openr1_sft_qwen3-8b / trainer_state.json
Chtholly17's picture
Upload trainer_state.json with huggingface_hub
adaa191 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 15000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002,
"grad_norm": 14.211126327514648,
"learning_rate": 6.000000000000001e-08,
"loss": 0.9806,
"step": 10
},
{
"epoch": 0.004,
"grad_norm": 8.715738296508789,
"learning_rate": 1.2666666666666666e-07,
"loss": 0.9899,
"step": 20
},
{
"epoch": 0.006,
"grad_norm": 10.858619689941406,
"learning_rate": 1.9333333333333337e-07,
"loss": 1.0494,
"step": 30
},
{
"epoch": 0.008,
"grad_norm": 11.522907257080078,
"learning_rate": 2.6e-07,
"loss": 0.9985,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 11.823868751525879,
"learning_rate": 3.266666666666667e-07,
"loss": 0.9687,
"step": 50
},
{
"epoch": 0.012,
"grad_norm": 10.01672077178955,
"learning_rate": 3.9333333333333336e-07,
"loss": 0.7783,
"step": 60
},
{
"epoch": 0.014,
"grad_norm": 8.285725593566895,
"learning_rate": 4.6000000000000004e-07,
"loss": 0.8029,
"step": 70
},
{
"epoch": 0.016,
"grad_norm": 3.7833917140960693,
"learning_rate": 5.266666666666667e-07,
"loss": 0.6355,
"step": 80
},
{
"epoch": 0.018,
"grad_norm": 3.166145086288452,
"learning_rate": 5.933333333333334e-07,
"loss": 0.5781,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 2.388826847076416,
"learning_rate": 6.6e-07,
"loss": 0.5602,
"step": 100
},
{
"epoch": 0.022,
"grad_norm": 2.683410167694092,
"learning_rate": 7.266666666666668e-07,
"loss": 0.5822,
"step": 110
},
{
"epoch": 0.024,
"grad_norm": 2.4587814807891846,
"learning_rate": 7.933333333333335e-07,
"loss": 0.4702,
"step": 120
},
{
"epoch": 0.026,
"grad_norm": 2.512497663497925,
"learning_rate": 8.6e-07,
"loss": 0.5497,
"step": 130
},
{
"epoch": 0.028,
"grad_norm": 1.4995685815811157,
"learning_rate": 9.266666666666667e-07,
"loss": 0.5195,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 2.7800076007843018,
"learning_rate": 9.933333333333333e-07,
"loss": 0.4665,
"step": 150
},
{
"epoch": 0.032,
"grad_norm": 1.491605281829834,
"learning_rate": 1.06e-06,
"loss": 0.5083,
"step": 160
},
{
"epoch": 0.034,
"grad_norm": 1.643656611442566,
"learning_rate": 1.1266666666666667e-06,
"loss": 0.4549,
"step": 170
},
{
"epoch": 0.036,
"grad_norm": 2.509424924850464,
"learning_rate": 1.1933333333333335e-06,
"loss": 0.4816,
"step": 180
},
{
"epoch": 0.038,
"grad_norm": 1.9511481523513794,
"learning_rate": 1.26e-06,
"loss": 0.5072,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 2.516787052154541,
"learning_rate": 1.3266666666666667e-06,
"loss": 0.497,
"step": 200
},
{
"epoch": 0.042,
"grad_norm": 2.568359851837158,
"learning_rate": 1.3933333333333335e-06,
"loss": 0.485,
"step": 210
},
{
"epoch": 0.044,
"grad_norm": 1.8086744546890259,
"learning_rate": 1.46e-06,
"loss": 0.5004,
"step": 220
},
{
"epoch": 0.046,
"grad_norm": 1.9711308479309082,
"learning_rate": 1.526666666666667e-06,
"loss": 0.502,
"step": 230
},
{
"epoch": 0.048,
"grad_norm": 1.787848949432373,
"learning_rate": 1.5933333333333335e-06,
"loss": 0.5269,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 1.7853062152862549,
"learning_rate": 1.6600000000000002e-06,
"loss": 0.4902,
"step": 250
},
{
"epoch": 0.052,
"grad_norm": 1.663499355316162,
"learning_rate": 1.7266666666666667e-06,
"loss": 0.5142,
"step": 260
},
{
"epoch": 0.054,
"grad_norm": 1.4939693212509155,
"learning_rate": 1.7933333333333337e-06,
"loss": 0.4769,
"step": 270
},
{
"epoch": 0.056,
"grad_norm": 1.9765783548355103,
"learning_rate": 1.8600000000000002e-06,
"loss": 0.4955,
"step": 280
},
{
"epoch": 0.058,
"grad_norm": 1.5771379470825195,
"learning_rate": 1.926666666666667e-06,
"loss": 0.4419,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 1.0956958532333374,
"learning_rate": 1.9933333333333334e-06,
"loss": 0.4221,
"step": 300
},
{
"epoch": 0.062,
"grad_norm": 2.1830596923828125,
"learning_rate": 2.06e-06,
"loss": 0.5702,
"step": 310
},
{
"epoch": 0.064,
"grad_norm": 2.9908502101898193,
"learning_rate": 2.126666666666667e-06,
"loss": 0.5233,
"step": 320
},
{
"epoch": 0.066,
"grad_norm": 1.752899169921875,
"learning_rate": 2.1933333333333332e-06,
"loss": 0.4822,
"step": 330
},
{
"epoch": 0.068,
"grad_norm": 1.7774080038070679,
"learning_rate": 2.2600000000000004e-06,
"loss": 0.4571,
"step": 340
},
{
"epoch": 0.07,
"grad_norm": 2.2225241661071777,
"learning_rate": 2.3266666666666667e-06,
"loss": 0.3769,
"step": 350
},
{
"epoch": 0.072,
"grad_norm": 1.7270737886428833,
"learning_rate": 2.3933333333333334e-06,
"loss": 0.4333,
"step": 360
},
{
"epoch": 0.074,
"grad_norm": 1.9639253616333008,
"learning_rate": 2.46e-06,
"loss": 0.494,
"step": 370
},
{
"epoch": 0.076,
"grad_norm": 1.4829277992248535,
"learning_rate": 2.526666666666667e-06,
"loss": 0.455,
"step": 380
},
{
"epoch": 0.078,
"grad_norm": 2.67972469329834,
"learning_rate": 2.5933333333333336e-06,
"loss": 0.4865,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 2.0642127990722656,
"learning_rate": 2.6600000000000004e-06,
"loss": 0.4665,
"step": 400
},
{
"epoch": 0.082,
"grad_norm": 1.5665631294250488,
"learning_rate": 2.726666666666667e-06,
"loss": 0.4586,
"step": 410
},
{
"epoch": 0.084,
"grad_norm": 1.1839872598648071,
"learning_rate": 2.7933333333333334e-06,
"loss": 0.4541,
"step": 420
},
{
"epoch": 0.086,
"grad_norm": 1.2557004690170288,
"learning_rate": 2.86e-06,
"loss": 0.4357,
"step": 430
},
{
"epoch": 0.088,
"grad_norm": 1.4764341115951538,
"learning_rate": 2.9266666666666673e-06,
"loss": 0.5011,
"step": 440
},
{
"epoch": 0.09,
"grad_norm": 1.9956254959106445,
"learning_rate": 2.9933333333333336e-06,
"loss": 0.4933,
"step": 450
},
{
"epoch": 0.092,
"grad_norm": 1.6207813024520874,
"learning_rate": 3.0600000000000003e-06,
"loss": 0.4402,
"step": 460
},
{
"epoch": 0.094,
"grad_norm": 1.7665313482284546,
"learning_rate": 3.1266666666666667e-06,
"loss": 0.4392,
"step": 470
},
{
"epoch": 0.096,
"grad_norm": 1.76856529712677,
"learning_rate": 3.193333333333334e-06,
"loss": 0.4521,
"step": 480
},
{
"epoch": 0.098,
"grad_norm": 1.7528122663497925,
"learning_rate": 3.2600000000000006e-06,
"loss": 0.4168,
"step": 490
},
{
"epoch": 0.1,
"grad_norm": 1.3086860179901123,
"learning_rate": 3.326666666666667e-06,
"loss": 0.4228,
"step": 500
},
{
"epoch": 0.102,
"grad_norm": 2.676274061203003,
"learning_rate": 3.3933333333333336e-06,
"loss": 0.4733,
"step": 510
},
{
"epoch": 0.104,
"grad_norm": 1.9702718257904053,
"learning_rate": 3.46e-06,
"loss": 0.4273,
"step": 520
},
{
"epoch": 0.106,
"grad_norm": 1.8276294469833374,
"learning_rate": 3.526666666666667e-06,
"loss": 0.491,
"step": 530
},
{
"epoch": 0.108,
"grad_norm": 3.189826726913452,
"learning_rate": 3.593333333333334e-06,
"loss": 0.5354,
"step": 540
},
{
"epoch": 0.11,
"grad_norm": 1.3039813041687012,
"learning_rate": 3.66e-06,
"loss": 0.4561,
"step": 550
},
{
"epoch": 0.112,
"grad_norm": 1.4156097173690796,
"learning_rate": 3.726666666666667e-06,
"loss": 0.3896,
"step": 560
},
{
"epoch": 0.114,
"grad_norm": 1.6464276313781738,
"learning_rate": 3.793333333333334e-06,
"loss": 0.4438,
"step": 570
},
{
"epoch": 0.116,
"grad_norm": 1.4432697296142578,
"learning_rate": 3.86e-06,
"loss": 0.5175,
"step": 580
},
{
"epoch": 0.118,
"grad_norm": 1.6621713638305664,
"learning_rate": 3.926666666666667e-06,
"loss": 0.4429,
"step": 590
},
{
"epoch": 0.12,
"grad_norm": 1.6414673328399658,
"learning_rate": 3.993333333333334e-06,
"loss": 0.4958,
"step": 600
},
{
"epoch": 0.122,
"grad_norm": 1.51468026638031,
"learning_rate": 4.060000000000001e-06,
"loss": 0.4636,
"step": 610
},
{
"epoch": 0.124,
"grad_norm": 1.9095144271850586,
"learning_rate": 4.126666666666667e-06,
"loss": 0.4545,
"step": 620
},
{
"epoch": 0.126,
"grad_norm": 1.6873503923416138,
"learning_rate": 4.1933333333333336e-06,
"loss": 0.4203,
"step": 630
},
{
"epoch": 0.128,
"grad_norm": 1.8741629123687744,
"learning_rate": 4.26e-06,
"loss": 0.4713,
"step": 640
},
{
"epoch": 0.13,
"grad_norm": 1.5286614894866943,
"learning_rate": 4.326666666666667e-06,
"loss": 0.3901,
"step": 650
},
{
"epoch": 0.132,
"grad_norm": 1.4989817142486572,
"learning_rate": 4.393333333333334e-06,
"loss": 0.4117,
"step": 660
},
{
"epoch": 0.134,
"grad_norm": 1.642307996749878,
"learning_rate": 4.4600000000000005e-06,
"loss": 0.5395,
"step": 670
},
{
"epoch": 0.136,
"grad_norm": 2.079261064529419,
"learning_rate": 4.526666666666667e-06,
"loss": 0.4314,
"step": 680
},
{
"epoch": 0.138,
"grad_norm": 1.440948486328125,
"learning_rate": 4.593333333333333e-06,
"loss": 0.451,
"step": 690
},
{
"epoch": 0.14,
"grad_norm": 1.213745355606079,
"learning_rate": 4.66e-06,
"loss": 0.4941,
"step": 700
},
{
"epoch": 0.142,
"grad_norm": 2.1743085384368896,
"learning_rate": 4.7266666666666674e-06,
"loss": 0.4749,
"step": 710
},
{
"epoch": 0.144,
"grad_norm": 1.6815190315246582,
"learning_rate": 4.793333333333334e-06,
"loss": 0.606,
"step": 720
},
{
"epoch": 0.146,
"grad_norm": 1.5993754863739014,
"learning_rate": 4.86e-06,
"loss": 0.4041,
"step": 730
},
{
"epoch": 0.148,
"grad_norm": 2.1879842281341553,
"learning_rate": 4.926666666666667e-06,
"loss": 0.482,
"step": 740
},
{
"epoch": 0.15,
"grad_norm": 1.3680864572525024,
"learning_rate": 4.9933333333333335e-06,
"loss": 0.4598,
"step": 750
},
{
"epoch": 0.152,
"grad_norm": 1.4133329391479492,
"learning_rate": 5.060000000000001e-06,
"loss": 0.4606,
"step": 760
},
{
"epoch": 0.154,
"grad_norm": 1.6556570529937744,
"learning_rate": 5.126666666666668e-06,
"loss": 0.3716,
"step": 770
},
{
"epoch": 0.156,
"grad_norm": 1.5880504846572876,
"learning_rate": 5.193333333333333e-06,
"loss": 0.5006,
"step": 780
},
{
"epoch": 0.158,
"grad_norm": 1.4626226425170898,
"learning_rate": 5.2600000000000005e-06,
"loss": 0.4802,
"step": 790
},
{
"epoch": 0.16,
"grad_norm": 1.9758862257003784,
"learning_rate": 5.326666666666667e-06,
"loss": 0.4725,
"step": 800
},
{
"epoch": 0.162,
"grad_norm": 1.7073544263839722,
"learning_rate": 5.393333333333334e-06,
"loss": 0.412,
"step": 810
},
{
"epoch": 0.164,
"grad_norm": 1.9127854108810425,
"learning_rate": 5.460000000000001e-06,
"loss": 0.4806,
"step": 820
},
{
"epoch": 0.166,
"grad_norm": 1.452903389930725,
"learning_rate": 5.5266666666666666e-06,
"loss": 0.4171,
"step": 830
},
{
"epoch": 0.168,
"grad_norm": 1.603830337524414,
"learning_rate": 5.593333333333334e-06,
"loss": 0.4348,
"step": 840
},
{
"epoch": 0.17,
"grad_norm": 1.7848249673843384,
"learning_rate": 5.66e-06,
"loss": 0.4739,
"step": 850
},
{
"epoch": 0.172,
"grad_norm": 1.8033798933029175,
"learning_rate": 5.726666666666667e-06,
"loss": 0.4251,
"step": 860
},
{
"epoch": 0.174,
"grad_norm": 1.6332205533981323,
"learning_rate": 5.793333333333334e-06,
"loss": 0.4422,
"step": 870
},
{
"epoch": 0.176,
"grad_norm": 1.8807281255722046,
"learning_rate": 5.86e-06,
"loss": 0.4224,
"step": 880
},
{
"epoch": 0.178,
"grad_norm": 1.7026886940002441,
"learning_rate": 5.926666666666667e-06,
"loss": 0.4299,
"step": 890
},
{
"epoch": 0.18,
"grad_norm": 1.69607412815094,
"learning_rate": 5.993333333333334e-06,
"loss": 0.5092,
"step": 900
},
{
"epoch": 0.182,
"grad_norm": 1.3589431047439575,
"learning_rate": 6.0600000000000004e-06,
"loss": 0.399,
"step": 910
},
{
"epoch": 0.184,
"grad_norm": 1.3969651460647583,
"learning_rate": 6.126666666666668e-06,
"loss": 0.5225,
"step": 920
},
{
"epoch": 0.186,
"grad_norm": 1.8621569871902466,
"learning_rate": 6.193333333333333e-06,
"loss": 0.4389,
"step": 930
},
{
"epoch": 0.188,
"grad_norm": 1.6221462488174438,
"learning_rate": 6.26e-06,
"loss": 0.4367,
"step": 940
},
{
"epoch": 0.19,
"grad_norm": 1.452337622642517,
"learning_rate": 6.326666666666667e-06,
"loss": 0.452,
"step": 950
},
{
"epoch": 0.192,
"grad_norm": 1.5459558963775635,
"learning_rate": 6.393333333333334e-06,
"loss": 0.5254,
"step": 960
},
{
"epoch": 0.194,
"grad_norm": 1.353761076927185,
"learning_rate": 6.460000000000001e-06,
"loss": 0.4755,
"step": 970
},
{
"epoch": 0.196,
"grad_norm": 1.3966904878616333,
"learning_rate": 6.526666666666666e-06,
"loss": 0.4842,
"step": 980
},
{
"epoch": 0.198,
"grad_norm": 1.3571563959121704,
"learning_rate": 6.5933333333333335e-06,
"loss": 0.4776,
"step": 990
},
{
"epoch": 0.2,
"grad_norm": 1.7380127906799316,
"learning_rate": 6.660000000000001e-06,
"loss": 0.5087,
"step": 1000
},
{
"epoch": 0.202,
"grad_norm": 1.30695378780365,
"learning_rate": 6.726666666666667e-06,
"loss": 0.4261,
"step": 1010
},
{
"epoch": 0.204,
"grad_norm": 1.4552329778671265,
"learning_rate": 6.793333333333334e-06,
"loss": 0.4485,
"step": 1020
},
{
"epoch": 0.206,
"grad_norm": 1.4665312767028809,
"learning_rate": 6.860000000000001e-06,
"loss": 0.4326,
"step": 1030
},
{
"epoch": 0.208,
"grad_norm": 1.511225938796997,
"learning_rate": 6.926666666666667e-06,
"loss": 0.4234,
"step": 1040
},
{
"epoch": 0.21,
"grad_norm": 1.2507603168487549,
"learning_rate": 6.993333333333334e-06,
"loss": 0.3983,
"step": 1050
},
{
"epoch": 0.212,
"grad_norm": 1.8667805194854736,
"learning_rate": 7.06e-06,
"loss": 0.4798,
"step": 1060
},
{
"epoch": 0.214,
"grad_norm": 1.400815725326538,
"learning_rate": 7.126666666666667e-06,
"loss": 0.4947,
"step": 1070
},
{
"epoch": 0.216,
"grad_norm": 1.2267810106277466,
"learning_rate": 7.1933333333333345e-06,
"loss": 0.4703,
"step": 1080
},
{
"epoch": 0.218,
"grad_norm": 1.2562822103500366,
"learning_rate": 7.260000000000001e-06,
"loss": 0.4272,
"step": 1090
},
{
"epoch": 0.22,
"grad_norm": 1.6255943775177002,
"learning_rate": 7.326666666666667e-06,
"loss": 0.5082,
"step": 1100
},
{
"epoch": 0.222,
"grad_norm": 1.6352773904800415,
"learning_rate": 7.393333333333333e-06,
"loss": 0.4679,
"step": 1110
},
{
"epoch": 0.224,
"grad_norm": 1.4105769395828247,
"learning_rate": 7.4600000000000006e-06,
"loss": 0.4716,
"step": 1120
},
{
"epoch": 0.226,
"grad_norm": 1.766946792602539,
"learning_rate": 7.526666666666668e-06,
"loss": 0.5016,
"step": 1130
},
{
"epoch": 0.228,
"grad_norm": 1.1533312797546387,
"learning_rate": 7.593333333333334e-06,
"loss": 0.4691,
"step": 1140
},
{
"epoch": 0.23,
"grad_norm": 1.6040184497833252,
"learning_rate": 7.660000000000001e-06,
"loss": 0.4123,
"step": 1150
},
{
"epoch": 0.232,
"grad_norm": 1.3271604776382446,
"learning_rate": 7.726666666666667e-06,
"loss": 0.4571,
"step": 1160
},
{
"epoch": 0.234,
"grad_norm": 1.7813302278518677,
"learning_rate": 7.793333333333334e-06,
"loss": 0.4868,
"step": 1170
},
{
"epoch": 0.236,
"grad_norm": 1.5846160650253296,
"learning_rate": 7.860000000000001e-06,
"loss": 0.4237,
"step": 1180
},
{
"epoch": 0.238,
"grad_norm": 1.4138127565383911,
"learning_rate": 7.926666666666666e-06,
"loss": 0.4485,
"step": 1190
},
{
"epoch": 0.24,
"grad_norm": 1.7159594297409058,
"learning_rate": 7.993333333333334e-06,
"loss": 0.5088,
"step": 1200
},
{
"epoch": 0.242,
"grad_norm": 1.5566740036010742,
"learning_rate": 8.06e-06,
"loss": 0.4248,
"step": 1210
},
{
"epoch": 0.244,
"grad_norm": 1.4056092500686646,
"learning_rate": 8.126666666666668e-06,
"loss": 0.4242,
"step": 1220
},
{
"epoch": 0.246,
"grad_norm": 1.2954442501068115,
"learning_rate": 8.193333333333335e-06,
"loss": 0.479,
"step": 1230
},
{
"epoch": 0.248,
"grad_norm": 1.576978325843811,
"learning_rate": 8.26e-06,
"loss": 0.4207,
"step": 1240
},
{
"epoch": 0.25,
"grad_norm": 1.4118748903274536,
"learning_rate": 8.326666666666668e-06,
"loss": 0.4451,
"step": 1250
},
{
"epoch": 0.252,
"grad_norm": 1.3957098722457886,
"learning_rate": 8.393333333333335e-06,
"loss": 0.4955,
"step": 1260
},
{
"epoch": 0.254,
"grad_norm": 1.2895923852920532,
"learning_rate": 8.46e-06,
"loss": 0.4241,
"step": 1270
},
{
"epoch": 0.256,
"grad_norm": 1.5065820217132568,
"learning_rate": 8.526666666666667e-06,
"loss": 0.4727,
"step": 1280
},
{
"epoch": 0.258,
"grad_norm": 1.086529016494751,
"learning_rate": 8.593333333333333e-06,
"loss": 0.4429,
"step": 1290
},
{
"epoch": 0.26,
"grad_norm": 1.2884867191314697,
"learning_rate": 8.66e-06,
"loss": 0.4496,
"step": 1300
},
{
"epoch": 0.262,
"grad_norm": 1.344300627708435,
"learning_rate": 8.726666666666667e-06,
"loss": 0.3874,
"step": 1310
},
{
"epoch": 0.264,
"grad_norm": 1.2245051860809326,
"learning_rate": 8.793333333333334e-06,
"loss": 0.5057,
"step": 1320
},
{
"epoch": 0.266,
"grad_norm": 1.3924754858016968,
"learning_rate": 8.860000000000002e-06,
"loss": 0.4008,
"step": 1330
},
{
"epoch": 0.268,
"grad_norm": 1.4217268228530884,
"learning_rate": 8.926666666666669e-06,
"loss": 0.4915,
"step": 1340
},
{
"epoch": 0.27,
"grad_norm": 0.9328693747520447,
"learning_rate": 8.993333333333334e-06,
"loss": 0.4044,
"step": 1350
},
{
"epoch": 0.272,
"grad_norm": 1.2598109245300293,
"learning_rate": 9.060000000000001e-06,
"loss": 0.4407,
"step": 1360
},
{
"epoch": 0.274,
"grad_norm": 1.253463625907898,
"learning_rate": 9.126666666666667e-06,
"loss": 0.4671,
"step": 1370
},
{
"epoch": 0.276,
"grad_norm": 1.41267728805542,
"learning_rate": 9.193333333333334e-06,
"loss": 0.5014,
"step": 1380
},
{
"epoch": 0.278,
"grad_norm": 1.285569190979004,
"learning_rate": 9.260000000000001e-06,
"loss": 0.5425,
"step": 1390
},
{
"epoch": 0.28,
"grad_norm": 1.9558253288269043,
"learning_rate": 9.326666666666667e-06,
"loss": 0.4851,
"step": 1400
},
{
"epoch": 0.282,
"grad_norm": 1.1378921270370483,
"learning_rate": 9.393333333333334e-06,
"loss": 0.5159,
"step": 1410
},
{
"epoch": 0.284,
"grad_norm": 1.152918815612793,
"learning_rate": 9.460000000000001e-06,
"loss": 0.4235,
"step": 1420
},
{
"epoch": 0.286,
"grad_norm": 1.405306339263916,
"learning_rate": 9.526666666666668e-06,
"loss": 0.4881,
"step": 1430
},
{
"epoch": 0.288,
"grad_norm": 1.4959778785705566,
"learning_rate": 9.593333333333335e-06,
"loss": 0.4729,
"step": 1440
},
{
"epoch": 0.29,
"grad_norm": 1.3399648666381836,
"learning_rate": 9.66e-06,
"loss": 0.4554,
"step": 1450
},
{
"epoch": 0.292,
"grad_norm": 1.1871027946472168,
"learning_rate": 9.726666666666668e-06,
"loss": 0.4591,
"step": 1460
},
{
"epoch": 0.294,
"grad_norm": 1.196593165397644,
"learning_rate": 9.793333333333333e-06,
"loss": 0.5592,
"step": 1470
},
{
"epoch": 0.296,
"grad_norm": 1.4280208349227905,
"learning_rate": 9.86e-06,
"loss": 0.4551,
"step": 1480
},
{
"epoch": 0.298,
"grad_norm": 1.1892995834350586,
"learning_rate": 9.926666666666668e-06,
"loss": 0.4864,
"step": 1490
},
{
"epoch": 0.3,
"grad_norm": 1.3978432416915894,
"learning_rate": 9.993333333333333e-06,
"loss": 0.5178,
"step": 1500
},
{
"epoch": 0.302,
"grad_norm": 1.2265063524246216,
"learning_rate": 9.999989033776898e-06,
"loss": 0.446,
"step": 1510
},
{
"epoch": 0.304,
"grad_norm": 1.7466825246810913,
"learning_rate": 9.999951125906936e-06,
"loss": 0.5599,
"step": 1520
},
{
"epoch": 0.306,
"grad_norm": 1.2893444299697876,
"learning_rate": 9.999886141209892e-06,
"loss": 0.5088,
"step": 1530
},
{
"epoch": 0.308,
"grad_norm": 1.459552526473999,
"learning_rate": 9.999794080037675e-06,
"loss": 0.4981,
"step": 1540
},
{
"epoch": 0.31,
"grad_norm": 1.654597520828247,
"learning_rate": 9.99967494288884e-06,
"loss": 0.561,
"step": 1550
},
{
"epoch": 0.312,
"grad_norm": 1.3994464874267578,
"learning_rate": 9.999528730408565e-06,
"loss": 0.4872,
"step": 1560
},
{
"epoch": 0.314,
"grad_norm": 1.6109237670898438,
"learning_rate": 9.999355443388649e-06,
"loss": 0.4748,
"step": 1570
},
{
"epoch": 0.316,
"grad_norm": 1.2756563425064087,
"learning_rate": 9.999155082767515e-06,
"loss": 0.4265,
"step": 1580
},
{
"epoch": 0.318,
"grad_norm": 1.2647457122802734,
"learning_rate": 9.998927649630202e-06,
"loss": 0.4898,
"step": 1590
},
{
"epoch": 0.32,
"grad_norm": 1.2949209213256836,
"learning_rate": 9.998673145208351e-06,
"loss": 0.4366,
"step": 1600
},
{
"epoch": 0.322,
"grad_norm": 1.2482858896255493,
"learning_rate": 9.998391570880212e-06,
"loss": 0.5014,
"step": 1610
},
{
"epoch": 0.324,
"grad_norm": 1.0791841745376587,
"learning_rate": 9.99808292817063e-06,
"loss": 0.4954,
"step": 1620
},
{
"epoch": 0.326,
"grad_norm": 1.4692087173461914,
"learning_rate": 9.997747218751032e-06,
"loss": 0.5007,
"step": 1630
},
{
"epoch": 0.328,
"grad_norm": 1.5948970317840576,
"learning_rate": 9.997384444439424e-06,
"loss": 0.4542,
"step": 1640
},
{
"epoch": 0.33,
"grad_norm": 1.2097396850585938,
"learning_rate": 9.996994607200382e-06,
"loss": 0.4525,
"step": 1650
},
{
"epoch": 0.332,
"grad_norm": 1.343247413635254,
"learning_rate": 9.99657770914504e-06,
"loss": 0.4434,
"step": 1660
},
{
"epoch": 0.334,
"grad_norm": 1.5379422903060913,
"learning_rate": 9.996133752531071e-06,
"loss": 0.5859,
"step": 1670
},
{
"epoch": 0.336,
"grad_norm": 1.158941388130188,
"learning_rate": 9.99566273976269e-06,
"loss": 0.495,
"step": 1680
},
{
"epoch": 0.338,
"grad_norm": 1.1630929708480835,
"learning_rate": 9.995164673390624e-06,
"loss": 0.5463,
"step": 1690
},
{
"epoch": 0.34,
"grad_norm": 1.256372094154358,
"learning_rate": 9.994639556112113e-06,
"loss": 0.4853,
"step": 1700
},
{
"epoch": 0.342,
"grad_norm": 1.8223704099655151,
"learning_rate": 9.994087390770887e-06,
"loss": 0.4836,
"step": 1710
},
{
"epoch": 0.344,
"grad_norm": 1.945887565612793,
"learning_rate": 9.993508180357154e-06,
"loss": 0.5115,
"step": 1720
},
{
"epoch": 0.346,
"grad_norm": 1.6152077913284302,
"learning_rate": 9.992901928007577e-06,
"loss": 0.4939,
"step": 1730
},
{
"epoch": 0.348,
"grad_norm": 1.2321640253067017,
"learning_rate": 9.992268637005268e-06,
"loss": 0.5277,
"step": 1740
},
{
"epoch": 0.35,
"grad_norm": 1.2367174625396729,
"learning_rate": 9.991608310779762e-06,
"loss": 0.4691,
"step": 1750
},
{
"epoch": 0.352,
"grad_norm": 1.744498372077942,
"learning_rate": 9.990920952907005e-06,
"loss": 0.497,
"step": 1760
},
{
"epoch": 0.354,
"grad_norm": 1.5234036445617676,
"learning_rate": 9.99020656710932e-06,
"loss": 0.4469,
"step": 1770
},
{
"epoch": 0.356,
"grad_norm": 1.0841994285583496,
"learning_rate": 9.989465157255413e-06,
"loss": 0.4317,
"step": 1780
},
{
"epoch": 0.358,
"grad_norm": 1.1552493572235107,
"learning_rate": 9.988696727360323e-06,
"loss": 0.461,
"step": 1790
},
{
"epoch": 0.36,
"grad_norm": 1.6440882682800293,
"learning_rate": 9.987901281585423e-06,
"loss": 0.5038,
"step": 1800
},
{
"epoch": 0.362,
"grad_norm": 1.3549693822860718,
"learning_rate": 9.987078824238384e-06,
"loss": 0.468,
"step": 1810
},
{
"epoch": 0.364,
"grad_norm": 1.3959366083145142,
"learning_rate": 9.986229359773154e-06,
"loss": 0.4709,
"step": 1820
},
{
"epoch": 0.366,
"grad_norm": 1.612743854522705,
"learning_rate": 9.985352892789941e-06,
"loss": 0.4717,
"step": 1830
},
{
"epoch": 0.368,
"grad_norm": 1.2929461002349854,
"learning_rate": 9.98444942803518e-06,
"loss": 0.4811,
"step": 1840
},
{
"epoch": 0.37,
"grad_norm": 1.450260877609253,
"learning_rate": 9.983518970401508e-06,
"loss": 0.5203,
"step": 1850
},
{
"epoch": 0.372,
"grad_norm": 1.1035138368606567,
"learning_rate": 9.982561524927749e-06,
"loss": 0.4759,
"step": 1860
},
{
"epoch": 0.374,
"grad_norm": 1.2472127676010132,
"learning_rate": 9.981577096798864e-06,
"loss": 0.5506,
"step": 1870
},
{
"epoch": 0.376,
"grad_norm": 1.1840726137161255,
"learning_rate": 9.980565691345945e-06,
"loss": 0.4878,
"step": 1880
},
{
"epoch": 0.378,
"grad_norm": 1.3959659337997437,
"learning_rate": 9.979527314046177e-06,
"loss": 0.4916,
"step": 1890
},
{
"epoch": 0.38,
"grad_norm": 1.154722809791565,
"learning_rate": 9.978461970522807e-06,
"loss": 0.4592,
"step": 1900
},
{
"epoch": 0.382,
"grad_norm": 1.2845226526260376,
"learning_rate": 9.977369666545114e-06,
"loss": 0.5465,
"step": 1910
},
{
"epoch": 0.384,
"grad_norm": 1.175721287727356,
"learning_rate": 9.976250408028383e-06,
"loss": 0.4742,
"step": 1920
},
{
"epoch": 0.386,
"grad_norm": 1.5375049114227295,
"learning_rate": 9.975104201033868e-06,
"loss": 0.4783,
"step": 1930
},
{
"epoch": 0.388,
"grad_norm": 1.305464506149292,
"learning_rate": 9.973931051768756e-06,
"loss": 0.4931,
"step": 1940
},
{
"epoch": 0.39,
"grad_norm": 1.2971112728118896,
"learning_rate": 9.972730966586144e-06,
"loss": 0.4617,
"step": 1950
},
{
"epoch": 0.392,
"grad_norm": 1.2412433624267578,
"learning_rate": 9.971503951984996e-06,
"loss": 0.5366,
"step": 1960
},
{
"epoch": 0.394,
"grad_norm": 1.3253568410873413,
"learning_rate": 9.970250014610105e-06,
"loss": 0.5186,
"step": 1970
},
{
"epoch": 0.396,
"grad_norm": 1.2338709831237793,
"learning_rate": 9.968969161252072e-06,
"loss": 0.5074,
"step": 1980
},
{
"epoch": 0.398,
"grad_norm": 1.0964592695236206,
"learning_rate": 9.96766139884725e-06,
"loss": 0.4096,
"step": 1990
},
{
"epoch": 0.4,
"grad_norm": 1.8194420337677002,
"learning_rate": 9.966326734477721e-06,
"loss": 0.4439,
"step": 2000
},
{
"epoch": 0.402,
"grad_norm": 1.0826818943023682,
"learning_rate": 9.96496517537125e-06,
"loss": 0.4602,
"step": 2010
},
{
"epoch": 0.404,
"grad_norm": 0.9985308647155762,
"learning_rate": 9.96357672890125e-06,
"loss": 0.429,
"step": 2020
},
{
"epoch": 0.406,
"grad_norm": 1.4715704917907715,
"learning_rate": 9.96216140258674e-06,
"loss": 0.5181,
"step": 2030
},
{
"epoch": 0.408,
"grad_norm": 1.606497049331665,
"learning_rate": 9.9607192040923e-06,
"loss": 0.4314,
"step": 2040
},
{
"epoch": 0.41,
"grad_norm": 1.4677903652191162,
"learning_rate": 9.959250141228046e-06,
"loss": 0.4204,
"step": 2050
},
{
"epoch": 0.412,
"grad_norm": 1.4426929950714111,
"learning_rate": 9.95775422194956e-06,
"loss": 0.4803,
"step": 2060
},
{
"epoch": 0.414,
"grad_norm": 1.2194904088974,
"learning_rate": 9.956231454357876e-06,
"loss": 0.4695,
"step": 2070
},
{
"epoch": 0.416,
"grad_norm": 1.1238853931427002,
"learning_rate": 9.954681846699414e-06,
"loss": 0.5755,
"step": 2080
},
{
"epoch": 0.418,
"grad_norm": 1.0031379461288452,
"learning_rate": 9.953105407365952e-06,
"loss": 0.4325,
"step": 2090
},
{
"epoch": 0.42,
"grad_norm": 1.6501160860061646,
"learning_rate": 9.951502144894566e-06,
"loss": 0.5403,
"step": 2100
},
{
"epoch": 0.422,
"grad_norm": 1.440741777420044,
"learning_rate": 9.94987206796759e-06,
"loss": 0.4725,
"step": 2110
},
{
"epoch": 0.424,
"grad_norm": 1.2382850646972656,
"learning_rate": 9.948215185412578e-06,
"loss": 0.4235,
"step": 2120
},
{
"epoch": 0.426,
"grad_norm": 1.3473109006881714,
"learning_rate": 9.94653150620224e-06,
"loss": 0.4388,
"step": 2130
},
{
"epoch": 0.428,
"grad_norm": 1.2151130437850952,
"learning_rate": 9.944821039454403e-06,
"loss": 0.5012,
"step": 2140
},
{
"epoch": 0.43,
"grad_norm": 1.1826070547103882,
"learning_rate": 9.943083794431959e-06,
"loss": 0.486,
"step": 2150
},
{
"epoch": 0.432,
"grad_norm": 1.0905817747116089,
"learning_rate": 9.941319780542817e-06,
"loss": 0.5423,
"step": 2160
},
{
"epoch": 0.434,
"grad_norm": 1.3496341705322266,
"learning_rate": 9.939529007339852e-06,
"loss": 0.5141,
"step": 2170
},
{
"epoch": 0.436,
"grad_norm": 1.33112633228302,
"learning_rate": 9.937711484520848e-06,
"loss": 0.4819,
"step": 2180
},
{
"epoch": 0.438,
"grad_norm": 1.1748610734939575,
"learning_rate": 9.935867221928454e-06,
"loss": 0.5097,
"step": 2190
},
{
"epoch": 0.44,
"grad_norm": 1.2363240718841553,
"learning_rate": 9.93399622955012e-06,
"loss": 0.4261,
"step": 2200
},
{
"epoch": 0.442,
"grad_norm": 1.667703628540039,
"learning_rate": 9.932098517518056e-06,
"loss": 0.4662,
"step": 2210
},
{
"epoch": 0.444,
"grad_norm": 1.446291446685791,
"learning_rate": 9.930174096109163e-06,
"loss": 0.445,
"step": 2220
},
{
"epoch": 0.446,
"grad_norm": 1.476731300354004,
"learning_rate": 9.928222975744992e-06,
"loss": 0.4666,
"step": 2230
},
{
"epoch": 0.448,
"grad_norm": 1.3054319620132446,
"learning_rate": 9.926245166991671e-06,
"loss": 0.4248,
"step": 2240
},
{
"epoch": 0.45,
"grad_norm": 1.5817550420761108,
"learning_rate": 9.924240680559867e-06,
"loss": 0.4325,
"step": 2250
},
{
"epoch": 0.452,
"grad_norm": 1.0748894214630127,
"learning_rate": 9.922209527304709e-06,
"loss": 0.4862,
"step": 2260
},
{
"epoch": 0.454,
"grad_norm": 1.6298894882202148,
"learning_rate": 9.920151718225743e-06,
"loss": 0.542,
"step": 2270
},
{
"epoch": 0.456,
"grad_norm": 1.4869043827056885,
"learning_rate": 9.918067264466867e-06,
"loss": 0.4627,
"step": 2280
},
{
"epoch": 0.458,
"grad_norm": 1.3258094787597656,
"learning_rate": 9.915956177316269e-06,
"loss": 0.4389,
"step": 2290
},
{
"epoch": 0.46,
"grad_norm": 1.586165189743042,
"learning_rate": 9.913818468206368e-06,
"loss": 0.6127,
"step": 2300
},
{
"epoch": 0.462,
"grad_norm": 1.3697926998138428,
"learning_rate": 9.911654148713757e-06,
"loss": 0.574,
"step": 2310
},
{
"epoch": 0.464,
"grad_norm": 1.212844729423523,
"learning_rate": 9.909463230559127e-06,
"loss": 0.5196,
"step": 2320
},
{
"epoch": 0.466,
"grad_norm": 0.9597765207290649,
"learning_rate": 9.907245725607217e-06,
"loss": 0.5269,
"step": 2330
},
{
"epoch": 0.468,
"grad_norm": 1.3319644927978516,
"learning_rate": 9.905001645866746e-06,
"loss": 0.4136,
"step": 2340
},
{
"epoch": 0.47,
"grad_norm": 1.0987292528152466,
"learning_rate": 9.902731003490344e-06,
"loss": 0.4325,
"step": 2350
},
{
"epoch": 0.472,
"grad_norm": 1.3473751544952393,
"learning_rate": 9.90043381077449e-06,
"loss": 0.464,
"step": 2360
},
{
"epoch": 0.474,
"grad_norm": 1.5998402833938599,
"learning_rate": 9.898110080159442e-06,
"loss": 0.6172,
"step": 2370
},
{
"epoch": 0.476,
"grad_norm": 0.9513059258460999,
"learning_rate": 9.895759824229176e-06,
"loss": 0.4431,
"step": 2380
},
{
"epoch": 0.478,
"grad_norm": 1.3348459005355835,
"learning_rate": 9.893383055711308e-06,
"loss": 0.4383,
"step": 2390
},
{
"epoch": 0.48,
"grad_norm": 1.3876981735229492,
"learning_rate": 9.890979787477036e-06,
"loss": 0.5066,
"step": 2400
},
{
"epoch": 0.482,
"grad_norm": 1.3132597208023071,
"learning_rate": 9.88855003254106e-06,
"loss": 0.4995,
"step": 2410
},
{
"epoch": 0.484,
"grad_norm": 1.3520206212997437,
"learning_rate": 9.886093804061523e-06,
"loss": 0.4658,
"step": 2420
},
{
"epoch": 0.486,
"grad_norm": 1.4986363649368286,
"learning_rate": 9.883611115339929e-06,
"loss": 0.4718,
"step": 2430
},
{
"epoch": 0.488,
"grad_norm": 1.3672336339950562,
"learning_rate": 9.881101979821075e-06,
"loss": 0.5289,
"step": 2440
},
{
"epoch": 0.49,
"grad_norm": 1.610236406326294,
"learning_rate": 9.87856641109298e-06,
"loss": 0.5065,
"step": 2450
},
{
"epoch": 0.492,
"grad_norm": 1.241811752319336,
"learning_rate": 9.876004422886809e-06,
"loss": 0.3937,
"step": 2460
},
{
"epoch": 0.494,
"grad_norm": 1.385197639465332,
"learning_rate": 9.873416029076801e-06,
"loss": 0.5553,
"step": 2470
},
{
"epoch": 0.496,
"grad_norm": 1.2933454513549805,
"learning_rate": 9.870801243680191e-06,
"loss": 0.4653,
"step": 2480
},
{
"epoch": 0.498,
"grad_norm": 1.2639875411987305,
"learning_rate": 9.868160080857134e-06,
"loss": 0.4908,
"step": 2490
},
{
"epoch": 0.5,
"grad_norm": 1.142906665802002,
"learning_rate": 9.865492554910634e-06,
"loss": 0.4884,
"step": 2500
},
{
"epoch": 0.502,
"grad_norm": 1.0850920677185059,
"learning_rate": 9.862798680286459e-06,
"loss": 0.5224,
"step": 2510
},
{
"epoch": 0.504,
"grad_norm": 1.3904608488082886,
"learning_rate": 9.860078471573066e-06,
"loss": 0.5273,
"step": 2520
},
{
"epoch": 0.506,
"grad_norm": 1.1105901002883911,
"learning_rate": 9.857331943501527e-06,
"loss": 0.4478,
"step": 2530
},
{
"epoch": 0.508,
"grad_norm": 1.278568983078003,
"learning_rate": 9.854559110945436e-06,
"loss": 0.5204,
"step": 2540
},
{
"epoch": 0.51,
"grad_norm": 0.9616843461990356,
"learning_rate": 9.851759988920843e-06,
"loss": 0.452,
"step": 2550
},
{
"epoch": 0.512,
"grad_norm": 1.7013401985168457,
"learning_rate": 9.848934592586165e-06,
"loss": 0.5092,
"step": 2560
},
{
"epoch": 0.514,
"grad_norm": 1.2884467840194702,
"learning_rate": 9.846082937242108e-06,
"loss": 0.4333,
"step": 2570
},
{
"epoch": 0.516,
"grad_norm": 1.0977705717086792,
"learning_rate": 9.843205038331574e-06,
"loss": 0.4188,
"step": 2580
},
{
"epoch": 0.518,
"grad_norm": 1.1589962244033813,
"learning_rate": 9.84030091143959e-06,
"loss": 0.5456,
"step": 2590
},
{
"epoch": 0.52,
"grad_norm": 1.4452130794525146,
"learning_rate": 9.837370572293221e-06,
"loss": 0.538,
"step": 2600
},
{
"epoch": 0.522,
"grad_norm": 0.8810553550720215,
"learning_rate": 9.834414036761477e-06,
"loss": 0.4697,
"step": 2610
},
{
"epoch": 0.524,
"grad_norm": 1.2297122478485107,
"learning_rate": 9.831431320855235e-06,
"loss": 0.4525,
"step": 2620
},
{
"epoch": 0.526,
"grad_norm": 1.4245611429214478,
"learning_rate": 9.828422440727152e-06,
"loss": 0.5304,
"step": 2630
},
{
"epoch": 0.528,
"grad_norm": 1.4869318008422852,
"learning_rate": 9.82538741267157e-06,
"loss": 0.5631,
"step": 2640
},
{
"epoch": 0.53,
"grad_norm": 1.0916733741760254,
"learning_rate": 9.822326253124436e-06,
"loss": 0.4249,
"step": 2650
},
{
"epoch": 0.532,
"grad_norm": 1.3281223773956299,
"learning_rate": 9.819238978663212e-06,
"loss": 0.4827,
"step": 2660
},
{
"epoch": 0.534,
"grad_norm": 1.1685035228729248,
"learning_rate": 9.816125606006777e-06,
"loss": 0.547,
"step": 2670
},
{
"epoch": 0.536,
"grad_norm": 1.309923768043518,
"learning_rate": 9.812986152015349e-06,
"loss": 0.4927,
"step": 2680
},
{
"epoch": 0.538,
"grad_norm": 1.3431212902069092,
"learning_rate": 9.809820633690383e-06,
"loss": 0.5215,
"step": 2690
},
{
"epoch": 0.54,
"grad_norm": 1.6138921976089478,
"learning_rate": 9.806629068174486e-06,
"loss": 0.4668,
"step": 2700
},
{
"epoch": 0.542,
"grad_norm": 1.1583821773529053,
"learning_rate": 9.803411472751321e-06,
"loss": 0.4882,
"step": 2710
},
{
"epoch": 0.544,
"grad_norm": 1.1183314323425293,
"learning_rate": 9.800167864845513e-06,
"loss": 0.5254,
"step": 2720
},
{
"epoch": 0.546,
"grad_norm": 1.4351534843444824,
"learning_rate": 9.796898262022555e-06,
"loss": 0.5631,
"step": 2730
},
{
"epoch": 0.548,
"grad_norm": 1.114702582359314,
"learning_rate": 9.793602681988714e-06,
"loss": 0.5087,
"step": 2740
},
{
"epoch": 0.55,
"grad_norm": 1.6921616792678833,
"learning_rate": 9.790281142590937e-06,
"loss": 0.5482,
"step": 2750
},
{
"epoch": 0.552,
"grad_norm": 1.3350164890289307,
"learning_rate": 9.786933661816747e-06,
"loss": 0.5148,
"step": 2760
},
{
"epoch": 0.554,
"grad_norm": 1.3733949661254883,
"learning_rate": 9.783560257794153e-06,
"loss": 0.4336,
"step": 2770
},
{
"epoch": 0.556,
"grad_norm": 1.232109546661377,
"learning_rate": 9.78016094879155e-06,
"loss": 0.4304,
"step": 2780
},
{
"epoch": 0.558,
"grad_norm": 1.4043456315994263,
"learning_rate": 9.776735753217618e-06,
"loss": 0.4891,
"step": 2790
},
{
"epoch": 0.56,
"grad_norm": 1.1592031717300415,
"learning_rate": 9.773284689621223e-06,
"loss": 0.4892,
"step": 2800
},
{
"epoch": 0.562,
"grad_norm": 1.1605182886123657,
"learning_rate": 9.76980777669132e-06,
"loss": 0.4433,
"step": 2810
},
{
"epoch": 0.564,
"grad_norm": 1.1218903064727783,
"learning_rate": 9.766305033256847e-06,
"loss": 0.5072,
"step": 2820
},
{
"epoch": 0.566,
"grad_norm": 1.4614591598510742,
"learning_rate": 9.762776478286622e-06,
"loss": 0.4988,
"step": 2830
},
{
"epoch": 0.568,
"grad_norm": 1.0388537645339966,
"learning_rate": 9.75922213088925e-06,
"loss": 0.4314,
"step": 2840
},
{
"epoch": 0.57,
"grad_norm": 1.3844202756881714,
"learning_rate": 9.755642010313007e-06,
"loss": 0.4516,
"step": 2850
},
{
"epoch": 0.572,
"grad_norm": 1.5695295333862305,
"learning_rate": 9.752036135945743e-06,
"loss": 0.4761,
"step": 2860
},
{
"epoch": 0.574,
"grad_norm": 1.3042584657669067,
"learning_rate": 9.748404527314782e-06,
"loss": 0.4851,
"step": 2870
},
{
"epoch": 0.576,
"grad_norm": 1.3485811948776245,
"learning_rate": 9.744747204086795e-06,
"loss": 0.4127,
"step": 2880
},
{
"epoch": 0.578,
"grad_norm": 1.074900507926941,
"learning_rate": 9.741064186067723e-06,
"loss": 0.4803,
"step": 2890
},
{
"epoch": 0.58,
"grad_norm": 1.3059568405151367,
"learning_rate": 9.73735549320265e-06,
"loss": 0.4279,
"step": 2900
},
{
"epoch": 0.582,
"grad_norm": 1.1361761093139648,
"learning_rate": 9.733621145575697e-06,
"loss": 0.5364,
"step": 2910
},
{
"epoch": 0.584,
"grad_norm": 1.4308760166168213,
"learning_rate": 9.72986116340992e-06,
"loss": 0.5978,
"step": 2920
},
{
"epoch": 0.586,
"grad_norm": 0.9736759066581726,
"learning_rate": 9.726075567067193e-06,
"loss": 0.4471,
"step": 2930
},
{
"epoch": 0.588,
"grad_norm": 1.555156946182251,
"learning_rate": 9.722264377048105e-06,
"loss": 0.526,
"step": 2940
},
{
"epoch": 0.59,
"grad_norm": 1.1218699216842651,
"learning_rate": 9.718427613991848e-06,
"loss": 0.4025,
"step": 2950
},
{
"epoch": 0.592,
"grad_norm": 1.1059517860412598,
"learning_rate": 9.714565298676093e-06,
"loss": 0.4503,
"step": 2960
},
{
"epoch": 0.594,
"grad_norm": 1.3590087890625,
"learning_rate": 9.710677452016898e-06,
"loss": 0.5018,
"step": 2970
},
{
"epoch": 0.596,
"grad_norm": 0.9605317115783691,
"learning_rate": 9.706764095068579e-06,
"loss": 0.474,
"step": 2980
},
{
"epoch": 0.598,
"grad_norm": 0.9634230136871338,
"learning_rate": 9.702825249023597e-06,
"loss": 0.4051,
"step": 2990
},
{
"epoch": 0.6,
"grad_norm": 1.0012608766555786,
"learning_rate": 9.698860935212455e-06,
"loss": 0.5104,
"step": 3000
},
{
"epoch": 0.602,
"grad_norm": 1.254647970199585,
"learning_rate": 9.69487117510357e-06,
"loss": 0.4357,
"step": 3010
},
{
"epoch": 0.604,
"grad_norm": 1.2579063177108765,
"learning_rate": 9.69085599030316e-06,
"loss": 0.4413,
"step": 3020
},
{
"epoch": 0.606,
"grad_norm": 1.1054672002792358,
"learning_rate": 9.686815402555133e-06,
"loss": 0.6412,
"step": 3030
},
{
"epoch": 0.608,
"grad_norm": 1.3997899293899536,
"learning_rate": 9.682749433740963e-06,
"loss": 0.4615,
"step": 3040
},
{
"epoch": 0.61,
"grad_norm": 1.025417685508728,
"learning_rate": 9.678658105879568e-06,
"loss": 0.4214,
"step": 3050
},
{
"epoch": 0.612,
"grad_norm": 1.0472135543823242,
"learning_rate": 9.674541441127202e-06,
"loss": 0.5116,
"step": 3060
},
{
"epoch": 0.614,
"grad_norm": 1.078364372253418,
"learning_rate": 9.670399461777328e-06,
"loss": 0.4569,
"step": 3070
},
{
"epoch": 0.616,
"grad_norm": 1.3852829933166504,
"learning_rate": 9.666232190260496e-06,
"loss": 0.53,
"step": 3080
},
{
"epoch": 0.618,
"grad_norm": 1.3731964826583862,
"learning_rate": 9.662039649144224e-06,
"loss": 0.4696,
"step": 3090
},
{
"epoch": 0.62,
"grad_norm": 1.291717290878296,
"learning_rate": 9.65782186113288e-06,
"loss": 0.4263,
"step": 3100
},
{
"epoch": 0.622,
"grad_norm": 1.3405389785766602,
"learning_rate": 9.653578849067542e-06,
"loss": 0.4435,
"step": 3110
},
{
"epoch": 0.624,
"grad_norm": 1.226496696472168,
"learning_rate": 9.649310635925904e-06,
"loss": 0.4492,
"step": 3120
},
{
"epoch": 0.626,
"grad_norm": 1.15779447555542,
"learning_rate": 9.645017244822124e-06,
"loss": 0.5387,
"step": 3130
},
{
"epoch": 0.628,
"grad_norm": 1.0545017719268799,
"learning_rate": 9.640698699006708e-06,
"loss": 0.4616,
"step": 3140
},
{
"epoch": 0.63,
"grad_norm": 1.0854800939559937,
"learning_rate": 9.636355021866388e-06,
"loss": 0.518,
"step": 3150
},
{
"epoch": 0.632,
"grad_norm": 1.292471170425415,
"learning_rate": 9.631986236923998e-06,
"loss": 0.4888,
"step": 3160
},
{
"epoch": 0.634,
"grad_norm": 1.2289882898330688,
"learning_rate": 9.62759236783833e-06,
"loss": 0.5193,
"step": 3170
},
{
"epoch": 0.636,
"grad_norm": 1.4421725273132324,
"learning_rate": 9.623173438404027e-06,
"loss": 0.5314,
"step": 3180
},
{
"epoch": 0.638,
"grad_norm": 1.4118566513061523,
"learning_rate": 9.61872947255144e-06,
"loss": 0.5525,
"step": 3190
},
{
"epoch": 0.64,
"grad_norm": 1.1490426063537598,
"learning_rate": 9.614260494346505e-06,
"loss": 0.4832,
"step": 3200
},
{
"epoch": 0.642,
"grad_norm": 1.3505995273590088,
"learning_rate": 9.609766527990604e-06,
"loss": 0.4742,
"step": 3210
},
{
"epoch": 0.644,
"grad_norm": 1.4278666973114014,
"learning_rate": 9.605247597820448e-06,
"loss": 0.458,
"step": 3220
},
{
"epoch": 0.646,
"grad_norm": 1.5824147462844849,
"learning_rate": 9.600703728307935e-06,
"loss": 0.4327,
"step": 3230
},
{
"epoch": 0.648,
"grad_norm": 1.1965423822402954,
"learning_rate": 9.596134944060018e-06,
"loss": 0.4241,
"step": 3240
},
{
"epoch": 0.65,
"grad_norm": 1.3452823162078857,
"learning_rate": 9.591541269818574e-06,
"loss": 0.4312,
"step": 3250
},
{
"epoch": 0.652,
"grad_norm": 1.6801984310150146,
"learning_rate": 9.586922730460273e-06,
"loss": 0.4858,
"step": 3260
},
{
"epoch": 0.654,
"grad_norm": 1.521422266960144,
"learning_rate": 9.582279350996437e-06,
"loss": 0.4886,
"step": 3270
},
{
"epoch": 0.656,
"grad_norm": 1.0848934650421143,
"learning_rate": 9.577611156572908e-06,
"loss": 0.5233,
"step": 3280
},
{
"epoch": 0.658,
"grad_norm": 1.131194829940796,
"learning_rate": 9.572918172469912e-06,
"loss": 0.5093,
"step": 3290
},
{
"epoch": 0.66,
"grad_norm": 0.7532334923744202,
"learning_rate": 9.568200424101918e-06,
"loss": 0.4439,
"step": 3300
},
{
"epoch": 0.662,
"grad_norm": 0.992387056350708,
"learning_rate": 9.563457937017514e-06,
"loss": 0.4405,
"step": 3310
},
{
"epoch": 0.664,
"grad_norm": 1.5561617612838745,
"learning_rate": 9.558690736899248e-06,
"loss": 0.5558,
"step": 3320
},
{
"epoch": 0.666,
"grad_norm": 0.9538534283638,
"learning_rate": 9.553898849563502e-06,
"loss": 0.4767,
"step": 3330
},
{
"epoch": 0.668,
"grad_norm": 1.5033385753631592,
"learning_rate": 9.549082300960351e-06,
"loss": 0.4721,
"step": 3340
},
{
"epoch": 0.67,
"grad_norm": 1.4979592561721802,
"learning_rate": 9.544241117173422e-06,
"loss": 0.4778,
"step": 3350
},
{
"epoch": 0.672,
"grad_norm": 1.2602815628051758,
"learning_rate": 9.539375324419748e-06,
"loss": 0.4715,
"step": 3360
},
{
"epoch": 0.674,
"grad_norm": 1.1525267362594604,
"learning_rate": 9.534484949049636e-06,
"loss": 0.5271,
"step": 3370
},
{
"epoch": 0.676,
"grad_norm": 1.18638014793396,
"learning_rate": 9.529570017546512e-06,
"loss": 0.5016,
"step": 3380
},
{
"epoch": 0.678,
"grad_norm": 1.3862594366073608,
"learning_rate": 9.524630556526788e-06,
"loss": 0.4872,
"step": 3390
},
{
"epoch": 0.68,
"grad_norm": 1.5191760063171387,
"learning_rate": 9.51966659273971e-06,
"loss": 0.4688,
"step": 3400
},
{
"epoch": 0.682,
"grad_norm": 1.176364541053772,
"learning_rate": 9.514678153067218e-06,
"loss": 0.4674,
"step": 3410
},
{
"epoch": 0.684,
"grad_norm": 1.2833201885223389,
"learning_rate": 9.509665264523803e-06,
"loss": 0.4414,
"step": 3420
},
{
"epoch": 0.686,
"grad_norm": 0.9362300634384155,
"learning_rate": 9.504627954256352e-06,
"loss": 0.368,
"step": 3430
},
{
"epoch": 0.688,
"grad_norm": 1.2998939752578735,
"learning_rate": 9.499566249544006e-06,
"loss": 0.4505,
"step": 3440
},
{
"epoch": 0.69,
"grad_norm": 0.7536768913269043,
"learning_rate": 9.494480177798013e-06,
"loss": 0.5055,
"step": 3450
},
{
"epoch": 0.692,
"grad_norm": 1.1865419149398804,
"learning_rate": 9.489369766561584e-06,
"loss": 0.4671,
"step": 3460
},
{
"epoch": 0.694,
"grad_norm": 1.128171443939209,
"learning_rate": 9.48423504350973e-06,
"loss": 0.4721,
"step": 3470
},
{
"epoch": 0.696,
"grad_norm": 1.1129778623580933,
"learning_rate": 9.479076036449125e-06,
"loss": 0.5855,
"step": 3480
},
{
"epoch": 0.698,
"grad_norm": 1.0635051727294922,
"learning_rate": 9.473892773317952e-06,
"loss": 0.4509,
"step": 3490
},
{
"epoch": 0.7,
"grad_norm": 0.9211987257003784,
"learning_rate": 9.468685282185745e-06,
"loss": 0.4543,
"step": 3500
},
{
"epoch": 0.702,
"grad_norm": 1.2037076950073242,
"learning_rate": 9.463453591253253e-06,
"loss": 0.4839,
"step": 3510
},
{
"epoch": 0.704,
"grad_norm": 0.9373881220817566,
"learning_rate": 9.458197728852268e-06,
"loss": 0.4215,
"step": 3520
},
{
"epoch": 0.706,
"grad_norm": 0.9381740093231201,
"learning_rate": 9.452917723445484e-06,
"loss": 0.4598,
"step": 3530
},
{
"epoch": 0.708,
"grad_norm": 1.6144822835922241,
"learning_rate": 9.447613603626337e-06,
"loss": 0.5762,
"step": 3540
},
{
"epoch": 0.71,
"grad_norm": 1.3809951543807983,
"learning_rate": 9.44228539811886e-06,
"loss": 0.4934,
"step": 3550
},
{
"epoch": 0.712,
"grad_norm": 1.1491771936416626,
"learning_rate": 9.436933135777511e-06,
"loss": 0.4639,
"step": 3560
},
{
"epoch": 0.714,
"grad_norm": 1.0632610321044922,
"learning_rate": 9.431556845587029e-06,
"loss": 0.4966,
"step": 3570
},
{
"epoch": 0.716,
"grad_norm": 1.2312216758728027,
"learning_rate": 9.426156556662276e-06,
"loss": 0.4287,
"step": 3580
},
{
"epoch": 0.718,
"grad_norm": 1.2142528295516968,
"learning_rate": 9.420732298248077e-06,
"loss": 0.4216,
"step": 3590
},
{
"epoch": 0.72,
"grad_norm": 0.7389311194419861,
"learning_rate": 9.41528409971906e-06,
"loss": 0.4317,
"step": 3600
},
{
"epoch": 0.722,
"grad_norm": 1.421856164932251,
"learning_rate": 9.409811990579498e-06,
"loss": 0.4798,
"step": 3610
},
{
"epoch": 0.724,
"grad_norm": 1.5812768936157227,
"learning_rate": 9.404316000463152e-06,
"loss": 0.5848,
"step": 3620
},
{
"epoch": 0.726,
"grad_norm": 0.9333869218826294,
"learning_rate": 9.398796159133108e-06,
"loss": 0.4232,
"step": 3630
},
{
"epoch": 0.728,
"grad_norm": 1.2076997756958008,
"learning_rate": 9.393252496481615e-06,
"loss": 0.5026,
"step": 3640
},
{
"epoch": 0.73,
"grad_norm": 1.4881305694580078,
"learning_rate": 9.387685042529926e-06,
"loss": 0.525,
"step": 3650
},
{
"epoch": 0.732,
"grad_norm": 1.2889773845672607,
"learning_rate": 9.382093827428135e-06,
"loss": 0.4729,
"step": 3660
},
{
"epoch": 0.734,
"grad_norm": 1.4050449132919312,
"learning_rate": 9.376478881455008e-06,
"loss": 0.4944,
"step": 3670
},
{
"epoch": 0.736,
"grad_norm": 0.8713739514350891,
"learning_rate": 9.370840235017829e-06,
"loss": 0.4217,
"step": 3680
},
{
"epoch": 0.738,
"grad_norm": 1.031435251235962,
"learning_rate": 9.365177918652226e-06,
"loss": 0.4611,
"step": 3690
},
{
"epoch": 0.74,
"grad_norm": 0.8381909132003784,
"learning_rate": 9.35949196302201e-06,
"loss": 0.4212,
"step": 3700
},
{
"epoch": 0.742,
"grad_norm": 1.1888693571090698,
"learning_rate": 9.353782398919012e-06,
"loss": 0.4543,
"step": 3710
},
{
"epoch": 0.744,
"grad_norm": 1.4899085760116577,
"learning_rate": 9.348049257262908e-06,
"loss": 0.5147,
"step": 3720
},
{
"epoch": 0.746,
"grad_norm": 1.2330344915390015,
"learning_rate": 9.342292569101061e-06,
"loss": 0.6145,
"step": 3730
},
{
"epoch": 0.748,
"grad_norm": 1.0226585865020752,
"learning_rate": 9.336512365608343e-06,
"loss": 0.4879,
"step": 3740
},
{
"epoch": 0.75,
"grad_norm": 1.1054939031600952,
"learning_rate": 9.330708678086975e-06,
"loss": 0.4206,
"step": 3750
},
{
"epoch": 0.752,
"grad_norm": 1.4774991273880005,
"learning_rate": 9.324881537966355e-06,
"loss": 0.4093,
"step": 3760
},
{
"epoch": 0.754,
"grad_norm": 1.1856415271759033,
"learning_rate": 9.319030976802881e-06,
"loss": 0.4614,
"step": 3770
},
{
"epoch": 0.756,
"grad_norm": 1.3986643552780151,
"learning_rate": 9.313157026279792e-06,
"loss": 0.5174,
"step": 3780
},
{
"epoch": 0.758,
"grad_norm": 1.2036128044128418,
"learning_rate": 9.307259718206984e-06,
"loss": 0.464,
"step": 3790
},
{
"epoch": 0.76,
"grad_norm": 0.8177152276039124,
"learning_rate": 9.301339084520853e-06,
"loss": 0.4331,
"step": 3800
},
{
"epoch": 0.762,
"grad_norm": 1.1826705932617188,
"learning_rate": 9.295395157284103e-06,
"loss": 0.4363,
"step": 3810
},
{
"epoch": 0.764,
"grad_norm": 1.0908697843551636,
"learning_rate": 9.289427968685588e-06,
"loss": 0.4087,
"step": 3820
},
{
"epoch": 0.766,
"grad_norm": 0.9723864793777466,
"learning_rate": 9.28343755104013e-06,
"loss": 0.3952,
"step": 3830
},
{
"epoch": 0.768,
"grad_norm": 1.1800845861434937,
"learning_rate": 9.277423936788348e-06,
"loss": 0.4966,
"step": 3840
},
{
"epoch": 0.77,
"grad_norm": 1.5162053108215332,
"learning_rate": 9.271387158496477e-06,
"loss": 0.5525,
"step": 3850
},
{
"epoch": 0.772,
"grad_norm": 1.1737457513809204,
"learning_rate": 9.265327248856198e-06,
"loss": 0.4374,
"step": 3860
},
{
"epoch": 0.774,
"grad_norm": 1.1317731142044067,
"learning_rate": 9.259244240684457e-06,
"loss": 0.4321,
"step": 3870
},
{
"epoch": 0.776,
"grad_norm": 1.3863508701324463,
"learning_rate": 9.25313816692329e-06,
"loss": 0.5441,
"step": 3880
},
{
"epoch": 0.778,
"grad_norm": 1.0942466259002686,
"learning_rate": 9.247009060639637e-06,
"loss": 0.5124,
"step": 3890
},
{
"epoch": 0.78,
"grad_norm": 1.3743023872375488,
"learning_rate": 9.240856955025175e-06,
"loss": 0.4357,
"step": 3900
},
{
"epoch": 0.782,
"grad_norm": 1.1339491605758667,
"learning_rate": 9.234681883396129e-06,
"loss": 0.5066,
"step": 3910
},
{
"epoch": 0.784,
"grad_norm": 1.9541819095611572,
"learning_rate": 9.228483879193096e-06,
"loss": 0.4864,
"step": 3920
},
{
"epoch": 0.786,
"grad_norm": 1.257652759552002,
"learning_rate": 9.22226297598086e-06,
"loss": 0.4361,
"step": 3930
},
{
"epoch": 0.788,
"grad_norm": 1.4581412076950073,
"learning_rate": 9.216019207448216e-06,
"loss": 0.5005,
"step": 3940
},
{
"epoch": 0.79,
"grad_norm": 1.6676081418991089,
"learning_rate": 9.209752607407784e-06,
"loss": 0.4307,
"step": 3950
},
{
"epoch": 0.792,
"grad_norm": 0.898801863193512,
"learning_rate": 9.203463209795822e-06,
"loss": 0.487,
"step": 3960
},
{
"epoch": 0.794,
"grad_norm": 1.6811095476150513,
"learning_rate": 9.197151048672051e-06,
"loss": 0.5448,
"step": 3970
},
{
"epoch": 0.796,
"grad_norm": 1.222164511680603,
"learning_rate": 9.190816158219462e-06,
"loss": 0.4603,
"step": 3980
},
{
"epoch": 0.798,
"grad_norm": 1.161139726638794,
"learning_rate": 9.18445857274414e-06,
"loss": 0.5216,
"step": 3990
},
{
"epoch": 0.8,
"grad_norm": 1.3466241359710693,
"learning_rate": 9.178078326675069e-06,
"loss": 0.5273,
"step": 4000
},
{
"epoch": 0.802,
"grad_norm": 1.1372344493865967,
"learning_rate": 9.171675454563949e-06,
"loss": 0.4842,
"step": 4010
},
{
"epoch": 0.804,
"grad_norm": 1.10002601146698,
"learning_rate": 9.165249991085012e-06,
"loss": 0.5429,
"step": 4020
},
{
"epoch": 0.806,
"grad_norm": 0.8916090726852417,
"learning_rate": 9.158801971034832e-06,
"loss": 0.5703,
"step": 4030
},
{
"epoch": 0.808,
"grad_norm": 1.131230115890503,
"learning_rate": 9.152331429332136e-06,
"loss": 0.4555,
"step": 4040
},
{
"epoch": 0.81,
"grad_norm": 0.9842562079429626,
"learning_rate": 9.14583840101761e-06,
"loss": 0.4475,
"step": 4050
},
{
"epoch": 0.812,
"grad_norm": 1.0680824518203735,
"learning_rate": 9.139322921253724e-06,
"loss": 0.4471,
"step": 4060
},
{
"epoch": 0.814,
"grad_norm": 1.2112277746200562,
"learning_rate": 9.132785025324524e-06,
"loss": 0.5342,
"step": 4070
},
{
"epoch": 0.816,
"grad_norm": 1.359027624130249,
"learning_rate": 9.12622474863545e-06,
"loss": 0.5038,
"step": 4080
},
{
"epoch": 0.818,
"grad_norm": 1.2182432413101196,
"learning_rate": 9.119642126713147e-06,
"loss": 0.4761,
"step": 4090
},
{
"epoch": 0.82,
"grad_norm": 0.905654788017273,
"learning_rate": 9.113037195205267e-06,
"loss": 0.4434,
"step": 4100
},
{
"epoch": 0.822,
"grad_norm": 1.273230791091919,
"learning_rate": 9.106409989880274e-06,
"loss": 0.4119,
"step": 4110
},
{
"epoch": 0.824,
"grad_norm": 1.1936393976211548,
"learning_rate": 9.099760546627262e-06,
"loss": 0.5159,
"step": 4120
},
{
"epoch": 0.826,
"grad_norm": 1.196582555770874,
"learning_rate": 9.093088901455746e-06,
"loss": 0.4788,
"step": 4130
},
{
"epoch": 0.828,
"grad_norm": 1.420160174369812,
"learning_rate": 9.086395090495475e-06,
"loss": 0.4298,
"step": 4140
},
{
"epoch": 0.83,
"grad_norm": 1.2535393238067627,
"learning_rate": 9.079679149996235e-06,
"loss": 0.5149,
"step": 4150
},
{
"epoch": 0.832,
"grad_norm": 0.9435884356498718,
"learning_rate": 9.072941116327654e-06,
"loss": 0.4495,
"step": 4160
},
{
"epoch": 0.834,
"grad_norm": 1.1288723945617676,
"learning_rate": 9.066181025979006e-06,
"loss": 0.4988,
"step": 4170
},
{
"epoch": 0.836,
"grad_norm": 1.53179132938385,
"learning_rate": 9.059398915559005e-06,
"loss": 0.4834,
"step": 4180
},
{
"epoch": 0.838,
"grad_norm": 1.48374342918396,
"learning_rate": 9.052594821795616e-06,
"loss": 0.5016,
"step": 4190
},
{
"epoch": 0.84,
"grad_norm": 1.3857916593551636,
"learning_rate": 9.045768781535857e-06,
"loss": 0.5277,
"step": 4200
},
{
"epoch": 0.842,
"grad_norm": 1.1805408000946045,
"learning_rate": 9.038920831745587e-06,
"loss": 0.432,
"step": 4210
},
{
"epoch": 0.844,
"grad_norm": 1.1315293312072754,
"learning_rate": 9.032051009509324e-06,
"loss": 0.4377,
"step": 4220
},
{
"epoch": 0.846,
"grad_norm": 0.9386146068572998,
"learning_rate": 9.025159352030024e-06,
"loss": 0.4993,
"step": 4230
},
{
"epoch": 0.848,
"grad_norm": 1.187091588973999,
"learning_rate": 9.0182458966289e-06,
"loss": 0.5185,
"step": 4240
},
{
"epoch": 0.85,
"grad_norm": 1.0254029035568237,
"learning_rate": 9.011310680745203e-06,
"loss": 0.471,
"step": 4250
},
{
"epoch": 0.852,
"grad_norm": 1.1841859817504883,
"learning_rate": 9.004353741936028e-06,
"loss": 0.4047,
"step": 4260
},
{
"epoch": 0.854,
"grad_norm": 1.25779390335083,
"learning_rate": 8.99737511787611e-06,
"loss": 0.4827,
"step": 4270
},
{
"epoch": 0.856,
"grad_norm": 1.2962913513183594,
"learning_rate": 8.990374846357616e-06,
"loss": 0.4792,
"step": 4280
},
{
"epoch": 0.858,
"grad_norm": 1.1236214637756348,
"learning_rate": 8.98335296528995e-06,
"loss": 0.5092,
"step": 4290
},
{
"epoch": 0.86,
"grad_norm": 1.0570876598358154,
"learning_rate": 8.97630951269953e-06,
"loss": 0.4761,
"step": 4300
},
{
"epoch": 0.862,
"grad_norm": 1.0793334245681763,
"learning_rate": 8.969244526729599e-06,
"loss": 0.425,
"step": 4310
},
{
"epoch": 0.864,
"grad_norm": 1.0466865301132202,
"learning_rate": 8.962158045640014e-06,
"loss": 0.4867,
"step": 4320
},
{
"epoch": 0.866,
"grad_norm": 1.8099735975265503,
"learning_rate": 8.955050107807035e-06,
"loss": 0.453,
"step": 4330
},
{
"epoch": 0.868,
"grad_norm": 1.0396076440811157,
"learning_rate": 8.947920751723119e-06,
"loss": 0.4536,
"step": 4340
},
{
"epoch": 0.87,
"grad_norm": 1.1914063692092896,
"learning_rate": 8.940770015996707e-06,
"loss": 0.448,
"step": 4350
},
{
"epoch": 0.872,
"grad_norm": 1.0686973333358765,
"learning_rate": 8.933597939352031e-06,
"loss": 0.4932,
"step": 4360
},
{
"epoch": 0.874,
"grad_norm": 0.9669390320777893,
"learning_rate": 8.926404560628882e-06,
"loss": 0.4812,
"step": 4370
},
{
"epoch": 0.876,
"grad_norm": 1.4926239252090454,
"learning_rate": 8.919189918782419e-06,
"loss": 0.4337,
"step": 4380
},
{
"epoch": 0.878,
"grad_norm": 1.3908580541610718,
"learning_rate": 8.911954052882941e-06,
"loss": 0.4595,
"step": 4390
},
{
"epoch": 0.88,
"grad_norm": 1.088005781173706,
"learning_rate": 8.904697002115693e-06,
"loss": 0.4685,
"step": 4400
},
{
"epoch": 0.882,
"grad_norm": 1.2951703071594238,
"learning_rate": 8.89741880578064e-06,
"loss": 0.5138,
"step": 4410
},
{
"epoch": 0.884,
"grad_norm": 1.3422932624816895,
"learning_rate": 8.890119503292258e-06,
"loss": 0.5339,
"step": 4420
},
{
"epoch": 0.886,
"grad_norm": 1.6533243656158447,
"learning_rate": 8.882799134179326e-06,
"loss": 0.4854,
"step": 4430
},
{
"epoch": 0.888,
"grad_norm": 1.5386734008789062,
"learning_rate": 8.875457738084706e-06,
"loss": 0.4791,
"step": 4440
},
{
"epoch": 0.89,
"grad_norm": 1.3411436080932617,
"learning_rate": 8.868095354765125e-06,
"loss": 0.4922,
"step": 4450
},
{
"epoch": 0.892,
"grad_norm": 0.9799019694328308,
"learning_rate": 8.860712024090971e-06,
"loss": 0.4265,
"step": 4460
},
{
"epoch": 0.894,
"grad_norm": 1.7590001821517944,
"learning_rate": 8.853307786046073e-06,
"loss": 0.5033,
"step": 4470
},
{
"epoch": 0.896,
"grad_norm": 1.1440484523773193,
"learning_rate": 8.84588268072747e-06,
"loss": 0.5036,
"step": 4480
},
{
"epoch": 0.898,
"grad_norm": 1.0841706991195679,
"learning_rate": 8.838436748345217e-06,
"loss": 0.5344,
"step": 4490
},
{
"epoch": 0.9,
"grad_norm": 1.1884775161743164,
"learning_rate": 8.830970029222152e-06,
"loss": 0.3504,
"step": 4500
},
{
"epoch": 0.902,
"grad_norm": 0.9909444451332092,
"learning_rate": 8.823482563793687e-06,
"loss": 0.4838,
"step": 4510
},
{
"epoch": 0.904,
"grad_norm": 1.079696774482727,
"learning_rate": 8.815974392607573e-06,
"loss": 0.4601,
"step": 4520
},
{
"epoch": 0.906,
"grad_norm": 1.1453964710235596,
"learning_rate": 8.808445556323703e-06,
"loss": 0.4623,
"step": 4530
},
{
"epoch": 0.908,
"grad_norm": 1.1338945627212524,
"learning_rate": 8.80089609571387e-06,
"loss": 0.5288,
"step": 4540
},
{
"epoch": 0.91,
"grad_norm": 1.3115707635879517,
"learning_rate": 8.79332605166157e-06,
"loss": 0.5267,
"step": 4550
},
{
"epoch": 0.912,
"grad_norm": 1.3860828876495361,
"learning_rate": 8.785735465161752e-06,
"loss": 0.4985,
"step": 4560
},
{
"epoch": 0.914,
"grad_norm": 1.2067689895629883,
"learning_rate": 8.778124377320619e-06,
"loss": 0.4699,
"step": 4570
},
{
"epoch": 0.916,
"grad_norm": 1.4744197130203247,
"learning_rate": 8.770492829355395e-06,
"loss": 0.4483,
"step": 4580
},
{
"epoch": 0.918,
"grad_norm": 1.1251822710037231,
"learning_rate": 8.762840862594106e-06,
"loss": 0.4664,
"step": 4590
},
{
"epoch": 0.92,
"grad_norm": 0.8830546736717224,
"learning_rate": 8.755168518475351e-06,
"loss": 0.533,
"step": 4600
},
{
"epoch": 0.922,
"grad_norm": 1.2034662961959839,
"learning_rate": 8.747475838548088e-06,
"loss": 0.4262,
"step": 4610
},
{
"epoch": 0.924,
"grad_norm": 1.3047678470611572,
"learning_rate": 8.739762864471392e-06,
"loss": 0.5621,
"step": 4620
},
{
"epoch": 0.926,
"grad_norm": 1.0258435010910034,
"learning_rate": 8.732029638014249e-06,
"loss": 0.5337,
"step": 4630
},
{
"epoch": 0.928,
"grad_norm": 1.1698180437088013,
"learning_rate": 8.724276201055311e-06,
"loss": 0.4402,
"step": 4640
},
{
"epoch": 0.93,
"grad_norm": 1.0054242610931396,
"learning_rate": 8.716502595582685e-06,
"loss": 0.5391,
"step": 4650
},
{
"epoch": 0.932,
"grad_norm": 0.8926470875740051,
"learning_rate": 8.708708863693696e-06,
"loss": 0.4663,
"step": 4660
},
{
"epoch": 0.934,
"grad_norm": 1.3373093605041504,
"learning_rate": 8.700895047594664e-06,
"loss": 0.5086,
"step": 4670
},
{
"epoch": 0.936,
"grad_norm": 1.17959725856781,
"learning_rate": 8.693061189600671e-06,
"loss": 0.4284,
"step": 4680
},
{
"epoch": 0.938,
"grad_norm": 1.0325318574905396,
"learning_rate": 8.685207332135337e-06,
"loss": 0.5204,
"step": 4690
},
{
"epoch": 0.94,
"grad_norm": 0.8622048497200012,
"learning_rate": 8.677333517730582e-06,
"loss": 0.4318,
"step": 4700
},
{
"epoch": 0.942,
"grad_norm": 1.214130163192749,
"learning_rate": 8.669439789026409e-06,
"loss": 0.4886,
"step": 4710
},
{
"epoch": 0.944,
"grad_norm": 1.3919086456298828,
"learning_rate": 8.66152618877066e-06,
"loss": 0.4725,
"step": 4720
},
{
"epoch": 0.946,
"grad_norm": 1.263291835784912,
"learning_rate": 8.65359275981879e-06,
"loss": 0.4854,
"step": 4730
},
{
"epoch": 0.948,
"grad_norm": 1.5392141342163086,
"learning_rate": 8.645639545133638e-06,
"loss": 0.5634,
"step": 4740
},
{
"epoch": 0.95,
"grad_norm": 1.246302843093872,
"learning_rate": 8.637666587785185e-06,
"loss": 0.4597,
"step": 4750
},
{
"epoch": 0.952,
"grad_norm": 1.0350593328475952,
"learning_rate": 8.629673930950335e-06,
"loss": 0.4152,
"step": 4760
},
{
"epoch": 0.954,
"grad_norm": 1.295961618423462,
"learning_rate": 8.621661617912665e-06,
"loss": 0.5035,
"step": 4770
},
{
"epoch": 0.956,
"grad_norm": 1.3225280046463013,
"learning_rate": 8.613629692062204e-06,
"loss": 0.5491,
"step": 4780
},
{
"epoch": 0.958,
"grad_norm": 1.495888352394104,
"learning_rate": 8.60557819689519e-06,
"loss": 0.5184,
"step": 4790
},
{
"epoch": 0.96,
"grad_norm": 1.549146056175232,
"learning_rate": 8.597507176013839e-06,
"loss": 0.4976,
"step": 4800
},
{
"epoch": 0.962,
"grad_norm": 1.1670464277267456,
"learning_rate": 8.589416673126104e-06,
"loss": 0.4262,
"step": 4810
},
{
"epoch": 0.964,
"grad_norm": 1.1644515991210938,
"learning_rate": 8.581306732045443e-06,
"loss": 0.4646,
"step": 4820
},
{
"epoch": 0.966,
"grad_norm": 0.8502632975578308,
"learning_rate": 8.57317739669058e-06,
"loss": 0.4702,
"step": 4830
},
{
"epoch": 0.968,
"grad_norm": 1.3403635025024414,
"learning_rate": 8.565028711085266e-06,
"loss": 0.4864,
"step": 4840
},
{
"epoch": 0.97,
"grad_norm": 1.1791622638702393,
"learning_rate": 8.556860719358045e-06,
"loss": 0.41,
"step": 4850
},
{
"epoch": 0.972,
"grad_norm": 1.5249660015106201,
"learning_rate": 8.548673465742006e-06,
"loss": 0.4471,
"step": 4860
},
{
"epoch": 0.974,
"grad_norm": 1.3172943592071533,
"learning_rate": 8.540466994574556e-06,
"loss": 0.4734,
"step": 4870
},
{
"epoch": 0.976,
"grad_norm": 1.012715458869934,
"learning_rate": 8.532241350297167e-06,
"loss": 0.4765,
"step": 4880
},
{
"epoch": 0.978,
"grad_norm": 0.9505335092544556,
"learning_rate": 8.523996577455144e-06,
"loss": 0.4513,
"step": 4890
},
{
"epoch": 0.98,
"grad_norm": 0.9823141098022461,
"learning_rate": 8.515732720697383e-06,
"loss": 0.5406,
"step": 4900
},
{
"epoch": 0.982,
"grad_norm": 1.1936461925506592,
"learning_rate": 8.507449824776125e-06,
"loss": 0.4852,
"step": 4910
},
{
"epoch": 0.984,
"grad_norm": 1.1408801078796387,
"learning_rate": 8.499147934546716e-06,
"loss": 0.454,
"step": 4920
},
{
"epoch": 0.986,
"grad_norm": 1.1299476623535156,
"learning_rate": 8.490827094967364e-06,
"loss": 0.4875,
"step": 4930
},
{
"epoch": 0.988,
"grad_norm": 0.7930355668067932,
"learning_rate": 8.482487351098899e-06,
"loss": 0.4449,
"step": 4940
},
{
"epoch": 0.99,
"grad_norm": 1.4795210361480713,
"learning_rate": 8.47412874810452e-06,
"loss": 0.5077,
"step": 4950
},
{
"epoch": 0.992,
"grad_norm": 1.3691011667251587,
"learning_rate": 8.465751331249558e-06,
"loss": 0.4822,
"step": 4960
},
{
"epoch": 0.994,
"grad_norm": 1.6543487310409546,
"learning_rate": 8.457355145901235e-06,
"loss": 0.4765,
"step": 4970
},
{
"epoch": 0.996,
"grad_norm": 1.147316813468933,
"learning_rate": 8.448940237528404e-06,
"loss": 0.4662,
"step": 4980
},
{
"epoch": 0.998,
"grad_norm": 1.3053276538848877,
"learning_rate": 8.440506651701315e-06,
"loss": 0.4694,
"step": 4990
},
{
"epoch": 1.0,
"grad_norm": 0.914472758769989,
"learning_rate": 8.43205443409136e-06,
"loss": 0.3919,
"step": 5000
},
{
"epoch": 1.002,
"grad_norm": 1.1292088031768799,
"learning_rate": 8.42358363047084e-06,
"loss": 0.2872,
"step": 5010
},
{
"epoch": 1.004,
"grad_norm": 1.1585907936096191,
"learning_rate": 8.415094286712694e-06,
"loss": 0.3777,
"step": 5020
},
{
"epoch": 1.006,
"grad_norm": 1.4778828620910645,
"learning_rate": 8.406586448790277e-06,
"loss": 0.3282,
"step": 5030
},
{
"epoch": 1.008,
"grad_norm": 1.3356106281280518,
"learning_rate": 8.398060162777084e-06,
"loss": 0.3194,
"step": 5040
},
{
"epoch": 1.01,
"grad_norm": 1.343839168548584,
"learning_rate": 8.389515474846522e-06,
"loss": 0.3194,
"step": 5050
},
{
"epoch": 1.012,
"grad_norm": 1.3251718282699585,
"learning_rate": 8.380952431271653e-06,
"loss": 0.3459,
"step": 5060
},
{
"epoch": 1.014,
"grad_norm": 1.7042008638381958,
"learning_rate": 8.372371078424941e-06,
"loss": 0.3321,
"step": 5070
},
{
"epoch": 1.016,
"grad_norm": 1.3194146156311035,
"learning_rate": 8.363771462778e-06,
"loss": 0.3652,
"step": 5080
},
{
"epoch": 1.018,
"grad_norm": 1.481685996055603,
"learning_rate": 8.355153630901344e-06,
"loss": 0.3853,
"step": 5090
},
{
"epoch": 1.02,
"grad_norm": 1.6290403604507446,
"learning_rate": 8.346517629464141e-06,
"loss": 0.2857,
"step": 5100
},
{
"epoch": 1.022,
"grad_norm": 0.9795701503753662,
"learning_rate": 8.337863505233954e-06,
"loss": 0.3388,
"step": 5110
},
{
"epoch": 1.024,
"grad_norm": 1.4000308513641357,
"learning_rate": 8.32919130507648e-06,
"loss": 0.3922,
"step": 5120
},
{
"epoch": 1.026,
"grad_norm": 1.081754446029663,
"learning_rate": 8.320501075955317e-06,
"loss": 0.2852,
"step": 5130
},
{
"epoch": 1.028,
"grad_norm": 1.0564137697219849,
"learning_rate": 8.311792864931686e-06,
"loss": 0.3219,
"step": 5140
},
{
"epoch": 1.03,
"grad_norm": 1.1438813209533691,
"learning_rate": 8.303066719164195e-06,
"loss": 0.2861,
"step": 5150
},
{
"epoch": 1.032,
"grad_norm": 1.4927905797958374,
"learning_rate": 8.294322685908576e-06,
"loss": 0.3144,
"step": 5160
},
{
"epoch": 1.034,
"grad_norm": 1.9864319562911987,
"learning_rate": 8.285560812517423e-06,
"loss": 0.3573,
"step": 5170
},
{
"epoch": 1.036,
"grad_norm": 1.1521250009536743,
"learning_rate": 8.27678114643995e-06,
"loss": 0.3696,
"step": 5180
},
{
"epoch": 1.038,
"grad_norm": 1.2286995649337769,
"learning_rate": 8.26798373522172e-06,
"loss": 0.3882,
"step": 5190
},
{
"epoch": 1.04,
"grad_norm": 1.3970285654067993,
"learning_rate": 8.259168626504395e-06,
"loss": 0.3408,
"step": 5200
},
{
"epoch": 1.042,
"grad_norm": 1.3147468566894531,
"learning_rate": 8.250335868025477e-06,
"loss": 0.3673,
"step": 5210
},
{
"epoch": 1.044,
"grad_norm": 1.1939131021499634,
"learning_rate": 8.241485507618046e-06,
"loss": 0.3329,
"step": 5220
},
{
"epoch": 1.046,
"grad_norm": 1.4440771341323853,
"learning_rate": 8.232617593210512e-06,
"loss": 0.3474,
"step": 5230
},
{
"epoch": 1.048,
"grad_norm": 1.0242334604263306,
"learning_rate": 8.223732172826336e-06,
"loss": 0.361,
"step": 5240
},
{
"epoch": 1.05,
"grad_norm": 1.4275943040847778,
"learning_rate": 8.214829294583786e-06,
"loss": 0.3043,
"step": 5250
},
{
"epoch": 1.052,
"grad_norm": 1.5001543760299683,
"learning_rate": 8.205909006695679e-06,
"loss": 0.3676,
"step": 5260
},
{
"epoch": 1.054,
"grad_norm": 1.3494583368301392,
"learning_rate": 8.196971357469098e-06,
"loss": 0.3486,
"step": 5270
},
{
"epoch": 1.056,
"grad_norm": 1.2720623016357422,
"learning_rate": 8.188016395305156e-06,
"loss": 0.3145,
"step": 5280
},
{
"epoch": 1.058,
"grad_norm": 1.5187513828277588,
"learning_rate": 8.179044168698722e-06,
"loss": 0.3581,
"step": 5290
},
{
"epoch": 1.06,
"grad_norm": 2.041038990020752,
"learning_rate": 8.170054726238152e-06,
"loss": 0.3858,
"step": 5300
},
{
"epoch": 1.062,
"grad_norm": 1.375152587890625,
"learning_rate": 8.161048116605039e-06,
"loss": 0.3401,
"step": 5310
},
{
"epoch": 1.064,
"grad_norm": 0.8547447323799133,
"learning_rate": 8.152024388573945e-06,
"loss": 0.3006,
"step": 5320
},
{
"epoch": 1.066,
"grad_norm": 2.131930351257324,
"learning_rate": 8.142983591012128e-06,
"loss": 0.3521,
"step": 5330
},
{
"epoch": 1.068,
"grad_norm": 1.6276785135269165,
"learning_rate": 8.133925772879292e-06,
"loss": 0.3877,
"step": 5340
},
{
"epoch": 1.07,
"grad_norm": 1.4579194784164429,
"learning_rate": 8.124850983227313e-06,
"loss": 0.3793,
"step": 5350
},
{
"epoch": 1.072,
"grad_norm": 1.182732105255127,
"learning_rate": 8.115759271199967e-06,
"loss": 0.3425,
"step": 5360
},
{
"epoch": 1.074,
"grad_norm": 1.5831196308135986,
"learning_rate": 8.106650686032687e-06,
"loss": 0.3424,
"step": 5370
},
{
"epoch": 1.076,
"grad_norm": 0.9246402978897095,
"learning_rate": 8.097525277052265e-06,
"loss": 0.3245,
"step": 5380
},
{
"epoch": 1.078,
"grad_norm": 1.0254276990890503,
"learning_rate": 8.08838309367661e-06,
"loss": 0.3218,
"step": 5390
},
{
"epoch": 1.08,
"grad_norm": 0.9792593717575073,
"learning_rate": 8.079224185414471e-06,
"loss": 0.3681,
"step": 5400
},
{
"epoch": 1.082,
"grad_norm": 1.3363689184188843,
"learning_rate": 8.07004860186517e-06,
"loss": 0.3799,
"step": 5410
},
{
"epoch": 1.084,
"grad_norm": 1.571494698524475,
"learning_rate": 8.060856392718326e-06,
"loss": 0.2975,
"step": 5420
},
{
"epoch": 1.086,
"grad_norm": 1.1808347702026367,
"learning_rate": 8.051647607753598e-06,
"loss": 0.3532,
"step": 5430
},
{
"epoch": 1.088,
"grad_norm": 2.0380117893218994,
"learning_rate": 8.04242229684041e-06,
"loss": 0.3543,
"step": 5440
},
{
"epoch": 1.09,
"grad_norm": 1.3502197265625,
"learning_rate": 8.033180509937683e-06,
"loss": 0.3566,
"step": 5450
},
{
"epoch": 1.092,
"grad_norm": 1.074385404586792,
"learning_rate": 8.023922297093557e-06,
"loss": 0.3152,
"step": 5460
},
{
"epoch": 1.094,
"grad_norm": 1.133074402809143,
"learning_rate": 8.014647708445124e-06,
"loss": 0.3784,
"step": 5470
},
{
"epoch": 1.096,
"grad_norm": 1.1202969551086426,
"learning_rate": 8.005356794218167e-06,
"loss": 0.4327,
"step": 5480
},
{
"epoch": 1.098,
"grad_norm": 1.4087947607040405,
"learning_rate": 7.99604960472687e-06,
"loss": 0.3206,
"step": 5490
},
{
"epoch": 1.1,
"grad_norm": 1.140956163406372,
"learning_rate": 7.986726190373562e-06,
"loss": 0.2804,
"step": 5500
},
{
"epoch": 1.102,
"grad_norm": 0.9047802686691284,
"learning_rate": 7.977386601648427e-06,
"loss": 0.3162,
"step": 5510
},
{
"epoch": 1.104,
"grad_norm": 0.9571928381919861,
"learning_rate": 7.968030889129247e-06,
"loss": 0.2635,
"step": 5520
},
{
"epoch": 1.106,
"grad_norm": 1.3515325784683228,
"learning_rate": 7.95865910348112e-06,
"loss": 0.3598,
"step": 5530
},
{
"epoch": 1.108,
"grad_norm": 1.0597070455551147,
"learning_rate": 7.949271295456187e-06,
"loss": 0.2918,
"step": 5540
},
{
"epoch": 1.11,
"grad_norm": 1.3461328744888306,
"learning_rate": 7.939867515893353e-06,
"loss": 0.2928,
"step": 5550
},
{
"epoch": 1.112,
"grad_norm": 1.0447543859481812,
"learning_rate": 7.930447815718022e-06,
"loss": 0.3584,
"step": 5560
},
{
"epoch": 1.114,
"grad_norm": 1.611246109008789,
"learning_rate": 7.921012245941809e-06,
"loss": 0.3816,
"step": 5570
},
{
"epoch": 1.116,
"grad_norm": 0.886353611946106,
"learning_rate": 7.91156085766227e-06,
"loss": 0.322,
"step": 5580
},
{
"epoch": 1.1179999999999999,
"grad_norm": 1.2039692401885986,
"learning_rate": 7.90209370206263e-06,
"loss": 0.2804,
"step": 5590
},
{
"epoch": 1.12,
"grad_norm": 1.0296542644500732,
"learning_rate": 7.892610830411496e-06,
"loss": 0.3007,
"step": 5600
},
{
"epoch": 1.1219999999999999,
"grad_norm": 1.3383187055587769,
"learning_rate": 7.883112294062585e-06,
"loss": 0.38,
"step": 5610
},
{
"epoch": 1.124,
"grad_norm": 1.2576732635498047,
"learning_rate": 7.873598144454444e-06,
"loss": 0.3637,
"step": 5620
},
{
"epoch": 1.126,
"grad_norm": 0.6940869688987732,
"learning_rate": 7.864068433110176e-06,
"loss": 0.2982,
"step": 5630
},
{
"epoch": 1.1280000000000001,
"grad_norm": 1.15080726146698,
"learning_rate": 7.854523211637152e-06,
"loss": 0.31,
"step": 5640
},
{
"epoch": 1.13,
"grad_norm": 1.8969014883041382,
"learning_rate": 7.844962531726742e-06,
"loss": 0.3375,
"step": 5650
},
{
"epoch": 1.1320000000000001,
"grad_norm": 1.5290457010269165,
"learning_rate": 7.835386445154023e-06,
"loss": 0.3467,
"step": 5660
},
{
"epoch": 1.134,
"grad_norm": 1.6540489196777344,
"learning_rate": 7.825795003777515e-06,
"loss": 0.3403,
"step": 5670
},
{
"epoch": 1.1360000000000001,
"grad_norm": 1.4303158521652222,
"learning_rate": 7.816188259538885e-06,
"loss": 0.3727,
"step": 5680
},
{
"epoch": 1.138,
"grad_norm": 1.2462937831878662,
"learning_rate": 7.806566264462668e-06,
"loss": 0.3325,
"step": 5690
},
{
"epoch": 1.1400000000000001,
"grad_norm": 1.2328237295150757,
"learning_rate": 7.796929070655994e-06,
"loss": 0.3572,
"step": 5700
},
{
"epoch": 1.142,
"grad_norm": 1.305349349975586,
"learning_rate": 7.787276730308304e-06,
"loss": 0.3046,
"step": 5710
},
{
"epoch": 1.144,
"grad_norm": 1.3444890975952148,
"learning_rate": 7.777609295691055e-06,
"loss": 0.3464,
"step": 5720
},
{
"epoch": 1.146,
"grad_norm": 0.8992980718612671,
"learning_rate": 7.767926819157452e-06,
"loss": 0.3742,
"step": 5730
},
{
"epoch": 1.148,
"grad_norm": 1.571289300918579,
"learning_rate": 7.758229353142153e-06,
"loss": 0.3863,
"step": 5740
},
{
"epoch": 1.15,
"grad_norm": 1.2571865320205688,
"learning_rate": 7.748516950160993e-06,
"loss": 0.3611,
"step": 5750
},
{
"epoch": 1.152,
"grad_norm": 0.9523794651031494,
"learning_rate": 7.738789662810702e-06,
"loss": 0.3352,
"step": 5760
},
{
"epoch": 1.154,
"grad_norm": 1.1043280363082886,
"learning_rate": 7.729047543768608e-06,
"loss": 0.4024,
"step": 5770
},
{
"epoch": 1.156,
"grad_norm": 1.4577151536941528,
"learning_rate": 7.719290645792361e-06,
"loss": 0.3124,
"step": 5780
},
{
"epoch": 1.158,
"grad_norm": 1.7957823276519775,
"learning_rate": 7.709519021719644e-06,
"loss": 0.4165,
"step": 5790
},
{
"epoch": 1.16,
"grad_norm": 1.2084007263183594,
"learning_rate": 7.699732724467894e-06,
"loss": 0.3357,
"step": 5800
},
{
"epoch": 1.162,
"grad_norm": 1.2520586252212524,
"learning_rate": 7.689931807033999e-06,
"loss": 0.3114,
"step": 5810
},
{
"epoch": 1.164,
"grad_norm": 1.048161268234253,
"learning_rate": 7.68011632249403e-06,
"loss": 0.3216,
"step": 5820
},
{
"epoch": 1.166,
"grad_norm": 1.231397032737732,
"learning_rate": 7.670286324002943e-06,
"loss": 0.3458,
"step": 5830
},
{
"epoch": 1.168,
"grad_norm": 1.6105576753616333,
"learning_rate": 7.66044186479429e-06,
"loss": 0.3423,
"step": 5840
},
{
"epoch": 1.17,
"grad_norm": 1.3911272287368774,
"learning_rate": 7.650582998179939e-06,
"loss": 0.3088,
"step": 5850
},
{
"epoch": 1.172,
"grad_norm": 1.618174433708191,
"learning_rate": 7.640709777549773e-06,
"loss": 0.3516,
"step": 5860
},
{
"epoch": 1.174,
"grad_norm": 2.1823925971984863,
"learning_rate": 7.630822256371415e-06,
"loss": 0.3054,
"step": 5870
},
{
"epoch": 1.176,
"grad_norm": 1.236488938331604,
"learning_rate": 7.620920488189929e-06,
"loss": 0.3375,
"step": 5880
},
{
"epoch": 1.178,
"grad_norm": 1.020044207572937,
"learning_rate": 7.6110045266275305e-06,
"loss": 0.2757,
"step": 5890
},
{
"epoch": 1.18,
"grad_norm": 0.9718794822692871,
"learning_rate": 7.601074425383302e-06,
"loss": 0.3025,
"step": 5900
},
{
"epoch": 1.182,
"grad_norm": 1.4841420650482178,
"learning_rate": 7.591130238232892e-06,
"loss": 0.3117,
"step": 5910
},
{
"epoch": 1.184,
"grad_norm": 1.511138916015625,
"learning_rate": 7.581172019028238e-06,
"loss": 0.3936,
"step": 5920
},
{
"epoch": 1.186,
"grad_norm": 1.3772696256637573,
"learning_rate": 7.571199821697263e-06,
"loss": 0.3486,
"step": 5930
},
{
"epoch": 1.188,
"grad_norm": 1.3091078996658325,
"learning_rate": 7.561213700243584e-06,
"loss": 0.2936,
"step": 5940
},
{
"epoch": 1.19,
"grad_norm": 1.3505816459655762,
"learning_rate": 7.55121370874623e-06,
"loss": 0.3203,
"step": 5950
},
{
"epoch": 1.192,
"grad_norm": 1.3498326539993286,
"learning_rate": 7.541199901359335e-06,
"loss": 0.3734,
"step": 5960
},
{
"epoch": 1.194,
"grad_norm": 1.5176153182983398,
"learning_rate": 7.531172332311861e-06,
"loss": 0.2997,
"step": 5970
},
{
"epoch": 1.196,
"grad_norm": 0.9407598972320557,
"learning_rate": 7.521131055907283e-06,
"loss": 0.3204,
"step": 5980
},
{
"epoch": 1.198,
"grad_norm": 1.3535823822021484,
"learning_rate": 7.5110761265233156e-06,
"loss": 0.2981,
"step": 5990
},
{
"epoch": 1.2,
"grad_norm": 1.7967560291290283,
"learning_rate": 7.501007598611609e-06,
"loss": 0.3508,
"step": 6000
},
{
"epoch": 1.202,
"grad_norm": 1.682611107826233,
"learning_rate": 7.490925526697455e-06,
"loss": 0.3258,
"step": 6010
},
{
"epoch": 1.204,
"grad_norm": 1.2590186595916748,
"learning_rate": 7.480829965379489e-06,
"loss": 0.3376,
"step": 6020
},
{
"epoch": 1.206,
"grad_norm": 1.1302205324172974,
"learning_rate": 7.470720969329399e-06,
"loss": 0.3297,
"step": 6030
},
{
"epoch": 1.208,
"grad_norm": 1.1471021175384521,
"learning_rate": 7.460598593291628e-06,
"loss": 0.3566,
"step": 6040
},
{
"epoch": 1.21,
"grad_norm": 2.0590806007385254,
"learning_rate": 7.450462892083079e-06,
"loss": 0.3354,
"step": 6050
},
{
"epoch": 1.212,
"grad_norm": 1.6424793004989624,
"learning_rate": 7.44031392059281e-06,
"loss": 0.348,
"step": 6060
},
{
"epoch": 1.214,
"grad_norm": 1.1267763376235962,
"learning_rate": 7.430151733781752e-06,
"loss": 0.324,
"step": 6070
},
{
"epoch": 1.216,
"grad_norm": 1.049166202545166,
"learning_rate": 7.419976386682395e-06,
"loss": 0.3282,
"step": 6080
},
{
"epoch": 1.218,
"grad_norm": 1.6423687934875488,
"learning_rate": 7.409787934398502e-06,
"loss": 0.2872,
"step": 6090
},
{
"epoch": 1.22,
"grad_norm": 1.4791516065597534,
"learning_rate": 7.3995864321048036e-06,
"loss": 0.416,
"step": 6100
},
{
"epoch": 1.222,
"grad_norm": 1.1335197687149048,
"learning_rate": 7.389371935046703e-06,
"loss": 0.3191,
"step": 6110
},
{
"epoch": 1.224,
"grad_norm": 1.4202783107757568,
"learning_rate": 7.3791444985399755e-06,
"loss": 0.3429,
"step": 6120
},
{
"epoch": 1.226,
"grad_norm": 1.132926106452942,
"learning_rate": 7.368904177970466e-06,
"loss": 0.3469,
"step": 6130
},
{
"epoch": 1.228,
"grad_norm": 1.7757046222686768,
"learning_rate": 7.358651028793797e-06,
"loss": 0.333,
"step": 6140
},
{
"epoch": 1.23,
"grad_norm": 1.3629705905914307,
"learning_rate": 7.3483851065350595e-06,
"loss": 0.3802,
"step": 6150
},
{
"epoch": 1.232,
"grad_norm": 1.0564378499984741,
"learning_rate": 7.33810646678852e-06,
"loss": 0.3904,
"step": 6160
},
{
"epoch": 1.234,
"grad_norm": 1.6500481367111206,
"learning_rate": 7.327815165217309e-06,
"loss": 0.3675,
"step": 6170
},
{
"epoch": 1.236,
"grad_norm": 1.5274205207824707,
"learning_rate": 7.317511257553131e-06,
"loss": 0.3182,
"step": 6180
},
{
"epoch": 1.238,
"grad_norm": 1.1405715942382812,
"learning_rate": 7.307194799595958e-06,
"loss": 0.3103,
"step": 6190
},
{
"epoch": 1.24,
"grad_norm": 1.4158636331558228,
"learning_rate": 7.296865847213724e-06,
"loss": 0.3453,
"step": 6200
},
{
"epoch": 1.242,
"grad_norm": 1.440067172050476,
"learning_rate": 7.2865244563420304e-06,
"loss": 0.4289,
"step": 6210
},
{
"epoch": 1.244,
"grad_norm": 1.2499229907989502,
"learning_rate": 7.27617068298383e-06,
"loss": 0.3657,
"step": 6220
},
{
"epoch": 1.246,
"grad_norm": 1.599689245223999,
"learning_rate": 7.265804583209142e-06,
"loss": 0.339,
"step": 6230
},
{
"epoch": 1.248,
"grad_norm": 1.6074563264846802,
"learning_rate": 7.25542621315473e-06,
"loss": 0.3562,
"step": 6240
},
{
"epoch": 1.25,
"grad_norm": 1.6778024435043335,
"learning_rate": 7.245035629023812e-06,
"loss": 0.313,
"step": 6250
},
{
"epoch": 1.252,
"grad_norm": 1.2234922647476196,
"learning_rate": 7.2346328870857465e-06,
"loss": 0.3822,
"step": 6260
},
{
"epoch": 1.254,
"grad_norm": 1.2148653268814087,
"learning_rate": 7.224218043675735e-06,
"loss": 0.3567,
"step": 6270
},
{
"epoch": 1.256,
"grad_norm": 0.9696487188339233,
"learning_rate": 7.21379115519451e-06,
"loss": 0.3688,
"step": 6280
},
{
"epoch": 1.258,
"grad_norm": 1.371671438217163,
"learning_rate": 7.2033522781080325e-06,
"loss": 0.3409,
"step": 6290
},
{
"epoch": 1.26,
"grad_norm": 1.5369791984558105,
"learning_rate": 7.192901468947193e-06,
"loss": 0.3777,
"step": 6300
},
{
"epoch": 1.262,
"grad_norm": 1.8359071016311646,
"learning_rate": 7.182438784307495e-06,
"loss": 0.3179,
"step": 6310
},
{
"epoch": 1.264,
"grad_norm": 0.9479062557220459,
"learning_rate": 7.171964280848749e-06,
"loss": 0.2901,
"step": 6320
},
{
"epoch": 1.266,
"grad_norm": 1.220046877861023,
"learning_rate": 7.161478015294778e-06,
"loss": 0.3478,
"step": 6330
},
{
"epoch": 1.268,
"grad_norm": 1.2605116367340088,
"learning_rate": 7.150980044433094e-06,
"loss": 0.2351,
"step": 6340
},
{
"epoch": 1.27,
"grad_norm": 1.6339104175567627,
"learning_rate": 7.140470425114603e-06,
"loss": 0.3123,
"step": 6350
},
{
"epoch": 1.272,
"grad_norm": 1.1417044401168823,
"learning_rate": 7.1299492142532876e-06,
"loss": 0.3129,
"step": 6360
},
{
"epoch": 1.274,
"grad_norm": 0.994657039642334,
"learning_rate": 7.119416468825908e-06,
"loss": 0.3247,
"step": 6370
},
{
"epoch": 1.276,
"grad_norm": 1.182531714439392,
"learning_rate": 7.108872245871687e-06,
"loss": 0.2929,
"step": 6380
},
{
"epoch": 1.278,
"grad_norm": 1.1829276084899902,
"learning_rate": 7.098316602492004e-06,
"loss": 0.3174,
"step": 6390
},
{
"epoch": 1.28,
"grad_norm": 1.1945512294769287,
"learning_rate": 7.087749595850084e-06,
"loss": 0.3341,
"step": 6400
},
{
"epoch": 1.282,
"grad_norm": 1.33237886428833,
"learning_rate": 7.0771712831706855e-06,
"loss": 0.3453,
"step": 6410
},
{
"epoch": 1.284,
"grad_norm": 1.4239740371704102,
"learning_rate": 7.066581721739801e-06,
"loss": 0.3416,
"step": 6420
},
{
"epoch": 1.286,
"grad_norm": 1.329602837562561,
"learning_rate": 7.0559809689043325e-06,
"loss": 0.3434,
"step": 6430
},
{
"epoch": 1.288,
"grad_norm": 1.4223501682281494,
"learning_rate": 7.045369082071793e-06,
"loss": 0.3664,
"step": 6440
},
{
"epoch": 1.29,
"grad_norm": 1.5064311027526855,
"learning_rate": 7.034746118709989e-06,
"loss": 0.3446,
"step": 6450
},
{
"epoch": 1.292,
"grad_norm": 1.0884929895401,
"learning_rate": 7.024112136346713e-06,
"loss": 0.2676,
"step": 6460
},
{
"epoch": 1.294,
"grad_norm": 1.623273491859436,
"learning_rate": 7.013467192569427e-06,
"loss": 0.4001,
"step": 6470
},
{
"epoch": 1.296,
"grad_norm": 1.3760849237442017,
"learning_rate": 7.002811345024951e-06,
"loss": 0.3415,
"step": 6480
},
{
"epoch": 1.298,
"grad_norm": 1.071457862854004,
"learning_rate": 6.992144651419163e-06,
"loss": 0.3584,
"step": 6490
},
{
"epoch": 1.3,
"grad_norm": 1.245360255241394,
"learning_rate": 6.981467169516671e-06,
"loss": 0.341,
"step": 6500
},
{
"epoch": 1.302,
"grad_norm": 1.653593897819519,
"learning_rate": 6.9707789571405025e-06,
"loss": 0.3626,
"step": 6510
},
{
"epoch": 1.304,
"grad_norm": 0.9302300214767456,
"learning_rate": 6.960080072171802e-06,
"loss": 0.2995,
"step": 6520
},
{
"epoch": 1.306,
"grad_norm": 1.358490228652954,
"learning_rate": 6.949370572549506e-06,
"loss": 0.3312,
"step": 6530
},
{
"epoch": 1.308,
"grad_norm": 1.5318645238876343,
"learning_rate": 6.938650516270038e-06,
"loss": 0.3552,
"step": 6540
},
{
"epoch": 1.31,
"grad_norm": 1.6989809274673462,
"learning_rate": 6.927919961386984e-06,
"loss": 0.3347,
"step": 6550
},
{
"epoch": 1.312,
"grad_norm": 1.2937415838241577,
"learning_rate": 6.9171789660107876e-06,
"loss": 0.3167,
"step": 6560
},
{
"epoch": 1.314,
"grad_norm": 1.2662570476531982,
"learning_rate": 6.906427588308436e-06,
"loss": 0.374,
"step": 6570
},
{
"epoch": 1.316,
"grad_norm": 1.152688980102539,
"learning_rate": 6.895665886503136e-06,
"loss": 0.2952,
"step": 6580
},
{
"epoch": 1.318,
"grad_norm": 1.0898213386535645,
"learning_rate": 6.8848939188740034e-06,
"loss": 0.3529,
"step": 6590
},
{
"epoch": 1.32,
"grad_norm": 1.2499762773513794,
"learning_rate": 6.874111743755751e-06,
"loss": 0.335,
"step": 6600
},
{
"epoch": 1.322,
"grad_norm": 1.5297431945800781,
"learning_rate": 6.863319419538366e-06,
"loss": 0.3896,
"step": 6610
},
{
"epoch": 1.324,
"grad_norm": 1.6625542640686035,
"learning_rate": 6.852517004666801e-06,
"loss": 0.3455,
"step": 6620
},
{
"epoch": 1.326,
"grad_norm": 1.4104151725769043,
"learning_rate": 6.84170455764065e-06,
"loss": 0.3535,
"step": 6630
},
{
"epoch": 1.328,
"grad_norm": 1.2529921531677246,
"learning_rate": 6.830882137013839e-06,
"loss": 0.3219,
"step": 6640
},
{
"epoch": 1.33,
"grad_norm": 1.6299757957458496,
"learning_rate": 6.820049801394303e-06,
"loss": 0.4037,
"step": 6650
},
{
"epoch": 1.332,
"grad_norm": 1.215126872062683,
"learning_rate": 6.80920760944367e-06,
"loss": 0.3307,
"step": 6660
},
{
"epoch": 1.334,
"grad_norm": 1.0372707843780518,
"learning_rate": 6.798355619876944e-06,
"loss": 0.3326,
"step": 6670
},
{
"epoch": 1.336,
"grad_norm": 1.1610356569290161,
"learning_rate": 6.787493891462191e-06,
"loss": 0.3467,
"step": 6680
},
{
"epoch": 1.338,
"grad_norm": 1.7156941890716553,
"learning_rate": 6.776622483020214e-06,
"loss": 0.4261,
"step": 6690
},
{
"epoch": 1.34,
"grad_norm": 0.9658931493759155,
"learning_rate": 6.765741453424237e-06,
"loss": 0.307,
"step": 6700
},
{
"epoch": 1.342,
"grad_norm": 0.8651520013809204,
"learning_rate": 6.754850861599589e-06,
"loss": 0.3802,
"step": 6710
},
{
"epoch": 1.3439999999999999,
"grad_norm": 1.15030837059021,
"learning_rate": 6.743950766523377e-06,
"loss": 0.3026,
"step": 6720
},
{
"epoch": 1.346,
"grad_norm": 1.2045297622680664,
"learning_rate": 6.733041227224182e-06,
"loss": 0.3673,
"step": 6730
},
{
"epoch": 1.3479999999999999,
"grad_norm": 1.098310112953186,
"learning_rate": 6.722122302781716e-06,
"loss": 0.3305,
"step": 6740
},
{
"epoch": 1.35,
"grad_norm": 1.5491714477539062,
"learning_rate": 6.711194052326528e-06,
"loss": 0.3063,
"step": 6750
},
{
"epoch": 1.3519999999999999,
"grad_norm": 1.581486463546753,
"learning_rate": 6.700256535039665e-06,
"loss": 0.3534,
"step": 6760
},
{
"epoch": 1.354,
"grad_norm": 1.3409374952316284,
"learning_rate": 6.689309810152359e-06,
"loss": 0.3866,
"step": 6770
},
{
"epoch": 1.3559999999999999,
"grad_norm": 1.0301975011825562,
"learning_rate": 6.678353936945704e-06,
"loss": 0.3391,
"step": 6780
},
{
"epoch": 1.358,
"grad_norm": 1.1352406740188599,
"learning_rate": 6.6673889747503364e-06,
"loss": 0.2808,
"step": 6790
},
{
"epoch": 1.3599999999999999,
"grad_norm": 1.2140257358551025,
"learning_rate": 6.656414982946115e-06,
"loss": 0.3619,
"step": 6800
},
{
"epoch": 1.362,
"grad_norm": 1.853940725326538,
"learning_rate": 6.645432020961796e-06,
"loss": 0.3014,
"step": 6810
},
{
"epoch": 1.3639999999999999,
"grad_norm": 0.9603062272071838,
"learning_rate": 6.634440148274712e-06,
"loss": 0.3011,
"step": 6820
},
{
"epoch": 1.366,
"grad_norm": 0.9446805715560913,
"learning_rate": 6.623439424410456e-06,
"loss": 0.3335,
"step": 6830
},
{
"epoch": 1.3679999999999999,
"grad_norm": 1.3330634832382202,
"learning_rate": 6.612429908942546e-06,
"loss": 0.3776,
"step": 6840
},
{
"epoch": 1.37,
"grad_norm": 1.3779356479644775,
"learning_rate": 6.601411661492114e-06,
"loss": 0.3175,
"step": 6850
},
{
"epoch": 1.3719999999999999,
"grad_norm": 1.492652416229248,
"learning_rate": 6.590384741727583e-06,
"loss": 0.3377,
"step": 6860
},
{
"epoch": 1.374,
"grad_norm": 1.719750165939331,
"learning_rate": 6.579349209364332e-06,
"loss": 0.3711,
"step": 6870
},
{
"epoch": 1.376,
"grad_norm": 1.331905722618103,
"learning_rate": 6.5683051241643894e-06,
"loss": 0.3052,
"step": 6880
},
{
"epoch": 1.3780000000000001,
"grad_norm": 1.1688127517700195,
"learning_rate": 6.557252545936095e-06,
"loss": 0.301,
"step": 6890
},
{
"epoch": 1.38,
"grad_norm": 0.8940510153770447,
"learning_rate": 6.546191534533783e-06,
"loss": 0.3122,
"step": 6900
},
{
"epoch": 1.3820000000000001,
"grad_norm": 1.363988995552063,
"learning_rate": 6.53512214985746e-06,
"loss": 0.3079,
"step": 6910
},
{
"epoch": 1.384,
"grad_norm": 0.9122200608253479,
"learning_rate": 6.5240444518524736e-06,
"loss": 0.3486,
"step": 6920
},
{
"epoch": 1.3860000000000001,
"grad_norm": 1.0334652662277222,
"learning_rate": 6.512958500509193e-06,
"loss": 0.2696,
"step": 6930
},
{
"epoch": 1.388,
"grad_norm": 1.257965326309204,
"learning_rate": 6.501864355862682e-06,
"loss": 0.3648,
"step": 6940
},
{
"epoch": 1.3900000000000001,
"grad_norm": 1.1571115255355835,
"learning_rate": 6.490762077992376e-06,
"loss": 0.3474,
"step": 6950
},
{
"epoch": 1.392,
"grad_norm": 1.2089656591415405,
"learning_rate": 6.479651727021754e-06,
"loss": 0.3313,
"step": 6960
},
{
"epoch": 1.3940000000000001,
"grad_norm": 1.4407920837402344,
"learning_rate": 6.4685333631180145e-06,
"loss": 0.3695,
"step": 6970
},
{
"epoch": 1.396,
"grad_norm": 2.0778284072875977,
"learning_rate": 6.457407046491748e-06,
"loss": 0.2625,
"step": 6980
},
{
"epoch": 1.3980000000000001,
"grad_norm": 1.36530339717865,
"learning_rate": 6.4462728373966165e-06,
"loss": 0.3015,
"step": 6990
},
{
"epoch": 1.4,
"grad_norm": 1.8104184865951538,
"learning_rate": 6.435130796129019e-06,
"loss": 0.4064,
"step": 7000
},
{
"epoch": 1.4020000000000001,
"grad_norm": 1.3507591485977173,
"learning_rate": 6.423980983027769e-06,
"loss": 0.3225,
"step": 7010
},
{
"epoch": 1.404,
"grad_norm": 0.7463653087615967,
"learning_rate": 6.412823458473772e-06,
"loss": 0.2916,
"step": 7020
},
{
"epoch": 1.4060000000000001,
"grad_norm": 1.5296553373336792,
"learning_rate": 6.401658282889689e-06,
"loss": 0.3413,
"step": 7030
},
{
"epoch": 1.408,
"grad_norm": 1.1751571893692017,
"learning_rate": 6.390485516739616e-06,
"loss": 0.3598,
"step": 7040
},
{
"epoch": 1.41,
"grad_norm": 1.090745449066162,
"learning_rate": 6.379305220528758e-06,
"loss": 0.3398,
"step": 7050
},
{
"epoch": 1.412,
"grad_norm": 1.3860448598861694,
"learning_rate": 6.368117454803093e-06,
"loss": 0.2953,
"step": 7060
},
{
"epoch": 1.414,
"grad_norm": 1.4662338495254517,
"learning_rate": 6.356922280149058e-06,
"loss": 0.3662,
"step": 7070
},
{
"epoch": 1.416,
"grad_norm": 1.0763154029846191,
"learning_rate": 6.345719757193203e-06,
"loss": 0.3182,
"step": 7080
},
{
"epoch": 1.418,
"grad_norm": 1.2442153692245483,
"learning_rate": 6.334509946601879e-06,
"loss": 0.3177,
"step": 7090
},
{
"epoch": 1.42,
"grad_norm": 1.304867148399353,
"learning_rate": 6.323292909080897e-06,
"loss": 0.3322,
"step": 7100
},
{
"epoch": 1.422,
"grad_norm": 1.5023126602172852,
"learning_rate": 6.3120687053752114e-06,
"loss": 0.3425,
"step": 7110
},
{
"epoch": 1.424,
"grad_norm": 1.539007306098938,
"learning_rate": 6.3008373962685785e-06,
"loss": 0.3376,
"step": 7120
},
{
"epoch": 1.426,
"grad_norm": 1.6825087070465088,
"learning_rate": 6.289599042583237e-06,
"loss": 0.364,
"step": 7130
},
{
"epoch": 1.428,
"grad_norm": 1.3176345825195312,
"learning_rate": 6.278353705179572e-06,
"loss": 0.3753,
"step": 7140
},
{
"epoch": 1.43,
"grad_norm": 1.2803112268447876,
"learning_rate": 6.267101444955792e-06,
"loss": 0.3295,
"step": 7150
},
{
"epoch": 1.432,
"grad_norm": 1.4513520002365112,
"learning_rate": 6.255842322847594e-06,
"loss": 0.3179,
"step": 7160
},
{
"epoch": 1.434,
"grad_norm": 1.314857006072998,
"learning_rate": 6.244576399827831e-06,
"loss": 0.3537,
"step": 7170
},
{
"epoch": 1.436,
"grad_norm": 1.1115436553955078,
"learning_rate": 6.233303736906193e-06,
"loss": 0.313,
"step": 7180
},
{
"epoch": 1.438,
"grad_norm": 1.1615487337112427,
"learning_rate": 6.222024395128864e-06,
"loss": 0.3293,
"step": 7190
},
{
"epoch": 1.44,
"grad_norm": 1.4486738443374634,
"learning_rate": 6.210738435578198e-06,
"loss": 0.3793,
"step": 7200
},
{
"epoch": 1.442,
"grad_norm": 1.143863558769226,
"learning_rate": 6.199445919372388e-06,
"loss": 0.4045,
"step": 7210
},
{
"epoch": 1.444,
"grad_norm": 1.4991508722305298,
"learning_rate": 6.1881469076651336e-06,
"loss": 0.3216,
"step": 7220
},
{
"epoch": 1.446,
"grad_norm": 1.1230875253677368,
"learning_rate": 6.176841461645311e-06,
"loss": 0.3679,
"step": 7230
},
{
"epoch": 1.448,
"grad_norm": 1.141588568687439,
"learning_rate": 6.16552964253664e-06,
"loss": 0.3378,
"step": 7240
},
{
"epoch": 1.45,
"grad_norm": 1.8605092763900757,
"learning_rate": 6.15421151159735e-06,
"loss": 0.3412,
"step": 7250
},
{
"epoch": 1.452,
"grad_norm": 1.14920175075531,
"learning_rate": 6.14288713011986e-06,
"loss": 0.2663,
"step": 7260
},
{
"epoch": 1.454,
"grad_norm": 1.7807815074920654,
"learning_rate": 6.13155655943043e-06,
"loss": 0.3288,
"step": 7270
},
{
"epoch": 1.456,
"grad_norm": 1.6375200748443604,
"learning_rate": 6.1202198608888416e-06,
"loss": 0.4102,
"step": 7280
},
{
"epoch": 1.458,
"grad_norm": 1.5959913730621338,
"learning_rate": 6.1088770958880595e-06,
"loss": 0.3615,
"step": 7290
},
{
"epoch": 1.46,
"grad_norm": 1.452133297920227,
"learning_rate": 6.097528325853903e-06,
"loss": 0.2841,
"step": 7300
},
{
"epoch": 1.462,
"grad_norm": 1.8519127368927002,
"learning_rate": 6.086173612244708e-06,
"loss": 0.3331,
"step": 7310
},
{
"epoch": 1.464,
"grad_norm": 1.0498560667037964,
"learning_rate": 6.074813016550998e-06,
"loss": 0.2939,
"step": 7320
},
{
"epoch": 1.466,
"grad_norm": 1.369956612586975,
"learning_rate": 6.0634466002951545e-06,
"loss": 0.3859,
"step": 7330
},
{
"epoch": 1.468,
"grad_norm": 1.7278661727905273,
"learning_rate": 6.052074425031075e-06,
"loss": 0.3599,
"step": 7340
},
{
"epoch": 1.47,
"grad_norm": 1.1767780780792236,
"learning_rate": 6.040696552343845e-06,
"loss": 0.3392,
"step": 7350
},
{
"epoch": 1.472,
"grad_norm": 1.4453753232955933,
"learning_rate": 6.029313043849407e-06,
"loss": 0.4165,
"step": 7360
},
{
"epoch": 1.474,
"grad_norm": 1.0843770503997803,
"learning_rate": 6.017923961194221e-06,
"loss": 0.334,
"step": 7370
},
{
"epoch": 1.476,
"grad_norm": 1.1147040128707886,
"learning_rate": 6.006529366054935e-06,
"loss": 0.3453,
"step": 7380
},
{
"epoch": 1.478,
"grad_norm": 1.7290093898773193,
"learning_rate": 5.995129320138047e-06,
"loss": 0.3127,
"step": 7390
},
{
"epoch": 1.48,
"grad_norm": 1.4090754985809326,
"learning_rate": 5.983723885179576e-06,
"loss": 0.3197,
"step": 7400
},
{
"epoch": 1.482,
"grad_norm": 1.1394073963165283,
"learning_rate": 5.972313122944724e-06,
"loss": 0.3234,
"step": 7410
},
{
"epoch": 1.484,
"grad_norm": 1.281721591949463,
"learning_rate": 5.960897095227541e-06,
"loss": 0.3311,
"step": 7420
},
{
"epoch": 1.486,
"grad_norm": 0.9098526239395142,
"learning_rate": 5.949475863850595e-06,
"loss": 0.3342,
"step": 7430
},
{
"epoch": 1.488,
"grad_norm": 0.939958930015564,
"learning_rate": 5.938049490664629e-06,
"loss": 0.3139,
"step": 7440
},
{
"epoch": 1.49,
"grad_norm": 1.2372125387191772,
"learning_rate": 5.926618037548237e-06,
"loss": 0.3121,
"step": 7450
},
{
"epoch": 1.492,
"grad_norm": 1.0505400896072388,
"learning_rate": 5.915181566407519e-06,
"loss": 0.3165,
"step": 7460
},
{
"epoch": 1.494,
"grad_norm": 1.4748669862747192,
"learning_rate": 5.903740139175752e-06,
"loss": 0.3584,
"step": 7470
},
{
"epoch": 1.496,
"grad_norm": 1.09158456325531,
"learning_rate": 5.892293817813048e-06,
"loss": 0.3162,
"step": 7480
},
{
"epoch": 1.498,
"grad_norm": 1.4084733724594116,
"learning_rate": 5.8808426643060265e-06,
"loss": 0.3474,
"step": 7490
},
{
"epoch": 1.5,
"grad_norm": 1.2051564455032349,
"learning_rate": 5.869386740667478e-06,
"loss": 0.3169,
"step": 7500
},
{
"epoch": 1.502,
"grad_norm": 1.2120217084884644,
"learning_rate": 5.857926108936015e-06,
"loss": 0.3545,
"step": 7510
},
{
"epoch": 1.504,
"grad_norm": 1.084529161453247,
"learning_rate": 5.84646083117576e-06,
"loss": 0.3407,
"step": 7520
},
{
"epoch": 1.506,
"grad_norm": 2.2102389335632324,
"learning_rate": 5.834990969475984e-06,
"loss": 0.3275,
"step": 7530
},
{
"epoch": 1.508,
"grad_norm": 1.6285516023635864,
"learning_rate": 5.823516585950787e-06,
"loss": 0.2969,
"step": 7540
},
{
"epoch": 1.51,
"grad_norm": 1.2168736457824707,
"learning_rate": 5.812037742738759e-06,
"loss": 0.3845,
"step": 7550
},
{
"epoch": 1.512,
"grad_norm": 1.2020785808563232,
"learning_rate": 5.800554502002635e-06,
"loss": 0.2706,
"step": 7560
},
{
"epoch": 1.514,
"grad_norm": 1.3338371515274048,
"learning_rate": 5.78906692592897e-06,
"loss": 0.3217,
"step": 7570
},
{
"epoch": 1.516,
"grad_norm": 1.2562854290008545,
"learning_rate": 5.77757507672779e-06,
"loss": 0.338,
"step": 7580
},
{
"epoch": 1.518,
"grad_norm": 1.2902873754501343,
"learning_rate": 5.766079016632272e-06,
"loss": 0.3115,
"step": 7590
},
{
"epoch": 1.52,
"grad_norm": 1.3528056144714355,
"learning_rate": 5.7545788078983875e-06,
"loss": 0.3526,
"step": 7600
},
{
"epoch": 1.522,
"grad_norm": 0.9707136750221252,
"learning_rate": 5.743074512804579e-06,
"loss": 0.3693,
"step": 7610
},
{
"epoch": 1.524,
"grad_norm": 1.6493264436721802,
"learning_rate": 5.731566193651416e-06,
"loss": 0.3304,
"step": 7620
},
{
"epoch": 1.526,
"grad_norm": 1.095731258392334,
"learning_rate": 5.720053912761261e-06,
"loss": 0.3102,
"step": 7630
},
{
"epoch": 1.528,
"grad_norm": 1.691691279411316,
"learning_rate": 5.708537732477934e-06,
"loss": 0.3019,
"step": 7640
},
{
"epoch": 1.53,
"grad_norm": 1.4168593883514404,
"learning_rate": 5.697017715166366e-06,
"loss": 0.3338,
"step": 7650
},
{
"epoch": 1.532,
"grad_norm": 1.5588029623031616,
"learning_rate": 5.685493923212273e-06,
"loss": 0.2633,
"step": 7660
},
{
"epoch": 1.534,
"grad_norm": 1.6496397256851196,
"learning_rate": 5.673966419021806e-06,
"loss": 0.4173,
"step": 7670
},
{
"epoch": 1.536,
"grad_norm": 1.0448615550994873,
"learning_rate": 5.662435265021225e-06,
"loss": 0.2849,
"step": 7680
},
{
"epoch": 1.538,
"grad_norm": 1.2294347286224365,
"learning_rate": 5.650900523656553e-06,
"loss": 0.2609,
"step": 7690
},
{
"epoch": 1.54,
"grad_norm": 1.5806496143341064,
"learning_rate": 5.63936225739324e-06,
"loss": 0.3771,
"step": 7700
},
{
"epoch": 1.542,
"grad_norm": 1.1037795543670654,
"learning_rate": 5.627820528715824e-06,
"loss": 0.3149,
"step": 7710
},
{
"epoch": 1.544,
"grad_norm": 0.8963010907173157,
"learning_rate": 5.616275400127594e-06,
"loss": 0.3073,
"step": 7720
},
{
"epoch": 1.546,
"grad_norm": 1.5788958072662354,
"learning_rate": 5.604726934150253e-06,
"loss": 0.3628,
"step": 7730
},
{
"epoch": 1.548,
"grad_norm": 1.0473259687423706,
"learning_rate": 5.593175193323574e-06,
"loss": 0.3114,
"step": 7740
},
{
"epoch": 1.55,
"grad_norm": 1.295811414718628,
"learning_rate": 5.581620240205068e-06,
"loss": 0.4363,
"step": 7750
},
{
"epoch": 1.552,
"grad_norm": 1.9268922805786133,
"learning_rate": 5.57006213736964e-06,
"loss": 0.2928,
"step": 7760
},
{
"epoch": 1.554,
"grad_norm": 1.4225900173187256,
"learning_rate": 5.558500947409249e-06,
"loss": 0.3989,
"step": 7770
},
{
"epoch": 1.556,
"grad_norm": 1.2886687517166138,
"learning_rate": 5.546936732932578e-06,
"loss": 0.3041,
"step": 7780
},
{
"epoch": 1.558,
"grad_norm": 1.265717625617981,
"learning_rate": 5.535369556564687e-06,
"loss": 0.3384,
"step": 7790
},
{
"epoch": 1.56,
"grad_norm": 1.2660807371139526,
"learning_rate": 5.523799480946673e-06,
"loss": 0.294,
"step": 7800
},
{
"epoch": 1.562,
"grad_norm": 1.319225788116455,
"learning_rate": 5.512226568735338e-06,
"loss": 0.3159,
"step": 7810
},
{
"epoch": 1.564,
"grad_norm": 1.193955659866333,
"learning_rate": 5.500650882602842e-06,
"loss": 0.3044,
"step": 7820
},
{
"epoch": 1.5659999999999998,
"grad_norm": 1.2253046035766602,
"learning_rate": 5.489072485236368e-06,
"loss": 0.3376,
"step": 7830
},
{
"epoch": 1.568,
"grad_norm": 1.0734641551971436,
"learning_rate": 5.477491439337782e-06,
"loss": 0.3555,
"step": 7840
},
{
"epoch": 1.5699999999999998,
"grad_norm": 1.7126284837722778,
"learning_rate": 5.4659078076232906e-06,
"loss": 0.3871,
"step": 7850
},
{
"epoch": 1.572,
"grad_norm": 1.323228359222412,
"learning_rate": 5.45432165282311e-06,
"loss": 0.3337,
"step": 7860
},
{
"epoch": 1.5739999999999998,
"grad_norm": 1.7261728048324585,
"learning_rate": 5.442733037681112e-06,
"loss": 0.2912,
"step": 7870
},
{
"epoch": 1.576,
"grad_norm": 1.790432095527649,
"learning_rate": 5.431142024954496e-06,
"loss": 0.3271,
"step": 7880
},
{
"epoch": 1.5779999999999998,
"grad_norm": 1.6568429470062256,
"learning_rate": 5.419548677413445e-06,
"loss": 0.3535,
"step": 7890
},
{
"epoch": 1.58,
"grad_norm": 1.545526146888733,
"learning_rate": 5.4079530578407895e-06,
"loss": 0.348,
"step": 7900
},
{
"epoch": 1.5819999999999999,
"grad_norm": 1.8154616355895996,
"learning_rate": 5.396355229031656e-06,
"loss": 0.2892,
"step": 7910
},
{
"epoch": 1.584,
"grad_norm": 0.9787377715110779,
"learning_rate": 5.3847552537931395e-06,
"loss": 0.2931,
"step": 7920
},
{
"epoch": 1.5859999999999999,
"grad_norm": 1.3805267810821533,
"learning_rate": 5.373153194943962e-06,
"loss": 0.3615,
"step": 7930
},
{
"epoch": 1.588,
"grad_norm": 2.3790717124938965,
"learning_rate": 5.361549115314123e-06,
"loss": 0.3537,
"step": 7940
},
{
"epoch": 1.5899999999999999,
"grad_norm": 1.5367940664291382,
"learning_rate": 5.349943077744573e-06,
"loss": 0.3413,
"step": 7950
},
{
"epoch": 1.592,
"grad_norm": 1.2089755535125732,
"learning_rate": 5.338335145086855e-06,
"loss": 0.3632,
"step": 7960
},
{
"epoch": 1.5939999999999999,
"grad_norm": 1.4796146154403687,
"learning_rate": 5.3267253802027826e-06,
"loss": 0.4095,
"step": 7970
},
{
"epoch": 1.596,
"grad_norm": 1.4473693370819092,
"learning_rate": 5.315113845964091e-06,
"loss": 0.3714,
"step": 7980
},
{
"epoch": 1.5979999999999999,
"grad_norm": 1.5862455368041992,
"learning_rate": 5.303500605252095e-06,
"loss": 0.3283,
"step": 7990
},
{
"epoch": 1.6,
"grad_norm": 1.4115406274795532,
"learning_rate": 5.291885720957351e-06,
"loss": 0.3127,
"step": 8000
},
{
"epoch": 1.6019999999999999,
"grad_norm": 1.6145530939102173,
"learning_rate": 5.2802692559793175e-06,
"loss": 0.4085,
"step": 8010
},
{
"epoch": 1.604,
"grad_norm": 1.8569374084472656,
"learning_rate": 5.268651273226011e-06,
"loss": 0.2898,
"step": 8020
},
{
"epoch": 1.6059999999999999,
"grad_norm": 1.736876368522644,
"learning_rate": 5.2570318356136705e-06,
"loss": 0.3503,
"step": 8030
},
{
"epoch": 1.608,
"grad_norm": 1.5102406740188599,
"learning_rate": 5.2454110060664075e-06,
"loss": 0.3111,
"step": 8040
},
{
"epoch": 1.6099999999999999,
"grad_norm": 1.7225403785705566,
"learning_rate": 5.233788847515882e-06,
"loss": 0.3586,
"step": 8050
},
{
"epoch": 1.612,
"grad_norm": 1.0642247200012207,
"learning_rate": 5.222165422900939e-06,
"loss": 0.3721,
"step": 8060
},
{
"epoch": 1.6139999999999999,
"grad_norm": 1.3807940483093262,
"learning_rate": 5.210540795167287e-06,
"loss": 0.398,
"step": 8070
},
{
"epoch": 1.616,
"grad_norm": 1.396348237991333,
"learning_rate": 5.19891502726715e-06,
"loss": 0.3657,
"step": 8080
},
{
"epoch": 1.6179999999999999,
"grad_norm": 0.8553020358085632,
"learning_rate": 5.187288182158924e-06,
"loss": 0.3018,
"step": 8090
},
{
"epoch": 1.62,
"grad_norm": 1.608389973640442,
"learning_rate": 5.175660322806838e-06,
"loss": 0.3563,
"step": 8100
},
{
"epoch": 1.6219999999999999,
"grad_norm": 1.3906745910644531,
"learning_rate": 5.164031512180616e-06,
"loss": 0.3186,
"step": 8110
},
{
"epoch": 1.624,
"grad_norm": 1.6109449863433838,
"learning_rate": 5.152401813255134e-06,
"loss": 0.3028,
"step": 8120
},
{
"epoch": 1.626,
"grad_norm": 1.3031402826309204,
"learning_rate": 5.140771289010073e-06,
"loss": 0.3089,
"step": 8130
},
{
"epoch": 1.6280000000000001,
"grad_norm": 0.8983224034309387,
"learning_rate": 5.1291400024295946e-06,
"loss": 0.3388,
"step": 8140
},
{
"epoch": 1.63,
"grad_norm": 1.6017338037490845,
"learning_rate": 5.117508016501973e-06,
"loss": 0.3465,
"step": 8150
},
{
"epoch": 1.6320000000000001,
"grad_norm": 1.4374866485595703,
"learning_rate": 5.105875394219283e-06,
"loss": 0.3152,
"step": 8160
},
{
"epoch": 1.634,
"grad_norm": 1.3967483043670654,
"learning_rate": 5.0942421985770415e-06,
"loss": 0.3158,
"step": 8170
},
{
"epoch": 1.6360000000000001,
"grad_norm": 1.0798684358596802,
"learning_rate": 5.0826084925738675e-06,
"loss": 0.3273,
"step": 8180
},
{
"epoch": 1.638,
"grad_norm": 1.1431955099105835,
"learning_rate": 5.070974339211148e-06,
"loss": 0.4185,
"step": 8190
},
{
"epoch": 1.6400000000000001,
"grad_norm": 1.4295889139175415,
"learning_rate": 5.059339801492687e-06,
"loss": 0.3667,
"step": 8200
},
{
"epoch": 1.642,
"grad_norm": 1.2767932415008545,
"learning_rate": 5.047704942424377e-06,
"loss": 0.3993,
"step": 8210
},
{
"epoch": 1.6440000000000001,
"grad_norm": 1.4808293581008911,
"learning_rate": 5.0360698250138465e-06,
"loss": 0.3618,
"step": 8220
},
{
"epoch": 1.646,
"grad_norm": 1.2203787565231323,
"learning_rate": 5.024434512270123e-06,
"loss": 0.3237,
"step": 8230
},
{
"epoch": 1.6480000000000001,
"grad_norm": 1.2480844259262085,
"learning_rate": 5.0127990672032945e-06,
"loss": 0.3361,
"step": 8240
},
{
"epoch": 1.65,
"grad_norm": 1.269690990447998,
"learning_rate": 5.001163552824162e-06,
"loss": 0.2754,
"step": 8250
},
{
"epoch": 1.6520000000000001,
"grad_norm": 1.5688775777816772,
"learning_rate": 4.9895280321439036e-06,
"loss": 0.3312,
"step": 8260
},
{
"epoch": 1.654,
"grad_norm": 1.5578945875167847,
"learning_rate": 4.977892568173733e-06,
"loss": 0.3141,
"step": 8270
},
{
"epoch": 1.6560000000000001,
"grad_norm": 1.0505149364471436,
"learning_rate": 4.966257223924554e-06,
"loss": 0.2919,
"step": 8280
},
{
"epoch": 1.658,
"grad_norm": 1.2484036684036255,
"learning_rate": 4.954622062406623e-06,
"loss": 0.3513,
"step": 8290
},
{
"epoch": 1.6600000000000001,
"grad_norm": 1.8447239398956299,
"learning_rate": 4.94298714662921e-06,
"loss": 0.3111,
"step": 8300
},
{
"epoch": 1.662,
"grad_norm": 1.375136137008667,
"learning_rate": 4.931352539600248e-06,
"loss": 0.2827,
"step": 8310
},
{
"epoch": 1.6640000000000001,
"grad_norm": 1.4478310346603394,
"learning_rate": 4.9197183043260035e-06,
"loss": 0.3177,
"step": 8320
},
{
"epoch": 1.666,
"grad_norm": 1.2656853199005127,
"learning_rate": 4.9080845038107264e-06,
"loss": 0.2888,
"step": 8330
},
{
"epoch": 1.6680000000000001,
"grad_norm": 1.177017092704773,
"learning_rate": 4.896451201056315e-06,
"loss": 0.365,
"step": 8340
},
{
"epoch": 1.67,
"grad_norm": 1.2763571739196777,
"learning_rate": 4.88481845906197e-06,
"loss": 0.2697,
"step": 8350
},
{
"epoch": 1.6720000000000002,
"grad_norm": 1.6193398237228394,
"learning_rate": 4.873186340823854e-06,
"loss": 0.2913,
"step": 8360
},
{
"epoch": 1.674,
"grad_norm": 1.353211760520935,
"learning_rate": 4.861554909334757e-06,
"loss": 0.3652,
"step": 8370
},
{
"epoch": 1.6760000000000002,
"grad_norm": 1.4687409400939941,
"learning_rate": 4.8499242275837444e-06,
"loss": 0.3417,
"step": 8380
},
{
"epoch": 1.678,
"grad_norm": 1.644424557685852,
"learning_rate": 4.838294358555824e-06,
"loss": 0.3228,
"step": 8390
},
{
"epoch": 1.6800000000000002,
"grad_norm": 1.8202049732208252,
"learning_rate": 4.826665365231601e-06,
"loss": 0.2984,
"step": 8400
},
{
"epoch": 1.682,
"grad_norm": 0.9253039360046387,
"learning_rate": 4.815037310586941e-06,
"loss": 0.3807,
"step": 8410
},
{
"epoch": 1.6840000000000002,
"grad_norm": 1.1849908828735352,
"learning_rate": 4.803410257592625e-06,
"loss": 0.2896,
"step": 8420
},
{
"epoch": 1.686,
"grad_norm": 1.0804787874221802,
"learning_rate": 4.791784269214004e-06,
"loss": 0.3129,
"step": 8430
},
{
"epoch": 1.688,
"grad_norm": 1.3081520795822144,
"learning_rate": 4.780159408410677e-06,
"loss": 0.3055,
"step": 8440
},
{
"epoch": 1.69,
"grad_norm": 1.2547225952148438,
"learning_rate": 4.7685357381361224e-06,
"loss": 0.3321,
"step": 8450
},
{
"epoch": 1.692,
"grad_norm": 1.228356957435608,
"learning_rate": 4.75691332133738e-06,
"loss": 0.3556,
"step": 8460
},
{
"epoch": 1.694,
"grad_norm": 1.2856372594833374,
"learning_rate": 4.745292220954696e-06,
"loss": 0.3334,
"step": 8470
},
{
"epoch": 1.696,
"grad_norm": 1.2076910734176636,
"learning_rate": 4.733672499921195e-06,
"loss": 0.2957,
"step": 8480
},
{
"epoch": 1.698,
"grad_norm": 1.3459007740020752,
"learning_rate": 4.722054221162528e-06,
"loss": 0.3326,
"step": 8490
},
{
"epoch": 1.7,
"grad_norm": 1.899490475654602,
"learning_rate": 4.710437447596528e-06,
"loss": 0.3721,
"step": 8500
},
{
"epoch": 1.702,
"grad_norm": 1.0619953870773315,
"learning_rate": 4.698822242132891e-06,
"loss": 0.3474,
"step": 8510
},
{
"epoch": 1.704,
"grad_norm": 1.1741877794265747,
"learning_rate": 4.687208667672812e-06,
"loss": 0.3206,
"step": 8520
},
{
"epoch": 1.706,
"grad_norm": 1.295904517173767,
"learning_rate": 4.675596787108652e-06,
"loss": 0.3445,
"step": 8530
},
{
"epoch": 1.708,
"grad_norm": 1.2269822359085083,
"learning_rate": 4.6639866633236066e-06,
"loss": 0.32,
"step": 8540
},
{
"epoch": 1.71,
"grad_norm": 1.142608404159546,
"learning_rate": 4.652378359191352e-06,
"loss": 0.3178,
"step": 8550
},
{
"epoch": 1.712,
"grad_norm": 1.4112498760223389,
"learning_rate": 4.6407719375757095e-06,
"loss": 0.3598,
"step": 8560
},
{
"epoch": 1.714,
"grad_norm": 1.3373513221740723,
"learning_rate": 4.629167461330308e-06,
"loss": 0.3895,
"step": 8570
},
{
"epoch": 1.716,
"grad_norm": 1.5300648212432861,
"learning_rate": 4.617564993298244e-06,
"loss": 0.3675,
"step": 8580
},
{
"epoch": 1.718,
"grad_norm": 1.3708034753799438,
"learning_rate": 4.605964596311733e-06,
"loss": 0.3543,
"step": 8590
},
{
"epoch": 1.72,
"grad_norm": 1.1198241710662842,
"learning_rate": 4.594366333191778e-06,
"loss": 0.4109,
"step": 8600
},
{
"epoch": 1.722,
"grad_norm": 1.1231240034103394,
"learning_rate": 4.5827702667478275e-06,
"loss": 0.3643,
"step": 8610
},
{
"epoch": 1.724,
"grad_norm": 1.6655538082122803,
"learning_rate": 4.571176459777431e-06,
"loss": 0.3271,
"step": 8620
},
{
"epoch": 1.726,
"grad_norm": 1.260842204093933,
"learning_rate": 4.559584975065905e-06,
"loss": 0.3342,
"step": 8630
},
{
"epoch": 1.728,
"grad_norm": 1.2399426698684692,
"learning_rate": 4.547995875385986e-06,
"loss": 0.2852,
"step": 8640
},
{
"epoch": 1.73,
"grad_norm": 1.537028193473816,
"learning_rate": 4.5364092234975e-06,
"loss": 0.4146,
"step": 8650
},
{
"epoch": 1.732,
"grad_norm": 1.4281028509140015,
"learning_rate": 4.524825082147013e-06,
"loss": 0.2961,
"step": 8660
},
{
"epoch": 1.734,
"grad_norm": 1.294998049736023,
"learning_rate": 4.513243514067495e-06,
"loss": 0.2937,
"step": 8670
},
{
"epoch": 1.736,
"grad_norm": 1.0920134782791138,
"learning_rate": 4.5016645819779865e-06,
"loss": 0.3434,
"step": 8680
},
{
"epoch": 1.738,
"grad_norm": 1.160057544708252,
"learning_rate": 4.490088348583246e-06,
"loss": 0.2608,
"step": 8690
},
{
"epoch": 1.74,
"grad_norm": 1.2526758909225464,
"learning_rate": 4.47851487657342e-06,
"loss": 0.3652,
"step": 8700
},
{
"epoch": 1.742,
"grad_norm": 1.275640845298767,
"learning_rate": 4.466944228623701e-06,
"loss": 0.3862,
"step": 8710
},
{
"epoch": 1.744,
"grad_norm": 1.3032732009887695,
"learning_rate": 4.455376467393991e-06,
"loss": 0.2721,
"step": 8720
},
{
"epoch": 1.746,
"grad_norm": 1.2573343515396118,
"learning_rate": 4.443811655528553e-06,
"loss": 0.3276,
"step": 8730
},
{
"epoch": 1.748,
"grad_norm": 1.478424310684204,
"learning_rate": 4.432249855655681e-06,
"loss": 0.3695,
"step": 8740
},
{
"epoch": 1.75,
"grad_norm": 1.3663966655731201,
"learning_rate": 4.420691130387365e-06,
"loss": 0.3233,
"step": 8750
},
{
"epoch": 1.752,
"grad_norm": 1.5956822633743286,
"learning_rate": 4.409135542318931e-06,
"loss": 0.3789,
"step": 8760
},
{
"epoch": 1.754,
"grad_norm": 1.566262125968933,
"learning_rate": 4.397583154028725e-06,
"loss": 0.3546,
"step": 8770
},
{
"epoch": 1.756,
"grad_norm": 1.0735946893692017,
"learning_rate": 4.38603402807776e-06,
"loss": 0.3638,
"step": 8780
},
{
"epoch": 1.758,
"grad_norm": 1.182511329650879,
"learning_rate": 4.374488227009391e-06,
"loss": 0.3271,
"step": 8790
},
{
"epoch": 1.76,
"grad_norm": 1.5073894262313843,
"learning_rate": 4.362945813348956e-06,
"loss": 0.2934,
"step": 8800
},
{
"epoch": 1.762,
"grad_norm": 1.476199746131897,
"learning_rate": 4.351406849603452e-06,
"loss": 0.3143,
"step": 8810
},
{
"epoch": 1.764,
"grad_norm": 1.2955067157745361,
"learning_rate": 4.3398713982612e-06,
"loss": 0.3358,
"step": 8820
},
{
"epoch": 1.766,
"grad_norm": 1.4470880031585693,
"learning_rate": 4.328339521791493e-06,
"loss": 0.3117,
"step": 8830
},
{
"epoch": 1.768,
"grad_norm": 1.264491319656372,
"learning_rate": 4.316811282644265e-06,
"loss": 0.3316,
"step": 8840
},
{
"epoch": 1.77,
"grad_norm": 0.9871441125869751,
"learning_rate": 4.305286743249756e-06,
"loss": 0.3224,
"step": 8850
},
{
"epoch": 1.772,
"grad_norm": 1.7321498394012451,
"learning_rate": 4.293765966018167e-06,
"loss": 0.3177,
"step": 8860
},
{
"epoch": 1.774,
"grad_norm": 1.4253642559051514,
"learning_rate": 4.282249013339328e-06,
"loss": 0.396,
"step": 8870
},
{
"epoch": 1.776,
"grad_norm": 1.1400525569915771,
"learning_rate": 4.270735947582352e-06,
"loss": 0.3207,
"step": 8880
},
{
"epoch": 1.778,
"grad_norm": 1.3001306056976318,
"learning_rate": 4.259226831095311e-06,
"loss": 0.3579,
"step": 8890
},
{
"epoch": 1.78,
"grad_norm": 0.9259517788887024,
"learning_rate": 4.247721726204883e-06,
"loss": 0.3522,
"step": 8900
},
{
"epoch": 1.782,
"grad_norm": 1.352211594581604,
"learning_rate": 4.236220695216024e-06,
"loss": 0.3184,
"step": 8910
},
{
"epoch": 1.784,
"grad_norm": 1.4260315895080566,
"learning_rate": 4.224723800411631e-06,
"loss": 0.3632,
"step": 8920
},
{
"epoch": 1.786,
"grad_norm": 1.3724855184555054,
"learning_rate": 4.2132311040521975e-06,
"loss": 0.314,
"step": 8930
},
{
"epoch": 1.788,
"grad_norm": 1.4513239860534668,
"learning_rate": 4.201742668375481e-06,
"loss": 0.3224,
"step": 8940
},
{
"epoch": 1.79,
"grad_norm": 1.2112324237823486,
"learning_rate": 4.190258555596168e-06,
"loss": 0.3068,
"step": 8950
},
{
"epoch": 1.792,
"grad_norm": 1.5572727918624878,
"learning_rate": 4.178778827905535e-06,
"loss": 0.3771,
"step": 8960
},
{
"epoch": 1.794,
"grad_norm": 0.9895251989364624,
"learning_rate": 4.167303547471108e-06,
"loss": 0.3205,
"step": 8970
},
{
"epoch": 1.796,
"grad_norm": 1.2107610702514648,
"learning_rate": 4.155832776436331e-06,
"loss": 0.353,
"step": 8980
},
{
"epoch": 1.798,
"grad_norm": 1.2437814474105835,
"learning_rate": 4.14436657692023e-06,
"loss": 0.3566,
"step": 8990
},
{
"epoch": 1.8,
"grad_norm": 1.1405911445617676,
"learning_rate": 4.132905011017071e-06,
"loss": 0.3001,
"step": 9000
},
{
"epoch": 1.802,
"grad_norm": 1.2086176872253418,
"learning_rate": 4.121448140796029e-06,
"loss": 0.3399,
"step": 9010
},
{
"epoch": 1.804,
"grad_norm": 1.3747698068618774,
"learning_rate": 4.109996028300847e-06,
"loss": 0.3545,
"step": 9020
},
{
"epoch": 1.806,
"grad_norm": 1.1939738988876343,
"learning_rate": 4.098548735549508e-06,
"loss": 0.3397,
"step": 9030
},
{
"epoch": 1.808,
"grad_norm": 1.5216190814971924,
"learning_rate": 4.087106324533891e-06,
"loss": 0.3071,
"step": 9040
},
{
"epoch": 1.81,
"grad_norm": 1.557185411453247,
"learning_rate": 4.075668857219436e-06,
"loss": 0.2898,
"step": 9050
},
{
"epoch": 1.812,
"grad_norm": 1.694498896598816,
"learning_rate": 4.0642363955448175e-06,
"loss": 0.3705,
"step": 9060
},
{
"epoch": 1.814,
"grad_norm": 1.7340755462646484,
"learning_rate": 4.052809001421595e-06,
"loss": 0.3218,
"step": 9070
},
{
"epoch": 1.8159999999999998,
"grad_norm": 1.123704195022583,
"learning_rate": 4.041386736733889e-06,
"loss": 0.3743,
"step": 9080
},
{
"epoch": 1.818,
"grad_norm": 1.6737864017486572,
"learning_rate": 4.029969663338042e-06,
"loss": 0.3169,
"step": 9090
},
{
"epoch": 1.8199999999999998,
"grad_norm": 1.1411502361297607,
"learning_rate": 4.018557843062282e-06,
"loss": 0.2783,
"step": 9100
},
{
"epoch": 1.822,
"grad_norm": 1.4459763765335083,
"learning_rate": 4.007151337706391e-06,
"loss": 0.3288,
"step": 9110
},
{
"epoch": 1.8239999999999998,
"grad_norm": 1.1251758337020874,
"learning_rate": 3.995750209041365e-06,
"loss": 0.2966,
"step": 9120
},
{
"epoch": 1.826,
"grad_norm": 1.534656286239624,
"learning_rate": 3.98435451880909e-06,
"loss": 0.3306,
"step": 9130
},
{
"epoch": 1.8279999999999998,
"grad_norm": 1.2617437839508057,
"learning_rate": 3.972964328721992e-06,
"loss": 0.2749,
"step": 9140
},
{
"epoch": 1.83,
"grad_norm": 1.1563127040863037,
"learning_rate": 3.961579700462715e-06,
"loss": 0.348,
"step": 9150
},
{
"epoch": 1.8319999999999999,
"grad_norm": 1.5472575426101685,
"learning_rate": 3.950200695683788e-06,
"loss": 0.297,
"step": 9160
},
{
"epoch": 1.834,
"grad_norm": 1.342626690864563,
"learning_rate": 3.938827376007281e-06,
"loss": 0.3328,
"step": 9170
},
{
"epoch": 1.8359999999999999,
"grad_norm": 1.0215874910354614,
"learning_rate": 3.927459803024475e-06,
"loss": 0.3094,
"step": 9180
},
{
"epoch": 1.838,
"grad_norm": 1.6598362922668457,
"learning_rate": 3.9160980382955336e-06,
"loss": 0.3091,
"step": 9190
},
{
"epoch": 1.8399999999999999,
"grad_norm": 1.050584077835083,
"learning_rate": 3.904742143349169e-06,
"loss": 0.3078,
"step": 9200
},
{
"epoch": 1.842,
"grad_norm": 1.560577630996704,
"learning_rate": 3.893392179682304e-06,
"loss": 0.323,
"step": 9210
},
{
"epoch": 1.8439999999999999,
"grad_norm": 1.0825576782226562,
"learning_rate": 3.882048208759735e-06,
"loss": 0.3313,
"step": 9220
},
{
"epoch": 1.846,
"grad_norm": 1.0511835813522339,
"learning_rate": 3.870710292013815e-06,
"loss": 0.3295,
"step": 9230
},
{
"epoch": 1.8479999999999999,
"grad_norm": 1.3847706317901611,
"learning_rate": 3.859378490844104e-06,
"loss": 0.3417,
"step": 9240
},
{
"epoch": 1.85,
"grad_norm": 1.1455384492874146,
"learning_rate": 3.8480528666170495e-06,
"loss": 0.3394,
"step": 9250
},
{
"epoch": 1.8519999999999999,
"grad_norm": 1.5466185808181763,
"learning_rate": 3.836733480665637e-06,
"loss": 0.3222,
"step": 9260
},
{
"epoch": 1.854,
"grad_norm": 1.2878413200378418,
"learning_rate": 3.825420394289085e-06,
"loss": 0.3517,
"step": 9270
},
{
"epoch": 1.8559999999999999,
"grad_norm": 0.8397305011749268,
"learning_rate": 3.814113668752486e-06,
"loss": 0.3493,
"step": 9280
},
{
"epoch": 1.858,
"grad_norm": 1.3638249635696411,
"learning_rate": 3.8028133652864872e-06,
"loss": 0.4023,
"step": 9290
},
{
"epoch": 1.8599999999999999,
"grad_norm": 1.1861424446105957,
"learning_rate": 3.791519545086963e-06,
"loss": 0.3283,
"step": 9300
},
{
"epoch": 1.862,
"grad_norm": 1.501705527305603,
"learning_rate": 3.7802322693146726e-06,
"loss": 0.3947,
"step": 9310
},
{
"epoch": 1.8639999999999999,
"grad_norm": 1.1663841009140015,
"learning_rate": 3.7689515990949364e-06,
"loss": 0.3183,
"step": 9320
},
{
"epoch": 1.866,
"grad_norm": 1.418528437614441,
"learning_rate": 3.757677595517302e-06,
"loss": 0.2897,
"step": 9330
},
{
"epoch": 1.8679999999999999,
"grad_norm": 1.609471321105957,
"learning_rate": 3.7464103196352176e-06,
"loss": 0.3356,
"step": 9340
},
{
"epoch": 1.87,
"grad_norm": 1.3416111469268799,
"learning_rate": 3.7351498324656944e-06,
"loss": 0.3206,
"step": 9350
},
{
"epoch": 1.8719999999999999,
"grad_norm": 1.485053539276123,
"learning_rate": 3.7238961949889796e-06,
"loss": 0.3133,
"step": 9360
},
{
"epoch": 1.874,
"grad_norm": 1.3999130725860596,
"learning_rate": 3.7126494681482317e-06,
"loss": 0.367,
"step": 9370
},
{
"epoch": 1.876,
"grad_norm": 1.3190253973007202,
"learning_rate": 3.70140971284918e-06,
"loss": 0.3881,
"step": 9380
},
{
"epoch": 1.8780000000000001,
"grad_norm": 1.468934416770935,
"learning_rate": 3.690176989959801e-06,
"loss": 0.3039,
"step": 9390
},
{
"epoch": 1.88,
"grad_norm": 1.4125316143035889,
"learning_rate": 3.678951360309988e-06,
"loss": 0.3603,
"step": 9400
},
{
"epoch": 1.8820000000000001,
"grad_norm": 1.2523292303085327,
"learning_rate": 3.6677328846912237e-06,
"loss": 0.3139,
"step": 9410
},
{
"epoch": 1.884,
"grad_norm": 1.6310887336730957,
"learning_rate": 3.6565216238562464e-06,
"loss": 0.342,
"step": 9420
},
{
"epoch": 1.8860000000000001,
"grad_norm": 1.028517723083496,
"learning_rate": 3.645317638518721e-06,
"loss": 0.3088,
"step": 9430
},
{
"epoch": 1.888,
"grad_norm": 1.4186135530471802,
"learning_rate": 3.6341209893529195e-06,
"loss": 0.3238,
"step": 9440
},
{
"epoch": 1.8900000000000001,
"grad_norm": 1.2561564445495605,
"learning_rate": 3.6229317369933786e-06,
"loss": 0.3813,
"step": 9450
},
{
"epoch": 1.892,
"grad_norm": 1.3972420692443848,
"learning_rate": 3.61174994203458e-06,
"loss": 0.3561,
"step": 9460
},
{
"epoch": 1.8940000000000001,
"grad_norm": 1.1440085172653198,
"learning_rate": 3.6005756650306258e-06,
"loss": 0.2912,
"step": 9470
},
{
"epoch": 1.896,
"grad_norm": 1.243796944618225,
"learning_rate": 3.589408966494897e-06,
"loss": 0.2798,
"step": 9480
},
{
"epoch": 1.8980000000000001,
"grad_norm": 1.2843304872512817,
"learning_rate": 3.5782499068997386e-06,
"loss": 0.3649,
"step": 9490
},
{
"epoch": 1.9,
"grad_norm": 1.2538890838623047,
"learning_rate": 3.5670985466761243e-06,
"loss": 0.3573,
"step": 9500
},
{
"epoch": 1.9020000000000001,
"grad_norm": 1.5263408422470093,
"learning_rate": 3.5559549462133407e-06,
"loss": 0.3468,
"step": 9510
},
{
"epoch": 1.904,
"grad_norm": 1.2119961977005005,
"learning_rate": 3.5448191658586423e-06,
"loss": 0.2936,
"step": 9520
},
{
"epoch": 1.9060000000000001,
"grad_norm": 1.630476474761963,
"learning_rate": 3.5336912659169366e-06,
"loss": 0.3447,
"step": 9530
},
{
"epoch": 1.908,
"grad_norm": 1.7474474906921387,
"learning_rate": 3.522571306650462e-06,
"loss": 0.3666,
"step": 9540
},
{
"epoch": 1.9100000000000001,
"grad_norm": 2.313354730606079,
"learning_rate": 3.511459348278448e-06,
"loss": 0.3717,
"step": 9550
},
{
"epoch": 1.912,
"grad_norm": 1.660021185874939,
"learning_rate": 3.5003554509767966e-06,
"loss": 0.3335,
"step": 9560
},
{
"epoch": 1.9140000000000001,
"grad_norm": 1.404435396194458,
"learning_rate": 3.4892596748777563e-06,
"loss": 0.3466,
"step": 9570
},
{
"epoch": 1.916,
"grad_norm": 1.2691371440887451,
"learning_rate": 3.4781720800696006e-06,
"loss": 0.3342,
"step": 9580
},
{
"epoch": 1.9180000000000001,
"grad_norm": 1.8161771297454834,
"learning_rate": 3.4670927265962908e-06,
"loss": 0.3479,
"step": 9590
},
{
"epoch": 1.92,
"grad_norm": 1.5305116176605225,
"learning_rate": 3.4560216744571607e-06,
"loss": 0.3067,
"step": 9600
},
{
"epoch": 1.9220000000000002,
"grad_norm": 1.5921387672424316,
"learning_rate": 3.444958983606592e-06,
"loss": 0.3088,
"step": 9610
},
{
"epoch": 1.924,
"grad_norm": 1.490713119506836,
"learning_rate": 3.433904713953682e-06,
"loss": 0.3352,
"step": 9620
},
{
"epoch": 1.9260000000000002,
"grad_norm": 1.590400218963623,
"learning_rate": 3.4228589253619247e-06,
"loss": 0.3203,
"step": 9630
},
{
"epoch": 1.928,
"grad_norm": 1.140105128288269,
"learning_rate": 3.411821677648887e-06,
"loss": 0.3612,
"step": 9640
},
{
"epoch": 1.9300000000000002,
"grad_norm": 1.3013497591018677,
"learning_rate": 3.400793030585884e-06,
"loss": 0.277,
"step": 9650
},
{
"epoch": 1.932,
"grad_norm": 1.3757359981536865,
"learning_rate": 3.389773043897652e-06,
"loss": 0.3651,
"step": 9660
},
{
"epoch": 1.9340000000000002,
"grad_norm": 1.2634108066558838,
"learning_rate": 3.378761777262028e-06,
"loss": 0.3305,
"step": 9670
},
{
"epoch": 1.936,
"grad_norm": 1.5106260776519775,
"learning_rate": 3.36775929030963e-06,
"loss": 0.3067,
"step": 9680
},
{
"epoch": 1.938,
"grad_norm": 1.1745408773422241,
"learning_rate": 3.3567656426235275e-06,
"loss": 0.364,
"step": 9690
},
{
"epoch": 1.94,
"grad_norm": 1.7287566661834717,
"learning_rate": 3.34578089373892e-06,
"loss": 0.3663,
"step": 9700
},
{
"epoch": 1.942,
"grad_norm": 1.4090641736984253,
"learning_rate": 3.3348051031428184e-06,
"loss": 0.2921,
"step": 9710
},
{
"epoch": 1.944,
"grad_norm": 1.6581404209136963,
"learning_rate": 3.323838330273723e-06,
"loss": 0.4043,
"step": 9720
},
{
"epoch": 1.946,
"grad_norm": 1.7018346786499023,
"learning_rate": 3.312880634521295e-06,
"loss": 0.3818,
"step": 9730
},
{
"epoch": 1.948,
"grad_norm": 1.4111535549163818,
"learning_rate": 3.301932075226041e-06,
"loss": 0.33,
"step": 9740
},
{
"epoch": 1.95,
"grad_norm": 1.4237406253814697,
"learning_rate": 3.2909927116789908e-06,
"loss": 0.3326,
"step": 9750
},
{
"epoch": 1.952,
"grad_norm": 1.3221118450164795,
"learning_rate": 3.280062603121373e-06,
"loss": 0.2659,
"step": 9760
},
{
"epoch": 1.954,
"grad_norm": 1.3768141269683838,
"learning_rate": 3.2691418087442995e-06,
"loss": 0.3813,
"step": 9770
},
{
"epoch": 1.956,
"grad_norm": 1.5262187719345093,
"learning_rate": 3.2582303876884406e-06,
"loss": 0.3554,
"step": 9780
},
{
"epoch": 1.958,
"grad_norm": 1.184778094291687,
"learning_rate": 3.247328399043706e-06,
"loss": 0.2866,
"step": 9790
},
{
"epoch": 1.96,
"grad_norm": 1.3611551523208618,
"learning_rate": 3.2364359018489245e-06,
"loss": 0.3865,
"step": 9800
},
{
"epoch": 1.962,
"grad_norm": 1.296720027923584,
"learning_rate": 3.2255529550915242e-06,
"loss": 0.3272,
"step": 9810
},
{
"epoch": 1.964,
"grad_norm": 1.1612666845321655,
"learning_rate": 3.2146796177072183e-06,
"loss": 0.2417,
"step": 9820
},
{
"epoch": 1.966,
"grad_norm": 1.3343042135238647,
"learning_rate": 3.203815948579674e-06,
"loss": 0.2749,
"step": 9830
},
{
"epoch": 1.968,
"grad_norm": 1.5263704061508179,
"learning_rate": 3.192962006540205e-06,
"loss": 0.2788,
"step": 9840
},
{
"epoch": 1.97,
"grad_norm": 1.9184904098510742,
"learning_rate": 3.1821178503674515e-06,
"loss": 0.2875,
"step": 9850
},
{
"epoch": 1.972,
"grad_norm": 1.6000325679779053,
"learning_rate": 3.1712835387870527e-06,
"loss": 0.3293,
"step": 9860
},
{
"epoch": 1.974,
"grad_norm": 1.1482776403427124,
"learning_rate": 3.1604591304713394e-06,
"loss": 0.3441,
"step": 9870
},
{
"epoch": 1.976,
"grad_norm": 1.1263827085494995,
"learning_rate": 3.149644684039008e-06,
"loss": 0.3172,
"step": 9880
},
{
"epoch": 1.978,
"grad_norm": 1.3861206769943237,
"learning_rate": 3.1388402580548154e-06,
"loss": 0.3633,
"step": 9890
},
{
"epoch": 1.98,
"grad_norm": 1.7515361309051514,
"learning_rate": 3.1280459110292474e-06,
"loss": 0.3882,
"step": 9900
},
{
"epoch": 1.982,
"grad_norm": 1.409308671951294,
"learning_rate": 3.117261701418204e-06,
"loss": 0.3439,
"step": 9910
},
{
"epoch": 1.984,
"grad_norm": 1.6333212852478027,
"learning_rate": 3.106487687622697e-06,
"loss": 0.3062,
"step": 9920
},
{
"epoch": 1.986,
"grad_norm": 1.5096412897109985,
"learning_rate": 3.095723927988517e-06,
"loss": 0.3322,
"step": 9930
},
{
"epoch": 1.988,
"grad_norm": 0.9250149726867676,
"learning_rate": 3.0849704808059266e-06,
"loss": 0.3363,
"step": 9940
},
{
"epoch": 1.99,
"grad_norm": 1.4919676780700684,
"learning_rate": 3.074227404309336e-06,
"loss": 0.3459,
"step": 9950
},
{
"epoch": 1.992,
"grad_norm": 1.1677219867706299,
"learning_rate": 3.063494756677005e-06,
"loss": 0.3087,
"step": 9960
},
{
"epoch": 1.994,
"grad_norm": 1.3163172006607056,
"learning_rate": 3.0527725960307083e-06,
"loss": 0.3544,
"step": 9970
},
{
"epoch": 1.996,
"grad_norm": 1.1766602993011475,
"learning_rate": 3.0420609804354295e-06,
"loss": 0.2968,
"step": 9980
},
{
"epoch": 1.998,
"grad_norm": 1.754724144935608,
"learning_rate": 3.0313599678990514e-06,
"loss": 0.3324,
"step": 9990
},
{
"epoch": 2.0,
"grad_norm": 1.334686279296875,
"learning_rate": 3.0206696163720317e-06,
"loss": 0.3427,
"step": 10000
},
{
"epoch": 2.002,
"grad_norm": 1.149556279182434,
"learning_rate": 3.0099899837470976e-06,
"loss": 0.2296,
"step": 10010
},
{
"epoch": 2.004,
"grad_norm": 2.647596597671509,
"learning_rate": 2.999321127858925e-06,
"loss": 0.1997,
"step": 10020
},
{
"epoch": 2.006,
"grad_norm": 1.3816395998001099,
"learning_rate": 2.9886631064838355e-06,
"loss": 0.1688,
"step": 10030
},
{
"epoch": 2.008,
"grad_norm": 1.5958433151245117,
"learning_rate": 2.9780159773394713e-06,
"loss": 0.1862,
"step": 10040
},
{
"epoch": 2.01,
"grad_norm": 1.7773202657699585,
"learning_rate": 2.96737979808449e-06,
"loss": 0.1681,
"step": 10050
},
{
"epoch": 2.012,
"grad_norm": 1.0190283060073853,
"learning_rate": 2.9567546263182554e-06,
"loss": 0.1732,
"step": 10060
},
{
"epoch": 2.014,
"grad_norm": 1.5294930934906006,
"learning_rate": 2.9461405195805146e-06,
"loss": 0.1918,
"step": 10070
},
{
"epoch": 2.016,
"grad_norm": 1.375820517539978,
"learning_rate": 2.9355375353510973e-06,
"loss": 0.1856,
"step": 10080
},
{
"epoch": 2.018,
"grad_norm": 1.9150773286819458,
"learning_rate": 2.9249457310495994e-06,
"loss": 0.2228,
"step": 10090
},
{
"epoch": 2.02,
"grad_norm": 1.4141868352890015,
"learning_rate": 2.91436516403507e-06,
"loss": 0.2042,
"step": 10100
},
{
"epoch": 2.022,
"grad_norm": 1.665558099746704,
"learning_rate": 2.9037958916057104e-06,
"loss": 0.1844,
"step": 10110
},
{
"epoch": 2.024,
"grad_norm": 1.4332926273345947,
"learning_rate": 2.893237970998547e-06,
"loss": 0.2012,
"step": 10120
},
{
"epoch": 2.026,
"grad_norm": 1.575321078300476,
"learning_rate": 2.8826914593891396e-06,
"loss": 0.2173,
"step": 10130
},
{
"epoch": 2.028,
"grad_norm": 1.420453667640686,
"learning_rate": 2.872156413891263e-06,
"loss": 0.1859,
"step": 10140
},
{
"epoch": 2.03,
"grad_norm": 1.3123276233673096,
"learning_rate": 2.8616328915565907e-06,
"loss": 0.1539,
"step": 10150
},
{
"epoch": 2.032,
"grad_norm": 2.069629430770874,
"learning_rate": 2.8511209493744015e-06,
"loss": 0.1664,
"step": 10160
},
{
"epoch": 2.034,
"grad_norm": 1.5814684629440308,
"learning_rate": 2.8406206442712618e-06,
"loss": 0.2212,
"step": 10170
},
{
"epoch": 2.036,
"grad_norm": 1.5895538330078125,
"learning_rate": 2.830132033110713e-06,
"loss": 0.1419,
"step": 10180
},
{
"epoch": 2.038,
"grad_norm": 1.795087456703186,
"learning_rate": 2.8196551726929745e-06,
"loss": 0.1644,
"step": 10190
},
{
"epoch": 2.04,
"grad_norm": 1.6379607915878296,
"learning_rate": 2.80919011975463e-06,
"loss": 0.1866,
"step": 10200
},
{
"epoch": 2.042,
"grad_norm": 1.2147035598754883,
"learning_rate": 2.798736930968315e-06,
"loss": 0.1596,
"step": 10210
},
{
"epoch": 2.044,
"grad_norm": 1.3540210723876953,
"learning_rate": 2.788295662942423e-06,
"loss": 0.1673,
"step": 10220
},
{
"epoch": 2.046,
"grad_norm": 1.2441480159759521,
"learning_rate": 2.777866372220789e-06,
"loss": 0.1877,
"step": 10230
},
{
"epoch": 2.048,
"grad_norm": 2.1177780628204346,
"learning_rate": 2.7674491152823825e-06,
"loss": 0.2515,
"step": 10240
},
{
"epoch": 2.05,
"grad_norm": 1.9897186756134033,
"learning_rate": 2.7570439485410116e-06,
"loss": 0.1739,
"step": 10250
},
{
"epoch": 2.052,
"grad_norm": 2.053022861480713,
"learning_rate": 2.7466509283450026e-06,
"loss": 0.1839,
"step": 10260
},
{
"epoch": 2.054,
"grad_norm": 1.2909846305847168,
"learning_rate": 2.736270110976912e-06,
"loss": 0.1838,
"step": 10270
},
{
"epoch": 2.056,
"grad_norm": 1.4775480031967163,
"learning_rate": 2.7259015526532074e-06,
"loss": 0.1464,
"step": 10280
},
{
"epoch": 2.058,
"grad_norm": 2.2092342376708984,
"learning_rate": 2.7155453095239682e-06,
"loss": 0.1622,
"step": 10290
},
{
"epoch": 2.06,
"grad_norm": 1.2695140838623047,
"learning_rate": 2.705201437672585e-06,
"loss": 0.1665,
"step": 10300
},
{
"epoch": 2.062,
"grad_norm": 1.4544492959976196,
"learning_rate": 2.6948699931154533e-06,
"loss": 0.158,
"step": 10310
},
{
"epoch": 2.064,
"grad_norm": 1.6260836124420166,
"learning_rate": 2.684551031801662e-06,
"loss": 0.1697,
"step": 10320
},
{
"epoch": 2.066,
"grad_norm": 1.1283328533172607,
"learning_rate": 2.6742446096127086e-06,
"loss": 0.1621,
"step": 10330
},
{
"epoch": 2.068,
"grad_norm": 1.4716851711273193,
"learning_rate": 2.66395078236218e-06,
"loss": 0.1704,
"step": 10340
},
{
"epoch": 2.07,
"grad_norm": 1.6541157960891724,
"learning_rate": 2.6536696057954553e-06,
"loss": 0.2194,
"step": 10350
},
{
"epoch": 2.072,
"grad_norm": 1.5134286880493164,
"learning_rate": 2.6434011355894074e-06,
"loss": 0.185,
"step": 10360
},
{
"epoch": 2.074,
"grad_norm": 1.936950922012329,
"learning_rate": 2.633145427352102e-06,
"loss": 0.1546,
"step": 10370
},
{
"epoch": 2.076,
"grad_norm": 2.237762212753296,
"learning_rate": 2.6229025366224835e-06,
"loss": 0.1923,
"step": 10380
},
{
"epoch": 2.078,
"grad_norm": 2.4793312549591064,
"learning_rate": 2.612672518870093e-06,
"loss": 0.1842,
"step": 10390
},
{
"epoch": 2.08,
"grad_norm": 1.5138216018676758,
"learning_rate": 2.602455429494758e-06,
"loss": 0.1437,
"step": 10400
},
{
"epoch": 2.082,
"grad_norm": 1.5739482641220093,
"learning_rate": 2.5922513238262915e-06,
"loss": 0.1765,
"step": 10410
},
{
"epoch": 2.084,
"grad_norm": 1.1942251920700073,
"learning_rate": 2.582060257124195e-06,
"loss": 0.1793,
"step": 10420
},
{
"epoch": 2.086,
"grad_norm": 1.485691785812378,
"learning_rate": 2.5718822845773516e-06,
"loss": 0.1927,
"step": 10430
},
{
"epoch": 2.088,
"grad_norm": 1.794269323348999,
"learning_rate": 2.5617174613037503e-06,
"loss": 0.1779,
"step": 10440
},
{
"epoch": 2.09,
"grad_norm": 1.6613082885742188,
"learning_rate": 2.5515658423501573e-06,
"loss": 0.1955,
"step": 10450
},
{
"epoch": 2.092,
"grad_norm": 1.5060651302337646,
"learning_rate": 2.541427482691832e-06,
"loss": 0.1463,
"step": 10460
},
{
"epoch": 2.094,
"grad_norm": 1.6522201299667358,
"learning_rate": 2.5313024372322413e-06,
"loss": 0.2071,
"step": 10470
},
{
"epoch": 2.096,
"grad_norm": 2.138770818710327,
"learning_rate": 2.5211907608027366e-06,
"loss": 0.182,
"step": 10480
},
{
"epoch": 2.098,
"grad_norm": 2.109570264816284,
"learning_rate": 2.5110925081622796e-06,
"loss": 0.1743,
"step": 10490
},
{
"epoch": 2.1,
"grad_norm": 1.1957253217697144,
"learning_rate": 2.5010077339971283e-06,
"loss": 0.1715,
"step": 10500
},
{
"epoch": 2.102,
"grad_norm": 1.4688587188720703,
"learning_rate": 2.4909364929205575e-06,
"loss": 0.1838,
"step": 10510
},
{
"epoch": 2.104,
"grad_norm": 1.1424540281295776,
"learning_rate": 2.480878839472552e-06,
"loss": 0.1815,
"step": 10520
},
{
"epoch": 2.106,
"grad_norm": 1.2021138668060303,
"learning_rate": 2.470834828119509e-06,
"loss": 0.1657,
"step": 10530
},
{
"epoch": 2.108,
"grad_norm": 1.8013497591018677,
"learning_rate": 2.4608045132539536e-06,
"loss": 0.2006,
"step": 10540
},
{
"epoch": 2.11,
"grad_norm": 2.014941930770874,
"learning_rate": 2.4507879491942388e-06,
"loss": 0.1859,
"step": 10550
},
{
"epoch": 2.112,
"grad_norm": 1.798327922821045,
"learning_rate": 2.4407851901842465e-06,
"loss": 0.1798,
"step": 10560
},
{
"epoch": 2.114,
"grad_norm": 1.7324855327606201,
"learning_rate": 2.4307962903931025e-06,
"loss": 0.1823,
"step": 10570
},
{
"epoch": 2.116,
"grad_norm": 1.536121129989624,
"learning_rate": 2.4208213039148803e-06,
"loss": 0.1776,
"step": 10580
},
{
"epoch": 2.118,
"grad_norm": 2.203296422958374,
"learning_rate": 2.4108602847683012e-06,
"loss": 0.2172,
"step": 10590
},
{
"epoch": 2.12,
"grad_norm": 1.6374783515930176,
"learning_rate": 2.4009132868964525e-06,
"loss": 0.1755,
"step": 10600
},
{
"epoch": 2.122,
"grad_norm": 2.1419484615325928,
"learning_rate": 2.3909803641664907e-06,
"loss": 0.1998,
"step": 10610
},
{
"epoch": 2.124,
"grad_norm": 1.4426078796386719,
"learning_rate": 2.3810615703693446e-06,
"loss": 0.207,
"step": 10620
},
{
"epoch": 2.126,
"grad_norm": 1.7481943368911743,
"learning_rate": 2.3711569592194363e-06,
"loss": 0.1504,
"step": 10630
},
{
"epoch": 2.128,
"grad_norm": 1.914016604423523,
"learning_rate": 2.3612665843543737e-06,
"loss": 0.1743,
"step": 10640
},
{
"epoch": 2.13,
"grad_norm": 1.829076886177063,
"learning_rate": 2.3513904993346775e-06,
"loss": 0.2314,
"step": 10650
},
{
"epoch": 2.132,
"grad_norm": 1.4599560499191284,
"learning_rate": 2.3415287576434807e-06,
"loss": 0.1967,
"step": 10660
},
{
"epoch": 2.134,
"grad_norm": 1.5102792978286743,
"learning_rate": 2.3316814126862377e-06,
"loss": 0.1772,
"step": 10670
},
{
"epoch": 2.136,
"grad_norm": 1.5047998428344727,
"learning_rate": 2.321848517790442e-06,
"loss": 0.2231,
"step": 10680
},
{
"epoch": 2.138,
"grad_norm": 1.4379587173461914,
"learning_rate": 2.312030126205335e-06,
"loss": 0.1765,
"step": 10690
},
{
"epoch": 2.14,
"grad_norm": 2.076664447784424,
"learning_rate": 2.302226291101609e-06,
"loss": 0.1563,
"step": 10700
},
{
"epoch": 2.142,
"grad_norm": 1.7419641017913818,
"learning_rate": 2.2924370655711407e-06,
"loss": 0.1456,
"step": 10710
},
{
"epoch": 2.144,
"grad_norm": 1.5634517669677734,
"learning_rate": 2.282662502626678e-06,
"loss": 0.1508,
"step": 10720
},
{
"epoch": 2.146,
"grad_norm": 2.313948631286621,
"learning_rate": 2.2729026552015653e-06,
"loss": 0.1975,
"step": 10730
},
{
"epoch": 2.148,
"grad_norm": 1.3335062265396118,
"learning_rate": 2.263157576149463e-06,
"loss": 0.1611,
"step": 10740
},
{
"epoch": 2.15,
"grad_norm": 1.4061542749404907,
"learning_rate": 2.2534273182440515e-06,
"loss": 0.1353,
"step": 10750
},
{
"epoch": 2.152,
"grad_norm": 2.1555397510528564,
"learning_rate": 2.2437119341787444e-06,
"loss": 0.1941,
"step": 10760
},
{
"epoch": 2.154,
"grad_norm": 1.6966259479522705,
"learning_rate": 2.2340114765664137e-06,
"loss": 0.1828,
"step": 10770
},
{
"epoch": 2.156,
"grad_norm": 1.6515167951583862,
"learning_rate": 2.224325997939095e-06,
"loss": 0.1654,
"step": 10780
},
{
"epoch": 2.158,
"grad_norm": 1.68687105178833,
"learning_rate": 2.214655550747709e-06,
"loss": 0.1473,
"step": 10790
},
{
"epoch": 2.16,
"grad_norm": 1.4554318189620972,
"learning_rate": 2.2050001873617716e-06,
"loss": 0.1534,
"step": 10800
},
{
"epoch": 2.162,
"grad_norm": 1.4342589378356934,
"learning_rate": 2.19535996006911e-06,
"loss": 0.1541,
"step": 10810
},
{
"epoch": 2.164,
"grad_norm": 1.8129370212554932,
"learning_rate": 2.1857349210755956e-06,
"loss": 0.1739,
"step": 10820
},
{
"epoch": 2.166,
"grad_norm": 1.8690457344055176,
"learning_rate": 2.1761251225048385e-06,
"loss": 0.1665,
"step": 10830
},
{
"epoch": 2.168,
"grad_norm": 1.6199536323547363,
"learning_rate": 2.1665306163979132e-06,
"loss": 0.141,
"step": 10840
},
{
"epoch": 2.17,
"grad_norm": 1.7210586071014404,
"learning_rate": 2.156951454713093e-06,
"loss": 0.1729,
"step": 10850
},
{
"epoch": 2.172,
"grad_norm": 2.0658435821533203,
"learning_rate": 2.147387689325539e-06,
"loss": 0.2072,
"step": 10860
},
{
"epoch": 2.174,
"grad_norm": 2.763272762298584,
"learning_rate": 2.137839372027047e-06,
"loss": 0.2201,
"step": 10870
},
{
"epoch": 2.176,
"grad_norm": 1.9413820505142212,
"learning_rate": 2.1283065545257443e-06,
"loss": 0.1413,
"step": 10880
},
{
"epoch": 2.178,
"grad_norm": 1.5634864568710327,
"learning_rate": 2.118789288445829e-06,
"loss": 0.1497,
"step": 10890
},
{
"epoch": 2.18,
"grad_norm": 2.5157878398895264,
"learning_rate": 2.1092876253272793e-06,
"loss": 0.1919,
"step": 10900
},
{
"epoch": 2.182,
"grad_norm": 1.4105507135391235,
"learning_rate": 2.099801616625573e-06,
"loss": 0.1154,
"step": 10910
},
{
"epoch": 2.184,
"grad_norm": 1.6929118633270264,
"learning_rate": 2.090331313711417e-06,
"loss": 0.1904,
"step": 10920
},
{
"epoch": 2.186,
"grad_norm": 2.1146175861358643,
"learning_rate": 2.080876767870466e-06,
"loss": 0.191,
"step": 10930
},
{
"epoch": 2.188,
"grad_norm": 1.6462335586547852,
"learning_rate": 2.0714380303030373e-06,
"loss": 0.2233,
"step": 10940
},
{
"epoch": 2.19,
"grad_norm": 2.0763607025146484,
"learning_rate": 2.0620151521238453e-06,
"loss": 0.1847,
"step": 10950
},
{
"epoch": 2.192,
"grad_norm": 1.8710356950759888,
"learning_rate": 2.0526081843617183e-06,
"loss": 0.1867,
"step": 10960
},
{
"epoch": 2.194,
"grad_norm": 1.8661998510360718,
"learning_rate": 2.04321717795932e-06,
"loss": 0.1691,
"step": 10970
},
{
"epoch": 2.196,
"grad_norm": 1.706262230873108,
"learning_rate": 2.03384218377288e-06,
"loss": 0.1848,
"step": 10980
},
{
"epoch": 2.198,
"grad_norm": 1.395787239074707,
"learning_rate": 2.0244832525719155e-06,
"loss": 0.1661,
"step": 10990
},
{
"epoch": 2.2,
"grad_norm": 1.469756841659546,
"learning_rate": 2.015140435038951e-06,
"loss": 0.1701,
"step": 11000
},
{
"epoch": 2.202,
"grad_norm": 2.2943742275238037,
"learning_rate": 2.005813781769253e-06,
"loss": 0.159,
"step": 11010
},
{
"epoch": 2.204,
"grad_norm": 2.527089834213257,
"learning_rate": 1.996503343270554e-06,
"loss": 0.2051,
"step": 11020
},
{
"epoch": 2.206,
"grad_norm": 2.241828680038452,
"learning_rate": 1.987209169962769e-06,
"loss": 0.1856,
"step": 11030
},
{
"epoch": 2.208,
"grad_norm": 2.044684410095215,
"learning_rate": 1.9779313121777382e-06,
"loss": 0.1704,
"step": 11040
},
{
"epoch": 2.21,
"grad_norm": 1.545622706413269,
"learning_rate": 1.9686698201589395e-06,
"loss": 0.1666,
"step": 11050
},
{
"epoch": 2.212,
"grad_norm": 2.001255750656128,
"learning_rate": 1.9594247440612293e-06,
"loss": 0.1813,
"step": 11060
},
{
"epoch": 2.214,
"grad_norm": 2.0734426975250244,
"learning_rate": 1.9501961339505626e-06,
"loss": 0.2223,
"step": 11070
},
{
"epoch": 2.216,
"grad_norm": 1.9134544134140015,
"learning_rate": 1.94098403980372e-06,
"loss": 0.1919,
"step": 11080
},
{
"epoch": 2.218,
"grad_norm": 1.5950722694396973,
"learning_rate": 1.9317885115080514e-06,
"loss": 0.234,
"step": 11090
},
{
"epoch": 2.22,
"grad_norm": 1.8253746032714844,
"learning_rate": 1.922609598861187e-06,
"loss": 0.1471,
"step": 11100
},
{
"epoch": 2.222,
"grad_norm": 1.5433813333511353,
"learning_rate": 1.913447351570776e-06,
"loss": 0.1804,
"step": 11110
},
{
"epoch": 2.224,
"grad_norm": 2.1074371337890625,
"learning_rate": 1.9043018192542228e-06,
"loss": 0.2,
"step": 11120
},
{
"epoch": 2.226,
"grad_norm": 1.8115590810775757,
"learning_rate": 1.8951730514384103e-06,
"loss": 0.2082,
"step": 11130
},
{
"epoch": 2.228,
"grad_norm": 0.8639247417449951,
"learning_rate": 1.8860610975594384e-06,
"loss": 0.1944,
"step": 11140
},
{
"epoch": 2.23,
"grad_norm": 1.839328408241272,
"learning_rate": 1.8769660069623448e-06,
"loss": 0.212,
"step": 11150
},
{
"epoch": 2.232,
"grad_norm": 1.7485737800598145,
"learning_rate": 1.8678878289008511e-06,
"loss": 0.2095,
"step": 11160
},
{
"epoch": 2.234,
"grad_norm": 1.7731024026870728,
"learning_rate": 1.8588266125370929e-06,
"loss": 0.1836,
"step": 11170
},
{
"epoch": 2.2359999999999998,
"grad_norm": 1.8955568075180054,
"learning_rate": 1.8497824069413445e-06,
"loss": 0.2007,
"step": 11180
},
{
"epoch": 2.238,
"grad_norm": 1.6335498094558716,
"learning_rate": 1.84075526109176e-06,
"loss": 0.1826,
"step": 11190
},
{
"epoch": 2.24,
"grad_norm": 1.6790274381637573,
"learning_rate": 1.831745223874118e-06,
"loss": 0.1806,
"step": 11200
},
{
"epoch": 2.242,
"grad_norm": 1.6211411952972412,
"learning_rate": 1.8227523440815331e-06,
"loss": 0.1884,
"step": 11210
},
{
"epoch": 2.2439999999999998,
"grad_norm": 1.6547083854675293,
"learning_rate": 1.8137766704142141e-06,
"loss": 0.1747,
"step": 11220
},
{
"epoch": 2.246,
"grad_norm": 1.8069316148757935,
"learning_rate": 1.8048182514791901e-06,
"loss": 0.1676,
"step": 11230
},
{
"epoch": 2.248,
"grad_norm": 2.3242225646972656,
"learning_rate": 1.7958771357900446e-06,
"loss": 0.1868,
"step": 11240
},
{
"epoch": 2.25,
"grad_norm": 2.2176382541656494,
"learning_rate": 1.7869533717666626e-06,
"loss": 0.1645,
"step": 11250
},
{
"epoch": 2.252,
"grad_norm": 1.437696933746338,
"learning_rate": 1.7780470077349566e-06,
"loss": 0.1623,
"step": 11260
},
{
"epoch": 2.254,
"grad_norm": 2.0108892917633057,
"learning_rate": 1.769158091926615e-06,
"loss": 0.2113,
"step": 11270
},
{
"epoch": 2.2560000000000002,
"grad_norm": 1.2373582124710083,
"learning_rate": 1.760286672478837e-06,
"loss": 0.1587,
"step": 11280
},
{
"epoch": 2.258,
"grad_norm": 1.3458646535873413,
"learning_rate": 1.751432797434068e-06,
"loss": 0.2014,
"step": 11290
},
{
"epoch": 2.26,
"grad_norm": 1.4614745378494263,
"learning_rate": 1.7425965147397462e-06,
"loss": 0.1785,
"step": 11300
},
{
"epoch": 2.262,
"grad_norm": 2.939737558364868,
"learning_rate": 1.7337778722480413e-06,
"loss": 0.2342,
"step": 11310
},
{
"epoch": 2.2640000000000002,
"grad_norm": 1.6198670864105225,
"learning_rate": 1.7249769177155879e-06,
"loss": 0.1486,
"step": 11320
},
{
"epoch": 2.266,
"grad_norm": 2.2954392433166504,
"learning_rate": 1.7161936988032386e-06,
"loss": 0.196,
"step": 11330
},
{
"epoch": 2.268,
"grad_norm": 1.6478867530822754,
"learning_rate": 1.7074282630757998e-06,
"loss": 0.1795,
"step": 11340
},
{
"epoch": 2.27,
"grad_norm": 1.3108015060424805,
"learning_rate": 1.6986806580017695e-06,
"loss": 0.1213,
"step": 11350
},
{
"epoch": 2.2720000000000002,
"grad_norm": 1.4279754161834717,
"learning_rate": 1.689950930953091e-06,
"loss": 0.1663,
"step": 11360
},
{
"epoch": 2.274,
"grad_norm": 1.3529280424118042,
"learning_rate": 1.6812391292048897e-06,
"loss": 0.1478,
"step": 11370
},
{
"epoch": 2.276,
"grad_norm": 2.1252269744873047,
"learning_rate": 1.6725452999352137e-06,
"loss": 0.1675,
"step": 11380
},
{
"epoch": 2.278,
"grad_norm": 2.0060176849365234,
"learning_rate": 1.6638694902247866e-06,
"loss": 0.2113,
"step": 11390
},
{
"epoch": 2.2800000000000002,
"grad_norm": 2.1761856079101562,
"learning_rate": 1.655211747056749e-06,
"loss": 0.1726,
"step": 11400
},
{
"epoch": 2.282,
"grad_norm": 1.4057432413101196,
"learning_rate": 1.6465721173164e-06,
"loss": 0.1713,
"step": 11410
},
{
"epoch": 2.284,
"grad_norm": 1.271407961845398,
"learning_rate": 1.6379506477909518e-06,
"loss": 0.173,
"step": 11420
},
{
"epoch": 2.286,
"grad_norm": 1.5162497758865356,
"learning_rate": 1.629347385169263e-06,
"loss": 0.1821,
"step": 11430
},
{
"epoch": 2.288,
"grad_norm": 1.081526517868042,
"learning_rate": 1.6207623760416074e-06,
"loss": 0.1428,
"step": 11440
},
{
"epoch": 2.29,
"grad_norm": 1.603796362876892,
"learning_rate": 1.6121956668993977e-06,
"loss": 0.1669,
"step": 11450
},
{
"epoch": 2.292,
"grad_norm": 1.392375111579895,
"learning_rate": 1.6036473041349438e-06,
"loss": 0.1698,
"step": 11460
},
{
"epoch": 2.294,
"grad_norm": 1.9549520015716553,
"learning_rate": 1.5951173340412134e-06,
"loss": 0.2162,
"step": 11470
},
{
"epoch": 2.296,
"grad_norm": 1.8508301973342896,
"learning_rate": 1.58660580281156e-06,
"loss": 0.1534,
"step": 11480
},
{
"epoch": 2.298,
"grad_norm": 1.5918387174606323,
"learning_rate": 1.5781127565394838e-06,
"loss": 0.161,
"step": 11490
},
{
"epoch": 2.3,
"grad_norm": 2.0230064392089844,
"learning_rate": 1.5696382412183853e-06,
"loss": 0.1709,
"step": 11500
},
{
"epoch": 2.302,
"grad_norm": 2.3765511512756348,
"learning_rate": 1.5611823027413109e-06,
"loss": 0.1861,
"step": 11510
},
{
"epoch": 2.304,
"grad_norm": 1.9833528995513916,
"learning_rate": 1.5527449869007055e-06,
"loss": 0.2067,
"step": 11520
},
{
"epoch": 2.306,
"grad_norm": 1.4727070331573486,
"learning_rate": 1.5443263393881619e-06,
"loss": 0.1758,
"step": 11530
},
{
"epoch": 2.308,
"grad_norm": 1.8662854433059692,
"learning_rate": 1.535926405794179e-06,
"loss": 0.1892,
"step": 11540
},
{
"epoch": 2.31,
"grad_norm": 1.810899257659912,
"learning_rate": 1.5275452316079143e-06,
"loss": 0.1321,
"step": 11550
},
{
"epoch": 2.312,
"grad_norm": 2.3489248752593994,
"learning_rate": 1.519182862216929e-06,
"loss": 0.1976,
"step": 11560
},
{
"epoch": 2.314,
"grad_norm": 1.5641025304794312,
"learning_rate": 1.5108393429069501e-06,
"loss": 0.1718,
"step": 11570
},
{
"epoch": 2.316,
"grad_norm": 2.0649514198303223,
"learning_rate": 1.5025147188616308e-06,
"loss": 0.1905,
"step": 11580
},
{
"epoch": 2.318,
"grad_norm": 1.309147596359253,
"learning_rate": 1.4942090351622884e-06,
"loss": 0.185,
"step": 11590
},
{
"epoch": 2.32,
"grad_norm": 1.3974589109420776,
"learning_rate": 1.4859223367876762e-06,
"loss": 0.1564,
"step": 11600
},
{
"epoch": 2.322,
"grad_norm": 1.8548953533172607,
"learning_rate": 1.477654668613735e-06,
"loss": 0.1659,
"step": 11610
},
{
"epoch": 2.324,
"grad_norm": 1.7054270505905151,
"learning_rate": 1.469406075413342e-06,
"loss": 0.2003,
"step": 11620
},
{
"epoch": 2.326,
"grad_norm": 1.4067456722259521,
"learning_rate": 1.4611766018560835e-06,
"loss": 0.1569,
"step": 11630
},
{
"epoch": 2.328,
"grad_norm": 2.1007652282714844,
"learning_rate": 1.4529662925080023e-06,
"loss": 0.1882,
"step": 11640
},
{
"epoch": 2.33,
"grad_norm": 1.6421812772750854,
"learning_rate": 1.4447751918313552e-06,
"loss": 0.2184,
"step": 11650
},
{
"epoch": 2.332,
"grad_norm": 2.9512548446655273,
"learning_rate": 1.4366033441843823e-06,
"loss": 0.1994,
"step": 11660
},
{
"epoch": 2.334,
"grad_norm": 1.2109729051589966,
"learning_rate": 1.4284507938210545e-06,
"loss": 0.1702,
"step": 11670
},
{
"epoch": 2.336,
"grad_norm": 1.767602801322937,
"learning_rate": 1.420317584890844e-06,
"loss": 0.1508,
"step": 11680
},
{
"epoch": 2.338,
"grad_norm": 1.924599289894104,
"learning_rate": 1.4122037614384814e-06,
"loss": 0.2122,
"step": 11690
},
{
"epoch": 2.34,
"grad_norm": 1.8495938777923584,
"learning_rate": 1.404109367403712e-06,
"loss": 0.1816,
"step": 11700
},
{
"epoch": 2.342,
"grad_norm": 1.7605769634246826,
"learning_rate": 1.3960344466210669e-06,
"loss": 0.2041,
"step": 11710
},
{
"epoch": 2.344,
"grad_norm": 2.3613359928131104,
"learning_rate": 1.3879790428196226e-06,
"loss": 0.1848,
"step": 11720
},
{
"epoch": 2.346,
"grad_norm": 1.8871616125106812,
"learning_rate": 1.3799431996227569e-06,
"loss": 0.1771,
"step": 11730
},
{
"epoch": 2.348,
"grad_norm": 1.523779034614563,
"learning_rate": 1.3719269605479241e-06,
"loss": 0.1653,
"step": 11740
},
{
"epoch": 2.35,
"grad_norm": 1.7912616729736328,
"learning_rate": 1.363930369006415e-06,
"loss": 0.1834,
"step": 11750
},
{
"epoch": 2.352,
"grad_norm": 2.330554485321045,
"learning_rate": 1.3559534683031133e-06,
"loss": 0.1694,
"step": 11760
},
{
"epoch": 2.354,
"grad_norm": 1.3133292198181152,
"learning_rate": 1.3479963016362768e-06,
"loss": 0.1925,
"step": 11770
},
{
"epoch": 2.356,
"grad_norm": 1.9684919118881226,
"learning_rate": 1.3400589120972922e-06,
"loss": 0.1448,
"step": 11780
},
{
"epoch": 2.358,
"grad_norm": 1.2997586727142334,
"learning_rate": 1.3321413426704426e-06,
"loss": 0.2113,
"step": 11790
},
{
"epoch": 2.36,
"grad_norm": 2.1797378063201904,
"learning_rate": 1.3242436362326804e-06,
"loss": 0.1585,
"step": 11800
},
{
"epoch": 2.362,
"grad_norm": 1.4730124473571777,
"learning_rate": 1.3163658355533866e-06,
"loss": 0.1928,
"step": 11810
},
{
"epoch": 2.364,
"grad_norm": 1.4231460094451904,
"learning_rate": 1.3085079832941528e-06,
"loss": 0.2012,
"step": 11820
},
{
"epoch": 2.366,
"grad_norm": 1.8164182901382446,
"learning_rate": 1.3006701220085338e-06,
"loss": 0.1639,
"step": 11830
},
{
"epoch": 2.368,
"grad_norm": 2.293713092803955,
"learning_rate": 1.2928522941418241e-06,
"loss": 0.168,
"step": 11840
},
{
"epoch": 2.37,
"grad_norm": 1.6054863929748535,
"learning_rate": 1.2850545420308386e-06,
"loss": 0.1507,
"step": 11850
},
{
"epoch": 2.372,
"grad_norm": 2.180938959121704,
"learning_rate": 1.2772769079036639e-06,
"loss": 0.1563,
"step": 11860
},
{
"epoch": 2.374,
"grad_norm": 1.553425908088684,
"learning_rate": 1.2695194338794414e-06,
"loss": 0.1574,
"step": 11870
},
{
"epoch": 2.376,
"grad_norm": 1.3529037237167358,
"learning_rate": 1.2617821619681397e-06,
"loss": 0.1941,
"step": 11880
},
{
"epoch": 2.378,
"grad_norm": 1.9350522756576538,
"learning_rate": 1.2540651340703231e-06,
"loss": 0.1551,
"step": 11890
},
{
"epoch": 2.38,
"grad_norm": 2.1500189304351807,
"learning_rate": 1.246368391976927e-06,
"loss": 0.166,
"step": 11900
},
{
"epoch": 2.382,
"grad_norm": 1.851035237312317,
"learning_rate": 1.2386919773690276e-06,
"loss": 0.154,
"step": 11910
},
{
"epoch": 2.384,
"grad_norm": 1.200472116470337,
"learning_rate": 1.2310359318176229e-06,
"loss": 0.1441,
"step": 11920
},
{
"epoch": 2.386,
"grad_norm": 1.9997992515563965,
"learning_rate": 1.2234002967834036e-06,
"loss": 0.1886,
"step": 11930
},
{
"epoch": 2.388,
"grad_norm": 2.4099040031433105,
"learning_rate": 1.2157851136165243e-06,
"loss": 0.1631,
"step": 11940
},
{
"epoch": 2.39,
"grad_norm": 1.5515005588531494,
"learning_rate": 1.2081904235563908e-06,
"loss": 0.1468,
"step": 11950
},
{
"epoch": 2.392,
"grad_norm": 1.075179100036621,
"learning_rate": 1.2006162677314265e-06,
"loss": 0.1464,
"step": 11960
},
{
"epoch": 2.394,
"grad_norm": 1.8599796295166016,
"learning_rate": 1.1930626871588525e-06,
"loss": 0.1865,
"step": 11970
},
{
"epoch": 2.396,
"grad_norm": 1.7627993822097778,
"learning_rate": 1.185529722744469e-06,
"loss": 0.165,
"step": 11980
},
{
"epoch": 2.398,
"grad_norm": 1.7737939357757568,
"learning_rate": 1.1780174152824297e-06,
"loss": 0.1545,
"step": 11990
},
{
"epoch": 2.4,
"grad_norm": 0.9698889851570129,
"learning_rate": 1.1705258054550212e-06,
"loss": 0.1575,
"step": 12000
},
{
"epoch": 2.402,
"grad_norm": 1.0956158638000488,
"learning_rate": 1.1630549338324454e-06,
"loss": 0.1703,
"step": 12010
},
{
"epoch": 2.404,
"grad_norm": 1.4204548597335815,
"learning_rate": 1.155604840872599e-06,
"loss": 0.166,
"step": 12020
},
{
"epoch": 2.406,
"grad_norm": 1.6942260265350342,
"learning_rate": 1.1481755669208495e-06,
"loss": 0.1596,
"step": 12030
},
{
"epoch": 2.408,
"grad_norm": 1.5839786529541016,
"learning_rate": 1.1407671522098262e-06,
"loss": 0.1931,
"step": 12040
},
{
"epoch": 2.41,
"grad_norm": 2.013857841491699,
"learning_rate": 1.1333796368591915e-06,
"loss": 0.1791,
"step": 12050
},
{
"epoch": 2.412,
"grad_norm": 1.7067184448242188,
"learning_rate": 1.126013060875432e-06,
"loss": 0.1849,
"step": 12060
},
{
"epoch": 2.414,
"grad_norm": 1.8866993188858032,
"learning_rate": 1.1186674641516415e-06,
"loss": 0.1922,
"step": 12070
},
{
"epoch": 2.416,
"grad_norm": 2.4251067638397217,
"learning_rate": 1.1113428864672954e-06,
"loss": 0.1807,
"step": 12080
},
{
"epoch": 2.418,
"grad_norm": 1.515995740890503,
"learning_rate": 1.1040393674880478e-06,
"loss": 0.1896,
"step": 12090
},
{
"epoch": 2.42,
"grad_norm": 1.7295435667037964,
"learning_rate": 1.0967569467655104e-06,
"loss": 0.1473,
"step": 12100
},
{
"epoch": 2.422,
"grad_norm": 1.7556982040405273,
"learning_rate": 1.0894956637370363e-06,
"loss": 0.1475,
"step": 12110
},
{
"epoch": 2.424,
"grad_norm": 1.9276478290557861,
"learning_rate": 1.082255557725511e-06,
"loss": 0.1539,
"step": 12120
},
{
"epoch": 2.426,
"grad_norm": 1.5865824222564697,
"learning_rate": 1.0750366679391393e-06,
"loss": 0.1449,
"step": 12130
},
{
"epoch": 2.428,
"grad_norm": 1.8489198684692383,
"learning_rate": 1.0678390334712275e-06,
"loss": 0.1547,
"step": 12140
},
{
"epoch": 2.43,
"grad_norm": 1.6484555006027222,
"learning_rate": 1.0606626932999775e-06,
"loss": 0.1648,
"step": 12150
},
{
"epoch": 2.432,
"grad_norm": 1.7841017246246338,
"learning_rate": 1.053507686288276e-06,
"loss": 0.1557,
"step": 12160
},
{
"epoch": 2.434,
"grad_norm": 1.6266376972198486,
"learning_rate": 1.0463740511834759e-06,
"loss": 0.1449,
"step": 12170
},
{
"epoch": 2.436,
"grad_norm": 1.884961724281311,
"learning_rate": 1.0392618266171983e-06,
"loss": 0.1606,
"step": 12180
},
{
"epoch": 2.438,
"grad_norm": 2.732299566268921,
"learning_rate": 1.0321710511051108e-06,
"loss": 0.1637,
"step": 12190
},
{
"epoch": 2.44,
"grad_norm": 1.4075685739517212,
"learning_rate": 1.0251017630467347e-06,
"loss": 0.1798,
"step": 12200
},
{
"epoch": 2.442,
"grad_norm": 1.3892203569412231,
"learning_rate": 1.01805400072522e-06,
"loss": 0.1867,
"step": 12210
},
{
"epoch": 2.444,
"grad_norm": 1.6456845998764038,
"learning_rate": 1.0110278023071445e-06,
"loss": 0.1632,
"step": 12220
},
{
"epoch": 2.446,
"grad_norm": 1.5476148128509521,
"learning_rate": 1.0040232058423182e-06,
"loss": 0.1513,
"step": 12230
},
{
"epoch": 2.448,
"grad_norm": 1.5470598936080933,
"learning_rate": 9.970402492635583e-07,
"loss": 0.1741,
"step": 12240
},
{
"epoch": 2.45,
"grad_norm": 1.6543699502944946,
"learning_rate": 9.900789703864933e-07,
"loss": 0.1374,
"step": 12250
},
{
"epoch": 2.452,
"grad_norm": 1.6343821287155151,
"learning_rate": 9.831394069093663e-07,
"loss": 0.1526,
"step": 12260
},
{
"epoch": 2.454,
"grad_norm": 1.9217232465744019,
"learning_rate": 9.762215964128124e-07,
"loss": 0.1716,
"step": 12270
},
{
"epoch": 2.456,
"grad_norm": 2.077751874923706,
"learning_rate": 9.69325576359672e-07,
"loss": 0.1599,
"step": 12280
},
{
"epoch": 2.458,
"grad_norm": 1.3649940490722656,
"learning_rate": 9.624513840947764e-07,
"loss": 0.1499,
"step": 12290
},
{
"epoch": 2.46,
"grad_norm": 1.173978328704834,
"learning_rate": 9.555990568447538e-07,
"loss": 0.1685,
"step": 12300
},
{
"epoch": 2.462,
"grad_norm": 1.5035467147827148,
"learning_rate": 9.487686317178241e-07,
"loss": 0.1905,
"step": 12310
},
{
"epoch": 2.464,
"grad_norm": 1.3181893825531006,
"learning_rate": 9.419601457035943e-07,
"loss": 0.1994,
"step": 12320
},
{
"epoch": 2.466,
"grad_norm": 1.2825194597244263,
"learning_rate": 9.351736356728657e-07,
"loss": 0.161,
"step": 12330
},
{
"epoch": 2.468,
"grad_norm": 2.407343626022339,
"learning_rate": 9.284091383774313e-07,
"loss": 0.1929,
"step": 12340
},
{
"epoch": 2.4699999999999998,
"grad_norm": 1.4245587587356567,
"learning_rate": 9.2166669044987e-07,
"loss": 0.154,
"step": 12350
},
{
"epoch": 2.472,
"grad_norm": 2.4359230995178223,
"learning_rate": 9.149463284033605e-07,
"loss": 0.1854,
"step": 12360
},
{
"epoch": 2.474,
"grad_norm": 1.9102237224578857,
"learning_rate": 9.08248088631476e-07,
"loss": 0.2134,
"step": 12370
},
{
"epoch": 2.476,
"grad_norm": 1.268130898475647,
"learning_rate": 9.015720074079837e-07,
"loss": 0.2093,
"step": 12380
},
{
"epoch": 2.4779999999999998,
"grad_norm": 1.4418933391571045,
"learning_rate": 8.949181208866581e-07,
"loss": 0.1573,
"step": 12390
},
{
"epoch": 2.48,
"grad_norm": 1.8565024137496948,
"learning_rate": 8.882864651010798e-07,
"loss": 0.1778,
"step": 12400
},
{
"epoch": 2.482,
"grad_norm": 1.8131656646728516,
"learning_rate": 8.816770759644361e-07,
"loss": 0.1609,
"step": 12410
},
{
"epoch": 2.484,
"grad_norm": 1.573459506034851,
"learning_rate": 8.750899892693376e-07,
"loss": 0.155,
"step": 12420
},
{
"epoch": 2.4859999999999998,
"grad_norm": 1.8090794086456299,
"learning_rate": 8.685252406876116e-07,
"loss": 0.2023,
"step": 12430
},
{
"epoch": 2.488,
"grad_norm": 1.2365047931671143,
"learning_rate": 8.61982865770119e-07,
"loss": 0.1746,
"step": 12440
},
{
"epoch": 2.49,
"grad_norm": 1.4933933019638062,
"learning_rate": 8.554628999465592e-07,
"loss": 0.1726,
"step": 12450
},
{
"epoch": 2.492,
"grad_norm": 1.6978332996368408,
"learning_rate": 8.489653785252711e-07,
"loss": 0.1853,
"step": 12460
},
{
"epoch": 2.4939999999999998,
"grad_norm": 1.4669007062911987,
"learning_rate": 8.424903366930531e-07,
"loss": 0.22,
"step": 12470
},
{
"epoch": 2.496,
"grad_norm": 1.660845398902893,
"learning_rate": 8.360378095149674e-07,
"loss": 0.1646,
"step": 12480
},
{
"epoch": 2.498,
"grad_norm": 1.6666799783706665,
"learning_rate": 8.296078319341444e-07,
"loss": 0.1526,
"step": 12490
},
{
"epoch": 2.5,
"grad_norm": 1.4847021102905273,
"learning_rate": 8.232004387716053e-07,
"loss": 0.1516,
"step": 12500
},
{
"epoch": 2.502,
"grad_norm": 1.512980580329895,
"learning_rate": 8.16815664726065e-07,
"loss": 0.1886,
"step": 12510
},
{
"epoch": 2.504,
"grad_norm": 1.0014886856079102,
"learning_rate": 8.104535443737438e-07,
"loss": 0.151,
"step": 12520
},
{
"epoch": 2.5060000000000002,
"grad_norm": 1.831697940826416,
"learning_rate": 8.041141121681867e-07,
"loss": 0.1517,
"step": 12530
},
{
"epoch": 2.508,
"grad_norm": 1.6552419662475586,
"learning_rate": 7.977974024400703e-07,
"loss": 0.1864,
"step": 12540
},
{
"epoch": 2.51,
"grad_norm": 2.3965301513671875,
"learning_rate": 7.91503449397022e-07,
"loss": 0.1646,
"step": 12550
},
{
"epoch": 2.512,
"grad_norm": 2.3798716068267822,
"learning_rate": 7.852322871234286e-07,
"loss": 0.1702,
"step": 12560
},
{
"epoch": 2.5140000000000002,
"grad_norm": 1.1681225299835205,
"learning_rate": 7.789839495802581e-07,
"loss": 0.1726,
"step": 12570
},
{
"epoch": 2.516,
"grad_norm": 1.7020608186721802,
"learning_rate": 7.727584706048735e-07,
"loss": 0.1862,
"step": 12580
},
{
"epoch": 2.518,
"grad_norm": 1.7006206512451172,
"learning_rate": 7.665558839108467e-07,
"loss": 0.1638,
"step": 12590
},
{
"epoch": 2.52,
"grad_norm": 1.530461072921753,
"learning_rate": 7.603762230877776e-07,
"loss": 0.1511,
"step": 12600
},
{
"epoch": 2.5220000000000002,
"grad_norm": 1.7536654472351074,
"learning_rate": 7.542195216011188e-07,
"loss": 0.1693,
"step": 12610
},
{
"epoch": 2.524,
"grad_norm": 2.3601443767547607,
"learning_rate": 7.480858127919821e-07,
"loss": 0.1946,
"step": 12620
},
{
"epoch": 2.526,
"grad_norm": 1.876556158065796,
"learning_rate": 7.419751298769667e-07,
"loss": 0.1628,
"step": 12630
},
{
"epoch": 2.528,
"grad_norm": 1.4596962928771973,
"learning_rate": 7.358875059479792e-07,
"loss": 0.1703,
"step": 12640
},
{
"epoch": 2.5300000000000002,
"grad_norm": 1.8861957788467407,
"learning_rate": 7.29822973972047e-07,
"loss": 0.1865,
"step": 12650
},
{
"epoch": 2.532,
"grad_norm": 1.5087050199508667,
"learning_rate": 7.237815667911502e-07,
"loss": 0.1665,
"step": 12660
},
{
"epoch": 2.534,
"grad_norm": 1.47129487991333,
"learning_rate": 7.177633171220339e-07,
"loss": 0.1728,
"step": 12670
},
{
"epoch": 2.536,
"grad_norm": 2.511249542236328,
"learning_rate": 7.117682575560386e-07,
"loss": 0.2098,
"step": 12680
},
{
"epoch": 2.5380000000000003,
"grad_norm": 1.4766125679016113,
"learning_rate": 7.057964205589218e-07,
"loss": 0.1697,
"step": 12690
},
{
"epoch": 2.54,
"grad_norm": 2.0010459423065186,
"learning_rate": 6.99847838470677e-07,
"loss": 0.2231,
"step": 12700
},
{
"epoch": 2.542,
"grad_norm": 2.9428231716156006,
"learning_rate": 6.939225435053648e-07,
"loss": 0.2108,
"step": 12710
},
{
"epoch": 2.544,
"grad_norm": 2.4707274436950684,
"learning_rate": 6.880205677509383e-07,
"loss": 0.207,
"step": 12720
},
{
"epoch": 2.5460000000000003,
"grad_norm": 1.7522751092910767,
"learning_rate": 6.821419431690629e-07,
"loss": 0.1804,
"step": 12730
},
{
"epoch": 2.548,
"grad_norm": 2.5205798149108887,
"learning_rate": 6.762867015949514e-07,
"loss": 0.2039,
"step": 12740
},
{
"epoch": 2.55,
"grad_norm": 2.847193479537964,
"learning_rate": 6.704548747371869e-07,
"loss": 0.1945,
"step": 12750
},
{
"epoch": 2.552,
"grad_norm": 1.2266204357147217,
"learning_rate": 6.646464941775499e-07,
"loss": 0.178,
"step": 12760
},
{
"epoch": 2.5540000000000003,
"grad_norm": 1.5307732820510864,
"learning_rate": 6.588615913708524e-07,
"loss": 0.1616,
"step": 12770
},
{
"epoch": 2.556,
"grad_norm": 1.3956282138824463,
"learning_rate": 6.531001976447637e-07,
"loss": 0.1128,
"step": 12780
},
{
"epoch": 2.558,
"grad_norm": 1.526992678642273,
"learning_rate": 6.47362344199639e-07,
"loss": 0.2029,
"step": 12790
},
{
"epoch": 2.56,
"grad_norm": 1.8423537015914917,
"learning_rate": 6.416480621083582e-07,
"loss": 0.1969,
"step": 12800
},
{
"epoch": 2.5620000000000003,
"grad_norm": 1.3227084875106812,
"learning_rate": 6.359573823161457e-07,
"loss": 0.2046,
"step": 12810
},
{
"epoch": 2.564,
"grad_norm": 2.2461440563201904,
"learning_rate": 6.302903356404161e-07,
"loss": 0.2142,
"step": 12820
},
{
"epoch": 2.566,
"grad_norm": 0.9911752939224243,
"learning_rate": 6.246469527705978e-07,
"loss": 0.1574,
"step": 12830
},
{
"epoch": 2.568,
"grad_norm": 1.4631303548812866,
"learning_rate": 6.190272642679674e-07,
"loss": 0.1516,
"step": 12840
},
{
"epoch": 2.57,
"grad_norm": 1.7750544548034668,
"learning_rate": 6.134313005654929e-07,
"loss": 0.1454,
"step": 12850
},
{
"epoch": 2.572,
"grad_norm": 2.154233455657959,
"learning_rate": 6.078590919676575e-07,
"loss": 0.1687,
"step": 12860
},
{
"epoch": 2.574,
"grad_norm": 1.879696011543274,
"learning_rate": 6.023106686502988e-07,
"loss": 0.1999,
"step": 12870
},
{
"epoch": 2.576,
"grad_norm": 1.5651618242263794,
"learning_rate": 5.967860606604553e-07,
"loss": 0.1647,
"step": 12880
},
{
"epoch": 2.578,
"grad_norm": 1.1055610179901123,
"learning_rate": 5.912852979161876e-07,
"loss": 0.1461,
"step": 12890
},
{
"epoch": 2.58,
"grad_norm": 1.6748799085617065,
"learning_rate": 5.858084102064271e-07,
"loss": 0.1708,
"step": 12900
},
{
"epoch": 2.582,
"grad_norm": 1.5193976163864136,
"learning_rate": 5.803554271908124e-07,
"loss": 0.172,
"step": 12910
},
{
"epoch": 2.584,
"grad_norm": 1.606832504272461,
"learning_rate": 5.749263783995279e-07,
"loss": 0.1459,
"step": 12920
},
{
"epoch": 2.586,
"grad_norm": 2.012147903442383,
"learning_rate": 5.69521293233145e-07,
"loss": 0.1558,
"step": 12930
},
{
"epoch": 2.588,
"grad_norm": 1.5900416374206543,
"learning_rate": 5.641402009624591e-07,
"loss": 0.2086,
"step": 12940
},
{
"epoch": 2.59,
"grad_norm": 2.680753469467163,
"learning_rate": 5.587831307283375e-07,
"loss": 0.1668,
"step": 12950
},
{
"epoch": 2.592,
"grad_norm": 1.6326476335525513,
"learning_rate": 5.534501115415575e-07,
"loss": 0.2084,
"step": 12960
},
{
"epoch": 2.594,
"grad_norm": 1.771647572517395,
"learning_rate": 5.48141172282648e-07,
"loss": 0.1903,
"step": 12970
},
{
"epoch": 2.596,
"grad_norm": 1.6876575946807861,
"learning_rate": 5.428563417017335e-07,
"loss": 0.1602,
"step": 12980
},
{
"epoch": 2.598,
"grad_norm": 1.6764187812805176,
"learning_rate": 5.375956484183875e-07,
"loss": 0.2182,
"step": 12990
},
{
"epoch": 2.6,
"grad_norm": 2.2477219104766846,
"learning_rate": 5.323591209214612e-07,
"loss": 0.1858,
"step": 13000
},
{
"epoch": 2.602,
"grad_norm": 1.2577763795852661,
"learning_rate": 5.271467875689429e-07,
"loss": 0.1794,
"step": 13010
},
{
"epoch": 2.604,
"grad_norm": 2.3537323474884033,
"learning_rate": 5.219586765877998e-07,
"loss": 0.1936,
"step": 13020
},
{
"epoch": 2.606,
"grad_norm": 2.4193389415740967,
"learning_rate": 5.167948160738206e-07,
"loss": 0.193,
"step": 13030
},
{
"epoch": 2.608,
"grad_norm": 1.678281545639038,
"learning_rate": 5.116552339914726e-07,
"loss": 0.1599,
"step": 13040
},
{
"epoch": 2.61,
"grad_norm": 2.4545416831970215,
"learning_rate": 5.065399581737412e-07,
"loss": 0.2522,
"step": 13050
},
{
"epoch": 2.612,
"grad_norm": 1.6324633359909058,
"learning_rate": 5.014490163219854e-07,
"loss": 0.1548,
"step": 13060
},
{
"epoch": 2.614,
"grad_norm": 1.7150189876556396,
"learning_rate": 4.963824360057868e-07,
"loss": 0.1563,
"step": 13070
},
{
"epoch": 2.616,
"grad_norm": 1.821446418762207,
"learning_rate": 4.913402446627946e-07,
"loss": 0.1644,
"step": 13080
},
{
"epoch": 2.618,
"grad_norm": 1.215470552444458,
"learning_rate": 4.863224695985858e-07,
"loss": 0.1826,
"step": 13090
},
{
"epoch": 2.62,
"grad_norm": 1.6227540969848633,
"learning_rate": 4.813291379865126e-07,
"loss": 0.2281,
"step": 13100
},
{
"epoch": 2.622,
"grad_norm": 2.389894723892212,
"learning_rate": 4.763602768675529e-07,
"loss": 0.1783,
"step": 13110
},
{
"epoch": 2.624,
"grad_norm": 2.125441312789917,
"learning_rate": 4.714159131501689e-07,
"loss": 0.2042,
"step": 13120
},
{
"epoch": 2.626,
"grad_norm": 1.7118359804153442,
"learning_rate": 4.664960736101598e-07,
"loss": 0.1868,
"step": 13130
},
{
"epoch": 2.628,
"grad_norm": 2.2027170658111572,
"learning_rate": 4.61600784890513e-07,
"loss": 0.1808,
"step": 13140
},
{
"epoch": 2.63,
"grad_norm": 1.8995624780654907,
"learning_rate": 4.567300735012653e-07,
"loss": 0.1864,
"step": 13150
},
{
"epoch": 2.632,
"grad_norm": 1.7229032516479492,
"learning_rate": 4.5188396581935856e-07,
"loss": 0.1915,
"step": 13160
},
{
"epoch": 2.634,
"grad_norm": 2.252855062484741,
"learning_rate": 4.470624880884905e-07,
"loss": 0.2062,
"step": 13170
},
{
"epoch": 2.636,
"grad_norm": 1.9822123050689697,
"learning_rate": 4.4226566641898173e-07,
"loss": 0.1733,
"step": 13180
},
{
"epoch": 2.638,
"grad_norm": 1.441658854484558,
"learning_rate": 4.37493526787629e-07,
"loss": 0.1911,
"step": 13190
},
{
"epoch": 2.64,
"grad_norm": 1.633154273033142,
"learning_rate": 4.327460950375623e-07,
"loss": 0.1701,
"step": 13200
},
{
"epoch": 2.642,
"grad_norm": 1.7371437549591064,
"learning_rate": 4.280233968781139e-07,
"loss": 0.1534,
"step": 13210
},
{
"epoch": 2.644,
"grad_norm": 2.687547445297241,
"learning_rate": 4.233254578846657e-07,
"loss": 0.1954,
"step": 13220
},
{
"epoch": 2.646,
"grad_norm": 1.7198002338409424,
"learning_rate": 4.186523034985279e-07,
"loss": 0.172,
"step": 13230
},
{
"epoch": 2.648,
"grad_norm": 1.4830220937728882,
"learning_rate": 4.140039590267836e-07,
"loss": 0.2197,
"step": 13240
},
{
"epoch": 2.65,
"grad_norm": 2.4155113697052,
"learning_rate": 4.0938044964216164e-07,
"loss": 0.1562,
"step": 13250
},
{
"epoch": 2.652,
"grad_norm": 2.1205966472625732,
"learning_rate": 4.0478180038290296e-07,
"loss": 0.1917,
"step": 13260
},
{
"epoch": 2.654,
"grad_norm": 1.5603994131088257,
"learning_rate": 4.002080361526156e-07,
"loss": 0.1492,
"step": 13270
},
{
"epoch": 2.656,
"grad_norm": 2.272676467895508,
"learning_rate": 3.9565918172014495e-07,
"loss": 0.1571,
"step": 13280
},
{
"epoch": 2.658,
"grad_norm": 2.0802717208862305,
"learning_rate": 3.9113526171944195e-07,
"loss": 0.1503,
"step": 13290
},
{
"epoch": 2.66,
"grad_norm": 1.6078263521194458,
"learning_rate": 3.866363006494256e-07,
"loss": 0.1724,
"step": 13300
},
{
"epoch": 2.662,
"grad_norm": 2.419299364089966,
"learning_rate": 3.821623228738536e-07,
"loss": 0.188,
"step": 13310
},
{
"epoch": 2.664,
"grad_norm": 2.157850742340088,
"learning_rate": 3.777133526211857e-07,
"loss": 0.1876,
"step": 13320
},
{
"epoch": 2.666,
"grad_norm": 2.2567625045776367,
"learning_rate": 3.732894139844578e-07,
"loss": 0.1865,
"step": 13330
},
{
"epoch": 2.668,
"grad_norm": 1.955741047859192,
"learning_rate": 3.688905309211488e-07,
"loss": 0.1781,
"step": 13340
},
{
"epoch": 2.67,
"grad_norm": 1.4101120233535767,
"learning_rate": 3.6451672725304974e-07,
"loss": 0.1871,
"step": 13350
},
{
"epoch": 2.672,
"grad_norm": 2.0826516151428223,
"learning_rate": 3.601680266661367e-07,
"loss": 0.1913,
"step": 13360
},
{
"epoch": 2.674,
"grad_norm": 2.287477493286133,
"learning_rate": 3.5584445271044544e-07,
"loss": 0.2232,
"step": 13370
},
{
"epoch": 2.676,
"grad_norm": 2.551602602005005,
"learning_rate": 3.515460287999345e-07,
"loss": 0.1898,
"step": 13380
},
{
"epoch": 2.678,
"grad_norm": 1.6702704429626465,
"learning_rate": 3.472727782123697e-07,
"loss": 0.1873,
"step": 13390
},
{
"epoch": 2.68,
"grad_norm": 1.998603105545044,
"learning_rate": 3.430247240891904e-07,
"loss": 0.2089,
"step": 13400
},
{
"epoch": 2.682,
"grad_norm": 2.5154683589935303,
"learning_rate": 3.3880188943538617e-07,
"loss": 0.1813,
"step": 13410
},
{
"epoch": 2.684,
"grad_norm": 1.4025160074234009,
"learning_rate": 3.3460429711937417e-07,
"loss": 0.1896,
"step": 13420
},
{
"epoch": 2.686,
"grad_norm": 1.7359564304351807,
"learning_rate": 3.304319698728714e-07,
"loss": 0.1729,
"step": 13430
},
{
"epoch": 2.6879999999999997,
"grad_norm": 1.8465240001678467,
"learning_rate": 3.262849302907767e-07,
"loss": 0.1519,
"step": 13440
},
{
"epoch": 2.69,
"grad_norm": 1.9284135103225708,
"learning_rate": 3.2216320083104434e-07,
"loss": 0.2169,
"step": 13450
},
{
"epoch": 2.692,
"grad_norm": 1.5323783159255981,
"learning_rate": 3.180668038145629e-07,
"loss": 0.1498,
"step": 13460
},
{
"epoch": 2.694,
"grad_norm": 1.5375773906707764,
"learning_rate": 3.1399576142503606e-07,
"loss": 0.2163,
"step": 13470
},
{
"epoch": 2.6959999999999997,
"grad_norm": 2.2655889987945557,
"learning_rate": 3.0995009570886305e-07,
"loss": 0.1614,
"step": 13480
},
{
"epoch": 2.698,
"grad_norm": 2.173973560333252,
"learning_rate": 3.05929828575014e-07,
"loss": 0.155,
"step": 13490
},
{
"epoch": 2.7,
"grad_norm": 1.8570374250411987,
"learning_rate": 3.01934981794918e-07,
"loss": 0.1896,
"step": 13500
},
{
"epoch": 2.702,
"grad_norm": 1.7890677452087402,
"learning_rate": 2.9796557700234317e-07,
"loss": 0.1941,
"step": 13510
},
{
"epoch": 2.7039999999999997,
"grad_norm": 1.8128156661987305,
"learning_rate": 2.940216356932746e-07,
"loss": 0.2079,
"step": 13520
},
{
"epoch": 2.706,
"grad_norm": 1.5485886335372925,
"learning_rate": 2.901031792258058e-07,
"loss": 0.1603,
"step": 13530
},
{
"epoch": 2.708,
"grad_norm": 1.4785550832748413,
"learning_rate": 2.862102288200186e-07,
"loss": 0.2021,
"step": 13540
},
{
"epoch": 2.71,
"grad_norm": 2.2677414417266846,
"learning_rate": 2.823428055578664e-07,
"loss": 0.1853,
"step": 13550
},
{
"epoch": 2.7119999999999997,
"grad_norm": 1.6280418634414673,
"learning_rate": 2.7850093038306493e-07,
"loss": 0.164,
"step": 13560
},
{
"epoch": 2.714,
"grad_norm": 1.7581329345703125,
"learning_rate": 2.746846241009765e-07,
"loss": 0.2056,
"step": 13570
},
{
"epoch": 2.716,
"grad_norm": 1.453389286994934,
"learning_rate": 2.7089390737849386e-07,
"loss": 0.1762,
"step": 13580
},
{
"epoch": 2.718,
"grad_norm": 1.141204595565796,
"learning_rate": 2.67128800743936e-07,
"loss": 0.2031,
"step": 13590
},
{
"epoch": 2.7199999999999998,
"grad_norm": 1.8374614715576172,
"learning_rate": 2.6338932458692847e-07,
"loss": 0.1961,
"step": 13600
},
{
"epoch": 2.722,
"grad_norm": 1.7040131092071533,
"learning_rate": 2.596754991583017e-07,
"loss": 0.1614,
"step": 13610
},
{
"epoch": 2.724,
"grad_norm": 1.686630368232727,
"learning_rate": 2.5598734456997287e-07,
"loss": 0.1527,
"step": 13620
},
{
"epoch": 2.726,
"grad_norm": 1.6243407726287842,
"learning_rate": 2.523248807948403e-07,
"loss": 0.164,
"step": 13630
},
{
"epoch": 2.7279999999999998,
"grad_norm": 2.1291511058807373,
"learning_rate": 2.486881276666808e-07,
"loss": 0.1484,
"step": 13640
},
{
"epoch": 2.73,
"grad_norm": 1.8856449127197266,
"learning_rate": 2.4507710488003155e-07,
"loss": 0.2218,
"step": 13650
},
{
"epoch": 2.732,
"grad_norm": 1.4211268424987793,
"learning_rate": 2.414918319900922e-07,
"loss": 0.1666,
"step": 13660
},
{
"epoch": 2.734,
"grad_norm": 1.9689491987228394,
"learning_rate": 2.379323284126156e-07,
"loss": 0.2322,
"step": 13670
},
{
"epoch": 2.7359999999999998,
"grad_norm": 1.373671054840088,
"learning_rate": 2.3439861342380243e-07,
"loss": 0.1145,
"step": 13680
},
{
"epoch": 2.738,
"grad_norm": 1.6998653411865234,
"learning_rate": 2.3089070616019838e-07,
"loss": 0.1484,
"step": 13690
},
{
"epoch": 2.74,
"grad_norm": 2.0054614543914795,
"learning_rate": 2.2740862561858812e-07,
"loss": 0.1963,
"step": 13700
},
{
"epoch": 2.742,
"grad_norm": 2.4972550868988037,
"learning_rate": 2.2395239065589436e-07,
"loss": 0.1747,
"step": 13710
},
{
"epoch": 2.7439999999999998,
"grad_norm": 1.2388681173324585,
"learning_rate": 2.2052201998907673e-07,
"loss": 0.1661,
"step": 13720
},
{
"epoch": 2.746,
"grad_norm": 1.4141393899917603,
"learning_rate": 2.1711753219502584e-07,
"loss": 0.1864,
"step": 13730
},
{
"epoch": 2.748,
"grad_norm": 1.2588516473770142,
"learning_rate": 2.1373894571046772e-07,
"loss": 0.1865,
"step": 13740
},
{
"epoch": 2.75,
"grad_norm": 2.5091469287872314,
"learning_rate": 2.103862788318628e-07,
"loss": 0.1887,
"step": 13750
},
{
"epoch": 2.752,
"grad_norm": 1.595197081565857,
"learning_rate": 2.0705954971530385e-07,
"loss": 0.1678,
"step": 13760
},
{
"epoch": 2.754,
"grad_norm": 2.847097396850586,
"learning_rate": 2.0375877637642038e-07,
"loss": 0.1733,
"step": 13770
},
{
"epoch": 2.7560000000000002,
"grad_norm": 1.9094551801681519,
"learning_rate": 2.0048397669028164e-07,
"loss": 0.1525,
"step": 13780
},
{
"epoch": 2.758,
"grad_norm": 1.487987995147705,
"learning_rate": 1.9723516839129765e-07,
"loss": 0.1754,
"step": 13790
},
{
"epoch": 2.76,
"grad_norm": 2.1492459774017334,
"learning_rate": 1.9401236907312437e-07,
"loss": 0.1516,
"step": 13800
},
{
"epoch": 2.762,
"grad_norm": 1.7400575876235962,
"learning_rate": 1.9081559618856938e-07,
"loss": 0.1693,
"step": 13810
},
{
"epoch": 2.7640000000000002,
"grad_norm": 1.5171923637390137,
"learning_rate": 1.8764486704949402e-07,
"loss": 0.1957,
"step": 13820
},
{
"epoch": 2.766,
"grad_norm": 2.1869702339172363,
"learning_rate": 1.8450019882672366e-07,
"loss": 0.1919,
"step": 13830
},
{
"epoch": 2.768,
"grad_norm": 1.4110556840896606,
"learning_rate": 1.8138160854995145e-07,
"loss": 0.195,
"step": 13840
},
{
"epoch": 2.77,
"grad_norm": 1.9794977903366089,
"learning_rate": 1.7828911310764974e-07,
"loss": 0.1523,
"step": 13850
},
{
"epoch": 2.7720000000000002,
"grad_norm": 1.8616331815719604,
"learning_rate": 1.7522272924697438e-07,
"loss": 0.1648,
"step": 13860
},
{
"epoch": 2.774,
"grad_norm": 1.7763707637786865,
"learning_rate": 1.7218247357367656e-07,
"loss": 0.1844,
"step": 13870
},
{
"epoch": 2.776,
"grad_norm": 1.4819947481155396,
"learning_rate": 1.6916836255201297e-07,
"loss": 0.1744,
"step": 13880
},
{
"epoch": 2.778,
"grad_norm": 1.3867019414901733,
"learning_rate": 1.6618041250465623e-07,
"loss": 0.1908,
"step": 13890
},
{
"epoch": 2.7800000000000002,
"grad_norm": 2.0330920219421387,
"learning_rate": 1.6321863961260452e-07,
"loss": 0.1612,
"step": 13900
},
{
"epoch": 2.782,
"grad_norm": 2.5724823474884033,
"learning_rate": 1.602830599150984e-07,
"loss": 0.2228,
"step": 13910
},
{
"epoch": 2.784,
"grad_norm": 1.4262503385543823,
"learning_rate": 1.573736893095301e-07,
"loss": 0.1397,
"step": 13920
},
{
"epoch": 2.786,
"grad_norm": 1.5158369541168213,
"learning_rate": 1.5449054355135718e-07,
"loss": 0.1641,
"step": 13930
},
{
"epoch": 2.7880000000000003,
"grad_norm": 1.3711594343185425,
"learning_rate": 1.5163363825402121e-07,
"loss": 0.213,
"step": 13940
},
{
"epoch": 2.79,
"grad_norm": 1.3406068086624146,
"learning_rate": 1.488029888888598e-07,
"loss": 0.1991,
"step": 13950
},
{
"epoch": 2.792,
"grad_norm": 1.1743435859680176,
"learning_rate": 1.459986107850231e-07,
"loss": 0.1848,
"step": 13960
},
{
"epoch": 2.794,
"grad_norm": 1.5763418674468994,
"learning_rate": 1.4322051912939173e-07,
"loss": 0.166,
"step": 13970
},
{
"epoch": 2.7960000000000003,
"grad_norm": 1.492067813873291,
"learning_rate": 1.40468728966493e-07,
"loss": 0.1746,
"step": 13980
},
{
"epoch": 2.798,
"grad_norm": 1.7857205867767334,
"learning_rate": 1.3774325519842423e-07,
"loss": 0.2087,
"step": 13990
},
{
"epoch": 2.8,
"grad_norm": 2.416992425918579,
"learning_rate": 1.3504411258476514e-07,
"loss": 0.1844,
"step": 14000
},
{
"epoch": 2.802,
"grad_norm": 2.181525945663452,
"learning_rate": 1.323713157425005e-07,
"loss": 0.1949,
"step": 14010
},
{
"epoch": 2.8040000000000003,
"grad_norm": 1.5728259086608887,
"learning_rate": 1.29724879145946e-07,
"loss": 0.1979,
"step": 14020
},
{
"epoch": 2.806,
"grad_norm": 1.9201794862747192,
"learning_rate": 1.2710481712666144e-07,
"loss": 0.1512,
"step": 14030
},
{
"epoch": 2.808,
"grad_norm": 0.9020956158638,
"learning_rate": 1.245111438733798e-07,
"loss": 0.1687,
"step": 14040
},
{
"epoch": 2.81,
"grad_norm": 1.2932790517807007,
"learning_rate": 1.2194387343192504e-07,
"loss": 0.1652,
"step": 14050
},
{
"epoch": 2.8120000000000003,
"grad_norm": 1.8155263662338257,
"learning_rate": 1.194030197051421e-07,
"loss": 0.1753,
"step": 14060
},
{
"epoch": 2.814,
"grad_norm": 2.2060601711273193,
"learning_rate": 1.1688859645281659e-07,
"loss": 0.1405,
"step": 14070
},
{
"epoch": 2.816,
"grad_norm": 1.731437087059021,
"learning_rate": 1.1440061729160235e-07,
"loss": 0.1529,
"step": 14080
},
{
"epoch": 2.818,
"grad_norm": 2.3887579441070557,
"learning_rate": 1.1193909569494676e-07,
"loss": 0.2014,
"step": 14090
},
{
"epoch": 2.82,
"grad_norm": 1.5371918678283691,
"learning_rate": 1.0950404499302015e-07,
"loss": 0.1534,
"step": 14100
},
{
"epoch": 2.822,
"grad_norm": 1.737833857536316,
"learning_rate": 1.0709547837263967e-07,
"loss": 0.1727,
"step": 14110
},
{
"epoch": 2.824,
"grad_norm": 1.5639506578445435,
"learning_rate": 1.0471340887720171e-07,
"loss": 0.1899,
"step": 14120
},
{
"epoch": 2.826,
"grad_norm": 1.4468129873275757,
"learning_rate": 1.0235784940660965e-07,
"loss": 0.1617,
"step": 14130
},
{
"epoch": 2.828,
"grad_norm": 1.7118704319000244,
"learning_rate": 1.0002881271720222e-07,
"loss": 0.1679,
"step": 14140
},
{
"epoch": 2.83,
"grad_norm": 1.9047939777374268,
"learning_rate": 9.772631142168864e-08,
"loss": 0.196,
"step": 14150
},
{
"epoch": 2.832,
"grad_norm": 2.239691734313965,
"learning_rate": 9.545035798907642e-08,
"loss": 0.1941,
"step": 14160
},
{
"epoch": 2.834,
"grad_norm": 2.0404582023620605,
"learning_rate": 9.320096474460527e-08,
"loss": 0.1634,
"step": 14170
},
{
"epoch": 2.836,
"grad_norm": 2.4190192222595215,
"learning_rate": 9.097814386968052e-08,
"loss": 0.2195,
"step": 14180
},
{
"epoch": 2.838,
"grad_norm": 2.054330348968506,
"learning_rate": 8.878190740180759e-08,
"loss": 0.1953,
"step": 14190
},
{
"epoch": 2.84,
"grad_norm": 2.02219557762146,
"learning_rate": 8.661226723452542e-08,
"loss": 0.1761,
"step": 14200
},
{
"epoch": 2.842,
"grad_norm": 1.775745153427124,
"learning_rate": 8.446923511734317e-08,
"loss": 0.1449,
"step": 14210
},
{
"epoch": 2.844,
"grad_norm": 1.5100916624069214,
"learning_rate": 8.235282265567635e-08,
"loss": 0.1972,
"step": 14220
},
{
"epoch": 2.846,
"grad_norm": 2.3685286045074463,
"learning_rate": 8.02630413107841e-08,
"loss": 0.1673,
"step": 14230
},
{
"epoch": 2.848,
"grad_norm": 1.453798770904541,
"learning_rate": 7.819990239970654e-08,
"loss": 0.1724,
"step": 14240
},
{
"epoch": 2.85,
"grad_norm": 1.8765348196029663,
"learning_rate": 7.616341709520359e-08,
"loss": 0.1887,
"step": 14250
},
{
"epoch": 2.852,
"grad_norm": 1.9411275386810303,
"learning_rate": 7.415359642569564e-08,
"loss": 0.1774,
"step": 14260
},
{
"epoch": 2.854,
"grad_norm": 1.686895489692688,
"learning_rate": 7.21704512752025e-08,
"loss": 0.1713,
"step": 14270
},
{
"epoch": 2.856,
"grad_norm": 1.6647790670394897,
"learning_rate": 7.021399238328452e-08,
"loss": 0.1696,
"step": 14280
},
{
"epoch": 2.858,
"grad_norm": 2.324164390563965,
"learning_rate": 6.828423034498488e-08,
"loss": 0.1702,
"step": 14290
},
{
"epoch": 2.86,
"grad_norm": 1.610374093055725,
"learning_rate": 6.638117561077295e-08,
"loss": 0.1636,
"step": 14300
},
{
"epoch": 2.862,
"grad_norm": 1.3403874635696411,
"learning_rate": 6.450483848648547e-08,
"loss": 0.1701,
"step": 14310
},
{
"epoch": 2.864,
"grad_norm": 2.045170545578003,
"learning_rate": 6.265522913327326e-08,
"loss": 0.1838,
"step": 14320
},
{
"epoch": 2.866,
"grad_norm": 1.726438045501709,
"learning_rate": 6.083235756754513e-08,
"loss": 0.1493,
"step": 14330
},
{
"epoch": 2.868,
"grad_norm": 1.6481523513793945,
"learning_rate": 5.903623366091349e-08,
"loss": 0.1588,
"step": 14340
},
{
"epoch": 2.87,
"grad_norm": 1.7567453384399414,
"learning_rate": 5.726686714013996e-08,
"loss": 0.156,
"step": 14350
},
{
"epoch": 2.872,
"grad_norm": 1.5608904361724854,
"learning_rate": 5.552426758708429e-08,
"loss": 0.212,
"step": 14360
},
{
"epoch": 2.874,
"grad_norm": 1.747316837310791,
"learning_rate": 5.380844443865274e-08,
"loss": 0.1706,
"step": 14370
},
{
"epoch": 2.876,
"grad_norm": 2.100895881652832,
"learning_rate": 5.2119406986745336e-08,
"loss": 0.168,
"step": 14380
},
{
"epoch": 2.878,
"grad_norm": 1.1627360582351685,
"learning_rate": 5.0457164378205914e-08,
"loss": 0.1368,
"step": 14390
},
{
"epoch": 2.88,
"grad_norm": 1.7362346649169922,
"learning_rate": 4.882172561477438e-08,
"loss": 0.1627,
"step": 14400
},
{
"epoch": 2.882,
"grad_norm": 2.2970685958862305,
"learning_rate": 4.7213099553035655e-08,
"loss": 0.1751,
"step": 14410
},
{
"epoch": 2.884,
"grad_norm": 2.3094658851623535,
"learning_rate": 4.563129490437246e-08,
"loss": 0.2271,
"step": 14420
},
{
"epoch": 2.886,
"grad_norm": 1.8514689207077026,
"learning_rate": 4.4076320234919254e-08,
"loss": 0.162,
"step": 14430
},
{
"epoch": 2.888,
"grad_norm": 1.9428884983062744,
"learning_rate": 4.2548183965513415e-08,
"loss": 0.2024,
"step": 14440
},
{
"epoch": 2.89,
"grad_norm": 1.88101327419281,
"learning_rate": 4.104689437165354e-08,
"loss": 0.162,
"step": 14450
},
{
"epoch": 2.892,
"grad_norm": 3.1377224922180176,
"learning_rate": 3.957245958345013e-08,
"loss": 0.1987,
"step": 14460
},
{
"epoch": 2.894,
"grad_norm": 1.4738689661026,
"learning_rate": 3.812488758558386e-08,
"loss": 0.1797,
"step": 14470
},
{
"epoch": 2.896,
"grad_norm": 1.6967921257019043,
"learning_rate": 3.6704186217263457e-08,
"loss": 0.1748,
"step": 14480
},
{
"epoch": 2.898,
"grad_norm": 1.5632615089416504,
"learning_rate": 3.531036317218128e-08,
"loss": 0.1959,
"step": 14490
},
{
"epoch": 2.9,
"grad_norm": 1.3949593305587769,
"learning_rate": 3.394342599847111e-08,
"loss": 0.1612,
"step": 14500
},
{
"epoch": 2.902,
"grad_norm": 1.8362523317337036,
"learning_rate": 3.260338209867153e-08,
"loss": 0.1713,
"step": 14510
},
{
"epoch": 2.904,
"grad_norm": 1.7234903573989868,
"learning_rate": 3.129023872967874e-08,
"loss": 0.1514,
"step": 14520
},
{
"epoch": 2.906,
"grad_norm": 1.4591097831726074,
"learning_rate": 3.0004003002714886e-08,
"loss": 0.1996,
"step": 14530
},
{
"epoch": 2.908,
"grad_norm": 2.5706465244293213,
"learning_rate": 2.8744681883284274e-08,
"loss": 0.1624,
"step": 14540
},
{
"epoch": 2.91,
"grad_norm": 1.6069506406784058,
"learning_rate": 2.7512282191136663e-08,
"loss": 0.1696,
"step": 14550
},
{
"epoch": 2.912,
"grad_norm": 1.6182852983474731,
"learning_rate": 2.6306810600233435e-08,
"loss": 0.1736,
"step": 14560
},
{
"epoch": 2.914,
"grad_norm": 1.7584065198898315,
"learning_rate": 2.5128273638706513e-08,
"loss": 0.1879,
"step": 14570
},
{
"epoch": 2.916,
"grad_norm": 2.2820541858673096,
"learning_rate": 2.3976677688827276e-08,
"loss": 0.194,
"step": 14580
},
{
"epoch": 2.918,
"grad_norm": 1.8409901857376099,
"learning_rate": 2.285202898696881e-08,
"loss": 0.208,
"step": 14590
},
{
"epoch": 2.92,
"grad_norm": 2.1918840408325195,
"learning_rate": 2.175433362357482e-08,
"loss": 0.1869,
"step": 14600
},
{
"epoch": 2.922,
"grad_norm": 1.8916000127792358,
"learning_rate": 2.0683597543124655e-08,
"loss": 0.1531,
"step": 14610
},
{
"epoch": 2.924,
"grad_norm": 1.347509741783142,
"learning_rate": 1.963982654410279e-08,
"loss": 0.1879,
"step": 14620
},
{
"epoch": 2.926,
"grad_norm": 1.712788701057434,
"learning_rate": 1.862302627896495e-08,
"loss": 0.1711,
"step": 14630
},
{
"epoch": 2.928,
"grad_norm": 2.293809175491333,
"learning_rate": 1.763320225411036e-08,
"loss": 0.2065,
"step": 14640
},
{
"epoch": 2.93,
"grad_norm": 2.585679531097412,
"learning_rate": 1.6670359829850657e-08,
"loss": 0.2012,
"step": 14650
},
{
"epoch": 2.932,
"grad_norm": 2.2906439304351807,
"learning_rate": 1.573450422037992e-08,
"loss": 0.1803,
"step": 14660
},
{
"epoch": 2.934,
"grad_norm": 1.0971269607543945,
"learning_rate": 1.4825640493747462e-08,
"loss": 0.1916,
"step": 14670
},
{
"epoch": 2.936,
"grad_norm": 1.7314507961273193,
"learning_rate": 1.3943773571831188e-08,
"loss": 0.205,
"step": 14680
},
{
"epoch": 2.9379999999999997,
"grad_norm": 1.9683465957641602,
"learning_rate": 1.3088908230308728e-08,
"loss": 0.1355,
"step": 14690
},
{
"epoch": 2.94,
"grad_norm": 2.5472261905670166,
"learning_rate": 1.2261049098634125e-08,
"loss": 0.1713,
"step": 14700
},
{
"epoch": 2.942,
"grad_norm": 1.2666493654251099,
"learning_rate": 1.146020066001119e-08,
"loss": 0.1523,
"step": 14710
},
{
"epoch": 2.944,
"grad_norm": 1.82552170753479,
"learning_rate": 1.0686367251368512e-08,
"loss": 0.1705,
"step": 14720
},
{
"epoch": 2.9459999999999997,
"grad_norm": 1.3673129081726074,
"learning_rate": 9.939553063337826e-09,
"loss": 0.1737,
"step": 14730
},
{
"epoch": 2.948,
"grad_norm": 2.0833985805511475,
"learning_rate": 9.219762140231237e-09,
"loss": 0.1701,
"step": 14740
},
{
"epoch": 2.95,
"grad_norm": 1.2477442026138306,
"learning_rate": 8.526998380016804e-09,
"loss": 0.1864,
"step": 14750
},
{
"epoch": 2.952,
"grad_norm": 2.231452226638794,
"learning_rate": 7.861265534300222e-09,
"loss": 0.1928,
"step": 14760
},
{
"epoch": 2.9539999999999997,
"grad_norm": 1.7045234441757202,
"learning_rate": 7.222567208303721e-09,
"loss": 0.1471,
"step": 14770
},
{
"epoch": 2.956,
"grad_norm": 1.1955382823944092,
"learning_rate": 6.610906860845534e-09,
"loss": 0.1582,
"step": 14780
},
{
"epoch": 2.958,
"grad_norm": 1.9184893369674683,
"learning_rate": 6.026287804322683e-09,
"loss": 0.2123,
"step": 14790
},
{
"epoch": 2.96,
"grad_norm": 2.046898126602173,
"learning_rate": 5.468713204692111e-09,
"loss": 0.1694,
"step": 14800
},
{
"epoch": 2.9619999999999997,
"grad_norm": 1.4556955099105835,
"learning_rate": 4.9381860814529115e-09,
"loss": 0.1834,
"step": 14810
},
{
"epoch": 2.964,
"grad_norm": 1.5225237607955933,
"learning_rate": 4.434709307631901e-09,
"loss": 0.1509,
"step": 14820
},
{
"epoch": 2.966,
"grad_norm": 2.2594845294952393,
"learning_rate": 3.9582856097658554e-09,
"loss": 0.1882,
"step": 14830
},
{
"epoch": 2.968,
"grad_norm": 1.5588994026184082,
"learning_rate": 3.5089175678892916e-09,
"loss": 0.203,
"step": 14840
},
{
"epoch": 2.9699999999999998,
"grad_norm": 2.009692668914795,
"learning_rate": 3.0866076155172677e-09,
"loss": 0.1684,
"step": 14850
},
{
"epoch": 2.972,
"grad_norm": 1.583719253540039,
"learning_rate": 2.6913580396359384e-09,
"loss": 0.1863,
"step": 14860
},
{
"epoch": 2.974,
"grad_norm": 1.9162685871124268,
"learning_rate": 2.323170980685907e-09,
"loss": 0.1777,
"step": 14870
},
{
"epoch": 2.976,
"grad_norm": 1.2543405294418335,
"learning_rate": 1.982048432554451e-09,
"loss": 0.1686,
"step": 14880
},
{
"epoch": 2.9779999999999998,
"grad_norm": 2.3343191146850586,
"learning_rate": 1.6679922425638651e-09,
"loss": 0.2004,
"step": 14890
},
{
"epoch": 2.98,
"grad_norm": 2.2958972454071045,
"learning_rate": 1.3810041114581397e-09,
"loss": 0.2172,
"step": 14900
},
{
"epoch": 2.982,
"grad_norm": 1.3667978048324585,
"learning_rate": 1.1210855933996289e-09,
"loss": 0.1731,
"step": 14910
},
{
"epoch": 2.984,
"grad_norm": 2.113403558731079,
"learning_rate": 8.88238095955174e-10,
"loss": 0.1527,
"step": 14920
},
{
"epoch": 2.9859999999999998,
"grad_norm": 2.218931198120117,
"learning_rate": 6.824628800911059e-10,
"loss": 0.1829,
"step": 14930
},
{
"epoch": 2.988,
"grad_norm": 1.7202938795089722,
"learning_rate": 5.037610601665855e-10,
"loss": 0.148,
"step": 14940
},
{
"epoch": 2.99,
"grad_norm": 1.6666905879974365,
"learning_rate": 3.521336039263856e-10,
"loss": 0.1999,
"step": 14950
},
{
"epoch": 2.992,
"grad_norm": 1.587571620941162,
"learning_rate": 2.2758133249756175e-10,
"loss": 0.14,
"step": 14960
},
{
"epoch": 2.9939999999999998,
"grad_norm": 1.2970430850982666,
"learning_rate": 1.3010492038223465e-10,
"loss": 0.1418,
"step": 14970
},
{
"epoch": 2.996,
"grad_norm": 1.8305574655532837,
"learning_rate": 5.970489545537028e-11,
"loss": 0.1642,
"step": 14980
},
{
"epoch": 2.998,
"grad_norm": 1.9229685068130493,
"learning_rate": 1.63816389631144e-11,
"loss": 0.1686,
"step": 14990
},
{
"epoch": 3.0,
"grad_norm": 1.2971068620681763,
"learning_rate": 1.3538551946190582e-13,
"loss": 0.1871,
"step": 15000
}
],
"logging_steps": 10,
"max_steps": 15000,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.3591424743807386e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}