PEFT
Safetensors
tinyllama-conversational-v1 / trainer_state.json
Clark88's picture
Upload folder using huggingface_hub
8201d37 verified
Raw
History Blame Contribute Delete
131 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 125.0,
"eval_steps": 500,
"global_step": 8000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.15625,
"grad_norm": 0.6168534159660339,
"learning_rate": 0.00019999922893814705,
"loss": 1.9025,
"step": 10
},
{
"epoch": 0.3125,
"grad_norm": 0.6524689793586731,
"learning_rate": 0.00019999691576447898,
"loss": 1.6781,
"step": 20
},
{
"epoch": 0.46875,
"grad_norm": 0.7400289177894592,
"learning_rate": 0.00019999306051466772,
"loss": 1.5857,
"step": 30
},
{
"epoch": 0.625,
"grad_norm": 0.6026787161827087,
"learning_rate": 0.00019998766324816607,
"loss": 1.4631,
"step": 40
},
{
"epoch": 0.78125,
"grad_norm": 0.676242470741272,
"learning_rate": 0.0001999807240482065,
"loss": 1.4429,
"step": 50
},
{
"epoch": 0.9375,
"grad_norm": 0.5645259022712708,
"learning_rate": 0.00019997224302180006,
"loss": 1.4138,
"step": 60
},
{
"epoch": 1.09375,
"grad_norm": 0.6464071273803711,
"learning_rate": 0.0001999622202997347,
"loss": 1.4142,
"step": 70
},
{
"epoch": 1.25,
"grad_norm": 0.5914332270622253,
"learning_rate": 0.00019995065603657316,
"loss": 1.3459,
"step": 80
},
{
"epoch": 1.40625,
"grad_norm": 0.6979939341545105,
"learning_rate": 0.0001999375504106507,
"loss": 1.3659,
"step": 90
},
{
"epoch": 1.5625,
"grad_norm": 0.7593270540237427,
"learning_rate": 0.0001999229036240723,
"loss": 1.3414,
"step": 100
},
{
"epoch": 1.71875,
"grad_norm": 0.9171484112739563,
"learning_rate": 0.0001999067159027095,
"loss": 1.3622,
"step": 110
},
{
"epoch": 1.875,
"grad_norm": 0.9084861874580383,
"learning_rate": 0.00019988898749619702,
"loss": 1.3028,
"step": 120
},
{
"epoch": 2.03125,
"grad_norm": 0.7756249904632568,
"learning_rate": 0.00019986971867792875,
"loss": 1.2742,
"step": 130
},
{
"epoch": 2.1875,
"grad_norm": 0.813124418258667,
"learning_rate": 0.00019984890974505381,
"loss": 1.3174,
"step": 140
},
{
"epoch": 2.34375,
"grad_norm": 0.7856652140617371,
"learning_rate": 0.00019982656101847162,
"loss": 1.2747,
"step": 150
},
{
"epoch": 2.5,
"grad_norm": 0.8684581518173218,
"learning_rate": 0.00019980267284282717,
"loss": 1.2513,
"step": 160
},
{
"epoch": 2.65625,
"grad_norm": 1.0259562730789185,
"learning_rate": 0.00019977724558650572,
"loss": 1.2616,
"step": 170
},
{
"epoch": 2.8125,
"grad_norm": 0.9384151101112366,
"learning_rate": 0.00019975027964162702,
"loss": 1.2425,
"step": 180
},
{
"epoch": 2.96875,
"grad_norm": 0.9272423982620239,
"learning_rate": 0.0001997217754240393,
"loss": 1.1985,
"step": 190
},
{
"epoch": 3.125,
"grad_norm": 1.033215045928955,
"learning_rate": 0.0001996917333733128,
"loss": 1.2136,
"step": 200
},
{
"epoch": 3.28125,
"grad_norm": 1.1179872751235962,
"learning_rate": 0.00019966015395273318,
"loss": 1.1696,
"step": 210
},
{
"epoch": 3.4375,
"grad_norm": 1.0058550834655762,
"learning_rate": 0.00019962703764929413,
"loss": 1.2167,
"step": 220
},
{
"epoch": 3.59375,
"grad_norm": 1.0203720331192017,
"learning_rate": 0.00019959238497369003,
"loss": 1.2328,
"step": 230
},
{
"epoch": 3.75,
"grad_norm": 0.9134518504142761,
"learning_rate": 0.00019955619646030802,
"loss": 1.185,
"step": 240
},
{
"epoch": 3.90625,
"grad_norm": 1.0784626007080078,
"learning_rate": 0.0001995184726672197,
"loss": 1.1593,
"step": 250
},
{
"epoch": 4.0625,
"grad_norm": 1.117414951324463,
"learning_rate": 0.00019947921417617267,
"loss": 1.1788,
"step": 260
},
{
"epoch": 4.21875,
"grad_norm": 1.0805243253707886,
"learning_rate": 0.0001994384215925814,
"loss": 1.1486,
"step": 270
},
{
"epoch": 4.375,
"grad_norm": 1.2142518758773804,
"learning_rate": 0.000199396095545518,
"loss": 1.1243,
"step": 280
},
{
"epoch": 4.53125,
"grad_norm": 1.1736822128295898,
"learning_rate": 0.00019935223668770245,
"loss": 1.0945,
"step": 290
},
{
"epoch": 4.6875,
"grad_norm": 1.0896259546279907,
"learning_rate": 0.00019930684569549264,
"loss": 1.1531,
"step": 300
},
{
"epoch": 4.84375,
"grad_norm": 1.172074794769287,
"learning_rate": 0.00019925992326887379,
"loss": 1.15,
"step": 310
},
{
"epoch": 5.0,
"grad_norm": 1.1381478309631348,
"learning_rate": 0.0001992114701314478,
"loss": 1.1261,
"step": 320
},
{
"epoch": 5.15625,
"grad_norm": 1.2212597131729126,
"learning_rate": 0.00019916148703042193,
"loss": 1.0793,
"step": 330
},
{
"epoch": 5.3125,
"grad_norm": 1.1290854215621948,
"learning_rate": 0.0001991099747365975,
"loss": 1.1324,
"step": 340
},
{
"epoch": 5.46875,
"grad_norm": 1.298770785331726,
"learning_rate": 0.00019905693404435773,
"loss": 1.0982,
"step": 350
},
{
"epoch": 5.625,
"grad_norm": 1.1428149938583374,
"learning_rate": 0.00019900236577165576,
"loss": 1.0749,
"step": 360
},
{
"epoch": 5.78125,
"grad_norm": 1.2018159627914429,
"learning_rate": 0.00019894627076000185,
"loss": 1.0501,
"step": 370
},
{
"epoch": 5.9375,
"grad_norm": 1.2703092098236084,
"learning_rate": 0.0001988886498744505,
"loss": 1.077,
"step": 380
},
{
"epoch": 6.09375,
"grad_norm": 1.2956756353378296,
"learning_rate": 0.00019882950400358694,
"loss": 1.0571,
"step": 390
},
{
"epoch": 6.25,
"grad_norm": 1.324562668800354,
"learning_rate": 0.00019876883405951377,
"loss": 1.0494,
"step": 400
},
{
"epoch": 6.40625,
"grad_norm": 1.2973648309707642,
"learning_rate": 0.00019870664097783655,
"loss": 1.0362,
"step": 410
},
{
"epoch": 6.5625,
"grad_norm": 1.3334016799926758,
"learning_rate": 0.00019864292571764955,
"loss": 1.0537,
"step": 420
},
{
"epoch": 6.71875,
"grad_norm": 1.3302708864212036,
"learning_rate": 0.0001985776892615209,
"loss": 1.0082,
"step": 430
},
{
"epoch": 6.875,
"grad_norm": 1.346130609512329,
"learning_rate": 0.0001985109326154774,
"loss": 1.0633,
"step": 440
},
{
"epoch": 7.03125,
"grad_norm": 1.2094783782958984,
"learning_rate": 0.00019844265680898918,
"loss": 1.0505,
"step": 450
},
{
"epoch": 7.1875,
"grad_norm": 1.3239214420318604,
"learning_rate": 0.00019837286289495361,
"loss": 0.9989,
"step": 460
},
{
"epoch": 7.34375,
"grad_norm": 1.3971266746520996,
"learning_rate": 0.00019830155194967917,
"loss": 0.9816,
"step": 470
},
{
"epoch": 7.5,
"grad_norm": 1.3986828327178955,
"learning_rate": 0.0001982287250728689,
"loss": 0.9951,
"step": 480
},
{
"epoch": 7.65625,
"grad_norm": 1.4055757522583008,
"learning_rate": 0.00019815438338760327,
"loss": 1.0279,
"step": 490
},
{
"epoch": 7.8125,
"grad_norm": 1.5168254375457764,
"learning_rate": 0.00019807852804032305,
"loss": 1.0069,
"step": 500
},
{
"epoch": 7.96875,
"grad_norm": 1.3308908939361572,
"learning_rate": 0.00019800116020081158,
"loss": 0.9957,
"step": 510
},
{
"epoch": 8.125,
"grad_norm": 1.629763126373291,
"learning_rate": 0.00019792228106217658,
"loss": 0.9368,
"step": 520
},
{
"epoch": 8.28125,
"grad_norm": 1.487622618675232,
"learning_rate": 0.000197841891840832,
"loss": 1.0098,
"step": 530
},
{
"epoch": 8.4375,
"grad_norm": 1.3082317113876343,
"learning_rate": 0.0001977599937764791,
"loss": 0.9657,
"step": 540
},
{
"epoch": 8.59375,
"grad_norm": 1.4622434377670288,
"learning_rate": 0.00019767658813208726,
"loss": 0.9627,
"step": 550
},
{
"epoch": 8.75,
"grad_norm": 1.4137543439865112,
"learning_rate": 0.00019759167619387476,
"loss": 0.9818,
"step": 560
},
{
"epoch": 8.90625,
"grad_norm": 1.4448939561843872,
"learning_rate": 0.0001975052592712887,
"loss": 0.932,
"step": 570
},
{
"epoch": 9.0625,
"grad_norm": 1.4551522731781006,
"learning_rate": 0.00019741733869698495,
"loss": 0.9693,
"step": 580
},
{
"epoch": 9.21875,
"grad_norm": 1.4330525398254395,
"learning_rate": 0.0001973279158268075,
"loss": 0.9367,
"step": 590
},
{
"epoch": 9.375,
"grad_norm": 1.3818308115005493,
"learning_rate": 0.00019723699203976766,
"loss": 0.9344,
"step": 600
},
{
"epoch": 9.53125,
"grad_norm": 1.4964330196380615,
"learning_rate": 0.00019714456873802272,
"loss": 0.9046,
"step": 610
},
{
"epoch": 9.6875,
"grad_norm": 1.3817216157913208,
"learning_rate": 0.00019705064734685425,
"loss": 0.8933,
"step": 620
},
{
"epoch": 9.84375,
"grad_norm": 1.6543223857879639,
"learning_rate": 0.00019695522931464636,
"loss": 0.937,
"step": 630
},
{
"epoch": 10.0,
"grad_norm": 1.4744540452957153,
"learning_rate": 0.0001968583161128631,
"loss": 0.9789,
"step": 640
},
{
"epoch": 10.15625,
"grad_norm": 1.5337893962860107,
"learning_rate": 0.00019675990923602598,
"loss": 0.8901,
"step": 650
},
{
"epoch": 10.3125,
"grad_norm": 1.4886064529418945,
"learning_rate": 0.00019666001020169073,
"loss": 0.8747,
"step": 660
},
{
"epoch": 10.46875,
"grad_norm": 1.4878489971160889,
"learning_rate": 0.00019655862055042406,
"loss": 0.9148,
"step": 670
},
{
"epoch": 10.625,
"grad_norm": 1.537325143814087,
"learning_rate": 0.00019645574184577982,
"loss": 0.8928,
"step": 680
},
{
"epoch": 10.78125,
"grad_norm": 1.5436025857925415,
"learning_rate": 0.00019635137567427488,
"loss": 0.9109,
"step": 690
},
{
"epoch": 10.9375,
"grad_norm": 1.7200922966003418,
"learning_rate": 0.00019624552364536473,
"loss": 0.9218,
"step": 700
},
{
"epoch": 11.09375,
"grad_norm": 1.4200806617736816,
"learning_rate": 0.00019613818739141862,
"loss": 0.8591,
"step": 710
},
{
"epoch": 11.25,
"grad_norm": 1.618428349494934,
"learning_rate": 0.0001960293685676943,
"loss": 0.849,
"step": 720
},
{
"epoch": 11.40625,
"grad_norm": 1.538055658340454,
"learning_rate": 0.00019591906885231276,
"loss": 0.8597,
"step": 730
},
{
"epoch": 11.5625,
"grad_norm": 1.4808735847473145,
"learning_rate": 0.00019580728994623195,
"loss": 0.8691,
"step": 740
},
{
"epoch": 11.71875,
"grad_norm": 1.5934367179870605,
"learning_rate": 0.0001956940335732209,
"loss": 0.9137,
"step": 750
},
{
"epoch": 11.875,
"grad_norm": 1.584621787071228,
"learning_rate": 0.00019557930147983302,
"loss": 0.8624,
"step": 760
},
{
"epoch": 12.03125,
"grad_norm": 1.503152847290039,
"learning_rate": 0.0001954630954353791,
"loss": 0.8308,
"step": 770
},
{
"epoch": 12.1875,
"grad_norm": 1.3880295753479004,
"learning_rate": 0.0001953454172319001,
"loss": 0.7993,
"step": 780
},
{
"epoch": 12.34375,
"grad_norm": 1.665929913520813,
"learning_rate": 0.00019522626868413956,
"loss": 0.8547,
"step": 790
},
{
"epoch": 12.5,
"grad_norm": 1.6523503065109253,
"learning_rate": 0.00019510565162951537,
"loss": 0.8312,
"step": 800
},
{
"epoch": 12.65625,
"grad_norm": 1.761229157447815,
"learning_rate": 0.00019498356792809176,
"loss": 0.8257,
"step": 810
},
{
"epoch": 12.8125,
"grad_norm": 1.6496795415878296,
"learning_rate": 0.00019486001946255046,
"loss": 0.8549,
"step": 820
},
{
"epoch": 12.96875,
"grad_norm": 1.601778507232666,
"learning_rate": 0.00019473500813816164,
"loss": 0.8432,
"step": 830
},
{
"epoch": 13.125,
"grad_norm": 1.645071268081665,
"learning_rate": 0.00019460853588275454,
"loss": 0.8002,
"step": 840
},
{
"epoch": 13.28125,
"grad_norm": 1.704184889793396,
"learning_rate": 0.00019448060464668783,
"loss": 0.8224,
"step": 850
},
{
"epoch": 13.4375,
"grad_norm": 1.6473637819290161,
"learning_rate": 0.00019435121640281938,
"loss": 0.7799,
"step": 860
},
{
"epoch": 13.59375,
"grad_norm": 1.859934687614441,
"learning_rate": 0.00019422037314647598,
"loss": 0.7917,
"step": 870
},
{
"epoch": 13.75,
"grad_norm": 1.631324052810669,
"learning_rate": 0.00019408807689542257,
"loss": 0.8419,
"step": 880
},
{
"epoch": 13.90625,
"grad_norm": 1.794649362564087,
"learning_rate": 0.00019395432968983092,
"loss": 0.8033,
"step": 890
},
{
"epoch": 14.0625,
"grad_norm": 1.7658714056015015,
"learning_rate": 0.00019381913359224842,
"loss": 0.785,
"step": 900
},
{
"epoch": 14.21875,
"grad_norm": 1.7977893352508545,
"learning_rate": 0.00019368249068756613,
"loss": 0.7421,
"step": 910
},
{
"epoch": 14.375,
"grad_norm": 2.016810178756714,
"learning_rate": 0.00019354440308298675,
"loss": 0.7994,
"step": 920
},
{
"epoch": 14.53125,
"grad_norm": 1.9253318309783936,
"learning_rate": 0.00019340487290799184,
"loss": 0.7738,
"step": 930
},
{
"epoch": 14.6875,
"grad_norm": 1.7931846380233765,
"learning_rate": 0.00019326390231430942,
"loss": 0.7725,
"step": 940
},
{
"epoch": 14.84375,
"grad_norm": 1.9552454948425293,
"learning_rate": 0.00019312149347588037,
"loss": 0.7717,
"step": 950
},
{
"epoch": 15.0,
"grad_norm": 1.7500364780426025,
"learning_rate": 0.00019297764858882514,
"loss": 0.7987,
"step": 960
},
{
"epoch": 15.15625,
"grad_norm": 1.8716918230056763,
"learning_rate": 0.00019283236987140988,
"loss": 0.7371,
"step": 970
},
{
"epoch": 15.3125,
"grad_norm": 1.7724356651306152,
"learning_rate": 0.00019268565956401208,
"loss": 0.7553,
"step": 980
},
{
"epoch": 15.46875,
"grad_norm": 1.8796881437301636,
"learning_rate": 0.00019253751992908622,
"loss": 0.7531,
"step": 990
},
{
"epoch": 15.625,
"grad_norm": 1.8478586673736572,
"learning_rate": 0.0001923879532511287,
"loss": 0.7347,
"step": 1000
},
{
"epoch": 15.78125,
"grad_norm": 1.9052739143371582,
"learning_rate": 0.00019223696183664267,
"loss": 0.7333,
"step": 1010
},
{
"epoch": 15.9375,
"grad_norm": 2.0365710258483887,
"learning_rate": 0.00019208454801410266,
"loss": 0.7623,
"step": 1020
},
{
"epoch": 16.09375,
"grad_norm": 2.4890291690826416,
"learning_rate": 0.00019193071413391822,
"loss": 0.7259,
"step": 1030
},
{
"epoch": 16.25,
"grad_norm": 1.7967720031738281,
"learning_rate": 0.00019177546256839812,
"loss": 0.7239,
"step": 1040
},
{
"epoch": 16.40625,
"grad_norm": 1.8451858758926392,
"learning_rate": 0.00019161879571171362,
"loss": 0.7212,
"step": 1050
},
{
"epoch": 16.5625,
"grad_norm": 2.0714006423950195,
"learning_rate": 0.00019146071597986138,
"loss": 0.6777,
"step": 1060
},
{
"epoch": 16.71875,
"grad_norm": 1.8511077165603638,
"learning_rate": 0.00019130122581062644,
"loss": 0.7451,
"step": 1070
},
{
"epoch": 16.875,
"grad_norm": 1.961347222328186,
"learning_rate": 0.00019114032766354453,
"loss": 0.7201,
"step": 1080
},
{
"epoch": 17.03125,
"grad_norm": 1.7642933130264282,
"learning_rate": 0.0001909780240198641,
"loss": 0.7593,
"step": 1090
},
{
"epoch": 17.1875,
"grad_norm": 1.9495702981948853,
"learning_rate": 0.00019081431738250814,
"loss": 0.6792,
"step": 1100
},
{
"epoch": 17.34375,
"grad_norm": 2.0456838607788086,
"learning_rate": 0.00019064921027603547,
"loss": 0.6827,
"step": 1110
},
{
"epoch": 17.5,
"grad_norm": 2.0225956439971924,
"learning_rate": 0.00019048270524660196,
"loss": 0.7065,
"step": 1120
},
{
"epoch": 17.65625,
"grad_norm": 2.0372891426086426,
"learning_rate": 0.00019031480486192114,
"loss": 0.6961,
"step": 1130
},
{
"epoch": 17.8125,
"grad_norm": 1.9680888652801514,
"learning_rate": 0.00019014551171122457,
"loss": 0.6928,
"step": 1140
},
{
"epoch": 17.96875,
"grad_norm": 1.925260305404663,
"learning_rate": 0.00018997482840522217,
"loss": 0.7214,
"step": 1150
},
{
"epoch": 18.125,
"grad_norm": 2.222017526626587,
"learning_rate": 0.00018980275757606157,
"loss": 0.6755,
"step": 1160
},
{
"epoch": 18.28125,
"grad_norm": 1.8654993772506714,
"learning_rate": 0.0001896293018772879,
"loss": 0.6734,
"step": 1170
},
{
"epoch": 18.4375,
"grad_norm": 2.079108953475952,
"learning_rate": 0.0001894544639838025,
"loss": 0.6914,
"step": 1180
},
{
"epoch": 18.59375,
"grad_norm": 2.1051278114318848,
"learning_rate": 0.0001892782465918221,
"loss": 0.6685,
"step": 1190
},
{
"epoch": 18.75,
"grad_norm": 2.200761079788208,
"learning_rate": 0.0001891006524188368,
"loss": 0.6701,
"step": 1200
},
{
"epoch": 18.90625,
"grad_norm": 2.0247373580932617,
"learning_rate": 0.00018892168420356847,
"loss": 0.644,
"step": 1210
},
{
"epoch": 19.0625,
"grad_norm": 1.7870880365371704,
"learning_rate": 0.00018874134470592835,
"loss": 0.6479,
"step": 1220
},
{
"epoch": 19.21875,
"grad_norm": 2.407914638519287,
"learning_rate": 0.0001885596367069746,
"loss": 0.6514,
"step": 1230
},
{
"epoch": 19.375,
"grad_norm": 2.0643310546875,
"learning_rate": 0.00018837656300886937,
"loss": 0.623,
"step": 1240
},
{
"epoch": 19.53125,
"grad_norm": 1.943108081817627,
"learning_rate": 0.0001881921264348355,
"loss": 0.6305,
"step": 1250
},
{
"epoch": 19.6875,
"grad_norm": 1.9791898727416992,
"learning_rate": 0.00018800632982911322,
"loss": 0.6691,
"step": 1260
},
{
"epoch": 19.84375,
"grad_norm": 1.9491395950317383,
"learning_rate": 0.0001878191760569159,
"loss": 0.6466,
"step": 1270
},
{
"epoch": 20.0,
"grad_norm": 2.236191987991333,
"learning_rate": 0.00018763066800438636,
"loss": 0.6777,
"step": 1280
},
{
"epoch": 20.15625,
"grad_norm": 2.0054900646209717,
"learning_rate": 0.00018744080857855191,
"loss": 0.5978,
"step": 1290
},
{
"epoch": 20.3125,
"grad_norm": 1.9910144805908203,
"learning_rate": 0.00018724960070727972,
"loss": 0.616,
"step": 1300
},
{
"epoch": 20.46875,
"grad_norm": 2.0548858642578125,
"learning_rate": 0.00018705704733923176,
"loss": 0.6339,
"step": 1310
},
{
"epoch": 20.625,
"grad_norm": 2.190333843231201,
"learning_rate": 0.00018686315144381913,
"loss": 0.6166,
"step": 1320
},
{
"epoch": 20.78125,
"grad_norm": 2.069106101989746,
"learning_rate": 0.00018666791601115642,
"loss": 0.6425,
"step": 1330
},
{
"epoch": 20.9375,
"grad_norm": 1.9265080690383911,
"learning_rate": 0.0001864713440520155,
"loss": 0.633,
"step": 1340
},
{
"epoch": 21.09375,
"grad_norm": 2.2421185970306396,
"learning_rate": 0.0001862734385977792,
"loss": 0.6081,
"step": 1350
},
{
"epoch": 21.25,
"grad_norm": 2.1831634044647217,
"learning_rate": 0.0001860742027003944,
"loss": 0.5864,
"step": 1360
},
{
"epoch": 21.40625,
"grad_norm": 2.443437337875366,
"learning_rate": 0.00018587363943232504,
"loss": 0.6257,
"step": 1370
},
{
"epoch": 21.5625,
"grad_norm": 2.1479828357696533,
"learning_rate": 0.00018567175188650498,
"loss": 0.6078,
"step": 1380
},
{
"epoch": 21.71875,
"grad_norm": 2.1301865577697754,
"learning_rate": 0.0001854685431762898,
"loss": 0.5923,
"step": 1390
},
{
"epoch": 21.875,
"grad_norm": 2.476592779159546,
"learning_rate": 0.00018526401643540922,
"loss": 0.5948,
"step": 1400
},
{
"epoch": 22.03125,
"grad_norm": 1.8967773914337158,
"learning_rate": 0.00018505817481791862,
"loss": 0.603,
"step": 1410
},
{
"epoch": 22.1875,
"grad_norm": 2.07448148727417,
"learning_rate": 0.00018485102149815038,
"loss": 0.5644,
"step": 1420
},
{
"epoch": 22.34375,
"grad_norm": 2.0168447494506836,
"learning_rate": 0.00018464255967066493,
"loss": 0.5823,
"step": 1430
},
{
"epoch": 22.5,
"grad_norm": 2.0524818897247314,
"learning_rate": 0.00018443279255020152,
"loss": 0.5657,
"step": 1440
},
{
"epoch": 22.65625,
"grad_norm": 2.0055761337280273,
"learning_rate": 0.00018422172337162867,
"loss": 0.6089,
"step": 1450
},
{
"epoch": 22.8125,
"grad_norm": 2.2025387287139893,
"learning_rate": 0.0001840093553898942,
"loss": 0.5753,
"step": 1460
},
{
"epoch": 22.96875,
"grad_norm": 2.3673453330993652,
"learning_rate": 0.00018379569187997513,
"loss": 0.5791,
"step": 1470
},
{
"epoch": 23.125,
"grad_norm": 2.2405545711517334,
"learning_rate": 0.00018358073613682706,
"loss": 0.542,
"step": 1480
},
{
"epoch": 23.28125,
"grad_norm": 2.062706708908081,
"learning_rate": 0.00018336449147533345,
"loss": 0.582,
"step": 1490
},
{
"epoch": 23.4375,
"grad_norm": 2.4270834922790527,
"learning_rate": 0.00018314696123025454,
"loss": 0.5674,
"step": 1500
},
{
"epoch": 23.59375,
"grad_norm": 2.288257122039795,
"learning_rate": 0.00018292814875617576,
"loss": 0.551,
"step": 1510
},
{
"epoch": 23.75,
"grad_norm": 2.270174503326416,
"learning_rate": 0.00018270805742745617,
"loss": 0.5723,
"step": 1520
},
{
"epoch": 23.90625,
"grad_norm": 2.3951213359832764,
"learning_rate": 0.00018248669063817636,
"loss": 0.5529,
"step": 1530
},
{
"epoch": 24.0625,
"grad_norm": 2.5293216705322266,
"learning_rate": 0.000182264051802086,
"loss": 0.5531,
"step": 1540
},
{
"epoch": 24.21875,
"grad_norm": 2.5577666759490967,
"learning_rate": 0.00018204014435255135,
"loss": 0.5189,
"step": 1550
},
{
"epoch": 24.375,
"grad_norm": 2.211329460144043,
"learning_rate": 0.00018181497174250236,
"loss": 0.5425,
"step": 1560
},
{
"epoch": 24.53125,
"grad_norm": 2.0459985733032227,
"learning_rate": 0.00018158853744437914,
"loss": 0.5486,
"step": 1570
},
{
"epoch": 24.6875,
"grad_norm": 2.2284512519836426,
"learning_rate": 0.00018136084495007872,
"loss": 0.5268,
"step": 1580
},
{
"epoch": 24.84375,
"grad_norm": 2.029223918914795,
"learning_rate": 0.000181131897770901,
"loss": 0.5421,
"step": 1590
},
{
"epoch": 25.0,
"grad_norm": 2.371617078781128,
"learning_rate": 0.00018090169943749476,
"loss": 0.5612,
"step": 1600
},
{
"epoch": 25.15625,
"grad_norm": 2.1253976821899414,
"learning_rate": 0.000180670253499803,
"loss": 0.5142,
"step": 1610
},
{
"epoch": 25.3125,
"grad_norm": 2.6901040077209473,
"learning_rate": 0.00018043756352700846,
"loss": 0.5331,
"step": 1620
},
{
"epoch": 25.46875,
"grad_norm": 2.1748178005218506,
"learning_rate": 0.00018020363310747834,
"loss": 0.5031,
"step": 1630
},
{
"epoch": 25.625,
"grad_norm": 2.2645554542541504,
"learning_rate": 0.00017996846584870908,
"loss": 0.5171,
"step": 1640
},
{
"epoch": 25.78125,
"grad_norm": 2.1726343631744385,
"learning_rate": 0.00017973206537727073,
"loss": 0.5104,
"step": 1650
},
{
"epoch": 25.9375,
"grad_norm": 2.242905855178833,
"learning_rate": 0.000179494435338751,
"loss": 0.5543,
"step": 1660
},
{
"epoch": 26.09375,
"grad_norm": 2.870544195175171,
"learning_rate": 0.0001792555793976991,
"loss": 0.4795,
"step": 1670
},
{
"epoch": 26.25,
"grad_norm": 2.156977891921997,
"learning_rate": 0.00017901550123756906,
"loss": 0.4995,
"step": 1680
},
{
"epoch": 26.40625,
"grad_norm": 2.332562208175659,
"learning_rate": 0.0001787742045606631,
"loss": 0.4938,
"step": 1690
},
{
"epoch": 26.5625,
"grad_norm": 2.302238702774048,
"learning_rate": 0.00017853169308807448,
"loss": 0.5322,
"step": 1700
},
{
"epoch": 26.71875,
"grad_norm": 2.610508680343628,
"learning_rate": 0.00017828797055963018,
"loss": 0.5243,
"step": 1710
},
{
"epoch": 26.875,
"grad_norm": 2.429201126098633,
"learning_rate": 0.000178043040733833,
"loss": 0.4932,
"step": 1720
},
{
"epoch": 27.03125,
"grad_norm": 2.024339199066162,
"learning_rate": 0.00017779690738780387,
"loss": 0.516,
"step": 1730
},
{
"epoch": 27.1875,
"grad_norm": 2.3692007064819336,
"learning_rate": 0.00017754957431722346,
"loss": 0.4796,
"step": 1740
},
{
"epoch": 27.34375,
"grad_norm": 2.3139917850494385,
"learning_rate": 0.0001773010453362737,
"loss": 0.4633,
"step": 1750
},
{
"epoch": 27.5,
"grad_norm": 2.060307502746582,
"learning_rate": 0.00017705132427757895,
"loss": 0.4818,
"step": 1760
},
{
"epoch": 27.65625,
"grad_norm": 2.431605577468872,
"learning_rate": 0.00017680041499214678,
"loss": 0.4941,
"step": 1770
},
{
"epoch": 27.8125,
"grad_norm": 2.228679895401001,
"learning_rate": 0.00017654832134930882,
"loss": 0.4937,
"step": 1780
},
{
"epoch": 27.96875,
"grad_norm": 2.1984705924987793,
"learning_rate": 0.00017629504723666088,
"loss": 0.5022,
"step": 1790
},
{
"epoch": 28.125,
"grad_norm": 2.8257086277008057,
"learning_rate": 0.0001760405965600031,
"loss": 0.4586,
"step": 1800
},
{
"epoch": 28.28125,
"grad_norm": 2.157179594039917,
"learning_rate": 0.0001757849732432797,
"loss": 0.4596,
"step": 1810
},
{
"epoch": 28.4375,
"grad_norm": 2.935502290725708,
"learning_rate": 0.00017552818122851838,
"loss": 0.4741,
"step": 1820
},
{
"epoch": 28.59375,
"grad_norm": 2.6089301109313965,
"learning_rate": 0.0001752702244757697,
"loss": 0.4661,
"step": 1830
},
{
"epoch": 28.75,
"grad_norm": 2.3858306407928467,
"learning_rate": 0.00017501110696304596,
"loss": 0.4942,
"step": 1840
},
{
"epoch": 28.90625,
"grad_norm": 2.238586187362671,
"learning_rate": 0.0001747508326862597,
"loss": 0.461,
"step": 1850
},
{
"epoch": 29.0625,
"grad_norm": 2.3924996852874756,
"learning_rate": 0.00017448940565916222,
"loss": 0.478,
"step": 1860
},
{
"epoch": 29.21875,
"grad_norm": 2.1921300888061523,
"learning_rate": 0.0001742268299132817,
"loss": 0.4361,
"step": 1870
},
{
"epoch": 29.375,
"grad_norm": 2.3493051528930664,
"learning_rate": 0.000173963109497861,
"loss": 0.4497,
"step": 1880
},
{
"epoch": 29.53125,
"grad_norm": 2.3282814025878906,
"learning_rate": 0.0001736982484797951,
"loss": 0.4601,
"step": 1890
},
{
"epoch": 29.6875,
"grad_norm": 2.0696048736572266,
"learning_rate": 0.00017343225094356855,
"loss": 0.4692,
"step": 1900
},
{
"epoch": 29.84375,
"grad_norm": 2.1270716190338135,
"learning_rate": 0.0001731651209911925,
"loss": 0.4467,
"step": 1910
},
{
"epoch": 30.0,
"grad_norm": 2.4350905418395996,
"learning_rate": 0.00017289686274214118,
"loss": 0.4578,
"step": 1920
},
{
"epoch": 30.15625,
"grad_norm": 2.1730527877807617,
"learning_rate": 0.00017262748033328867,
"loss": 0.4147,
"step": 1930
},
{
"epoch": 30.3125,
"grad_norm": 2.2917675971984863,
"learning_rate": 0.00017235697791884494,
"loss": 0.4064,
"step": 1940
},
{
"epoch": 30.46875,
"grad_norm": 2.6113858222961426,
"learning_rate": 0.00017208535967029188,
"loss": 0.4486,
"step": 1950
},
{
"epoch": 30.625,
"grad_norm": 2.4307138919830322,
"learning_rate": 0.00017181262977631888,
"loss": 0.4359,
"step": 1960
},
{
"epoch": 30.78125,
"grad_norm": 2.257676601409912,
"learning_rate": 0.0001715387924427583,
"loss": 0.4759,
"step": 1970
},
{
"epoch": 30.9375,
"grad_norm": 2.196704864501953,
"learning_rate": 0.00017126385189252053,
"loss": 0.4618,
"step": 1980
},
{
"epoch": 31.09375,
"grad_norm": 2.23207950592041,
"learning_rate": 0.00017098781236552903,
"loss": 0.4165,
"step": 1990
},
{
"epoch": 31.25,
"grad_norm": 2.283047914505005,
"learning_rate": 0.00017071067811865476,
"loss": 0.4206,
"step": 2000
},
{
"epoch": 31.40625,
"grad_norm": 2.4019672870635986,
"learning_rate": 0.00017043245342565063,
"loss": 0.4286,
"step": 2010
},
{
"epoch": 31.5625,
"grad_norm": 2.395625591278076,
"learning_rate": 0.0001701531425770856,
"loss": 0.4219,
"step": 2020
},
{
"epoch": 31.71875,
"grad_norm": 2.2177236080169678,
"learning_rate": 0.00016987274988027843,
"loss": 0.4338,
"step": 2030
},
{
"epoch": 31.875,
"grad_norm": 2.526252269744873,
"learning_rate": 0.00016959127965923142,
"loss": 0.4337,
"step": 2040
},
{
"epoch": 32.03125,
"grad_norm": 2.5658786296844482,
"learning_rate": 0.0001693087362545636,
"loss": 0.4374,
"step": 2050
},
{
"epoch": 32.1875,
"grad_norm": 2.5486271381378174,
"learning_rate": 0.00016902512402344373,
"loss": 0.41,
"step": 2060
},
{
"epoch": 32.34375,
"grad_norm": 2.4342575073242188,
"learning_rate": 0.00016874044733952327,
"loss": 0.4051,
"step": 2070
},
{
"epoch": 32.5,
"grad_norm": 2.0787465572357178,
"learning_rate": 0.00016845471059286887,
"loss": 0.4065,
"step": 2080
},
{
"epoch": 32.65625,
"grad_norm": 2.413379430770874,
"learning_rate": 0.00016816791818989466,
"loss": 0.4121,
"step": 2090
},
{
"epoch": 32.8125,
"grad_norm": 2.255675792694092,
"learning_rate": 0.0001678800745532942,
"loss": 0.4199,
"step": 2100
},
{
"epoch": 32.96875,
"grad_norm": 2.7167301177978516,
"learning_rate": 0.00016759118412197247,
"loss": 0.4219,
"step": 2110
},
{
"epoch": 33.125,
"grad_norm": 2.5120389461517334,
"learning_rate": 0.00016730125135097735,
"loss": 0.4126,
"step": 2120
},
{
"epoch": 33.28125,
"grad_norm": 2.2318458557128906,
"learning_rate": 0.0001670102807114308,
"loss": 0.3766,
"step": 2130
},
{
"epoch": 33.4375,
"grad_norm": 2.6558902263641357,
"learning_rate": 0.00016671827669045998,
"loss": 0.3897,
"step": 2140
},
{
"epoch": 33.59375,
"grad_norm": 2.1000514030456543,
"learning_rate": 0.00016642524379112817,
"loss": 0.3809,
"step": 2150
},
{
"epoch": 33.75,
"grad_norm": 2.422773838043213,
"learning_rate": 0.00016613118653236518,
"loss": 0.4301,
"step": 2160
},
{
"epoch": 33.90625,
"grad_norm": 2.6482996940612793,
"learning_rate": 0.00016583610944889774,
"loss": 0.419,
"step": 2170
},
{
"epoch": 34.0625,
"grad_norm": 2.4564175605773926,
"learning_rate": 0.0001655400170911794,
"loss": 0.4099,
"step": 2180
},
{
"epoch": 34.21875,
"grad_norm": 2.3099935054779053,
"learning_rate": 0.0001652429140253207,
"loss": 0.3742,
"step": 2190
},
{
"epoch": 34.375,
"grad_norm": 2.5124337673187256,
"learning_rate": 0.00016494480483301836,
"loss": 0.3817,
"step": 2200
},
{
"epoch": 34.53125,
"grad_norm": 2.3638722896575928,
"learning_rate": 0.000164645694111485,
"loss": 0.3978,
"step": 2210
},
{
"epoch": 34.6875,
"grad_norm": 2.6030874252319336,
"learning_rate": 0.0001643455864733779,
"loss": 0.3908,
"step": 2220
},
{
"epoch": 34.84375,
"grad_norm": 2.0865495204925537,
"learning_rate": 0.0001640444865467281,
"loss": 0.3782,
"step": 2230
},
{
"epoch": 35.0,
"grad_norm": 2.1129355430603027,
"learning_rate": 0.000163742398974869,
"loss": 0.3998,
"step": 2240
},
{
"epoch": 35.15625,
"grad_norm": 2.4917526245117188,
"learning_rate": 0.00016343932841636456,
"loss": 0.3733,
"step": 2250
},
{
"epoch": 35.3125,
"grad_norm": 2.4784963130950928,
"learning_rate": 0.00016313527954493778,
"loss": 0.3557,
"step": 2260
},
{
"epoch": 35.46875,
"grad_norm": 2.292530059814453,
"learning_rate": 0.00016283025704939838,
"loss": 0.3821,
"step": 2270
},
{
"epoch": 35.625,
"grad_norm": 2.420898199081421,
"learning_rate": 0.00016252426563357055,
"loss": 0.382,
"step": 2280
},
{
"epoch": 35.78125,
"grad_norm": 2.4212636947631836,
"learning_rate": 0.0001622173100162204,
"loss": 0.3963,
"step": 2290
},
{
"epoch": 35.9375,
"grad_norm": 2.462033987045288,
"learning_rate": 0.00016190939493098344,
"loss": 0.3786,
"step": 2300
},
{
"epoch": 36.09375,
"grad_norm": 2.4041311740875244,
"learning_rate": 0.000161600525126291,
"loss": 0.3625,
"step": 2310
},
{
"epoch": 36.25,
"grad_norm": 2.265742540359497,
"learning_rate": 0.00016129070536529766,
"loss": 0.36,
"step": 2320
},
{
"epoch": 36.40625,
"grad_norm": 2.326496124267578,
"learning_rate": 0.0001609799404258074,
"loss": 0.3655,
"step": 2330
},
{
"epoch": 36.5625,
"grad_norm": 2.161198139190674,
"learning_rate": 0.00016066823510019998,
"loss": 0.3545,
"step": 2340
},
{
"epoch": 36.71875,
"grad_norm": 2.603285312652588,
"learning_rate": 0.00016035559419535716,
"loss": 0.3748,
"step": 2350
},
{
"epoch": 36.875,
"grad_norm": 2.2837438583374023,
"learning_rate": 0.00016004202253258842,
"loss": 0.3791,
"step": 2360
},
{
"epoch": 37.03125,
"grad_norm": 2.172354221343994,
"learning_rate": 0.00015972752494755672,
"loss": 0.3756,
"step": 2370
},
{
"epoch": 37.1875,
"grad_norm": 2.015866756439209,
"learning_rate": 0.00015941210629020388,
"loss": 0.344,
"step": 2380
},
{
"epoch": 37.34375,
"grad_norm": 2.2509257793426514,
"learning_rate": 0.00015909577142467573,
"loss": 0.3454,
"step": 2390
},
{
"epoch": 37.5,
"grad_norm": 2.574324131011963,
"learning_rate": 0.00015877852522924732,
"loss": 0.3518,
"step": 2400
},
{
"epoch": 37.65625,
"grad_norm": 2.231295108795166,
"learning_rate": 0.00015846037259624736,
"loss": 0.3607,
"step": 2410
},
{
"epoch": 37.8125,
"grad_norm": 2.4786765575408936,
"learning_rate": 0.00015814131843198308,
"loss": 0.3568,
"step": 2420
},
{
"epoch": 37.96875,
"grad_norm": 2.8630552291870117,
"learning_rate": 0.0001578213676566643,
"loss": 0.3787,
"step": 2430
},
{
"epoch": 38.125,
"grad_norm": 2.386781692504883,
"learning_rate": 0.00015750052520432787,
"loss": 0.3387,
"step": 2440
},
{
"epoch": 38.28125,
"grad_norm": 2.238036632537842,
"learning_rate": 0.00015717879602276122,
"loss": 0.3514,
"step": 2450
},
{
"epoch": 38.4375,
"grad_norm": 2.1005611419677734,
"learning_rate": 0.0001568561850734264,
"loss": 0.3522,
"step": 2460
},
{
"epoch": 38.59375,
"grad_norm": 2.509990930557251,
"learning_rate": 0.00015653269733138328,
"loss": 0.3429,
"step": 2470
},
{
"epoch": 38.75,
"grad_norm": 2.3400230407714844,
"learning_rate": 0.00015620833778521307,
"loss": 0.3429,
"step": 2480
},
{
"epoch": 38.90625,
"grad_norm": 2.3299190998077393,
"learning_rate": 0.00015588311143694116,
"loss": 0.3623,
"step": 2490
},
{
"epoch": 39.0625,
"grad_norm": 2.1369211673736572,
"learning_rate": 0.00015555702330196023,
"loss": 0.3222,
"step": 2500
},
{
"epoch": 39.21875,
"grad_norm": 2.5863115787506104,
"learning_rate": 0.0001552300784089527,
"loss": 0.3244,
"step": 2510
},
{
"epoch": 39.375,
"grad_norm": 2.0398120880126953,
"learning_rate": 0.0001549022817998132,
"loss": 0.336,
"step": 2520
},
{
"epoch": 39.53125,
"grad_norm": 2.0763356685638428,
"learning_rate": 0.000154573638529571,
"loss": 0.3179,
"step": 2530
},
{
"epoch": 39.6875,
"grad_norm": 2.39857816696167,
"learning_rate": 0.00015424415366631188,
"loss": 0.3508,
"step": 2540
},
{
"epoch": 39.84375,
"grad_norm": 2.5503716468811035,
"learning_rate": 0.00015391383229110007,
"loss": 0.3489,
"step": 2550
},
{
"epoch": 40.0,
"grad_norm": 2.2811882495880127,
"learning_rate": 0.00015358267949789966,
"loss": 0.3612,
"step": 2560
},
{
"epoch": 40.15625,
"grad_norm": 1.873091459274292,
"learning_rate": 0.00015325070039349655,
"loss": 0.3116,
"step": 2570
},
{
"epoch": 40.3125,
"grad_norm": 2.053217887878418,
"learning_rate": 0.00015291790009741907,
"loss": 0.3229,
"step": 2580
},
{
"epoch": 40.46875,
"grad_norm": 2.6282942295074463,
"learning_rate": 0.00015258428374185956,
"loss": 0.3257,
"step": 2590
},
{
"epoch": 40.625,
"grad_norm": 2.7164459228515625,
"learning_rate": 0.0001522498564715949,
"loss": 0.3414,
"step": 2600
},
{
"epoch": 40.78125,
"grad_norm": 1.9660325050354004,
"learning_rate": 0.0001519146234439073,
"loss": 0.3296,
"step": 2610
},
{
"epoch": 40.9375,
"grad_norm": 2.421337842941284,
"learning_rate": 0.00015157858982850475,
"loss": 0.3425,
"step": 2620
},
{
"epoch": 41.09375,
"grad_norm": 2.4952664375305176,
"learning_rate": 0.00015124176080744134,
"loss": 0.3221,
"step": 2630
},
{
"epoch": 41.25,
"grad_norm": 2.1518049240112305,
"learning_rate": 0.00015090414157503714,
"loss": 0.3121,
"step": 2640
},
{
"epoch": 41.40625,
"grad_norm": 2.6512484550476074,
"learning_rate": 0.00015056573733779848,
"loss": 0.321,
"step": 2650
},
{
"epoch": 41.5625,
"grad_norm": 2.3250184059143066,
"learning_rate": 0.00015022655331433727,
"loss": 0.3237,
"step": 2660
},
{
"epoch": 41.71875,
"grad_norm": 2.1326417922973633,
"learning_rate": 0.00014988659473529075,
"loss": 0.3228,
"step": 2670
},
{
"epoch": 41.875,
"grad_norm": 2.113276958465576,
"learning_rate": 0.00014954586684324078,
"loss": 0.3161,
"step": 2680
},
{
"epoch": 42.03125,
"grad_norm": 2.1591405868530273,
"learning_rate": 0.0001492043748926329,
"loss": 0.3217,
"step": 2690
},
{
"epoch": 42.1875,
"grad_norm": 2.3071820735931396,
"learning_rate": 0.00014886212414969553,
"loss": 0.3088,
"step": 2700
},
{
"epoch": 42.34375,
"grad_norm": 2.0190141201019287,
"learning_rate": 0.0001485191198923584,
"loss": 0.3015,
"step": 2710
},
{
"epoch": 42.5,
"grad_norm": 2.286022901535034,
"learning_rate": 0.00014817536741017152,
"loss": 0.3204,
"step": 2720
},
{
"epoch": 42.65625,
"grad_norm": 2.2908082008361816,
"learning_rate": 0.00014783087200422344,
"loss": 0.3051,
"step": 2730
},
{
"epoch": 42.8125,
"grad_norm": 2.825416326522827,
"learning_rate": 0.00014748563898705946,
"loss": 0.3305,
"step": 2740
},
{
"epoch": 42.96875,
"grad_norm": 2.2812416553497314,
"learning_rate": 0.0001471396736825998,
"loss": 0.3176,
"step": 2750
},
{
"epoch": 43.125,
"grad_norm": 2.5278165340423584,
"learning_rate": 0.00014679298142605734,
"loss": 0.316,
"step": 2760
},
{
"epoch": 43.28125,
"grad_norm": 2.1231706142425537,
"learning_rate": 0.00014644556756385565,
"loss": 0.2954,
"step": 2770
},
{
"epoch": 43.4375,
"grad_norm": 2.234170913696289,
"learning_rate": 0.00014609743745354624,
"loss": 0.3081,
"step": 2780
},
{
"epoch": 43.59375,
"grad_norm": 2.217315435409546,
"learning_rate": 0.00014574859646372605,
"loss": 0.2936,
"step": 2790
},
{
"epoch": 43.75,
"grad_norm": 2.164661407470703,
"learning_rate": 0.00014539904997395468,
"loss": 0.3105,
"step": 2800
},
{
"epoch": 43.90625,
"grad_norm": 2.3690619468688965,
"learning_rate": 0.00014504880337467145,
"loss": 0.321,
"step": 2810
},
{
"epoch": 44.0625,
"grad_norm": 2.202134370803833,
"learning_rate": 0.00014469786206711214,
"loss": 0.2908,
"step": 2820
},
{
"epoch": 44.21875,
"grad_norm": 2.483181953430176,
"learning_rate": 0.00014434623146322587,
"loss": 0.2933,
"step": 2830
},
{
"epoch": 44.375,
"grad_norm": 2.187701463699341,
"learning_rate": 0.00014399391698559152,
"loss": 0.3073,
"step": 2840
},
{
"epoch": 44.53125,
"grad_norm": 2.346254587173462,
"learning_rate": 0.0001436409240673342,
"loss": 0.3062,
"step": 2850
},
{
"epoch": 44.6875,
"grad_norm": 2.4697351455688477,
"learning_rate": 0.00014328725815204144,
"loss": 0.2915,
"step": 2860
},
{
"epoch": 44.84375,
"grad_norm": 2.093778371810913,
"learning_rate": 0.00014293292469367902,
"loss": 0.3092,
"step": 2870
},
{
"epoch": 45.0,
"grad_norm": 2.5177674293518066,
"learning_rate": 0.00014257792915650728,
"loss": 0.3092,
"step": 2880
},
{
"epoch": 45.15625,
"grad_norm": 1.9424992799758911,
"learning_rate": 0.00014222227701499656,
"loss": 0.2945,
"step": 2890
},
{
"epoch": 45.3125,
"grad_norm": 2.399775981903076,
"learning_rate": 0.0001418659737537428,
"loss": 0.2849,
"step": 2900
},
{
"epoch": 45.46875,
"grad_norm": 2.5698862075805664,
"learning_rate": 0.00014150902486738314,
"loss": 0.293,
"step": 2910
},
{
"epoch": 45.625,
"grad_norm": 2.2189462184906006,
"learning_rate": 0.00014115143586051088,
"loss": 0.3007,
"step": 2920
},
{
"epoch": 45.78125,
"grad_norm": 2.7480437755584717,
"learning_rate": 0.00014079321224759093,
"loss": 0.2906,
"step": 2930
},
{
"epoch": 45.9375,
"grad_norm": 2.392237663269043,
"learning_rate": 0.00014043435955287452,
"loss": 0.2981,
"step": 2940
},
{
"epoch": 46.09375,
"grad_norm": 2.056471109390259,
"learning_rate": 0.0001400748833103141,
"loss": 0.2773,
"step": 2950
},
{
"epoch": 46.25,
"grad_norm": 1.9949535131454468,
"learning_rate": 0.00013971478906347806,
"loss": 0.2852,
"step": 2960
},
{
"epoch": 46.40625,
"grad_norm": 2.4332056045532227,
"learning_rate": 0.00013935408236546515,
"loss": 0.2846,
"step": 2970
},
{
"epoch": 46.5625,
"grad_norm": 2.5237514972686768,
"learning_rate": 0.00013899276877881884,
"loss": 0.2864,
"step": 2980
},
{
"epoch": 46.71875,
"grad_norm": 2.497544050216675,
"learning_rate": 0.00013863085387544162,
"loss": 0.2883,
"step": 2990
},
{
"epoch": 46.875,
"grad_norm": 2.5128774642944336,
"learning_rate": 0.000138268343236509,
"loss": 0.2993,
"step": 3000
},
{
"epoch": 47.03125,
"grad_norm": 1.5527375936508179,
"learning_rate": 0.0001379052424523835,
"loss": 0.2926,
"step": 3010
},
{
"epoch": 47.1875,
"grad_norm": 2.4135637283325195,
"learning_rate": 0.00013754155712252832,
"loss": 0.2769,
"step": 3020
},
{
"epoch": 47.34375,
"grad_norm": 2.0572760105133057,
"learning_rate": 0.00013717729285542122,
"loss": 0.2841,
"step": 3030
},
{
"epoch": 47.5,
"grad_norm": 2.7703893184661865,
"learning_rate": 0.00013681245526846783,
"loss": 0.2704,
"step": 3040
},
{
"epoch": 47.65625,
"grad_norm": 2.430634021759033,
"learning_rate": 0.000136447049987915,
"loss": 0.2842,
"step": 3050
},
{
"epoch": 47.8125,
"grad_norm": 2.4779250621795654,
"learning_rate": 0.0001360810826487642,
"loss": 0.2922,
"step": 3060
},
{
"epoch": 47.96875,
"grad_norm": 2.3544280529022217,
"learning_rate": 0.00013571455889468457,
"loss": 0.2909,
"step": 3070
},
{
"epoch": 48.125,
"grad_norm": 1.854184627532959,
"learning_rate": 0.00013534748437792573,
"loss": 0.2672,
"step": 3080
},
{
"epoch": 48.28125,
"grad_norm": 2.352564573287964,
"learning_rate": 0.00013497986475923088,
"loss": 0.2725,
"step": 3090
},
{
"epoch": 48.4375,
"grad_norm": 2.332357883453369,
"learning_rate": 0.0001346117057077493,
"loss": 0.2736,
"step": 3100
},
{
"epoch": 48.59375,
"grad_norm": 1.9822356700897217,
"learning_rate": 0.000134243012900949,
"loss": 0.2735,
"step": 3110
},
{
"epoch": 48.75,
"grad_norm": 2.2452166080474854,
"learning_rate": 0.00013387379202452917,
"loss": 0.2793,
"step": 3120
},
{
"epoch": 48.90625,
"grad_norm": 1.6397210359573364,
"learning_rate": 0.0001335040487723324,
"loss": 0.2823,
"step": 3130
},
{
"epoch": 49.0625,
"grad_norm": 2.0738863945007324,
"learning_rate": 0.0001331337888462571,
"loss": 0.267,
"step": 3140
},
{
"epoch": 49.21875,
"grad_norm": 1.7931309938430786,
"learning_rate": 0.00013276301795616936,
"loss": 0.2685,
"step": 3150
},
{
"epoch": 49.375,
"grad_norm": 1.9530521631240845,
"learning_rate": 0.00013239174181981495,
"loss": 0.2606,
"step": 3160
},
{
"epoch": 49.53125,
"grad_norm": 2.3546273708343506,
"learning_rate": 0.00013201996616273118,
"loss": 0.269,
"step": 3170
},
{
"epoch": 49.6875,
"grad_norm": 2.2347261905670166,
"learning_rate": 0.00013164769671815862,
"loss": 0.274,
"step": 3180
},
{
"epoch": 49.84375,
"grad_norm": 2.118546485900879,
"learning_rate": 0.0001312749392269526,
"loss": 0.2775,
"step": 3190
},
{
"epoch": 50.0,
"grad_norm": 2.236036777496338,
"learning_rate": 0.00013090169943749476,
"loss": 0.28,
"step": 3200
},
{
"epoch": 50.15625,
"grad_norm": 1.9887173175811768,
"learning_rate": 0.00013052798310560438,
"loss": 0.2666,
"step": 3210
},
{
"epoch": 50.3125,
"grad_norm": 2.284052610397339,
"learning_rate": 0.00013015379599444957,
"loss": 0.2616,
"step": 3220
},
{
"epoch": 50.46875,
"grad_norm": 1.860603928565979,
"learning_rate": 0.00012977914387445855,
"loss": 0.2665,
"step": 3230
},
{
"epoch": 50.625,
"grad_norm": 1.7840187549591064,
"learning_rate": 0.0001294040325232304,
"loss": 0.2648,
"step": 3240
},
{
"epoch": 50.78125,
"grad_norm": 2.2570018768310547,
"learning_rate": 0.00012902846772544624,
"loss": 0.258,
"step": 3250
},
{
"epoch": 50.9375,
"grad_norm": 1.921676516532898,
"learning_rate": 0.00012865245527277986,
"loss": 0.269,
"step": 3260
},
{
"epoch": 51.09375,
"grad_norm": 2.2320163249969482,
"learning_rate": 0.00012827600096380833,
"loss": 0.2676,
"step": 3270
},
{
"epoch": 51.25,
"grad_norm": 1.9506572484970093,
"learning_rate": 0.00012789911060392294,
"loss": 0.2564,
"step": 3280
},
{
"epoch": 51.40625,
"grad_norm": 1.990869164466858,
"learning_rate": 0.00012752179000523917,
"loss": 0.2611,
"step": 3290
},
{
"epoch": 51.5625,
"grad_norm": 2.245129108428955,
"learning_rate": 0.00012714404498650743,
"loss": 0.2619,
"step": 3300
},
{
"epoch": 51.71875,
"grad_norm": 2.441230535507202,
"learning_rate": 0.00012676588137302327,
"loss": 0.2638,
"step": 3310
},
{
"epoch": 51.875,
"grad_norm": 1.967659831047058,
"learning_rate": 0.0001263873049965373,
"loss": 0.2574,
"step": 3320
},
{
"epoch": 52.03125,
"grad_norm": 1.8834177255630493,
"learning_rate": 0.00012600832169516567,
"loss": 0.2602,
"step": 3330
},
{
"epoch": 52.1875,
"grad_norm": 2.202791452407837,
"learning_rate": 0.00012562893731329967,
"loss": 0.2369,
"step": 3340
},
{
"epoch": 52.34375,
"grad_norm": 1.9373564720153809,
"learning_rate": 0.0001252491577015158,
"loss": 0.2615,
"step": 3350
},
{
"epoch": 52.5,
"grad_norm": 1.8653074502944946,
"learning_rate": 0.0001248689887164855,
"loss": 0.2513,
"step": 3360
},
{
"epoch": 52.65625,
"grad_norm": 2.0396597385406494,
"learning_rate": 0.00012448843622088476,
"loss": 0.2605,
"step": 3370
},
{
"epoch": 52.8125,
"grad_norm": 1.9073066711425781,
"learning_rate": 0.00012410750608330388,
"loss": 0.2611,
"step": 3380
},
{
"epoch": 52.96875,
"grad_norm": 2.09374737739563,
"learning_rate": 0.00012372620417815678,
"loss": 0.2665,
"step": 3390
},
{
"epoch": 53.125,
"grad_norm": 2.1771106719970703,
"learning_rate": 0.00012334453638559057,
"loss": 0.2334,
"step": 3400
},
{
"epoch": 53.28125,
"grad_norm": 2.0365121364593506,
"learning_rate": 0.0001229625085913947,
"loss": 0.2471,
"step": 3410
},
{
"epoch": 53.4375,
"grad_norm": 2.06864595413208,
"learning_rate": 0.0001225801266869104,
"loss": 0.2585,
"step": 3420
},
{
"epoch": 53.59375,
"grad_norm": 1.9580702781677246,
"learning_rate": 0.0001221973965689396,
"loss": 0.2543,
"step": 3430
},
{
"epoch": 53.75,
"grad_norm": 1.988387107849121,
"learning_rate": 0.00012181432413965428,
"loss": 0.2428,
"step": 3440
},
{
"epoch": 53.90625,
"grad_norm": 2.0155484676361084,
"learning_rate": 0.00012143091530650508,
"loss": 0.2635,
"step": 3450
},
{
"epoch": 54.0625,
"grad_norm": 1.7660094499588013,
"learning_rate": 0.00012104717598213056,
"loss": 0.2456,
"step": 3460
},
{
"epoch": 54.21875,
"grad_norm": 2.1635379791259766,
"learning_rate": 0.00012066311208426581,
"loss": 0.2319,
"step": 3470
},
{
"epoch": 54.375,
"grad_norm": 1.7314647436141968,
"learning_rate": 0.00012027872953565125,
"loss": 0.2486,
"step": 3480
},
{
"epoch": 54.53125,
"grad_norm": 1.8251374959945679,
"learning_rate": 0.00011989403426394125,
"loss": 0.2429,
"step": 3490
},
{
"epoch": 54.6875,
"grad_norm": 2.038282632827759,
"learning_rate": 0.00011950903220161285,
"loss": 0.249,
"step": 3500
},
{
"epoch": 54.84375,
"grad_norm": 2.0042312145233154,
"learning_rate": 0.00011912372928587406,
"loss": 0.2563,
"step": 3510
},
{
"epoch": 55.0,
"grad_norm": 2.0749430656433105,
"learning_rate": 0.00011873813145857249,
"loss": 0.2583,
"step": 3520
},
{
"epoch": 55.15625,
"grad_norm": 1.9021109342575073,
"learning_rate": 0.00011835224466610366,
"loss": 0.2374,
"step": 3530
},
{
"epoch": 55.3125,
"grad_norm": 2.4008102416992188,
"learning_rate": 0.00011796607485931928,
"loss": 0.2435,
"step": 3540
},
{
"epoch": 55.46875,
"grad_norm": 2.771571397781372,
"learning_rate": 0.00011757962799343547,
"loss": 0.2487,
"step": 3550
},
{
"epoch": 55.625,
"grad_norm": 1.8457822799682617,
"learning_rate": 0.00011719291002794096,
"loss": 0.2355,
"step": 3560
},
{
"epoch": 55.78125,
"grad_norm": 1.9695429801940918,
"learning_rate": 0.0001168059269265052,
"loss": 0.2493,
"step": 3570
},
{
"epoch": 55.9375,
"grad_norm": 2.004483461380005,
"learning_rate": 0.0001164186846568863,
"loss": 0.2493,
"step": 3580
},
{
"epoch": 56.09375,
"grad_norm": 1.6778466701507568,
"learning_rate": 0.00011603118919083913,
"loss": 0.2328,
"step": 3590
},
{
"epoch": 56.25,
"grad_norm": 1.9205918312072754,
"learning_rate": 0.0001156434465040231,
"loss": 0.2421,
"step": 3600
},
{
"epoch": 56.40625,
"grad_norm": 2.071448802947998,
"learning_rate": 0.0001152554625759101,
"loss": 0.2392,
"step": 3610
},
{
"epoch": 56.5625,
"grad_norm": 1.8244534730911255,
"learning_rate": 0.00011486724338969232,
"loss": 0.2348,
"step": 3620
},
{
"epoch": 56.71875,
"grad_norm": 2.3694911003112793,
"learning_rate": 0.00011447879493218979,
"loss": 0.2471,
"step": 3630
},
{
"epoch": 56.875,
"grad_norm": 1.7422678470611572,
"learning_rate": 0.00011409012319375827,
"loss": 0.244,
"step": 3640
},
{
"epoch": 57.03125,
"grad_norm": 1.80690336227417,
"learning_rate": 0.00011370123416819682,
"loss": 0.2417,
"step": 3650
},
{
"epoch": 57.1875,
"grad_norm": 1.9183917045593262,
"learning_rate": 0.00011331213385265524,
"loss": 0.2347,
"step": 3660
},
{
"epoch": 57.34375,
"grad_norm": 2.4743127822875977,
"learning_rate": 0.00011292282824754177,
"loss": 0.2265,
"step": 3670
},
{
"epoch": 57.5,
"grad_norm": 2.031683921813965,
"learning_rate": 0.00011253332335643043,
"loss": 0.2284,
"step": 3680
},
{
"epoch": 57.65625,
"grad_norm": 2.3016927242279053,
"learning_rate": 0.00011214362518596848,
"loss": 0.2426,
"step": 3690
},
{
"epoch": 57.8125,
"grad_norm": 2.0599136352539062,
"learning_rate": 0.00011175373974578378,
"loss": 0.2408,
"step": 3700
},
{
"epoch": 57.96875,
"grad_norm": 1.9112536907196045,
"learning_rate": 0.00011136367304839212,
"loss": 0.2428,
"step": 3710
},
{
"epoch": 58.125,
"grad_norm": 1.677746057510376,
"learning_rate": 0.00011097343110910452,
"loss": 0.2246,
"step": 3720
},
{
"epoch": 58.28125,
"grad_norm": 2.0791244506835938,
"learning_rate": 0.00011058301994593448,
"loss": 0.2356,
"step": 3730
},
{
"epoch": 58.4375,
"grad_norm": 1.9308363199234009,
"learning_rate": 0.000110192445579505,
"loss": 0.2258,
"step": 3740
},
{
"epoch": 58.59375,
"grad_norm": 1.9701329469680786,
"learning_rate": 0.0001098017140329561,
"loss": 0.2416,
"step": 3750
},
{
"epoch": 58.75,
"grad_norm": 1.7241312265396118,
"learning_rate": 0.00010941083133185146,
"loss": 0.2317,
"step": 3760
},
{
"epoch": 58.90625,
"grad_norm": 1.8373721837997437,
"learning_rate": 0.00010901980350408592,
"loss": 0.2345,
"step": 3770
},
{
"epoch": 59.0625,
"grad_norm": 2.176316022872925,
"learning_rate": 0.00010862863657979237,
"loss": 0.2296,
"step": 3780
},
{
"epoch": 59.21875,
"grad_norm": 1.9734270572662354,
"learning_rate": 0.00010823733659124857,
"loss": 0.2241,
"step": 3790
},
{
"epoch": 59.375,
"grad_norm": 2.1557183265686035,
"learning_rate": 0.0001078459095727845,
"loss": 0.2315,
"step": 3800
},
{
"epoch": 59.53125,
"grad_norm": 2.2887086868286133,
"learning_rate": 0.00010745436156068897,
"loss": 0.2287,
"step": 3810
},
{
"epoch": 59.6875,
"grad_norm": 1.7714214324951172,
"learning_rate": 0.00010706269859311669,
"loss": 0.2373,
"step": 3820
},
{
"epoch": 59.84375,
"grad_norm": 1.8595536947250366,
"learning_rate": 0.00010667092670999513,
"loss": 0.2343,
"step": 3830
},
{
"epoch": 60.0,
"grad_norm": 1.7251347303390503,
"learning_rate": 0.00010627905195293135,
"loss": 0.2254,
"step": 3840
},
{
"epoch": 60.15625,
"grad_norm": 2.3099634647369385,
"learning_rate": 0.0001058870803651189,
"loss": 0.2238,
"step": 3850
},
{
"epoch": 60.3125,
"grad_norm": 2.2596757411956787,
"learning_rate": 0.0001054950179912446,
"loss": 0.2289,
"step": 3860
},
{
"epoch": 60.46875,
"grad_norm": 1.6570988893508911,
"learning_rate": 0.00010510287087739517,
"loss": 0.2282,
"step": 3870
},
{
"epoch": 60.625,
"grad_norm": 2.195847749710083,
"learning_rate": 0.00010471064507096426,
"loss": 0.2286,
"step": 3880
},
{
"epoch": 60.78125,
"grad_norm": 2.108572006225586,
"learning_rate": 0.00010431834662055904,
"loss": 0.2246,
"step": 3890
},
{
"epoch": 60.9375,
"grad_norm": 1.8193440437316895,
"learning_rate": 0.00010392598157590688,
"loss": 0.2255,
"step": 3900
},
{
"epoch": 61.09375,
"grad_norm": 1.710485577583313,
"learning_rate": 0.00010353355598776214,
"loss": 0.222,
"step": 3910
},
{
"epoch": 61.25,
"grad_norm": 2.229994773864746,
"learning_rate": 0.00010314107590781284,
"loss": 0.2205,
"step": 3920
},
{
"epoch": 61.40625,
"grad_norm": 1.7697498798370361,
"learning_rate": 0.00010274854738858735,
"loss": 0.2208,
"step": 3930
},
{
"epoch": 61.5625,
"grad_norm": 2.168346405029297,
"learning_rate": 0.00010235597648336104,
"loss": 0.2235,
"step": 3940
},
{
"epoch": 61.71875,
"grad_norm": 1.8804610967636108,
"learning_rate": 0.00010196336924606283,
"loss": 0.2284,
"step": 3950
},
{
"epoch": 61.875,
"grad_norm": 1.934240698814392,
"learning_rate": 0.00010157073173118208,
"loss": 0.2255,
"step": 3960
},
{
"epoch": 62.03125,
"grad_norm": 1.456990361213684,
"learning_rate": 0.00010117806999367492,
"loss": 0.2226,
"step": 3970
},
{
"epoch": 62.1875,
"grad_norm": 1.7687050104141235,
"learning_rate": 0.00010078539008887114,
"loss": 0.2146,
"step": 3980
},
{
"epoch": 62.34375,
"grad_norm": 1.7442452907562256,
"learning_rate": 0.0001003926980723806,
"loss": 0.2209,
"step": 3990
},
{
"epoch": 62.5,
"grad_norm": 1.596010684967041,
"learning_rate": 0.0001,
"loss": 0.2199,
"step": 4000
},
{
"epoch": 62.65625,
"grad_norm": 1.7170418500900269,
"learning_rate": 9.960730192761942e-05,
"loss": 0.2191,
"step": 4010
},
{
"epoch": 62.8125,
"grad_norm": 1.9422141313552856,
"learning_rate": 9.921460991112891e-05,
"loss": 0.2157,
"step": 4020
},
{
"epoch": 62.96875,
"grad_norm": 1.6883125305175781,
"learning_rate": 9.882193000632506e-05,
"loss": 0.2239,
"step": 4030
},
{
"epoch": 63.125,
"grad_norm": 1.9851441383361816,
"learning_rate": 9.842926826881796e-05,
"loss": 0.2146,
"step": 4040
},
{
"epoch": 63.28125,
"grad_norm": 1.519945740699768,
"learning_rate": 9.807589324283746e-05,
"loss": 0.217,
"step": 4050
},
{
"epoch": 63.4375,
"grad_norm": 1.9202287197113037,
"learning_rate": 9.768328270532056e-05,
"loss": 0.2126,
"step": 4060
},
{
"epoch": 63.59375,
"grad_norm": 2.1956114768981934,
"learning_rate": 9.729070789445026e-05,
"loss": 0.2223,
"step": 4070
},
{
"epoch": 63.75,
"grad_norm": 1.8512564897537231,
"learning_rate": 9.68981748642158e-05,
"loss": 0.2162,
"step": 4080
},
{
"epoch": 63.90625,
"grad_norm": 2.0431482791900635,
"learning_rate": 9.650568966796203e-05,
"loss": 0.2195,
"step": 4090
},
{
"epoch": 64.0625,
"grad_norm": 1.8963639736175537,
"learning_rate": 9.61132583582963e-05,
"loss": 0.2226,
"step": 4100
},
{
"epoch": 64.21875,
"grad_norm": 1.5588335990905762,
"learning_rate": 9.572088698699478e-05,
"loss": 0.2116,
"step": 4110
},
{
"epoch": 64.375,
"grad_norm": 1.9125704765319824,
"learning_rate": 9.532858160490945e-05,
"loss": 0.2126,
"step": 4120
},
{
"epoch": 64.53125,
"grad_norm": 1.8598657846450806,
"learning_rate": 9.493634826187461e-05,
"loss": 0.214,
"step": 4130
},
{
"epoch": 64.6875,
"grad_norm": 1.7466522455215454,
"learning_rate": 9.454419300661355e-05,
"loss": 0.221,
"step": 4140
},
{
"epoch": 64.84375,
"grad_norm": 1.8226146697998047,
"learning_rate": 9.415212188664553e-05,
"loss": 0.2161,
"step": 4150
},
{
"epoch": 65.0,
"grad_norm": 1.7260197401046753,
"learning_rate": 9.376014094819217e-05,
"loss": 0.2147,
"step": 4160
},
{
"epoch": 65.15625,
"grad_norm": 1.7944152355194092,
"learning_rate": 9.336825623608448e-05,
"loss": 0.2085,
"step": 4170
},
{
"epoch": 65.3125,
"grad_norm": 2.3139121532440186,
"learning_rate": 9.297647379366944e-05,
"loss": 0.2112,
"step": 4180
},
{
"epoch": 65.46875,
"grad_norm": 1.478004813194275,
"learning_rate": 9.258479966271703e-05,
"loss": 0.2128,
"step": 4190
},
{
"epoch": 65.625,
"grad_norm": 1.8647255897521973,
"learning_rate": 9.219323988332688e-05,
"loss": 0.2055,
"step": 4200
},
{
"epoch": 65.78125,
"grad_norm": 1.883954405784607,
"learning_rate": 9.180180049383518e-05,
"loss": 0.2131,
"step": 4210
},
{
"epoch": 65.9375,
"grad_norm": 1.5575131177902222,
"learning_rate": 9.141048753072144e-05,
"loss": 0.2177,
"step": 4220
},
{
"epoch": 66.09375,
"grad_norm": 2.0962812900543213,
"learning_rate": 9.101930702851577e-05,
"loss": 0.2128,
"step": 4230
},
{
"epoch": 66.25,
"grad_norm": 1.675104022026062,
"learning_rate": 9.062826501970535e-05,
"loss": 0.2096,
"step": 4240
},
{
"epoch": 66.40625,
"grad_norm": 1.7287991046905518,
"learning_rate": 9.023736753464172e-05,
"loss": 0.2074,
"step": 4250
},
{
"epoch": 66.5625,
"grad_norm": 1.6228628158569336,
"learning_rate": 8.984662060144763e-05,
"loss": 0.2094,
"step": 4260
},
{
"epoch": 66.71875,
"grad_norm": 2.205287218093872,
"learning_rate": 8.94560302459242e-05,
"loss": 0.2129,
"step": 4270
},
{
"epoch": 66.875,
"grad_norm": 1.7029268741607666,
"learning_rate": 8.90656024914579e-05,
"loss": 0.2164,
"step": 4280
},
{
"epoch": 67.03125,
"grad_norm": 1.6045933961868286,
"learning_rate": 8.86753433589277e-05,
"loss": 0.2124,
"step": 4290
},
{
"epoch": 67.1875,
"grad_norm": 2.085911273956299,
"learning_rate": 8.82852588666121e-05,
"loss": 0.2074,
"step": 4300
},
{
"epoch": 67.34375,
"grad_norm": 2.3079230785369873,
"learning_rate": 8.789535503009667e-05,
"loss": 0.2066,
"step": 4310
},
{
"epoch": 67.5,
"grad_norm": 2.0734775066375732,
"learning_rate": 8.750563786218081e-05,
"loss": 0.2086,
"step": 4320
},
{
"epoch": 67.65625,
"grad_norm": 2.025387763977051,
"learning_rate": 8.711611337278537e-05,
"loss": 0.207,
"step": 4330
},
{
"epoch": 67.8125,
"grad_norm": 1.9657328128814697,
"learning_rate": 8.672678756885984e-05,
"loss": 0.2118,
"step": 4340
},
{
"epoch": 67.96875,
"grad_norm": 2.05079984664917,
"learning_rate": 8.633766645428974e-05,
"loss": 0.2097,
"step": 4350
},
{
"epoch": 68.125,
"grad_norm": 1.5424699783325195,
"learning_rate": 8.594875602980407e-05,
"loss": 0.2054,
"step": 4360
},
{
"epoch": 68.28125,
"grad_norm": 1.584788203239441,
"learning_rate": 8.55600622928826e-05,
"loss": 0.2083,
"step": 4370
},
{
"epoch": 68.4375,
"grad_norm": 1.837462067604065,
"learning_rate": 8.517159123766363e-05,
"loss": 0.2065,
"step": 4380
},
{
"epoch": 68.59375,
"grad_norm": 2.0221145153045654,
"learning_rate": 8.478334885485137e-05,
"loss": 0.202,
"step": 4390
},
{
"epoch": 68.75,
"grad_norm": 1.71314537525177,
"learning_rate": 8.439534113162368e-05,
"loss": 0.2086,
"step": 4400
},
{
"epoch": 68.90625,
"grad_norm": 1.6273434162139893,
"learning_rate": 8.40075740515396e-05,
"loss": 0.2107,
"step": 4410
},
{
"epoch": 69.0625,
"grad_norm": 1.656145691871643,
"learning_rate": 8.362005359444727e-05,
"loss": 0.2023,
"step": 4420
},
{
"epoch": 69.21875,
"grad_norm": 1.59807550907135,
"learning_rate": 8.323278573639139e-05,
"loss": 0.2037,
"step": 4430
},
{
"epoch": 69.375,
"grad_norm": 1.8091068267822266,
"learning_rate": 8.284577644952155e-05,
"loss": 0.2046,
"step": 4440
},
{
"epoch": 69.53125,
"grad_norm": 1.5284535884857178,
"learning_rate": 8.245903170199962e-05,
"loss": 0.2066,
"step": 4450
},
{
"epoch": 69.6875,
"grad_norm": 1.7114146947860718,
"learning_rate": 8.207255745790805e-05,
"loss": 0.2002,
"step": 4460
},
{
"epoch": 69.84375,
"grad_norm": 1.703376293182373,
"learning_rate": 8.168635967715776e-05,
"loss": 0.2054,
"step": 4470
},
{
"epoch": 70.0,
"grad_norm": 1.896020531654358,
"learning_rate": 8.130044431539634e-05,
"loss": 0.2059,
"step": 4480
},
{
"epoch": 70.15625,
"grad_norm": 1.8438818454742432,
"learning_rate": 8.091481732391599e-05,
"loss": 0.1955,
"step": 4490
},
{
"epoch": 70.3125,
"grad_norm": 1.7322238683700562,
"learning_rate": 8.052948464956205e-05,
"loss": 0.2014,
"step": 4500
},
{
"epoch": 70.46875,
"grad_norm": 1.8049304485321045,
"learning_rate": 8.01444522346409e-05,
"loss": 0.1978,
"step": 4510
},
{
"epoch": 70.625,
"grad_norm": 1.759329080581665,
"learning_rate": 7.975972601682884e-05,
"loss": 0.2057,
"step": 4520
},
{
"epoch": 70.78125,
"grad_norm": 1.6832956075668335,
"learning_rate": 7.937531192908e-05,
"loss": 0.2038,
"step": 4530
},
{
"epoch": 70.9375,
"grad_norm": 1.7098455429077148,
"learning_rate": 7.899121589953515e-05,
"loss": 0.2077,
"step": 4540
},
{
"epoch": 71.09375,
"grad_norm": 1.477565050125122,
"learning_rate": 7.860744385143022e-05,
"loss": 0.1924,
"step": 4550
},
{
"epoch": 71.25,
"grad_norm": 1.5800806283950806,
"learning_rate": 7.822400170300497e-05,
"loss": 0.1943,
"step": 4560
},
{
"epoch": 71.40625,
"grad_norm": 1.497495412826538,
"learning_rate": 7.784089536741164e-05,
"loss": 0.1944,
"step": 4570
},
{
"epoch": 71.5625,
"grad_norm": 1.9330024719238281,
"learning_rate": 7.745813075262389e-05,
"loss": 0.2039,
"step": 4580
},
{
"epoch": 71.71875,
"grad_norm": 2.2339158058166504,
"learning_rate": 7.707571376134548e-05,
"loss": 0.2052,
"step": 4590
},
{
"epoch": 71.875,
"grad_norm": 1.9479732513427734,
"learning_rate": 7.669365029091964e-05,
"loss": 0.2103,
"step": 4600
},
{
"epoch": 72.03125,
"grad_norm": 1.2796049118041992,
"learning_rate": 7.63119462332376e-05,
"loss": 0.2049,
"step": 4610
},
{
"epoch": 72.1875,
"grad_norm": 1.4597240686416626,
"learning_rate": 7.593060747464818e-05,
"loss": 0.1842,
"step": 4620
},
{
"epoch": 72.34375,
"grad_norm": 1.857862114906311,
"learning_rate": 7.554963989586675e-05,
"loss": 0.193,
"step": 4630
},
{
"epoch": 72.5,
"grad_norm": 1.5953953266143799,
"learning_rate": 7.51690493718846e-05,
"loss": 0.2033,
"step": 4640
},
{
"epoch": 72.65625,
"grad_norm": 1.7451122999191284,
"learning_rate": 7.478884177187855e-05,
"loss": 0.2082,
"step": 4650
},
{
"epoch": 72.8125,
"grad_norm": 1.7878004312515259,
"learning_rate": 7.440902295912005e-05,
"loss": 0.2025,
"step": 4660
},
{
"epoch": 72.96875,
"grad_norm": 1.911655306816101,
"learning_rate": 7.402959879088505e-05,
"loss": 0.2051,
"step": 4670
},
{
"epoch": 73.125,
"grad_norm": 1.783508062362671,
"learning_rate": 7.365057511836359e-05,
"loss": 0.1968,
"step": 4680
},
{
"epoch": 73.28125,
"grad_norm": 1.643697738647461,
"learning_rate": 7.327195778656962e-05,
"loss": 0.1925,
"step": 4690
},
{
"epoch": 73.4375,
"grad_norm": 1.5406553745269775,
"learning_rate": 7.289375263425073e-05,
"loss": 0.1969,
"step": 4700
},
{
"epoch": 73.59375,
"grad_norm": 1.8887746334075928,
"learning_rate": 7.251596549379827e-05,
"loss": 0.1971,
"step": 4710
},
{
"epoch": 73.75,
"grad_norm": 1.6867214441299438,
"learning_rate": 7.21386021911572e-05,
"loss": 0.197,
"step": 4720
},
{
"epoch": 73.90625,
"grad_norm": 1.8071234226226807,
"learning_rate": 7.176166854573659e-05,
"loss": 0.1978,
"step": 4730
},
{
"epoch": 74.0625,
"grad_norm": 1.795962929725647,
"learning_rate": 7.138517037031944e-05,
"loss": 0.2011,
"step": 4740
},
{
"epoch": 74.21875,
"grad_norm": 1.666642427444458,
"learning_rate": 7.100911347097343e-05,
"loss": 0.1908,
"step": 4750
},
{
"epoch": 74.375,
"grad_norm": 1.6308237314224243,
"learning_rate": 7.06335036469611e-05,
"loss": 0.1929,
"step": 4760
},
{
"epoch": 74.53125,
"grad_norm": 1.6244676113128662,
"learning_rate": 7.025834669065064e-05,
"loss": 0.1959,
"step": 4770
},
{
"epoch": 74.6875,
"grad_norm": 1.6163049936294556,
"learning_rate": 6.988364838742639e-05,
"loss": 0.1989,
"step": 4780
},
{
"epoch": 74.84375,
"grad_norm": 1.8937861919403076,
"learning_rate": 6.950941451559972e-05,
"loss": 0.1916,
"step": 4790
},
{
"epoch": 75.0,
"grad_norm": 1.4619686603546143,
"learning_rate": 6.913565084631979e-05,
"loss": 0.2031,
"step": 4800
},
{
"epoch": 75.15625,
"grad_norm": 1.46817147731781,
"learning_rate": 6.87623631434849e-05,
"loss": 0.1837,
"step": 4810
},
{
"epoch": 75.3125,
"grad_norm": 1.4131444692611694,
"learning_rate": 6.838955716365309e-05,
"loss": 0.1875,
"step": 4820
},
{
"epoch": 75.46875,
"grad_norm": 1.7622783184051514,
"learning_rate": 6.801723865595381e-05,
"loss": 0.1971,
"step": 4830
},
{
"epoch": 75.625,
"grad_norm": 1.4485961198806763,
"learning_rate": 6.764541336199899e-05,
"loss": 0.1921,
"step": 4840
},
{
"epoch": 75.78125,
"grad_norm": 1.8549201488494873,
"learning_rate": 6.727408701579467e-05,
"loss": 0.2013,
"step": 4850
},
{
"epoch": 75.9375,
"grad_norm": 1.809103012084961,
"learning_rate": 6.690326534365248e-05,
"loss": 0.1975,
"step": 4860
},
{
"epoch": 76.09375,
"grad_norm": 1.4220165014266968,
"learning_rate": 6.653295406410126e-05,
"loss": 0.1908,
"step": 4870
},
{
"epoch": 76.25,
"grad_norm": 1.5359421968460083,
"learning_rate": 6.616315888779907e-05,
"loss": 0.1903,
"step": 4880
},
{
"epoch": 76.40625,
"grad_norm": 2.1324145793914795,
"learning_rate": 6.579388551744499e-05,
"loss": 0.1874,
"step": 4890
},
{
"epoch": 76.5625,
"grad_norm": 1.7252235412597656,
"learning_rate": 6.542513964769122e-05,
"loss": 0.1929,
"step": 4900
},
{
"epoch": 76.71875,
"grad_norm": 1.6541646718978882,
"learning_rate": 6.50569269650552e-05,
"loss": 0.1902,
"step": 4910
},
{
"epoch": 76.875,
"grad_norm": 1.481459379196167,
"learning_rate": 6.468925314783205e-05,
"loss": 0.1958,
"step": 4920
},
{
"epoch": 77.03125,
"grad_norm": 1.6920355558395386,
"learning_rate": 6.432212386600676e-05,
"loss": 0.1952,
"step": 4930
},
{
"epoch": 77.1875,
"grad_norm": 1.567058801651001,
"learning_rate": 6.395554478116717e-05,
"loss": 0.1858,
"step": 4940
},
{
"epoch": 77.34375,
"grad_norm": 2.0057098865509033,
"learning_rate": 6.358952154641611e-05,
"loss": 0.1886,
"step": 4950
},
{
"epoch": 77.5,
"grad_norm": 2.1384148597717285,
"learning_rate": 6.322405980628472e-05,
"loss": 0.1943,
"step": 4960
},
{
"epoch": 77.65625,
"grad_norm": 1.9071811437606812,
"learning_rate": 6.28591651966451e-05,
"loss": 0.1923,
"step": 4970
},
{
"epoch": 77.8125,
"grad_norm": 1.9044047594070435,
"learning_rate": 6.249484334462356e-05,
"loss": 0.1967,
"step": 4980
},
{
"epoch": 77.96875,
"grad_norm": 1.7569500207901,
"learning_rate": 6.213109986851376e-05,
"loss": 0.1919,
"step": 4990
},
{
"epoch": 78.125,
"grad_norm": 1.7204627990722656,
"learning_rate": 6.176794037769003e-05,
"loss": 0.1844,
"step": 5000
},
{
"epoch": 78.28125,
"grad_norm": 1.8979413509368896,
"learning_rate": 6.140537047252092e-05,
"loss": 0.1893,
"step": 5010
},
{
"epoch": 78.4375,
"grad_norm": 1.5315955877304077,
"learning_rate": 6.104339574428301e-05,
"loss": 0.1856,
"step": 5020
},
{
"epoch": 78.59375,
"grad_norm": 1.5335980653762817,
"learning_rate": 6.0682021775074324e-05,
"loss": 0.1939,
"step": 5030
},
{
"epoch": 78.75,
"grad_norm": 1.9664844274520874,
"learning_rate": 6.032125413772851e-05,
"loss": 0.1925,
"step": 5040
},
{
"epoch": 78.90625,
"grad_norm": 1.4986895322799683,
"learning_rate": 5.996109839572881e-05,
"loss": 0.193,
"step": 5050
},
{
"epoch": 79.0625,
"grad_norm": 1.6273597478866577,
"learning_rate": 5.960156010312236e-05,
"loss": 0.1845,
"step": 5060
},
{
"epoch": 79.21875,
"grad_norm": 1.3411659002304077,
"learning_rate": 5.9242644804434354e-05,
"loss": 0.182,
"step": 5070
},
{
"epoch": 79.375,
"grad_norm": 1.550490379333496,
"learning_rate": 5.888435803458272e-05,
"loss": 0.1838,
"step": 5080
},
{
"epoch": 79.53125,
"grad_norm": 1.4418952465057373,
"learning_rate": 5.852670531879261e-05,
"loss": 0.1913,
"step": 5090
},
{
"epoch": 79.6875,
"grad_norm": 1.4645665884017944,
"learning_rate": 5.816969217251144e-05,
"loss": 0.1867,
"step": 5100
},
{
"epoch": 79.84375,
"grad_norm": 1.539569616317749,
"learning_rate": 5.781332410132352e-05,
"loss": 0.1927,
"step": 5110
},
{
"epoch": 80.0,
"grad_norm": 1.5742814540863037,
"learning_rate": 5.745760660086532e-05,
"loss": 0.1917,
"step": 5120
},
{
"epoch": 80.15625,
"grad_norm": 1.3147929906845093,
"learning_rate": 5.7102545156740805e-05,
"loss": 0.181,
"step": 5130
},
{
"epoch": 80.3125,
"grad_norm": 1.8531893491744995,
"learning_rate": 5.6748145244436615e-05,
"loss": 0.183,
"step": 5140
},
{
"epoch": 80.46875,
"grad_norm": 1.2582364082336426,
"learning_rate": 5.639441232923785e-05,
"loss": 0.1834,
"step": 5150
},
{
"epoch": 80.625,
"grad_norm": 1.6824560165405273,
"learning_rate": 5.604135186614368e-05,
"loss": 0.1859,
"step": 5160
},
{
"epoch": 80.78125,
"grad_norm": 1.5116562843322754,
"learning_rate": 5.568896929978316e-05,
"loss": 0.1908,
"step": 5170
},
{
"epoch": 80.9375,
"grad_norm": 1.6242319345474243,
"learning_rate": 5.5337270064331306e-05,
"loss": 0.1934,
"step": 5180
},
{
"epoch": 81.09375,
"grad_norm": 1.4753321409225464,
"learning_rate": 5.498625958342556e-05,
"loss": 0.1889,
"step": 5190
},
{
"epoch": 81.25,
"grad_norm": 1.5960818529129028,
"learning_rate": 5.463594327008168e-05,
"loss": 0.1798,
"step": 5200
},
{
"epoch": 81.40625,
"grad_norm": 1.2806315422058105,
"learning_rate": 5.4286326526610555e-05,
"loss": 0.1845,
"step": 5210
},
{
"epoch": 81.5625,
"grad_norm": 1.4679008722305298,
"learning_rate": 5.3937414744534874e-05,
"loss": 0.1858,
"step": 5220
},
{
"epoch": 81.71875,
"grad_norm": 1.5440151691436768,
"learning_rate": 5.3589213304506014e-05,
"loss": 0.1855,
"step": 5230
},
{
"epoch": 81.875,
"grad_norm": 1.5244407653808594,
"learning_rate": 5.324172757622081e-05,
"loss": 0.1876,
"step": 5240
},
{
"epoch": 82.03125,
"grad_norm": 1.654781460762024,
"learning_rate": 5.289496291833917e-05,
"loss": 0.1887,
"step": 5250
},
{
"epoch": 82.1875,
"grad_norm": 1.3543646335601807,
"learning_rate": 5.254892467840098e-05,
"loss": 0.1803,
"step": 5260
},
{
"epoch": 82.34375,
"grad_norm": 1.4349260330200195,
"learning_rate": 5.220361819274407e-05,
"loss": 0.1791,
"step": 5270
},
{
"epoch": 82.5,
"grad_norm": 1.5351791381835938,
"learning_rate": 5.18590487864215e-05,
"loss": 0.1851,
"step": 5280
},
{
"epoch": 82.65625,
"grad_norm": 1.2516671419143677,
"learning_rate": 5.151522177311989e-05,
"loss": 0.1861,
"step": 5290
},
{
"epoch": 82.8125,
"grad_norm": 1.4409736394882202,
"learning_rate": 5.1172142455077e-05,
"loss": 0.1861,
"step": 5300
},
{
"epoch": 82.96875,
"grad_norm": 1.476853847503662,
"learning_rate": 5.0829816123000395e-05,
"loss": 0.1862,
"step": 5310
},
{
"epoch": 83.125,
"grad_norm": 1.789379596710205,
"learning_rate": 5.048824805598562e-05,
"loss": 0.1841,
"step": 5320
},
{
"epoch": 83.28125,
"grad_norm": 1.3781448602676392,
"learning_rate": 5.014744352143477e-05,
"loss": 0.1853,
"step": 5330
},
{
"epoch": 83.4375,
"grad_norm": 1.5546520948410034,
"learning_rate": 4.980740777497529e-05,
"loss": 0.1787,
"step": 5340
},
{
"epoch": 83.59375,
"grad_norm": 1.8084760904312134,
"learning_rate": 4.946814606037917e-05,
"loss": 0.1791,
"step": 5350
},
{
"epoch": 83.75,
"grad_norm": 1.777557134628296,
"learning_rate": 4.9129663609481694e-05,
"loss": 0.1819,
"step": 5360
},
{
"epoch": 83.90625,
"grad_norm": 1.4378726482391357,
"learning_rate": 4.87919656421009e-05,
"loss": 0.1872,
"step": 5370
},
{
"epoch": 84.0625,
"grad_norm": 1.484693169593811,
"learning_rate": 4.845505736595729e-05,
"loss": 0.1887,
"step": 5380
},
{
"epoch": 84.21875,
"grad_norm": 1.4815410375595093,
"learning_rate": 4.811894397659319e-05,
"loss": 0.1784,
"step": 5390
},
{
"epoch": 84.375,
"grad_norm": 1.424680233001709,
"learning_rate": 4.778363065729292e-05,
"loss": 0.1821,
"step": 5400
},
{
"epoch": 84.53125,
"grad_norm": 1.4403337240219116,
"learning_rate": 4.7449122579002605e-05,
"loss": 0.1839,
"step": 5410
},
{
"epoch": 84.6875,
"grad_norm": 1.7450019121170044,
"learning_rate": 4.711542490025067e-05,
"loss": 0.1843,
"step": 5420
},
{
"epoch": 84.84375,
"grad_norm": 1.3429001569747925,
"learning_rate": 4.678254276706807e-05,
"loss": 0.181,
"step": 5430
},
{
"epoch": 85.0,
"grad_norm": 1.5108141899108887,
"learning_rate": 4.6450481312909136e-05,
"loss": 0.1815,
"step": 5440
},
{
"epoch": 85.15625,
"grad_norm": 1.2566031217575073,
"learning_rate": 4.6119245658572284e-05,
"loss": 0.1772,
"step": 5450
},
{
"epoch": 85.3125,
"grad_norm": 1.5348423719406128,
"learning_rate": 4.578884091212103e-05,
"loss": 0.1774,
"step": 5460
},
{
"epoch": 85.46875,
"grad_norm": 1.8427592515945435,
"learning_rate": 4.545927216880525e-05,
"loss": 0.1815,
"step": 5470
},
{
"epoch": 85.625,
"grad_norm": 1.6190025806427002,
"learning_rate": 4.513054451098276e-05,
"loss": 0.1862,
"step": 5480
},
{
"epoch": 85.78125,
"grad_norm": 1.5374351739883423,
"learning_rate": 4.4802663008040676e-05,
"loss": 0.1798,
"step": 5490
},
{
"epoch": 85.9375,
"grad_norm": 1.5920672416687012,
"learning_rate": 4.4475632716317304e-05,
"loss": 0.1816,
"step": 5500
},
{
"epoch": 86.09375,
"grad_norm": 1.5544840097427368,
"learning_rate": 4.414945867902437e-05,
"loss": 0.176,
"step": 5510
},
{
"epoch": 86.25,
"grad_norm": 1.6388013362884521,
"learning_rate": 4.3824145926169036e-05,
"loss": 0.1752,
"step": 5520
},
{
"epoch": 86.40625,
"grad_norm": 1.461263656616211,
"learning_rate": 4.349969947447632e-05,
"loss": 0.1797,
"step": 5530
},
{
"epoch": 86.5625,
"grad_norm": 1.5496662855148315,
"learning_rate": 4.317612432731193e-05,
"loss": 0.1817,
"step": 5540
},
{
"epoch": 86.71875,
"grad_norm": 1.4612733125686646,
"learning_rate": 4.2853425474604846e-05,
"loss": 0.178,
"step": 5550
},
{
"epoch": 86.875,
"grad_norm": 1.3939430713653564,
"learning_rate": 4.253160789277066e-05,
"loss": 0.1816,
"step": 5560
},
{
"epoch": 87.03125,
"grad_norm": 1.5148475170135498,
"learning_rate": 4.221067654463452e-05,
"loss": 0.1889,
"step": 5570
},
{
"epoch": 87.1875,
"grad_norm": 1.7056833505630493,
"learning_rate": 4.189063637935487e-05,
"loss": 0.1691,
"step": 5580
},
{
"epoch": 87.34375,
"grad_norm": 1.3451744318008423,
"learning_rate": 4.157149233234693e-05,
"loss": 0.181,
"step": 5590
},
{
"epoch": 87.5,
"grad_norm": 1.4934422969818115,
"learning_rate": 4.1253249325206713e-05,
"loss": 0.1784,
"step": 5600
},
{
"epoch": 87.65625,
"grad_norm": 1.5852150917053223,
"learning_rate": 4.0935912265635124e-05,
"loss": 0.1848,
"step": 5610
},
{
"epoch": 87.8125,
"grad_norm": 1.593066930770874,
"learning_rate": 4.061948604736214e-05,
"loss": 0.1769,
"step": 5620
},
{
"epoch": 87.96875,
"grad_norm": 1.4550515413284302,
"learning_rate": 4.0303975550071474e-05,
"loss": 0.1803,
"step": 5630
},
{
"epoch": 88.125,
"grad_norm": 1.4054372310638428,
"learning_rate": 3.99893856393253e-05,
"loss": 0.1781,
"step": 5640
},
{
"epoch": 88.28125,
"grad_norm": 1.4826765060424805,
"learning_rate": 3.967572116648927e-05,
"loss": 0.1786,
"step": 5650
},
{
"epoch": 88.4375,
"grad_norm": 1.4517465829849243,
"learning_rate": 3.936298696865749e-05,
"loss": 0.1795,
"step": 5660
},
{
"epoch": 88.59375,
"grad_norm": 2.170179843902588,
"learning_rate": 3.905118786857824e-05,
"loss": 0.1755,
"step": 5670
},
{
"epoch": 88.75,
"grad_norm": 1.5206856727600098,
"learning_rate": 3.87403286745793e-05,
"loss": 0.1725,
"step": 5680
},
{
"epoch": 88.90625,
"grad_norm": 1.4970860481262207,
"learning_rate": 3.843041418049402e-05,
"loss": 0.1818,
"step": 5690
},
{
"epoch": 89.0625,
"grad_norm": 1.3487545251846313,
"learning_rate": 3.8121449165587255e-05,
"loss": 0.1792,
"step": 5700
},
{
"epoch": 89.21875,
"grad_norm": 1.6100648641586304,
"learning_rate": 3.7813438394481794e-05,
"loss": 0.1764,
"step": 5710
},
{
"epoch": 89.375,
"grad_norm": 1.2975910902023315,
"learning_rate": 3.750638661708469e-05,
"loss": 0.1744,
"step": 5720
},
{
"epoch": 89.53125,
"grad_norm": 1.2627249956130981,
"learning_rate": 3.7200298568514214e-05,
"loss": 0.1763,
"step": 5730
},
{
"epoch": 89.6875,
"grad_norm": 1.3773270845413208,
"learning_rate": 3.689517896902678e-05,
"loss": 0.1803,
"step": 5740
},
{
"epoch": 89.84375,
"grad_norm": 1.8265244960784912,
"learning_rate": 3.659103252394401e-05,
"loss": 0.1772,
"step": 5750
},
{
"epoch": 90.0,
"grad_norm": 1.4150211811065674,
"learning_rate": 3.628786392358028e-05,
"loss": 0.1778,
"step": 5760
},
{
"epoch": 90.15625,
"grad_norm": 1.5570526123046875,
"learning_rate": 3.598567784317056e-05,
"loss": 0.173,
"step": 5770
},
{
"epoch": 90.3125,
"grad_norm": 1.424486756324768,
"learning_rate": 3.5684478942797984e-05,
"loss": 0.179,
"step": 5780
},
{
"epoch": 90.46875,
"grad_norm": 1.4627338647842407,
"learning_rate": 3.538427186732216e-05,
"loss": 0.1776,
"step": 5790
},
{
"epoch": 90.625,
"grad_norm": 1.7183500528335571,
"learning_rate": 3.508506124630759e-05,
"loss": 0.1795,
"step": 5800
},
{
"epoch": 90.78125,
"grad_norm": 1.4211575984954834,
"learning_rate": 3.478685169395224e-05,
"loss": 0.1769,
"step": 5810
},
{
"epoch": 90.9375,
"grad_norm": 1.5053857564926147,
"learning_rate": 3.448964780901622e-05,
"loss": 0.1751,
"step": 5820
},
{
"epoch": 91.09375,
"grad_norm": 1.6190037727355957,
"learning_rate": 3.4193454174751196e-05,
"loss": 0.1752,
"step": 5830
},
{
"epoch": 91.25,
"grad_norm": 1.623235821723938,
"learning_rate": 3.389827535882931e-05,
"loss": 0.17,
"step": 5840
},
{
"epoch": 91.40625,
"grad_norm": 1.6231571435928345,
"learning_rate": 3.3604115913273146e-05,
"loss": 0.1808,
"step": 5850
},
{
"epoch": 91.5625,
"grad_norm": 1.3666538000106812,
"learning_rate": 3.331098037438518e-05,
"loss": 0.1769,
"step": 5860
},
{
"epoch": 91.71875,
"grad_norm": 1.6779783964157104,
"learning_rate": 3.301887326267812e-05,
"loss": 0.1757,
"step": 5870
},
{
"epoch": 91.875,
"grad_norm": 1.3442100286483765,
"learning_rate": 3.272779908280493e-05,
"loss": 0.1735,
"step": 5880
},
{
"epoch": 92.03125,
"grad_norm": 1.3602097034454346,
"learning_rate": 3.2437762323489485e-05,
"loss": 0.1739,
"step": 5890
},
{
"epoch": 92.1875,
"grad_norm": 1.383625864982605,
"learning_rate": 3.214876745745754e-05,
"loss": 0.1732,
"step": 5900
},
{
"epoch": 92.34375,
"grad_norm": 1.433128833770752,
"learning_rate": 3.1860818941367367e-05,
"loss": 0.1716,
"step": 5910
},
{
"epoch": 92.5,
"grad_norm": 1.3733683824539185,
"learning_rate": 3.1573921215741254e-05,
"loss": 0.1743,
"step": 5920
},
{
"epoch": 92.65625,
"grad_norm": 1.2076114416122437,
"learning_rate": 3.1288078704897074e-05,
"loss": 0.1717,
"step": 5930
},
{
"epoch": 92.8125,
"grad_norm": 1.4948092699050903,
"learning_rate": 3.1003295816879985e-05,
"loss": 0.1784,
"step": 5940
},
{
"epoch": 92.96875,
"grad_norm": 1.3194231986999512,
"learning_rate": 3.071957694339435e-05,
"loss": 0.1767,
"step": 5950
},
{
"epoch": 93.125,
"grad_norm": 1.4278184175491333,
"learning_rate": 3.0436926459736227e-05,
"loss": 0.171,
"step": 5960
},
{
"epoch": 93.28125,
"grad_norm": 2.0391736030578613,
"learning_rate": 3.0155348724725673e-05,
"loss": 0.1763,
"step": 5970
},
{
"epoch": 93.4375,
"grad_norm": 1.4166723489761353,
"learning_rate": 2.987484808063976e-05,
"loss": 0.1738,
"step": 5980
},
{
"epoch": 93.59375,
"grad_norm": 1.4654409885406494,
"learning_rate": 2.959542885314536e-05,
"loss": 0.1715,
"step": 5990
},
{
"epoch": 93.75,
"grad_norm": 1.3458061218261719,
"learning_rate": 2.9317095351232648e-05,
"loss": 0.1758,
"step": 6000
},
{
"epoch": 93.90625,
"grad_norm": 1.6262662410736084,
"learning_rate": 2.9039851867148505e-05,
"loss": 0.1776,
"step": 6010
},
{
"epoch": 94.0625,
"grad_norm": 1.4164749383926392,
"learning_rate": 2.876370267633045e-05,
"loss": 0.1693,
"step": 6020
},
{
"epoch": 94.21875,
"grad_norm": 1.3470439910888672,
"learning_rate": 2.8488652037340634e-05,
"loss": 0.1681,
"step": 6030
},
{
"epoch": 94.375,
"grad_norm": 1.382950782775879,
"learning_rate": 2.8214704191800145e-05,
"loss": 0.1719,
"step": 6040
},
{
"epoch": 94.53125,
"grad_norm": 1.6302028894424438,
"learning_rate": 2.7941863364323595e-05,
"loss": 0.1738,
"step": 6050
},
{
"epoch": 94.6875,
"grad_norm": 1.4697049856185913,
"learning_rate": 2.7670133762454175e-05,
"loss": 0.1753,
"step": 6060
},
{
"epoch": 94.84375,
"grad_norm": 1.3718417882919312,
"learning_rate": 2.739951957659843e-05,
"loss": 0.1774,
"step": 6070
},
{
"epoch": 95.0,
"grad_norm": 1.6622495651245117,
"learning_rate": 2.715692393953073e-05,
"loss": 0.1724,
"step": 6080
},
{
"epoch": 95.15625,
"grad_norm": 1.3040485382080078,
"learning_rate": 2.688844052701359e-05,
"loss": 0.164,
"step": 6090
},
{
"epoch": 95.3125,
"grad_norm": 1.5511151552200317,
"learning_rate": 2.6621084585186828e-05,
"loss": 0.1725,
"step": 6100
},
{
"epoch": 95.46875,
"grad_norm": 1.6277883052825928,
"learning_rate": 2.635486023700976e-05,
"loss": 0.1736,
"step": 6110
},
{
"epoch": 95.625,
"grad_norm": 1.9094816446304321,
"learning_rate": 2.6089771587991206e-05,
"loss": 0.1749,
"step": 6120
},
{
"epoch": 95.78125,
"grad_norm": 1.1980408430099487,
"learning_rate": 2.582582272612609e-05,
"loss": 0.1786,
"step": 6130
},
{
"epoch": 95.9375,
"grad_norm": 1.4505863189697266,
"learning_rate": 2.556301772183233e-05,
"loss": 0.1739,
"step": 6140
},
{
"epoch": 96.09375,
"grad_norm": 1.252650260925293,
"learning_rate": 2.5301360627888236e-05,
"loss": 0.1676,
"step": 6150
},
{
"epoch": 96.25,
"grad_norm": 1.5206024646759033,
"learning_rate": 2.5040855479369853e-05,
"loss": 0.1677,
"step": 6160
},
{
"epoch": 96.40625,
"grad_norm": 1.4388213157653809,
"learning_rate": 2.4781506293588873e-05,
"loss": 0.1748,
"step": 6170
},
{
"epoch": 96.5625,
"grad_norm": 1.395816683769226,
"learning_rate": 2.4523317070030515e-05,
"loss": 0.1763,
"step": 6180
},
{
"epoch": 96.71875,
"grad_norm": 1.2511937618255615,
"learning_rate": 2.426629179029206e-05,
"loss": 0.1644,
"step": 6190
},
{
"epoch": 96.875,
"grad_norm": 1.137650489807129,
"learning_rate": 2.401043441802121e-05,
"loss": 0.1748,
"step": 6200
},
{
"epoch": 97.03125,
"grad_norm": 1.197763442993164,
"learning_rate": 2.37557488988552e-05,
"loss": 0.1788,
"step": 6210
},
{
"epoch": 97.1875,
"grad_norm": 1.7034637928009033,
"learning_rate": 2.350223916035983e-05,
"loss": 0.1681,
"step": 6220
},
{
"epoch": 97.34375,
"grad_norm": 1.2496026754379272,
"learning_rate": 2.3249909111968815e-05,
"loss": 0.1741,
"step": 6230
},
{
"epoch": 97.5,
"grad_norm": 1.8080981969833374,
"learning_rate": 2.2998762644923656e-05,
"loss": 0.1697,
"step": 6240
},
{
"epoch": 97.65625,
"grad_norm": 1.5005358457565308,
"learning_rate": 2.2748803632213557e-05,
"loss": 0.1709,
"step": 6250
},
{
"epoch": 97.8125,
"grad_norm": 1.2584110498428345,
"learning_rate": 2.250003592851575e-05,
"loss": 0.1712,
"step": 6260
},
{
"epoch": 97.96875,
"grad_norm": 1.4121007919311523,
"learning_rate": 2.225246337013591e-05,
"loss": 0.1725,
"step": 6270
},
{
"epoch": 98.125,
"grad_norm": 1.5540642738342285,
"learning_rate": 2.2006089774949203e-05,
"loss": 0.1735,
"step": 6280
},
{
"epoch": 98.28125,
"grad_norm": 1.2959460020065308,
"learning_rate": 2.1760918942341192e-05,
"loss": 0.1646,
"step": 6290
},
{
"epoch": 98.4375,
"grad_norm": 1.627122402191162,
"learning_rate": 2.151695465314946e-05,
"loss": 0.1715,
"step": 6300
},
{
"epoch": 98.59375,
"grad_norm": 1.5968302488327026,
"learning_rate": 2.127420066960508e-05,
"loss": 0.1754,
"step": 6310
},
{
"epoch": 98.75,
"grad_norm": 1.3676623106002808,
"learning_rate": 2.1032660735274858e-05,
"loss": 0.1746,
"step": 6320
},
{
"epoch": 98.90625,
"grad_norm": 1.2594794034957886,
"learning_rate": 2.07923385750033e-05,
"loss": 0.1695,
"step": 6330
},
{
"epoch": 99.0625,
"grad_norm": 1.6086690425872803,
"learning_rate": 2.0553237894855458e-05,
"loss": 0.1728,
"step": 6340
},
{
"epoch": 99.21875,
"grad_norm": 1.2404333353042603,
"learning_rate": 2.031536238205961e-05,
"loss": 0.1709,
"step": 6350
},
{
"epoch": 99.375,
"grad_norm": 1.3804398775100708,
"learning_rate": 2.0078715704950423e-05,
"loss": 0.1739,
"step": 6360
},
{
"epoch": 99.53125,
"grad_norm": 1.37823486328125,
"learning_rate": 1.9843301512912327e-05,
"loss": 0.1676,
"step": 6370
},
{
"epoch": 99.6875,
"grad_norm": 1.1960835456848145,
"learning_rate": 1.9609123436323508e-05,
"loss": 0.1678,
"step": 6380
},
{
"epoch": 99.84375,
"grad_norm": 1.3781687021255493,
"learning_rate": 1.937618508649954e-05,
"loss": 0.1673,
"step": 6390
},
{
"epoch": 100.0,
"grad_norm": 1.2113115787506104,
"learning_rate": 1.914449005563791e-05,
"loss": 0.1747,
"step": 6400
},
{
"epoch": 100.15625,
"grad_norm": 1.298387885093689,
"learning_rate": 1.891404191676265e-05,
"loss": 0.1637,
"step": 6410
},
{
"epoch": 100.3125,
"grad_norm": 1.1972380876541138,
"learning_rate": 1.868484422366914e-05,
"loss": 0.1721,
"step": 6420
},
{
"epoch": 100.46875,
"grad_norm": 2.0465595722198486,
"learning_rate": 1.8456900510869334e-05,
"loss": 0.1723,
"step": 6430
},
{
"epoch": 100.625,
"grad_norm": 1.6506719589233398,
"learning_rate": 1.823021429353724e-05,
"loss": 0.1689,
"step": 6440
},
{
"epoch": 100.78125,
"grad_norm": 1.431828498840332,
"learning_rate": 1.8004789067454764e-05,
"loss": 0.1688,
"step": 6450
},
{
"epoch": 100.9375,
"grad_norm": 1.4944453239440918,
"learning_rate": 1.7780628308957792e-05,
"loss": 0.1753,
"step": 6460
},
{
"epoch": 101.09375,
"grad_norm": 1.3024934530258179,
"learning_rate": 1.7557735474882465e-05,
"loss": 0.1659,
"step": 6470
},
{
"epoch": 101.25,
"grad_norm": 1.1864546537399292,
"learning_rate": 1.733611400251206e-05,
"loss": 0.1686,
"step": 6480
},
{
"epoch": 101.40625,
"grad_norm": 1.250319480895996,
"learning_rate": 1.7115767309523812e-05,
"loss": 0.1684,
"step": 6490
},
{
"epoch": 101.5625,
"grad_norm": 1.1592826843261719,
"learning_rate": 1.6896698793936316e-05,
"loss": 0.1734,
"step": 6500
},
{
"epoch": 101.71875,
"grad_norm": 1.490527629852295,
"learning_rate": 1.6678911834057098e-05,
"loss": 0.1719,
"step": 6510
},
{
"epoch": 101.875,
"grad_norm": 1.3874304294586182,
"learning_rate": 1.64624097884305e-05,
"loss": 0.1703,
"step": 6520
},
{
"epoch": 102.03125,
"grad_norm": 1.2031797170639038,
"learning_rate": 1.6247195995785837e-05,
"loss": 0.1707,
"step": 6530
},
{
"epoch": 102.1875,
"grad_norm": 1.4145324230194092,
"learning_rate": 1.6033273774986067e-05,
"loss": 0.1727,
"step": 6540
},
{
"epoch": 102.34375,
"grad_norm": 1.491633415222168,
"learning_rate": 1.5820646424976482e-05,
"loss": 0.167,
"step": 6550
},
{
"epoch": 102.5,
"grad_norm": 1.3200361728668213,
"learning_rate": 1.560931722473381e-05,
"loss": 0.1665,
"step": 6560
},
{
"epoch": 102.65625,
"grad_norm": 1.3420644998550415,
"learning_rate": 1.539928943321579e-05,
"loss": 0.1697,
"step": 6570
},
{
"epoch": 102.8125,
"grad_norm": 1.4412531852722168,
"learning_rate": 1.5190566289310747e-05,
"loss": 0.1711,
"step": 6580
},
{
"epoch": 102.96875,
"grad_norm": 1.22170090675354,
"learning_rate": 1.4983151011787788e-05,
"loss": 0.1681,
"step": 6590
},
{
"epoch": 103.125,
"grad_norm": 1.2543760538101196,
"learning_rate": 1.4777046799247052e-05,
"loss": 0.1647,
"step": 6600
},
{
"epoch": 103.28125,
"grad_norm": 1.4133225679397583,
"learning_rate": 1.4572256830070497e-05,
"loss": 0.1702,
"step": 6610
},
{
"epoch": 103.4375,
"grad_norm": 1.3233059644699097,
"learning_rate": 1.4368784262372736e-05,
"loss": 0.169,
"step": 6620
},
{
"epoch": 103.59375,
"grad_norm": 1.228870153427124,
"learning_rate": 1.4166632233952504e-05,
"loss": 0.1717,
"step": 6630
},
{
"epoch": 103.75,
"grad_norm": 1.2790390253067017,
"learning_rate": 1.3965803862244164e-05,
"loss": 0.1655,
"step": 6640
},
{
"epoch": 103.90625,
"grad_norm": 1.3549610376358032,
"learning_rate": 1.3766302244269624e-05,
"loss": 0.167,
"step": 6650
},
{
"epoch": 104.0625,
"grad_norm": 1.2854844331741333,
"learning_rate": 1.3568130456590588e-05,
"loss": 0.1663,
"step": 6660
},
{
"epoch": 104.21875,
"grad_norm": 1.270037055015564,
"learning_rate": 1.3371291555261256e-05,
"loss": 0.1627,
"step": 6670
},
{
"epoch": 104.375,
"grad_norm": 1.276854395866394,
"learning_rate": 1.3175788575780945e-05,
"loss": 0.1732,
"step": 6680
},
{
"epoch": 104.53125,
"grad_norm": 1.3133665323257446,
"learning_rate": 1.2981624533047432e-05,
"loss": 0.1683,
"step": 6690
},
{
"epoch": 104.6875,
"grad_norm": 1.3067399263381958,
"learning_rate": 1.2788802421310463e-05,
"loss": 0.1683,
"step": 6700
},
{
"epoch": 104.84375,
"grad_norm": 1.3023086786270142,
"learning_rate": 1.2597325214125566e-05,
"loss": 0.1711,
"step": 6710
},
{
"epoch": 105.0,
"grad_norm": 1.4500195980072021,
"learning_rate": 1.2407195864308096e-05,
"loss": 0.1714,
"step": 6720
},
{
"epoch": 105.15625,
"grad_norm": 1.4866911172866821,
"learning_rate": 1.2218417303887842e-05,
"loss": 0.1693,
"step": 6730
},
{
"epoch": 105.3125,
"grad_norm": 1.0444018840789795,
"learning_rate": 1.2030992444063726e-05,
"loss": 0.1649,
"step": 6740
},
{
"epoch": 105.46875,
"grad_norm": 1.5279264450073242,
"learning_rate": 1.184492417515899e-05,
"loss": 0.1683,
"step": 6750
},
{
"epoch": 105.625,
"grad_norm": 1.2571183443069458,
"learning_rate": 1.1660215366576454e-05,
"loss": 0.1691,
"step": 6760
},
{
"epoch": 105.78125,
"grad_norm": 1.3707107305526733,
"learning_rate": 1.1476868866754486e-05,
"loss": 0.1662,
"step": 6770
},
{
"epoch": 105.9375,
"grad_norm": 1.2116854190826416,
"learning_rate": 1.1294887503122919e-05,
"loss": 0.1664,
"step": 6780
},
{
"epoch": 106.09375,
"grad_norm": 1.4501864910125732,
"learning_rate": 1.1114274082059451e-05,
"loss": 0.1663,
"step": 6790
},
{
"epoch": 106.25,
"grad_norm": 1.7519021034240723,
"learning_rate": 1.093503138884654e-05,
"loss": 0.167,
"step": 6800
},
{
"epoch": 106.40625,
"grad_norm": 1.4105887413024902,
"learning_rate": 1.0757162187628222e-05,
"loss": 0.1688,
"step": 6810
},
{
"epoch": 106.5625,
"grad_norm": 1.1597462892532349,
"learning_rate": 1.0580669221367589e-05,
"loss": 0.1639,
"step": 6820
},
{
"epoch": 106.71875,
"grad_norm": 1.3421756029129028,
"learning_rate": 1.0405555211804518e-05,
"loss": 0.1689,
"step": 6830
},
{
"epoch": 106.875,
"grad_norm": 1.3345783948898315,
"learning_rate": 1.0231822859413709e-05,
"loss": 0.1717,
"step": 6840
},
{
"epoch": 107.03125,
"grad_norm": 1.2460522651672363,
"learning_rate": 1.0059474843362892e-05,
"loss": 0.1649,
"step": 6850
},
{
"epoch": 107.1875,
"grad_norm": 1.328639268875122,
"learning_rate": 9.888513821471735e-06,
"loss": 0.1623,
"step": 6860
},
{
"epoch": 107.34375,
"grad_norm": 1.4876681566238403,
"learning_rate": 9.71894243017063e-06,
"loss": 0.1668,
"step": 6870
},
{
"epoch": 107.5,
"grad_norm": 1.2804096937179565,
"learning_rate": 9.55076328446024e-06,
"loss": 0.1693,
"step": 6880
},
{
"epoch": 107.65625,
"grad_norm": 1.2431553602218628,
"learning_rate": 9.383978977871021e-06,
"loss": 0.1669,
"step": 6890
},
{
"epoch": 107.8125,
"grad_norm": 1.6604621410369873,
"learning_rate": 9.218592082423316e-06,
"loss": 0.1686,
"step": 6900
},
{
"epoch": 107.96875,
"grad_norm": 1.3399462699890137,
"learning_rate": 9.054605148587624e-06,
"loss": 0.1689,
"step": 6910
},
{
"epoch": 108.125,
"grad_norm": 1.9574614763259888,
"learning_rate": 8.892020705245341e-06,
"loss": 0.1635,
"step": 6920
},
{
"epoch": 108.28125,
"grad_norm": 1.5866326093673706,
"learning_rate": 8.730841259649725e-06,
"loss": 0.1645,
"step": 6930
},
{
"epoch": 108.4375,
"grad_norm": 1.3083440065383911,
"learning_rate": 8.571069297387202e-06,
"loss": 0.1682,
"step": 6940
},
{
"epoch": 108.59375,
"grad_norm": 1.4477092027664185,
"learning_rate": 8.41270728233905e-06,
"loss": 0.1634,
"step": 6950
},
{
"epoch": 108.75,
"grad_norm": 1.584891676902771,
"learning_rate": 8.255757656643482e-06,
"loss": 0.1685,
"step": 6960
},
{
"epoch": 108.90625,
"grad_norm": 1.29928457736969,
"learning_rate": 8.100222840657878e-06,
"loss": 0.1703,
"step": 6970
},
{
"epoch": 109.0625,
"grad_norm": 1.2276784181594849,
"learning_rate": 7.946105232921474e-06,
"loss": 0.1695,
"step": 6980
},
{
"epoch": 109.21875,
"grad_norm": 1.2223650217056274,
"learning_rate": 7.793407210118464e-06,
"loss": 0.1651,
"step": 6990
},
{
"epoch": 109.375,
"grad_norm": 1.1992684602737427,
"learning_rate": 7.642131127041275e-06,
"loss": 0.1693,
"step": 7000
},
{
"epoch": 109.53125,
"grad_norm": 1.429261565208435,
"learning_rate": 7.492279316554207e-06,
"loss": 0.1676,
"step": 7010
},
{
"epoch": 109.6875,
"grad_norm": 1.4679069519042969,
"learning_rate": 7.343854089557556e-06,
"loss": 0.1687,
"step": 7020
},
{
"epoch": 109.84375,
"grad_norm": 1.261810541152954,
"learning_rate": 7.1968577349519565e-06,
"loss": 0.162,
"step": 7030
},
{
"epoch": 110.0,
"grad_norm": 1.4870808124542236,
"learning_rate": 7.051292519603014e-06,
"loss": 0.1682,
"step": 7040
},
{
"epoch": 110.15625,
"grad_norm": 1.2893967628479004,
"learning_rate": 6.907160688306425e-06,
"loss": 0.1637,
"step": 7050
},
{
"epoch": 110.3125,
"grad_norm": 1.3684417009353638,
"learning_rate": 6.764464463753362e-06,
"loss": 0.1674,
"step": 7060
},
{
"epoch": 110.46875,
"grad_norm": 1.3469291925430298,
"learning_rate": 6.6232060464961e-06,
"loss": 0.1628,
"step": 7070
},
{
"epoch": 110.625,
"grad_norm": 1.1284126043319702,
"learning_rate": 6.48338761491416e-06,
"loss": 0.1654,
"step": 7080
},
{
"epoch": 110.78125,
"grad_norm": 1.8623775243759155,
"learning_rate": 6.345011325180772e-06,
"loss": 0.1677,
"step": 7090
},
{
"epoch": 110.9375,
"grad_norm": 1.4487367868423462,
"learning_rate": 6.2080793112294755e-06,
"loss": 0.1674,
"step": 7100
},
{
"epoch": 111.09375,
"grad_norm": 1.5421063899993896,
"learning_rate": 6.072593684721295e-06,
"loss": 0.1657,
"step": 7110
},
{
"epoch": 111.25,
"grad_norm": 1.2256017923355103,
"learning_rate": 5.938556535012229e-06,
"loss": 0.1626,
"step": 7120
},
{
"epoch": 111.40625,
"grad_norm": 1.5549722909927368,
"learning_rate": 5.805969929120947e-06,
"loss": 0.1684,
"step": 7130
},
{
"epoch": 111.5625,
"grad_norm": 1.4856703281402588,
"learning_rate": 5.674835911696885e-06,
"loss": 0.1642,
"step": 7140
},
{
"epoch": 111.71875,
"grad_norm": 1.4706385135650635,
"learning_rate": 5.545156504988857e-06,
"loss": 0.1689,
"step": 7150
},
{
"epoch": 111.875,
"grad_norm": 1.3578001260757446,
"learning_rate": 5.4169337088136984e-06,
"loss": 0.1656,
"step": 7160
},
{
"epoch": 112.03125,
"grad_norm": 1.2759300470352173,
"learning_rate": 5.290169500525577e-06,
"loss": 0.1676,
"step": 7170
},
{
"epoch": 112.1875,
"grad_norm": 1.3984161615371704,
"learning_rate": 5.164865834985377e-06,
"loss": 0.1641,
"step": 7180
},
{
"epoch": 112.34375,
"grad_norm": 1.5849649906158447,
"learning_rate": 5.0410246445306455e-06,
"loss": 0.1674,
"step": 7190
},
{
"epoch": 112.5,
"grad_norm": 1.6095918416976929,
"learning_rate": 4.918647838945711e-06,
"loss": 0.1676,
"step": 7200
},
{
"epoch": 112.65625,
"grad_norm": 1.3017394542694092,
"learning_rate": 4.797737305432337e-06,
"loss": 0.1637,
"step": 7210
},
{
"epoch": 112.8125,
"grad_norm": 1.5304924249649048,
"learning_rate": 4.678294908580505e-06,
"loss": 0.1664,
"step": 7220
},
{
"epoch": 112.96875,
"grad_norm": 1.3150688409805298,
"learning_rate": 4.560322490339741e-06,
"loss": 0.1651,
"step": 7230
},
{
"epoch": 113.125,
"grad_norm": 1.2478370666503906,
"learning_rate": 4.443821869990661e-06,
"loss": 0.1627,
"step": 7240
},
{
"epoch": 113.28125,
"grad_norm": 1.1985414028167725,
"learning_rate": 4.328794844116946e-06,
"loss": 0.1673,
"step": 7250
},
{
"epoch": 113.4375,
"grad_norm": 1.6331043243408203,
"learning_rate": 4.215243186577656e-06,
"loss": 0.1677,
"step": 7260
},
{
"epoch": 113.59375,
"grad_norm": 1.3105155229568481,
"learning_rate": 4.103168648479794e-06,
"loss": 0.1633,
"step": 7270
},
{
"epoch": 113.75,
"grad_norm": 1.6450676918029785,
"learning_rate": 3.992572958151397e-06,
"loss": 0.1698,
"step": 7280
},
{
"epoch": 113.90625,
"grad_norm": 1.1760913133621216,
"learning_rate": 3.883457821114811e-06,
"loss": 0.1667,
"step": 7290
},
{
"epoch": 114.0625,
"grad_norm": 1.5002760887145996,
"learning_rate": 3.7758249200604444e-06,
"loss": 0.1661,
"step": 7300
},
{
"epoch": 114.21875,
"grad_norm": 1.1639865636825562,
"learning_rate": 3.6696759148207583e-06,
"loss": 0.1633,
"step": 7310
},
{
"epoch": 114.375,
"grad_norm": 1.173621654510498,
"learning_rate": 3.5650124423447507e-06,
"loss": 0.1643,
"step": 7320
},
{
"epoch": 114.53125,
"grad_norm": 1.1538360118865967,
"learning_rate": 3.461836116672612e-06,
"loss": 0.1675,
"step": 7330
},
{
"epoch": 114.6875,
"grad_norm": 1.4228068590164185,
"learning_rate": 3.3601485289109446e-06,
"loss": 0.1661,
"step": 7340
},
{
"epoch": 114.84375,
"grad_norm": 1.2065403461456299,
"learning_rate": 3.259951247208148e-06,
"loss": 0.1649,
"step": 7350
},
{
"epoch": 115.0,
"grad_norm": 1.54328191280365,
"learning_rate": 3.1612458167302518e-06,
"loss": 0.1614,
"step": 7360
},
{
"epoch": 115.15625,
"grad_norm": 1.7460312843322754,
"learning_rate": 3.064033759637064e-06,
"loss": 0.165,
"step": 7370
},
{
"epoch": 115.3125,
"grad_norm": 1.4038829803466797,
"learning_rate": 2.968316575058816e-06,
"loss": 0.1602,
"step": 7380
},
{
"epoch": 115.46875,
"grad_norm": 1.490884780883789,
"learning_rate": 2.8740957390728796e-06,
"loss": 0.1662,
"step": 7390
},
{
"epoch": 115.625,
"grad_norm": 1.2959210872650146,
"learning_rate": 2.7813727046810846e-06,
"loss": 0.1649,
"step": 7400
},
{
"epoch": 115.78125,
"grad_norm": 1.3744455575942993,
"learning_rate": 2.690148901787337e-06,
"loss": 0.1706,
"step": 7410
},
{
"epoch": 115.9375,
"grad_norm": 1.475809931755066,
"learning_rate": 2.600425737175549e-06,
"loss": 0.1661,
"step": 7420
},
{
"epoch": 116.09375,
"grad_norm": 1.2936959266662598,
"learning_rate": 2.5122045944878882e-06,
"loss": 0.171,
"step": 7430
},
{
"epoch": 116.25,
"grad_norm": 1.2538620233535767,
"learning_rate": 2.4254868342035054e-06,
"loss": 0.1666,
"step": 7440
},
{
"epoch": 116.40625,
"grad_norm": 1.0937186479568481,
"learning_rate": 2.3402737936175425e-06,
"loss": 0.1633,
"step": 7450
},
{
"epoch": 116.5625,
"grad_norm": 1.490662932395935,
"learning_rate": 2.2565667868205133e-06,
"loss": 0.1652,
"step": 7460
},
{
"epoch": 116.71875,
"grad_norm": 1.2698343992233276,
"learning_rate": 2.174367104677999e-06,
"loss": 0.1676,
"step": 7470
},
{
"epoch": 116.875,
"grad_norm": 1.3303751945495605,
"learning_rate": 2.093676014810797e-06,
"loss": 0.1623,
"step": 7480
},
{
"epoch": 117.03125,
"grad_norm": 1.4380319118499756,
"learning_rate": 2.014494761575314e-06,
"loss": 0.1603,
"step": 7490
},
{
"epoch": 117.1875,
"grad_norm": 1.2588881254196167,
"learning_rate": 1.9368245660444372e-06,
"loss": 0.1652,
"step": 7500
},
{
"epoch": 117.34375,
"grad_norm": 1.4556334018707275,
"learning_rate": 1.8606666259886719e-06,
"loss": 0.1623,
"step": 7510
},
{
"epoch": 117.5,
"grad_norm": 1.451319932937622,
"learning_rate": 1.7860221158576662e-06,
"loss": 0.1696,
"step": 7520
},
{
"epoch": 117.65625,
"grad_norm": 1.4228541851043701,
"learning_rate": 1.712892186762083e-06,
"loss": 0.1616,
"step": 7530
},
{
"epoch": 117.8125,
"grad_norm": 1.5599255561828613,
"learning_rate": 1.6412779664559119e-06,
"loss": 0.1641,
"step": 7540
},
{
"epoch": 117.96875,
"grad_norm": 1.4440757036209106,
"learning_rate": 1.5711805593190188e-06,
"loss": 0.1676,
"step": 7550
},
{
"epoch": 118.125,
"grad_norm": 1.2332494258880615,
"learning_rate": 1.5026010463401352e-06,
"loss": 0.1689,
"step": 7560
},
{
"epoch": 118.28125,
"grad_norm": 1.2784188985824585,
"learning_rate": 1.4355404851001952e-06,
"loss": 0.1606,
"step": 7570
},
{
"epoch": 118.4375,
"grad_norm": 1.3040153980255127,
"learning_rate": 1.3699999097560034e-06,
"loss": 0.1652,
"step": 7580
},
{
"epoch": 118.59375,
"grad_norm": 1.331621766090393,
"learning_rate": 1.3059803310243147e-06,
"loss": 0.165,
"step": 7590
},
{
"epoch": 118.75,
"grad_norm": 1.4870508909225464,
"learning_rate": 1.2434827361662237e-06,
"loss": 0.1654,
"step": 7600
},
{
"epoch": 118.90625,
"grad_norm": 1.3456453084945679,
"learning_rate": 1.1825080889719563e-06,
"loss": 0.1638,
"step": 7610
},
{
"epoch": 119.0625,
"grad_norm": 1.5864441394805908,
"learning_rate": 1.1230573297460022e-06,
"loss": 0.1668,
"step": 7620
},
{
"epoch": 119.21875,
"grad_norm": 1.2966445684432983,
"learning_rate": 1.0651313752926052e-06,
"loss": 0.1644,
"step": 7630
},
{
"epoch": 119.375,
"grad_norm": 1.3712732791900635,
"learning_rate": 1.0087311189016623e-06,
"loss": 0.1641,
"step": 7640
},
{
"epoch": 119.53125,
"grad_norm": 1.4837236404418945,
"learning_rate": 9.538574303348813e-07,
"loss": 0.1639,
"step": 7650
},
{
"epoch": 119.6875,
"grad_norm": 1.2135469913482666,
"learning_rate": 9.005111558124223e-07,
"loss": 0.1648,
"step": 7660
},
{
"epoch": 119.84375,
"grad_norm": 1.3776029348373413,
"learning_rate": 8.486931179998325e-07,
"loss": 0.1646,
"step": 7670
},
{
"epoch": 120.0,
"grad_norm": 1.459794521331787,
"learning_rate": 7.984041159953659e-07,
"loss": 0.1681,
"step": 7680
},
{
"epoch": 120.15625,
"grad_norm": 1.116540551185608,
"learning_rate": 7.496449253176274e-07,
"loss": 0.1627,
"step": 7690
},
{
"epoch": 120.3125,
"grad_norm": 1.203869104385376,
"learning_rate": 7.0241629789366e-07,
"loss": 0.1655,
"step": 7700
},
{
"epoch": 120.46875,
"grad_norm": 1.3655177354812622,
"learning_rate": 6.567189620473424e-07,
"loss": 0.1689,
"step": 7710
},
{
"epoch": 120.625,
"grad_norm": 1.3210086822509766,
"learning_rate": 6.12553622488099e-07,
"loss": 0.164,
"step": 7720
},
{
"epoch": 120.78125,
"grad_norm": 1.2926894426345825,
"learning_rate": 5.699209603001076e-07,
"loss": 0.1651,
"step": 7730
},
{
"epoch": 120.9375,
"grad_norm": 1.445736050605774,
"learning_rate": 5.288216329317641e-07,
"loss": 0.1626,
"step": 7740
},
{
"epoch": 121.09375,
"grad_norm": 1.3684810400009155,
"learning_rate": 4.892562741855455e-07,
"loss": 0.1661,
"step": 7750
},
{
"epoch": 121.25,
"grad_norm": 1.2554900646209717,
"learning_rate": 4.512254942082073e-07,
"loss": 0.1561,
"step": 7760
},
{
"epoch": 121.40625,
"grad_norm": 1.4690603017807007,
"learning_rate": 4.1472987948143473e-07,
"loss": 0.1674,
"step": 7770
},
{
"epoch": 121.5625,
"grad_norm": 1.4324357509613037,
"learning_rate": 3.797699928127507e-07,
"loss": 0.1698,
"step": 7780
},
{
"epoch": 121.71875,
"grad_norm": 1.0271071195602417,
"learning_rate": 3.463463733268557e-07,
"loss": 0.1661,
"step": 7790
},
{
"epoch": 121.875,
"grad_norm": 1.151157021522522,
"learning_rate": 3.144595364573233e-07,
"loss": 0.16,
"step": 7800
},
{
"epoch": 122.03125,
"grad_norm": 1.1873737573623657,
"learning_rate": 2.841099739386066e-07,
"loss": 0.1667,
"step": 7810
},
{
"epoch": 122.1875,
"grad_norm": 1.3843321800231934,
"learning_rate": 2.552981537985111e-07,
"loss": 0.1608,
"step": 7820
},
{
"epoch": 122.34375,
"grad_norm": 1.2830843925476074,
"learning_rate": 2.2802452035093346e-07,
"loss": 0.1639,
"step": 7830
},
{
"epoch": 122.5,
"grad_norm": 1.2948521375656128,
"learning_rate": 2.022894941890674e-07,
"loss": 0.167,
"step": 7840
},
{
"epoch": 122.65625,
"grad_norm": 1.2428685426712036,
"learning_rate": 1.7809347217881966e-07,
"loss": 0.161,
"step": 7850
},
{
"epoch": 122.8125,
"grad_norm": 1.55159592628479,
"learning_rate": 1.5543682745280396e-07,
"loss": 0.1648,
"step": 7860
},
{
"epoch": 122.96875,
"grad_norm": 1.3191789388656616,
"learning_rate": 1.3431990940450113e-07,
"loss": 0.1698,
"step": 7870
},
{
"epoch": 123.125,
"grad_norm": 1.4506255388259888,
"learning_rate": 1.1474304368291888e-07,
"loss": 0.1676,
"step": 7880
},
{
"epoch": 123.28125,
"grad_norm": 1.2544479370117188,
"learning_rate": 9.670653218752934e-08,
"loss": 0.1649,
"step": 7890
},
{
"epoch": 123.4375,
"grad_norm": 1.184103012084961,
"learning_rate": 8.021065306366149e-08,
"loss": 0.16,
"step": 7900
},
{
"epoch": 123.59375,
"grad_norm": 1.0644958019256592,
"learning_rate": 6.525566069817135e-08,
"loss": 0.1632,
"step": 7910
},
{
"epoch": 123.75,
"grad_norm": 1.4110705852508545,
"learning_rate": 5.18417857155562e-08,
"loss": 0.1688,
"step": 7920
},
{
"epoch": 123.90625,
"grad_norm": 1.2266552448272705,
"learning_rate": 3.996923497434635e-08,
"loss": 0.1702,
"step": 7930
},
{
"epoch": 124.0625,
"grad_norm": 1.2383413314819336,
"learning_rate": 2.9638191563974294e-08,
"loss": 0.1597,
"step": 7940
},
{
"epoch": 124.21875,
"grad_norm": 1.2518082857131958,
"learning_rate": 2.084881480188816e-08,
"loss": 0.1639,
"step": 7950
},
{
"epoch": 124.375,
"grad_norm": 1.429856777191162,
"learning_rate": 1.3601240231175816e-08,
"loss": 0.1697,
"step": 7960
},
{
"epoch": 124.53125,
"grad_norm": 1.521837592124939,
"learning_rate": 7.895579618388827e-09,
"loss": 0.1649,
"step": 7970
},
{
"epoch": 124.6875,
"grad_norm": 1.2289235591888428,
"learning_rate": 3.731920951877133e-09,
"loss": 0.1648,
"step": 7980
},
{
"epoch": 124.84375,
"grad_norm": 1.900389313697815,
"learning_rate": 1.1103284404012648e-09,
"loss": 0.1635,
"step": 7990
},
{
"epoch": 125.0,
"grad_norm": 1.170914888381958,
"learning_rate": 3.084251216645129e-11,
"loss": 0.1648,
"step": 8000
}
],
"logging_steps": 10,
"max_steps": 8000,
"num_input_tokens_seen": 0,
"num_train_epochs": 125,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.7740235738629325e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}