rqadri's picture
Upload folder using huggingface_hub
3ea8a82 verified
Raw History Blame Contribute Delete
49.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.18621837549933423,
"eval_steps": 500,
"global_step": 2797,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006657789613848203,
"grad_norm": 235.8969256890802,
"learning_rate": 1.1061797752808989e-06,
"loss": 1.7364160537719726,
"step": 10
},
{
"epoch": 0.0013315579227696406,
"grad_norm": 454.87323539275144,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.213594627380371,
"step": 20
},
{
"epoch": 0.0019973368841544607,
"grad_norm": 56.07729488657429,
"learning_rate": 3.342134831460675e-06,
"loss": 0.7420229911804199,
"step": 30
},
{
"epoch": 0.002663115845539281,
"grad_norm": 150.0094098857343,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.5900703430175781,
"step": 40
},
{
"epoch": 0.003328894806924101,
"grad_norm": 10.492010241968945,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.6019559383392334,
"step": 50
},
{
"epoch": 0.0039946737683089215,
"grad_norm": 58.9158199879992,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.5498143672943115,
"step": 60
},
{
"epoch": 0.004660452729693742,
"grad_norm": 38.7751489526521,
"learning_rate": 7.814044943820226e-06,
"loss": 0.5471227645874024,
"step": 70
},
{
"epoch": 0.005326231691078562,
"grad_norm": 44.65639947244615,
"learning_rate": 8.932022471910112e-06,
"loss": 0.5947897911071778,
"step": 80
},
{
"epoch": 0.005992010652463382,
"grad_norm": 42.021501768665296,
"learning_rate": 1.005e-05,
"loss": 0.5396684169769287,
"step": 90
},
{
"epoch": 0.006657789613848202,
"grad_norm": 102.55835870245248,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.5591165065765381,
"step": 100
},
{
"epoch": 0.007323568575233023,
"grad_norm": 63.550219661252996,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.5836419582366943,
"step": 110
},
{
"epoch": 0.007989347536617843,
"grad_norm": 29.408506432576182,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.5985694885253906,
"step": 120
},
{
"epoch": 0.008655126498002663,
"grad_norm": 8.348962303851348,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.6041747093200683,
"step": 130
},
{
"epoch": 0.009320905459387484,
"grad_norm": 6.192722737720432,
"learning_rate": 1.563988764044944e-05,
"loss": 0.5608590602874756,
"step": 140
},
{
"epoch": 0.009986684420772303,
"grad_norm": 9.660859490705201,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.5558311462402343,
"step": 150
},
{
"epoch": 0.010652463382157125,
"grad_norm": 4.046195620530419,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.5265848636627197,
"step": 160
},
{
"epoch": 0.011318242343541944,
"grad_norm": 7.900205661847944,
"learning_rate": 1.89938202247191e-05,
"loss": 0.5326771259307861,
"step": 170
},
{
"epoch": 0.011984021304926764,
"grad_norm": 41.222719067968995,
"learning_rate": 2e-05,
"loss": 0.5493779182434082,
"step": 180
},
{
"epoch": 0.012649800266311585,
"grad_norm": 11.51285962029523,
"learning_rate": 2e-05,
"loss": 0.5718745708465576,
"step": 190
},
{
"epoch": 0.013315579227696404,
"grad_norm": 69.2723017393699,
"learning_rate": 2e-05,
"loss": 0.6055091381072998,
"step": 200
},
{
"epoch": 0.013981358189081226,
"grad_norm": 31.148285201072877,
"learning_rate": 2e-05,
"loss": 0.5791210174560547,
"step": 210
},
{
"epoch": 0.014647137150466045,
"grad_norm": 14.988316338719642,
"learning_rate": 2e-05,
"loss": 0.5553981781005859,
"step": 220
},
{
"epoch": 0.015312916111850865,
"grad_norm": 5.657204329585924,
"learning_rate": 2e-05,
"loss": 0.5496438026428223,
"step": 230
},
{
"epoch": 0.015978695073235686,
"grad_norm": 4.013199968748617,
"learning_rate": 2e-05,
"loss": 0.5337262630462647,
"step": 240
},
{
"epoch": 0.016644474034620507,
"grad_norm": 39.164202345517396,
"learning_rate": 2e-05,
"loss": 0.5856839179992676,
"step": 250
},
{
"epoch": 0.017310252996005325,
"grad_norm": 8.999791467884044,
"learning_rate": 2e-05,
"loss": 0.6439942836761474,
"step": 260
},
{
"epoch": 0.017976031957390146,
"grad_norm": 12.92963394726804,
"learning_rate": 2e-05,
"loss": 0.6030735015869141,
"step": 270
},
{
"epoch": 0.018641810918774968,
"grad_norm": 10.165083476794843,
"learning_rate": 2e-05,
"loss": 0.5539159297943115,
"step": 280
},
{
"epoch": 0.019307589880159785,
"grad_norm": 7.211149855639102,
"learning_rate": 2e-05,
"loss": 0.5592308044433594,
"step": 290
},
{
"epoch": 0.019973368841544607,
"grad_norm": 5.551546545660331,
"learning_rate": 2e-05,
"loss": 0.5586291790008545,
"step": 300
},
{
"epoch": 0.020639147802929428,
"grad_norm": 14.196903064174798,
"learning_rate": 2e-05,
"loss": 0.5297722816467285,
"step": 310
},
{
"epoch": 0.02130492676431425,
"grad_norm": 8.695901179045977,
"learning_rate": 2e-05,
"loss": 0.5313222885131836,
"step": 320
},
{
"epoch": 0.021970705725699067,
"grad_norm": 73.2391576714464,
"learning_rate": 2e-05,
"loss": 0.5299333572387696,
"step": 330
},
{
"epoch": 0.022636484687083888,
"grad_norm": 10.173641738816025,
"learning_rate": 2e-05,
"loss": 0.5360995292663574,
"step": 340
},
{
"epoch": 0.02330226364846871,
"grad_norm": 6.198967892661863,
"learning_rate": 2e-05,
"loss": 0.5845535278320313,
"step": 350
},
{
"epoch": 0.023968042609853527,
"grad_norm": 5.164050381832549,
"learning_rate": 2e-05,
"loss": 0.5637627601623535,
"step": 360
},
{
"epoch": 0.02463382157123835,
"grad_norm": 4.33608376434225,
"learning_rate": 2e-05,
"loss": 0.5674751281738282,
"step": 370
},
{
"epoch": 0.02529960053262317,
"grad_norm": 6.676449836472397,
"learning_rate": 2e-05,
"loss": 0.5269466400146484,
"step": 380
},
{
"epoch": 0.02596537949400799,
"grad_norm": 34.60400985665482,
"learning_rate": 2e-05,
"loss": 0.5582229614257812,
"step": 390
},
{
"epoch": 0.02663115845539281,
"grad_norm": 2.0382084950493686,
"learning_rate": 2e-05,
"loss": 0.533026123046875,
"step": 400
},
{
"epoch": 0.02729693741677763,
"grad_norm": 12.318088970720998,
"learning_rate": 2e-05,
"loss": 0.5565634727478027,
"step": 410
},
{
"epoch": 0.02796271637816245,
"grad_norm": 5.5824838209854395,
"learning_rate": 2e-05,
"loss": 0.5633725643157959,
"step": 420
},
{
"epoch": 0.02862849533954727,
"grad_norm": 2.2430627781253163,
"learning_rate": 2e-05,
"loss": 0.5287482738494873,
"step": 430
},
{
"epoch": 0.02929427430093209,
"grad_norm": 3.386109702013143,
"learning_rate": 2e-05,
"loss": 0.5164429187774658,
"step": 440
},
{
"epoch": 0.02996005326231691,
"grad_norm": 4.52797743621058,
"learning_rate": 2e-05,
"loss": 0.49767842292785647,
"step": 450
},
{
"epoch": 0.03062583222370173,
"grad_norm": 8.559435723184958,
"learning_rate": 2e-05,
"loss": 0.545814323425293,
"step": 460
},
{
"epoch": 0.031291611185086554,
"grad_norm": 2.4213157821392652,
"learning_rate": 2e-05,
"loss": 0.5681285858154297,
"step": 470
},
{
"epoch": 0.03195739014647137,
"grad_norm": 5.002642882306833,
"learning_rate": 2e-05,
"loss": 0.5424331188201904,
"step": 480
},
{
"epoch": 0.03262316910785619,
"grad_norm": 11.069858537054463,
"learning_rate": 2e-05,
"loss": 0.5557833671569824,
"step": 490
},
{
"epoch": 0.033288948069241014,
"grad_norm": 10.790349593013875,
"learning_rate": 2e-05,
"loss": 0.5815642833709717,
"step": 500
},
{
"epoch": 0.03395472703062583,
"grad_norm": 3.3611249244779615,
"learning_rate": 2e-05,
"loss": 0.5631533622741699,
"step": 510
},
{
"epoch": 0.03462050599201065,
"grad_norm": 5.726005842469412,
"learning_rate": 2e-05,
"loss": 0.569805908203125,
"step": 520
},
{
"epoch": 0.035286284953395475,
"grad_norm": 14.583572065182688,
"learning_rate": 2e-05,
"loss": 0.5803721904754638,
"step": 530
},
{
"epoch": 0.03595206391478029,
"grad_norm": 4.297489848956763,
"learning_rate": 2e-05,
"loss": 0.5814414024353027,
"step": 540
},
{
"epoch": 0.03661784287616511,
"grad_norm": 5.944128076013417,
"learning_rate": 2e-05,
"loss": 0.564124059677124,
"step": 550
},
{
"epoch": 0.037283621837549935,
"grad_norm": 3.2423104393351307,
"learning_rate": 2e-05,
"loss": 0.6266448974609375,
"step": 560
},
{
"epoch": 0.03794940079893475,
"grad_norm": 43.85395650355969,
"learning_rate": 2e-05,
"loss": 0.572756004333496,
"step": 570
},
{
"epoch": 0.03861517976031957,
"grad_norm": 3.9530294075080237,
"learning_rate": 2e-05,
"loss": 0.5774184226989746,
"step": 580
},
{
"epoch": 0.039280958721704395,
"grad_norm": 11.062180432200206,
"learning_rate": 2e-05,
"loss": 0.5977572441101074,
"step": 590
},
{
"epoch": 0.03994673768308921,
"grad_norm": 6.443399893752391,
"learning_rate": 2e-05,
"loss": 0.5534317016601562,
"step": 600
},
{
"epoch": 0.04061251664447404,
"grad_norm": 18.52097082878789,
"learning_rate": 2e-05,
"loss": 0.554855489730835,
"step": 610
},
{
"epoch": 0.041278295605858856,
"grad_norm": 3.3934097540968753,
"learning_rate": 2e-05,
"loss": 0.5600467681884765,
"step": 620
},
{
"epoch": 0.041944074567243674,
"grad_norm": 7.706610686679725,
"learning_rate": 2e-05,
"loss": 0.5719536781311035,
"step": 630
},
{
"epoch": 0.0426098535286285,
"grad_norm": 4.425720098798294,
"learning_rate": 2e-05,
"loss": 0.5739123344421386,
"step": 640
},
{
"epoch": 0.043275632490013316,
"grad_norm": 1.7190184550483938,
"learning_rate": 2e-05,
"loss": 0.546561336517334,
"step": 650
},
{
"epoch": 0.043941411451398134,
"grad_norm": 8.48706166477749,
"learning_rate": 2e-05,
"loss": 0.5902390480041504,
"step": 660
},
{
"epoch": 0.04460719041278296,
"grad_norm": 2.034988576266325,
"learning_rate": 2e-05,
"loss": 0.5392104148864746,
"step": 670
},
{
"epoch": 0.045272969374167776,
"grad_norm": 2.8908851748120794,
"learning_rate": 2e-05,
"loss": 0.5352708339691162,
"step": 680
},
{
"epoch": 0.045938748335552594,
"grad_norm": 11.225417918014127,
"learning_rate": 2e-05,
"loss": 0.5520500183105469,
"step": 690
},
{
"epoch": 0.04660452729693742,
"grad_norm": 3.063402372041763,
"learning_rate": 2e-05,
"loss": 0.5145657539367676,
"step": 700
},
{
"epoch": 0.04727030625832224,
"grad_norm": 6.11767198069841,
"learning_rate": 2e-05,
"loss": 0.5366698265075683,
"step": 710
},
{
"epoch": 0.047936085219707054,
"grad_norm": 2.25307348329916,
"learning_rate": 2e-05,
"loss": 0.551693344116211,
"step": 720
},
{
"epoch": 0.04860186418109188,
"grad_norm": 4.55732420618067,
"learning_rate": 2e-05,
"loss": 0.5258326530456543,
"step": 730
},
{
"epoch": 0.0492676431424767,
"grad_norm": 3.4702813032613635,
"learning_rate": 2e-05,
"loss": 0.5708011627197266,
"step": 740
},
{
"epoch": 0.049933422103861515,
"grad_norm": 2.195262021450343,
"learning_rate": 2e-05,
"loss": 0.5412505149841309,
"step": 750
},
{
"epoch": 0.05059920106524634,
"grad_norm": 4.418015821922213,
"learning_rate": 2e-05,
"loss": 0.5207402229309082,
"step": 760
},
{
"epoch": 0.05126498002663116,
"grad_norm": 2.6654030789993564,
"learning_rate": 2e-05,
"loss": 0.5646546840667724,
"step": 770
},
{
"epoch": 0.05193075898801598,
"grad_norm": 1.463701773082979,
"learning_rate": 2e-05,
"loss": 0.5652810573577881,
"step": 780
},
{
"epoch": 0.0525965379494008,
"grad_norm": 1.1644755512746217,
"learning_rate": 2e-05,
"loss": 0.5259516716003418,
"step": 790
},
{
"epoch": 0.05326231691078562,
"grad_norm": 1.4724936122706427,
"learning_rate": 2e-05,
"loss": 0.5653295040130615,
"step": 800
},
{
"epoch": 0.05392809587217044,
"grad_norm": 1.451947446736679,
"learning_rate": 2e-05,
"loss": 0.523100757598877,
"step": 810
},
{
"epoch": 0.05459387483355526,
"grad_norm": 1.2453818129937126,
"learning_rate": 2e-05,
"loss": 0.5161214351654053,
"step": 820
},
{
"epoch": 0.05525965379494008,
"grad_norm": 8.366800471437534,
"learning_rate": 2e-05,
"loss": 0.5262998580932617,
"step": 830
},
{
"epoch": 0.0559254327563249,
"grad_norm": 2.239038325903991,
"learning_rate": 2e-05,
"loss": 0.5443816184997559,
"step": 840
},
{
"epoch": 0.05659121171770972,
"grad_norm": 2.1435511292161107,
"learning_rate": 2e-05,
"loss": 0.5617143154144287,
"step": 850
},
{
"epoch": 0.05725699067909454,
"grad_norm": 1.6208532172828776,
"learning_rate": 2e-05,
"loss": 0.5223365783691406,
"step": 860
},
{
"epoch": 0.05792276964047936,
"grad_norm": 1.552594266356934,
"learning_rate": 2e-05,
"loss": 0.5362037181854248,
"step": 870
},
{
"epoch": 0.05858854860186418,
"grad_norm": 1.732568906745257,
"learning_rate": 2e-05,
"loss": 0.5471210479736328,
"step": 880
},
{
"epoch": 0.059254327563249,
"grad_norm": 3.922703835213154,
"learning_rate": 2e-05,
"loss": 0.5224350929260254,
"step": 890
},
{
"epoch": 0.05992010652463382,
"grad_norm": 1.8443327131294107,
"learning_rate": 2e-05,
"loss": 0.5521405220031739,
"step": 900
},
{
"epoch": 0.06058588548601864,
"grad_norm": 18.041405872499528,
"learning_rate": 2e-05,
"loss": 0.5715356349945069,
"step": 910
},
{
"epoch": 0.06125166444740346,
"grad_norm": 1.975134294116238,
"learning_rate": 2e-05,
"loss": 0.5748860836029053,
"step": 920
},
{
"epoch": 0.061917443408788284,
"grad_norm": 1.0758923518754975,
"learning_rate": 2e-05,
"loss": 0.5490294456481933,
"step": 930
},
{
"epoch": 0.06258322237017311,
"grad_norm": 3.1323788152588383,
"learning_rate": 2e-05,
"loss": 0.5143556118011474,
"step": 940
},
{
"epoch": 0.06324900133155792,
"grad_norm": 2.312493834654343,
"learning_rate": 2e-05,
"loss": 0.4982307434082031,
"step": 950
},
{
"epoch": 0.06391478029294274,
"grad_norm": 1.8957287997713355,
"learning_rate": 2e-05,
"loss": 0.5262045383453369,
"step": 960
},
{
"epoch": 0.06458055925432757,
"grad_norm": 1.180103969220926,
"learning_rate": 2e-05,
"loss": 0.5141486167907715,
"step": 970
},
{
"epoch": 0.06524633821571238,
"grad_norm": 1.5035058067710745,
"learning_rate": 2e-05,
"loss": 0.5460874080657959,
"step": 980
},
{
"epoch": 0.0659121171770972,
"grad_norm": 6.3846590160328445,
"learning_rate": 2e-05,
"loss": 0.5823369026184082,
"step": 990
},
{
"epoch": 0.06657789613848203,
"grad_norm": 3.231925320560546,
"learning_rate": 2e-05,
"loss": 0.5262383460998535,
"step": 1000
},
{
"epoch": 0.06724367509986684,
"grad_norm": 2.315135540989524,
"learning_rate": 2e-05,
"loss": 0.5371553421020507,
"step": 1010
},
{
"epoch": 0.06790945406125166,
"grad_norm": 4.496108973654741,
"learning_rate": 2e-05,
"loss": 0.5573309898376465,
"step": 1020
},
{
"epoch": 0.06857523302263649,
"grad_norm": 2.166198851365719,
"learning_rate": 2e-05,
"loss": 0.5329337120056152,
"step": 1030
},
{
"epoch": 0.0692410119840213,
"grad_norm": 1.0425981603986465,
"learning_rate": 2e-05,
"loss": 0.5511478424072266,
"step": 1040
},
{
"epoch": 0.06990679094540612,
"grad_norm": 2.2014938720266937,
"learning_rate": 2e-05,
"loss": 0.5591570854187011,
"step": 1050
},
{
"epoch": 0.07057256990679095,
"grad_norm": 4.876753934559851,
"learning_rate": 2e-05,
"loss": 0.5555380821228028,
"step": 1060
},
{
"epoch": 0.07123834886817576,
"grad_norm": 2.857934925368923,
"learning_rate": 2e-05,
"loss": 0.5619023323059082,
"step": 1070
},
{
"epoch": 0.07190412782956059,
"grad_norm": 2.3266365189381832,
"learning_rate": 2e-05,
"loss": 0.5243755340576172,
"step": 1080
},
{
"epoch": 0.07256990679094541,
"grad_norm": 1.9277942910270909,
"learning_rate": 2e-05,
"loss": 0.5400153160095215,
"step": 1090
},
{
"epoch": 0.07323568575233022,
"grad_norm": 1.1764647055530073,
"learning_rate": 2e-05,
"loss": 0.5243360519409179,
"step": 1100
},
{
"epoch": 0.07390146471371505,
"grad_norm": 1.1058092372935702,
"learning_rate": 2e-05,
"loss": 0.544226598739624,
"step": 1110
},
{
"epoch": 0.07456724367509987,
"grad_norm": 2.841445800892787,
"learning_rate": 2e-05,
"loss": 0.550119161605835,
"step": 1120
},
{
"epoch": 0.07523302263648468,
"grad_norm": 1.4562183729193323,
"learning_rate": 2e-05,
"loss": 0.5335268020629883,
"step": 1130
},
{
"epoch": 0.0758988015978695,
"grad_norm": 2.5489855904207768,
"learning_rate": 2e-05,
"loss": 0.5286331176757812,
"step": 1140
},
{
"epoch": 0.07656458055925433,
"grad_norm": 4.365595677900958,
"learning_rate": 2e-05,
"loss": 0.5628453254699707,
"step": 1150
},
{
"epoch": 0.07723035952063914,
"grad_norm": 1.8349543210000938,
"learning_rate": 2e-05,
"loss": 0.5462607860565185,
"step": 1160
},
{
"epoch": 0.07789613848202397,
"grad_norm": 2.5076678969731416,
"learning_rate": 2e-05,
"loss": 0.5364606380462646,
"step": 1170
},
{
"epoch": 0.07856191744340879,
"grad_norm": 8.268429157439165,
"learning_rate": 2e-05,
"loss": 0.5362617492675781,
"step": 1180
},
{
"epoch": 0.0792276964047936,
"grad_norm": 2.4583852803700648,
"learning_rate": 2e-05,
"loss": 0.5119704246520996,
"step": 1190
},
{
"epoch": 0.07989347536617843,
"grad_norm": 2.1784687391986437,
"learning_rate": 2e-05,
"loss": 0.5308552742004394,
"step": 1200
},
{
"epoch": 0.08055925432756325,
"grad_norm": 1.2303988845257494,
"learning_rate": 2e-05,
"loss": 0.539813232421875,
"step": 1210
},
{
"epoch": 0.08122503328894808,
"grad_norm": 8.727296596355615,
"learning_rate": 2e-05,
"loss": 0.5063223838806152,
"step": 1220
},
{
"epoch": 0.08189081225033289,
"grad_norm": 1.6512049866144545,
"learning_rate": 2e-05,
"loss": 0.5685141563415528,
"step": 1230
},
{
"epoch": 0.08255659121171771,
"grad_norm": 1.8297074024323168,
"learning_rate": 2e-05,
"loss": 0.5602605819702149,
"step": 1240
},
{
"epoch": 0.08322237017310254,
"grad_norm": 1.050479235459917,
"learning_rate": 2e-05,
"loss": 0.5343076229095459,
"step": 1250
},
{
"epoch": 0.08388814913448735,
"grad_norm": 0.9525187920106077,
"learning_rate": 2e-05,
"loss": 0.5467718124389649,
"step": 1260
},
{
"epoch": 0.08455392809587217,
"grad_norm": 5.619490448508517,
"learning_rate": 2e-05,
"loss": 0.5230856895446777,
"step": 1270
},
{
"epoch": 0.085219707057257,
"grad_norm": 31.785133935281873,
"learning_rate": 2e-05,
"loss": 0.5371414184570312,
"step": 1280
},
{
"epoch": 0.08588548601864181,
"grad_norm": 1.6841285905458097,
"learning_rate": 2e-05,
"loss": 0.5382813930511474,
"step": 1290
},
{
"epoch": 0.08655126498002663,
"grad_norm": 3.074425583552512,
"learning_rate": 2e-05,
"loss": 0.5496634483337403,
"step": 1300
},
{
"epoch": 0.08721704394141146,
"grad_norm": 1.1493550672820814,
"learning_rate": 2e-05,
"loss": 0.515593433380127,
"step": 1310
},
{
"epoch": 0.08788282290279627,
"grad_norm": 5.47637808249577,
"learning_rate": 2e-05,
"loss": 0.5221123695373535,
"step": 1320
},
{
"epoch": 0.08854860186418109,
"grad_norm": 2.043477169717248,
"learning_rate": 2e-05,
"loss": 0.5497914314270019,
"step": 1330
},
{
"epoch": 0.08921438082556592,
"grad_norm": 2.055175898286977,
"learning_rate": 2e-05,
"loss": 0.5801389694213868,
"step": 1340
},
{
"epoch": 0.08988015978695073,
"grad_norm": 2.1668021346153115,
"learning_rate": 2e-05,
"loss": 0.517332124710083,
"step": 1350
},
{
"epoch": 0.09054593874833555,
"grad_norm": 1.5460307704723255,
"learning_rate": 2e-05,
"loss": 0.5286785125732422,
"step": 1360
},
{
"epoch": 0.09121171770972038,
"grad_norm": 7.339714501931053,
"learning_rate": 2e-05,
"loss": 0.5441049098968506,
"step": 1370
},
{
"epoch": 0.09187749667110519,
"grad_norm": 2.9991980124421977,
"learning_rate": 2e-05,
"loss": 0.5296726226806641,
"step": 1380
},
{
"epoch": 0.09254327563249001,
"grad_norm": 1.9704344709203743,
"learning_rate": 2e-05,
"loss": 0.5755397796630859,
"step": 1390
},
{
"epoch": 0.09320905459387484,
"grad_norm": 3.1974312777990006,
"learning_rate": 2e-05,
"loss": 0.5350383758544922,
"step": 1400
},
{
"epoch": 0.09387483355525965,
"grad_norm": 4.835254285595909,
"learning_rate": 2e-05,
"loss": 0.5583718299865723,
"step": 1410
},
{
"epoch": 0.09454061251664447,
"grad_norm": 4.086077808375664,
"learning_rate": 2e-05,
"loss": 0.5181577682495118,
"step": 1420
},
{
"epoch": 0.0952063914780293,
"grad_norm": 1.3712201211608543,
"learning_rate": 2e-05,
"loss": 0.5456793785095215,
"step": 1430
},
{
"epoch": 0.09587217043941411,
"grad_norm": 2.3710137180040323,
"learning_rate": 2e-05,
"loss": 0.6178105354309082,
"step": 1440
},
{
"epoch": 0.09653794940079893,
"grad_norm": 1.6819343647716947,
"learning_rate": 2e-05,
"loss": 0.5533603191375732,
"step": 1450
},
{
"epoch": 0.09720372836218376,
"grad_norm": 1.2365582424751604,
"learning_rate": 2e-05,
"loss": 0.5364840984344482,
"step": 1460
},
{
"epoch": 0.09786950732356857,
"grad_norm": 2.296453401394828,
"learning_rate": 2e-05,
"loss": 0.5314805030822753,
"step": 1470
},
{
"epoch": 0.0985352862849534,
"grad_norm": 1.24283562111146,
"learning_rate": 2e-05,
"loss": 0.5186245918273926,
"step": 1480
},
{
"epoch": 0.09920106524633822,
"grad_norm": 1.3611004524534043,
"learning_rate": 2e-05,
"loss": 0.524735975265503,
"step": 1490
},
{
"epoch": 0.09986684420772303,
"grad_norm": 3.9363806116623707,
"learning_rate": 2e-05,
"loss": 0.550562858581543,
"step": 1500
},
{
"epoch": 0.10053262316910785,
"grad_norm": 1.972796502107073,
"learning_rate": 2e-05,
"loss": 0.528165340423584,
"step": 1510
},
{
"epoch": 0.10119840213049268,
"grad_norm": 1.1667369026127106,
"learning_rate": 2e-05,
"loss": 0.5359246730804443,
"step": 1520
},
{
"epoch": 0.10186418109187749,
"grad_norm": 1.8610540585135436,
"learning_rate": 2e-05,
"loss": 0.52705078125,
"step": 1530
},
{
"epoch": 0.10252996005326231,
"grad_norm": 1.324687718260285,
"learning_rate": 2e-05,
"loss": 0.5163713455200195,
"step": 1540
},
{
"epoch": 0.10319573901464714,
"grad_norm": 2.232586255903306,
"learning_rate": 2e-05,
"loss": 0.5380361557006836,
"step": 1550
},
{
"epoch": 0.10386151797603196,
"grad_norm": 2.8377026825388247,
"learning_rate": 2e-05,
"loss": 0.5271590232849122,
"step": 1560
},
{
"epoch": 0.10452729693741677,
"grad_norm": 1.0232432789770685,
"learning_rate": 2e-05,
"loss": 0.5482112884521484,
"step": 1570
},
{
"epoch": 0.1051930758988016,
"grad_norm": 3.0682327172834913,
"learning_rate": 2e-05,
"loss": 0.5437918663024902,
"step": 1580
},
{
"epoch": 0.10585885486018642,
"grad_norm": 1.3849474396156232,
"learning_rate": 2e-05,
"loss": 0.49279470443725587,
"step": 1590
},
{
"epoch": 0.10652463382157124,
"grad_norm": 2.0299726086559993,
"learning_rate": 2e-05,
"loss": 0.5641605377197265,
"step": 1600
},
{
"epoch": 0.10719041278295606,
"grad_norm": 1.817550197200932,
"learning_rate": 2e-05,
"loss": 0.5258097648620605,
"step": 1610
},
{
"epoch": 0.10785619174434088,
"grad_norm": 1.1651824761649607,
"learning_rate": 2e-05,
"loss": 0.5535667419433594,
"step": 1620
},
{
"epoch": 0.1085219707057257,
"grad_norm": 1.5674773668466062,
"learning_rate": 2e-05,
"loss": 0.5567864894866943,
"step": 1630
},
{
"epoch": 0.10918774966711052,
"grad_norm": 1.1591412095258105,
"learning_rate": 2e-05,
"loss": 0.5203337669372559,
"step": 1640
},
{
"epoch": 0.10985352862849534,
"grad_norm": 1.1505534469163556,
"learning_rate": 2e-05,
"loss": 0.5513369560241699,
"step": 1650
},
{
"epoch": 0.11051930758988016,
"grad_norm": 1.0506019078549658,
"learning_rate": 2e-05,
"loss": 0.5552059173583984,
"step": 1660
},
{
"epoch": 0.11118508655126498,
"grad_norm": 1.2459158335355898,
"learning_rate": 2e-05,
"loss": 0.5355972290039063,
"step": 1670
},
{
"epoch": 0.1118508655126498,
"grad_norm": 1.4295425078712645,
"learning_rate": 2e-05,
"loss": 0.54208984375,
"step": 1680
},
{
"epoch": 0.11251664447403462,
"grad_norm": 1.005803323010581,
"learning_rate": 2e-05,
"loss": 0.5317649841308594,
"step": 1690
},
{
"epoch": 0.11318242343541944,
"grad_norm": 1.2565701592686467,
"learning_rate": 2e-05,
"loss": 0.5282378196716309,
"step": 1700
},
{
"epoch": 0.11384820239680427,
"grad_norm": 0.9550909473074799,
"learning_rate": 2e-05,
"loss": 0.5165553092956543,
"step": 1710
},
{
"epoch": 0.11451398135818908,
"grad_norm": 1.3278097810505283,
"learning_rate": 2e-05,
"loss": 0.5109647750854492,
"step": 1720
},
{
"epoch": 0.1151797603195739,
"grad_norm": 0.9808078433806386,
"learning_rate": 2e-05,
"loss": 0.5323241233825684,
"step": 1730
},
{
"epoch": 0.11584553928095873,
"grad_norm": 1.0288764189211852,
"learning_rate": 2e-05,
"loss": 0.5097640037536622,
"step": 1740
},
{
"epoch": 0.11651131824234354,
"grad_norm": 1.0455186608218316,
"learning_rate": 2e-05,
"loss": 0.5255751609802246,
"step": 1750
},
{
"epoch": 0.11717709720372836,
"grad_norm": 0.9659796277901321,
"learning_rate": 2e-05,
"loss": 0.5367124080657959,
"step": 1760
},
{
"epoch": 0.11784287616511319,
"grad_norm": 2.3356253200269204,
"learning_rate": 2e-05,
"loss": 0.5488951683044434,
"step": 1770
},
{
"epoch": 0.118508655126498,
"grad_norm": 1.1896876383401835,
"learning_rate": 2e-05,
"loss": 0.5306941509246826,
"step": 1780
},
{
"epoch": 0.11917443408788282,
"grad_norm": 1.2306037178037514,
"learning_rate": 2e-05,
"loss": 0.5404845237731933,
"step": 1790
},
{
"epoch": 0.11984021304926765,
"grad_norm": 1.6449345630034806,
"learning_rate": 2e-05,
"loss": 0.5113681316375732,
"step": 1800
},
{
"epoch": 0.12050599201065246,
"grad_norm": 13.948346139257183,
"learning_rate": 2e-05,
"loss": 0.5406979560852051,
"step": 1810
},
{
"epoch": 0.12117177097203728,
"grad_norm": 0.8897567907562994,
"learning_rate": 2e-05,
"loss": 0.545926856994629,
"step": 1820
},
{
"epoch": 0.1218375499334221,
"grad_norm": 1.2112299469668055,
"learning_rate": 2e-05,
"loss": 0.5686315059661865,
"step": 1830
},
{
"epoch": 0.12250332889480692,
"grad_norm": 1.2807189268351256,
"learning_rate": 2e-05,
"loss": 0.53879714012146,
"step": 1840
},
{
"epoch": 0.12316910785619174,
"grad_norm": 1.227450535129884,
"learning_rate": 2e-05,
"loss": 0.5089795112609863,
"step": 1850
},
{
"epoch": 0.12383488681757657,
"grad_norm": 1.2772771188177927,
"learning_rate": 2e-05,
"loss": 0.5080512523651123,
"step": 1860
},
{
"epoch": 0.12450066577896138,
"grad_norm": 0.9281113939706479,
"learning_rate": 2e-05,
"loss": 0.5448731899261474,
"step": 1870
},
{
"epoch": 0.12516644474034622,
"grad_norm": 1.6719732819004125,
"learning_rate": 2e-05,
"loss": 0.5150301933288575,
"step": 1880
},
{
"epoch": 0.12583222370173103,
"grad_norm": 1.133321175275307,
"learning_rate": 2e-05,
"loss": 0.5294013977050781,
"step": 1890
},
{
"epoch": 0.12649800266311584,
"grad_norm": 1.8063664874074499,
"learning_rate": 2e-05,
"loss": 0.5121081352233887,
"step": 1900
},
{
"epoch": 0.12716378162450068,
"grad_norm": 1.7628751698806506,
"learning_rate": 2e-05,
"loss": 0.5407076835632324,
"step": 1910
},
{
"epoch": 0.1278295605858855,
"grad_norm": 1.7994852581180745,
"learning_rate": 2e-05,
"loss": 0.5252192974090576,
"step": 1920
},
{
"epoch": 0.1284953395472703,
"grad_norm": 1.0915149354977984,
"learning_rate": 2e-05,
"loss": 0.525769853591919,
"step": 1930
},
{
"epoch": 0.12916111850865514,
"grad_norm": 2.151838757501649,
"learning_rate": 2e-05,
"loss": 0.51082444190979,
"step": 1940
},
{
"epoch": 0.12982689747003995,
"grad_norm": 1.1402975756842395,
"learning_rate": 2e-05,
"loss": 0.5432845115661621,
"step": 1950
},
{
"epoch": 0.13049267643142476,
"grad_norm": 2.2552774036378183,
"learning_rate": 2e-05,
"loss": 0.5111236572265625,
"step": 1960
},
{
"epoch": 0.1311584553928096,
"grad_norm": 2.3875590430821876,
"learning_rate": 2e-05,
"loss": 0.509479284286499,
"step": 1970
},
{
"epoch": 0.1318242343541944,
"grad_norm": 1.1594488373781606,
"learning_rate": 2e-05,
"loss": 0.5075606822967529,
"step": 1980
},
{
"epoch": 0.13249001331557922,
"grad_norm": 1.1364466542686973,
"learning_rate": 2e-05,
"loss": 0.5161935329437256,
"step": 1990
},
{
"epoch": 0.13315579227696406,
"grad_norm": 33.5151639217467,
"learning_rate": 2e-05,
"loss": 0.5208932876586914,
"step": 2000
},
{
"epoch": 0.13382157123834887,
"grad_norm": 3.718913798025406,
"learning_rate": 2e-05,
"loss": 0.533729362487793,
"step": 2010
},
{
"epoch": 0.13448735019973368,
"grad_norm": 1.1708432227005658,
"learning_rate": 2e-05,
"loss": 0.4863370418548584,
"step": 2020
},
{
"epoch": 0.13515312916111852,
"grad_norm": 1.5381854574243472,
"learning_rate": 2e-05,
"loss": 0.5670979499816895,
"step": 2030
},
{
"epoch": 0.13581890812250333,
"grad_norm": 1.5413767152488844,
"learning_rate": 2e-05,
"loss": 0.5272467613220215,
"step": 2040
},
{
"epoch": 0.13648468708388814,
"grad_norm": 2.4149932246072727,
"learning_rate": 2e-05,
"loss": 0.5259864330291748,
"step": 2050
},
{
"epoch": 0.13715046604527298,
"grad_norm": 0.8576063037502878,
"learning_rate": 2e-05,
"loss": 0.5189313888549805,
"step": 2060
},
{
"epoch": 0.1378162450066578,
"grad_norm": 2.0944398013918635,
"learning_rate": 2e-05,
"loss": 0.5383773803710937,
"step": 2070
},
{
"epoch": 0.1384820239680426,
"grad_norm": 2.202299198114209,
"learning_rate": 2e-05,
"loss": 0.5349685668945312,
"step": 2080
},
{
"epoch": 0.13914780292942744,
"grad_norm": 0.8804253733660655,
"learning_rate": 2e-05,
"loss": 0.5294415473937988,
"step": 2090
},
{
"epoch": 0.13981358189081225,
"grad_norm": 1.1171759751806474,
"learning_rate": 2e-05,
"loss": 0.5218812942504882,
"step": 2100
},
{
"epoch": 0.14047936085219706,
"grad_norm": 1.6634271061653607,
"learning_rate": 2e-05,
"loss": 0.5050858974456787,
"step": 2110
},
{
"epoch": 0.1411451398135819,
"grad_norm": 1.1835409358711206,
"learning_rate": 2e-05,
"loss": 0.5161876678466797,
"step": 2120
},
{
"epoch": 0.1418109187749667,
"grad_norm": 2.101176098222384,
"learning_rate": 2e-05,
"loss": 0.5150922775268555,
"step": 2130
},
{
"epoch": 0.14247669773635152,
"grad_norm": 1.0781328275401703,
"learning_rate": 2e-05,
"loss": 0.5243738174438477,
"step": 2140
},
{
"epoch": 0.14314247669773636,
"grad_norm": 2.6622883554564107,
"learning_rate": 2e-05,
"loss": 0.5393617630004883,
"step": 2150
},
{
"epoch": 0.14380825565912117,
"grad_norm": 3.249905568049039,
"learning_rate": 2e-05,
"loss": 0.5379310131072998,
"step": 2160
},
{
"epoch": 0.14447403462050598,
"grad_norm": 1.1442728323685893,
"learning_rate": 2e-05,
"loss": 0.5616377830505371,
"step": 2170
},
{
"epoch": 0.14513981358189082,
"grad_norm": 1.126585913850443,
"learning_rate": 2e-05,
"loss": 0.49822397232055665,
"step": 2180
},
{
"epoch": 0.14580559254327563,
"grad_norm": 1.3884430019967056,
"learning_rate": 2e-05,
"loss": 0.5449071407318116,
"step": 2190
},
{
"epoch": 0.14647137150466044,
"grad_norm": 1.146409501415657,
"learning_rate": 2e-05,
"loss": 0.5238983154296875,
"step": 2200
},
{
"epoch": 0.14713715046604528,
"grad_norm": 1.4908765281795415,
"learning_rate": 2e-05,
"loss": 0.5086498737335206,
"step": 2210
},
{
"epoch": 0.1478029294274301,
"grad_norm": 1.4052241706860102,
"learning_rate": 2e-05,
"loss": 0.4959442138671875,
"step": 2220
},
{
"epoch": 0.1484687083888149,
"grad_norm": 0.9232977271364534,
"learning_rate": 2e-05,
"loss": 0.49895877838134767,
"step": 2230
},
{
"epoch": 0.14913448735019974,
"grad_norm": 1.251500253657333,
"learning_rate": 2e-05,
"loss": 0.5154998779296875,
"step": 2240
},
{
"epoch": 0.14980026631158455,
"grad_norm": 0.9654701095863362,
"learning_rate": 2e-05,
"loss": 0.5073034286499023,
"step": 2250
},
{
"epoch": 0.15046604527296936,
"grad_norm": 0.9584873628963919,
"learning_rate": 2e-05,
"loss": 0.5323206901550293,
"step": 2260
},
{
"epoch": 0.1511318242343542,
"grad_norm": 1.115220481547035,
"learning_rate": 2e-05,
"loss": 0.5065470695495605,
"step": 2270
},
{
"epoch": 0.151797603195739,
"grad_norm": 4.662097485396481,
"learning_rate": 2e-05,
"loss": 0.4943705081939697,
"step": 2280
},
{
"epoch": 0.15246338215712382,
"grad_norm": 11.103524498016363,
"learning_rate": 2e-05,
"loss": 0.5245217323303223,
"step": 2290
},
{
"epoch": 0.15312916111850866,
"grad_norm": 1.4598282653889072,
"learning_rate": 2e-05,
"loss": 0.5193574905395508,
"step": 2300
},
{
"epoch": 0.15379494007989347,
"grad_norm": 1.7782516999908409,
"learning_rate": 2e-05,
"loss": 0.48776745796203613,
"step": 2310
},
{
"epoch": 0.15446071904127828,
"grad_norm": 3.2706999654913154,
"learning_rate": 2e-05,
"loss": 0.5390383720397949,
"step": 2320
},
{
"epoch": 0.15512649800266312,
"grad_norm": 45.974392184868876,
"learning_rate": 2e-05,
"loss": 0.5336198329925537,
"step": 2330
},
{
"epoch": 0.15579227696404793,
"grad_norm": 2.7586767382582815,
"learning_rate": 2e-05,
"loss": 0.5713801383972168,
"step": 2340
},
{
"epoch": 0.15645805592543274,
"grad_norm": 2.9092759262358094,
"learning_rate": 2e-05,
"loss": 0.5405255794525147,
"step": 2350
},
{
"epoch": 0.15712383488681758,
"grad_norm": 2.1258359511475198,
"learning_rate": 2e-05,
"loss": 0.5386243343353272,
"step": 2360
},
{
"epoch": 0.1577896138482024,
"grad_norm": 1.2381595039295767,
"learning_rate": 2e-05,
"loss": 0.5679190635681153,
"step": 2370
},
{
"epoch": 0.1584553928095872,
"grad_norm": 17.18583576572219,
"learning_rate": 2e-05,
"loss": 0.5169041633605957,
"step": 2380
},
{
"epoch": 0.15912117177097204,
"grad_norm": 5.3082457989157215,
"learning_rate": 2e-05,
"loss": 0.5436855316162109,
"step": 2390
},
{
"epoch": 0.15978695073235685,
"grad_norm": 1.227956735412768,
"learning_rate": 2e-05,
"loss": 0.555968189239502,
"step": 2400
},
{
"epoch": 0.1604527296937417,
"grad_norm": 1.4804230692740696,
"learning_rate": 2e-05,
"loss": 0.5193206787109375,
"step": 2410
},
{
"epoch": 0.1611185086551265,
"grad_norm": 4.436062371833058,
"learning_rate": 2e-05,
"loss": 0.5349526882171631,
"step": 2420
},
{
"epoch": 0.1617842876165113,
"grad_norm": 2.3524883548432056,
"learning_rate": 2e-05,
"loss": 0.5464076519012451,
"step": 2430
},
{
"epoch": 0.16245006657789615,
"grad_norm": 2.4186439262098625,
"learning_rate": 2e-05,
"loss": 0.5067841053009033,
"step": 2440
},
{
"epoch": 0.16311584553928096,
"grad_norm": 1.9343779365891003,
"learning_rate": 2e-05,
"loss": 0.5334929943084716,
"step": 2450
},
{
"epoch": 0.16378162450066577,
"grad_norm": 1.1199215401132792,
"learning_rate": 2e-05,
"loss": 0.5157291412353515,
"step": 2460
},
{
"epoch": 0.1644474034620506,
"grad_norm": 2.8193397360794696,
"learning_rate": 2e-05,
"loss": 0.5031748294830323,
"step": 2470
},
{
"epoch": 0.16511318242343542,
"grad_norm": 1.5928757383633196,
"learning_rate": 2e-05,
"loss": 0.5512802124023437,
"step": 2480
},
{
"epoch": 0.16577896138482023,
"grad_norm": 1.9522306023391771,
"learning_rate": 2e-05,
"loss": 0.547333812713623,
"step": 2490
},
{
"epoch": 0.16644474034620507,
"grad_norm": 2.172031447099043,
"learning_rate": 2e-05,
"loss": 0.5480151176452637,
"step": 2500
},
{
"epoch": 0.16711051930758988,
"grad_norm": 1.0415649470614736,
"learning_rate": 2e-05,
"loss": 0.5497314453125,
"step": 2510
},
{
"epoch": 0.1677762982689747,
"grad_norm": 1.2241977445131844,
"learning_rate": 2e-05,
"loss": 0.5218514919281005,
"step": 2520
},
{
"epoch": 0.16844207723035953,
"grad_norm": 1.46895120578611,
"learning_rate": 2e-05,
"loss": 0.5450241088867187,
"step": 2530
},
{
"epoch": 0.16910785619174434,
"grad_norm": 1.2668093618694478,
"learning_rate": 2e-05,
"loss": 0.5406510353088378,
"step": 2540
},
{
"epoch": 0.16977363515312915,
"grad_norm": 4.8126265158741175,
"learning_rate": 1.997261400182901e-05,
"loss": 0.5544777938297817,
"step": 2550
},
{
"epoch": 0.170439414114514,
"grad_norm": 1.741154798134385,
"learning_rate": 1.9805800746810423e-05,
"loss": 0.48458089828491213,
"step": 2560
},
{
"epoch": 0.1711051930758988,
"grad_norm": 1.7637979885777835,
"learning_rate": 1.9489934220110274e-05,
"loss": 0.5370548248291016,
"step": 2570
},
{
"epoch": 0.17177097203728361,
"grad_norm": 2.6959793981121107,
"learning_rate": 1.9029840364820392e-05,
"loss": 0.5050186157226563,
"step": 2580
},
{
"epoch": 0.17243675099866845,
"grad_norm": 1.4680790005517546,
"learning_rate": 1.8432548690479908e-05,
"loss": 0.5289761543273925,
"step": 2590
},
{
"epoch": 0.17310252996005326,
"grad_norm": 3.5843171763824606,
"learning_rate": 1.7707184873239772e-05,
"loss": 0.5158669471740722,
"step": 2600
},
{
"epoch": 0.17376830892143807,
"grad_norm": 2.2565157921241585,
"learning_rate": 1.6864831329902565e-05,
"loss": 0.5329915046691894,
"step": 2610
},
{
"epoch": 0.1744340878828229,
"grad_norm": 10.477332836589401,
"learning_rate": 1.591835789604667e-05,
"loss": 0.5508148193359375,
"step": 2620
},
{
"epoch": 0.17509986684420772,
"grad_norm": 1.5131459904033961,
"learning_rate": 1.4882225195206459e-05,
"loss": 0.5336699485778809,
"step": 2630
},
{
"epoch": 0.17576564580559254,
"grad_norm": 2.738022578340676,
"learning_rate": 1.3772263703317695e-05,
"loss": 0.4890913009643555,
"step": 2640
},
{
"epoch": 0.17643142476697737,
"grad_norm": 1.3916094420262712,
"learning_rate": 1.2605431883975547e-05,
"loss": 0.5570073604583741,
"step": 2650
},
{
"epoch": 0.17709720372836218,
"grad_norm": 1.2926526029787748,
"learning_rate": 1.1399557089817574e-05,
"loss": 0.4922811508178711,
"step": 2660
},
{
"epoch": 0.177762982689747,
"grad_norm": 2.19501331831183,
"learning_rate": 1.0173063188650463e-05,
"loss": 0.4790042400360107,
"step": 2670
},
{
"epoch": 0.17842876165113183,
"grad_norm": 1.256771429900528,
"learning_rate": 8.944689075764396e-06,
"loss": 0.5132005214691162,
"step": 2680
},
{
"epoch": 0.17909454061251665,
"grad_norm": 2.518695013104305,
"learning_rate": 7.73320237312362e-06,
"loss": 0.4863430976867676,
"step": 2690
},
{
"epoch": 0.17976031957390146,
"grad_norm": 2.029716657179604,
"learning_rate": 6.557112689658629e-06,
"loss": 0.5218551635742188,
"step": 2700
},
{
"epoch": 0.1804260985352863,
"grad_norm": 6.0348548394149155,
"learning_rate": 5.434388823591213e-06,
"loss": 0.5029655456542969,
"step": 2710
},
{
"epoch": 0.1810918774966711,
"grad_norm": 1.6032986061682017,
"learning_rate": 4.382184227495424e-06,
"loss": 0.5225400924682617,
"step": 2720
},
{
"epoch": 0.18175765645805592,
"grad_norm": 2.0209173822278594,
"learning_rate": 3.4165749305560203e-06,
"loss": 0.4966257095336914,
"step": 2730
},
{
"epoch": 0.18242343541944075,
"grad_norm": 1.6619982042537933,
"learning_rate": 2.5523139221595556e-06,
"loss": 0.4775592803955078,
"step": 2740
},
{
"epoch": 0.18308921438082557,
"grad_norm": 1.7316977246588428,
"learning_rate": 1.8026057494499821e-06,
"loss": 0.49677023887634275,
"step": 2750
},
{
"epoch": 0.18375499334221038,
"grad_norm": 1.692372792380116,
"learning_rate": 1.178904772643223e-06,
"loss": 0.49402785301208496,
"step": 2760
},
{
"epoch": 0.18442077230359522,
"grad_norm": 1.583424150846687,
"learning_rate": 6.90740160441754e-07,
"loss": 0.5309676647186279,
"step": 2770
},
{
"epoch": 0.18508655126498003,
"grad_norm": 1.4153865772628256,
"learning_rate": 3.45570299343632e-07,
"loss": 0.4863434314727783,
"step": 2780
},
{
"epoch": 0.18575233022636484,
"grad_norm": 2.138917425641507,
"learning_rate": 1.486688412425152e-07,
"loss": 0.50497407913208,
"step": 2790
},
{
"epoch": 0.18621837549933423,
"step": 2797,
"total_flos": 1324435843842048.0,
"train_loss": 0.04657678660384919,
"train_runtime": 3709.242,
"train_samples_per_second": 9.049,
"train_steps_per_second": 0.754
}
],
"logging_steps": 10,
"max_steps": 2797,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2543,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1324435843842048.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}