 4.0574445724487305,
"learning_rate": 3.1186440677966104e-07,
"loss": 1.5269,
"step": 29540
},
{
"epoch": 0.74,
"grad_norm": 2.609175443649292,
"learning_rate": 3.050847457627119e-07,
"loss": 1.3574,
"step": 29550
},
{
"epoch": 0.74,
"grad_norm": 5.445046901702881,
"learning_rate": 2.9830508474576273e-07,
"loss": 1.3081,
"step": 29560
},
{
"epoch": 0.74,
"grad_norm": 8.432710647583008,
"learning_rate": 2.9152542372881355e-07,
"loss": 1.2343,
"step": 29570
},
{
"epoch": 0.74,
"grad_norm": 10.050646781921387,
"learning_rate": 2.847457627118644e-07,
"loss": 1.2665,
"step": 29580
},
{
"epoch": 0.74,
"grad_norm": 6.895621299743652,
"learning_rate": 2.779661016949153e-07,
"loss": 1.1077,
"step": 29590
},
{
"epoch": 0.74,
"grad_norm": 6.045702934265137,
"learning_rate": 2.711864406779661e-07,
"loss": 1.0306,
"step": 29600
},
{
"epoch": 0.74,
"grad_norm": 2.18096661567688,
"learning_rate": 2.64406779661017e-07,
"loss": 1.167,
"step": 29610
},
{
"epoch": 0.74,
"grad_norm": 8.721917152404785,
"learning_rate": 2.576271186440678e-07,
"loss": 1.2476,
"step": 29620
},
{
"epoch": 0.74,
"grad_norm": 5.546863079071045,
"learning_rate": 2.5084745762711867e-07,
"loss": 1.3273,
"step": 29630
},
{
"epoch": 0.74,
"grad_norm": 7.264005661010742,
"learning_rate": 2.4406779661016954e-07,
"loss": 1.4333,
"step": 29640
},
{
"epoch": 0.74,
"grad_norm": 4.217822551727295,
"learning_rate": 2.3728813559322036e-07,
"loss": 1.5513,
"step": 29650
},
{
"epoch": 0.74,
"grad_norm": 21.425151824951172,
"learning_rate": 2.305084745762712e-07,
"loss": 1.15,
"step": 29660
},
{
"epoch": 0.74,
"grad_norm": 17.36193084716797,
"learning_rate": 2.2372881355932207e-07,
"loss": 1.4419,
"step": 29670
},
{
"epoch": 0.74,
"grad_norm": 9.74802017211914,
"learning_rate": 2.1694915254237292e-07,
"loss": 1.2543,
"step": 29680
},
{
"epoch": 0.74,
"grad_norm": 12.033125877380371,
"learning_rate": 2.1016949152542376e-07,
"loss": 1.378,
"step": 29690
},
{
"epoch": 0.74,
"grad_norm": 14.735278129577637,
"learning_rate": 2.0338983050847458e-07,
"loss": 1.2712,
"step": 29700
},
{
"epoch": 0.74,
"grad_norm": 11.541913986206055,
"learning_rate": 1.9661016949152543e-07,
"loss": 1.0699,
"step": 29710
},
{
"epoch": 0.74,
"grad_norm": 3.0062718391418457,
"learning_rate": 1.8983050847457627e-07,
"loss": 1.3865,
"step": 29720
},
{
"epoch": 0.74,
"grad_norm": 7.821793079376221,
"learning_rate": 1.8305084745762712e-07,
"loss": 1.1679,
"step": 29730
},
{
"epoch": 0.74,
"grad_norm": 12.209012031555176,
"learning_rate": 1.76271186440678e-07,
"loss": 1.1567,
"step": 29740
},
{
"epoch": 0.74,
"grad_norm": 3.285269021987915,
"learning_rate": 1.6949152542372883e-07,
"loss": 1.3667,
"step": 29750
},
{
"epoch": 0.74,
"grad_norm": 5.812708377838135,
"learning_rate": 1.6271186440677968e-07,
"loss": 1.2521,
"step": 29760
},
{
"epoch": 0.74,
"grad_norm": 3.4106059074401855,
"learning_rate": 1.5593220338983052e-07,
"loss": 1.3847,
"step": 29770
},
{
"epoch": 0.74,
"grad_norm": 9.550897598266602,
"learning_rate": 1.4915254237288137e-07,
"loss": 1.2853,
"step": 29780
},
{
"epoch": 0.74,
"grad_norm": 13.644095420837402,
"learning_rate": 1.423728813559322e-07,
"loss": 1.3305,
"step": 29790
},
{
"epoch": 0.74,
"grad_norm": 6.250187397003174,
"learning_rate": 1.3559322033898305e-07,
"loss": 1.2896,
"step": 29800
},
{
"epoch": 0.75,
"grad_norm": 11.07441234588623,
"learning_rate": 1.288135593220339e-07,
"loss": 1.2263,
"step": 29810
},
{
"epoch": 0.75,
"grad_norm": 13.597912788391113,
"learning_rate": 1.2203389830508477e-07,
"loss": 1.4195,
"step": 29820
},
{
"epoch": 0.75,
"grad_norm": 4.825064659118652,
"learning_rate": 1.152542372881356e-07,
"loss": 1.3549,
"step": 29830
},
{
"epoch": 0.75,
"grad_norm": 3.6931655406951904,
"learning_rate": 1.0847457627118646e-07,
"loss": 1.3675,
"step": 29840
},
{
"epoch": 0.75,
"grad_norm": 5.718669414520264,
"learning_rate": 1.0169491525423729e-07,
"loss": 1.4715,
"step": 29850
},
{
"epoch": 0.75,
"grad_norm": 6.913152694702148,
"learning_rate": 9.491525423728814e-08,
"loss": 1.4853,
"step": 29860
},
{
"epoch": 0.75,
"grad_norm": 5.6454644203186035,
"learning_rate": 8.8135593220339e-08,
"loss": 1.3638,
"step": 29870
},
{
"epoch": 0.75,
"grad_norm": 7.006107330322266,
"learning_rate": 8.135593220338984e-08,
"loss": 1.387,
"step": 29880
},
{
"epoch": 0.75,
"grad_norm": 7.425577163696289,
"learning_rate": 7.457627118644068e-08,
"loss": 1.314,
"step": 29890
},
{
"epoch": 0.75,
"grad_norm": 12.527627944946289,
"learning_rate": 6.779661016949153e-08,
"loss": 1.3822,
"step": 29900
},
{
"epoch": 0.75,
"grad_norm": 7.276275634765625,
"learning_rate": 6.101694915254239e-08,
"loss": 1.1887,
"step": 29910
},
{
"epoch": 0.75,
"grad_norm": 1.8828225135803223,
"learning_rate": 5.423728813559323e-08,
"loss": 1.3262,
"step": 29920
},
{
"epoch": 0.75,
"grad_norm": 3.2337329387664795,
"learning_rate": 4.745762711864407e-08,
"loss": 1.3487,
"step": 29930
},
{
"epoch": 0.75,
"grad_norm": 6.47509765625,
"learning_rate": 4.067796610169492e-08,
"loss": 1.3835,
"step": 29940
},
{
"epoch": 0.75,
"grad_norm": 8.018141746520996,
"learning_rate": 3.3898305084745764e-08,
"loss": 1.1818,
"step": 29950
},
{
"epoch": 0.75,
"grad_norm": 6.676591396331787,
"learning_rate": 2.7118644067796615e-08,
"loss": 1.2067,
"step": 29960
},
{
"epoch": 0.75,
"grad_norm": 3.872598648071289,
"learning_rate": 2.033898305084746e-08,
"loss": 1.2989,
"step": 29970
},
{
"epoch": 0.75,
"grad_norm": 2.7663283348083496,
"learning_rate": 1.3559322033898307e-08,
"loss": 1.321,
"step": 29980
},
{
"epoch": 0.75,
"grad_norm": 6.447133541107178,
"learning_rate": 6.779661016949154e-09,
"loss": 1.255,
"step": 29990
},
{
"epoch": 0.75,
"grad_norm": 5.224142074584961,
"learning_rate": 0.0,
"loss": 1.2453,
"step": 30000
},
{
"epoch": 0.75,
"eval_loss": 1.307248592376709,
"eval_runtime": 66.2202,
"eval_samples_per_second": 15.101,
"eval_steps_per_second": 15.101,
"step": 30000
}
],
"logging_steps": 10,
"max_steps": 30000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2500,
"total_flos": 4.8306377981952e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}