Buffer / v2v3_from_sample10_step2991 /trainer_state.json
CongWei1230's picture
Upload checkpoint-2991 (32B v2v3 from sample10, step 2991/3391, ~88%)
a5bfbd8 verified
Raw
History Blame Contribute Delete
56.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8821057288210573,
"eval_steps": 500,
"global_step": 2991,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0029492000294920003,
"grad_norm": 0.4438488175170118,
"learning_rate": 2.647058823529412e-07,
"loss": 0.49045677185058595,
"step": 10
},
{
"epoch": 0.005898400058984001,
"grad_norm": 0.510395268011925,
"learning_rate": 5.588235294117648e-07,
"loss": 0.4855191111564636,
"step": 20
},
{
"epoch": 0.008847600088476001,
"grad_norm": 0.5703078157632056,
"learning_rate": 8.529411764705883e-07,
"loss": 0.4901909828186035,
"step": 30
},
{
"epoch": 0.011796800117968001,
"grad_norm": 0.454332417722825,
"learning_rate": 1.1470588235294117e-06,
"loss": 0.49544334411621094,
"step": 40
},
{
"epoch": 0.014746000147460001,
"grad_norm": 0.5589271583735398,
"learning_rate": 1.4411764705882352e-06,
"loss": 0.49580230712890627,
"step": 50
},
{
"epoch": 0.017695200176952003,
"grad_norm": 0.4631646645227775,
"learning_rate": 1.735294117647059e-06,
"loss": 0.47724552154541017,
"step": 60
},
{
"epoch": 0.020644400206444,
"grad_norm": 0.43267675548046375,
"learning_rate": 2.0294117647058824e-06,
"loss": 0.4778195449284145,
"step": 70
},
{
"epoch": 0.023593600235936003,
"grad_norm": 0.411312707929396,
"learning_rate": 2.323529411764706e-06,
"loss": 0.48408994674682615,
"step": 80
},
{
"epoch": 0.026542800265428004,
"grad_norm": 0.4097006060327255,
"learning_rate": 2.61764705882353e-06,
"loss": 0.46933441162109374,
"step": 90
},
{
"epoch": 0.029492000294920002,
"grad_norm": 0.4196926934679535,
"learning_rate": 2.9117647058823534e-06,
"loss": 0.4877605438232422,
"step": 100
},
{
"epoch": 0.032441200324412,
"grad_norm": 0.760053533328545,
"learning_rate": 3.205882352941177e-06,
"loss": 0.47942724227905276,
"step": 110
},
{
"epoch": 0.035390400353904006,
"grad_norm": 0.46010290292368705,
"learning_rate": 3.5e-06,
"loss": 0.47931214173634845,
"step": 120
},
{
"epoch": 0.038339600383396004,
"grad_norm": 0.8247891534858383,
"learning_rate": 3.794117647058824e-06,
"loss": 0.47976908683776853,
"step": 130
},
{
"epoch": 0.041288800412888,
"grad_norm": 0.3959025132149403,
"learning_rate": 4.088235294117647e-06,
"loss": 0.47607669830322263,
"step": 140
},
{
"epoch": 0.04423800044238001,
"grad_norm": 0.4964738394557261,
"learning_rate": 4.382352941176471e-06,
"loss": 0.48482370376586914,
"step": 150
},
{
"epoch": 0.047187200471872005,
"grad_norm": 0.5454531923723247,
"learning_rate": 4.676470588235294e-06,
"loss": 0.4825325012207031,
"step": 160
},
{
"epoch": 0.050136400501364004,
"grad_norm": 0.5674627930340099,
"learning_rate": 4.970588235294118e-06,
"loss": 0.48376989364624023,
"step": 170
},
{
"epoch": 0.05308560053085601,
"grad_norm": 0.3842610886782812,
"learning_rate": 5.264705882352941e-06,
"loss": 0.4791515138414171,
"step": 180
},
{
"epoch": 0.05603480056034801,
"grad_norm": 0.4036085754825109,
"learning_rate": 5.558823529411766e-06,
"loss": 0.47362470626831055,
"step": 190
},
{
"epoch": 0.058984000589840005,
"grad_norm": 0.6114410512883032,
"learning_rate": 5.852941176470589e-06,
"loss": 0.4872968792915344,
"step": 200
},
{
"epoch": 0.061933200619332,
"grad_norm": 0.4530550752117145,
"learning_rate": 6.147058823529413e-06,
"loss": 0.47545700073242186,
"step": 210
},
{
"epoch": 0.064882400648824,
"grad_norm": 0.4077731305692506,
"learning_rate": 6.441176470588236e-06,
"loss": 0.47246241569519043,
"step": 220
},
{
"epoch": 0.067831600678316,
"grad_norm": 0.5622020487235674,
"learning_rate": 6.73529411764706e-06,
"loss": 0.47187185287475586,
"step": 230
},
{
"epoch": 0.07078080070780801,
"grad_norm": 0.4520676392505154,
"learning_rate": 7.029411764705882e-06,
"loss": 0.4745026588439941,
"step": 240
},
{
"epoch": 0.0737300007373,
"grad_norm": 0.665449972915322,
"learning_rate": 7.323529411764706e-06,
"loss": 0.47572927474975585,
"step": 250
},
{
"epoch": 0.07667920076679201,
"grad_norm": 0.4306799374275605,
"learning_rate": 7.617647058823529e-06,
"loss": 0.4793105125427246,
"step": 260
},
{
"epoch": 0.07962840079628401,
"grad_norm": 0.5067455358311616,
"learning_rate": 7.911764705882353e-06,
"loss": 0.4786632537841797,
"step": 270
},
{
"epoch": 0.082577600825776,
"grad_norm": 0.8311091310377784,
"learning_rate": 8.205882352941176e-06,
"loss": 0.45581644773483276,
"step": 280
},
{
"epoch": 0.08552680085526801,
"grad_norm": 0.5583480929067662,
"learning_rate": 8.5e-06,
"loss": 0.46883134841918944,
"step": 290
},
{
"epoch": 0.08847600088476001,
"grad_norm": 0.45584720235658543,
"learning_rate": 8.794117647058823e-06,
"loss": 0.4727641105651855,
"step": 300
},
{
"epoch": 0.091425200914252,
"grad_norm": 0.4353892064346544,
"learning_rate": 9.088235294117647e-06,
"loss": 0.46644264459609985,
"step": 310
},
{
"epoch": 0.09437440094374401,
"grad_norm": 0.4261195142712069,
"learning_rate": 9.382352941176472e-06,
"loss": 0.47017822265625,
"step": 320
},
{
"epoch": 0.09732360097323602,
"grad_norm": 0.40157213423946264,
"learning_rate": 9.676470588235296e-06,
"loss": 0.4639726161956787,
"step": 330
},
{
"epoch": 0.10027280100272801,
"grad_norm": 0.4640197567821238,
"learning_rate": 9.970588235294119e-06,
"loss": 0.46640992164611816,
"step": 340
},
{
"epoch": 0.10322200103222001,
"grad_norm": 0.4449377213834547,
"learning_rate": 9.999785297426788e-06,
"loss": 0.47008252143859863,
"step": 350
},
{
"epoch": 0.10617120106171202,
"grad_norm": 0.3938918579823758,
"learning_rate": 9.99904313936493e-06,
"loss": 0.46800899505615234,
"step": 360
},
{
"epoch": 0.10912040109120401,
"grad_norm": 0.441132137531731,
"learning_rate": 9.99777095383453e-06,
"loss": 0.4682627201080322,
"step": 370
},
{
"epoch": 0.11206960112069601,
"grad_norm": 0.4654074796615517,
"learning_rate": 9.995968875720052e-06,
"loss": 0.4647488594055176,
"step": 380
},
{
"epoch": 0.115018801150188,
"grad_norm": 0.5103009841666833,
"learning_rate": 9.993637096088248e-06,
"loss": 0.46791942914326984,
"step": 390
},
{
"epoch": 0.11796800117968001,
"grad_norm": 0.4737505346965037,
"learning_rate": 9.990775862167875e-06,
"loss": 0.4649852752685547,
"step": 400
},
{
"epoch": 0.12091720120917201,
"grad_norm": 0.47097212044710535,
"learning_rate": 9.987385477323507e-06,
"loss": 0.46214942932128905,
"step": 410
},
{
"epoch": 0.123866401238664,
"grad_norm": 0.44957761382004113,
"learning_rate": 9.983466301023355e-06,
"loss": 0.4626101652781169,
"step": 420
},
{
"epoch": 0.12681560126815603,
"grad_norm": 0.43782935492777864,
"learning_rate": 9.979018748801159e-06,
"loss": 0.4633960247039795,
"step": 430
},
{
"epoch": 0.129764801297648,
"grad_norm": 0.7105024998414807,
"learning_rate": 9.974043292212129e-06,
"loss": 0.4577487468719482,
"step": 440
},
{
"epoch": 0.13271400132714,
"grad_norm": 0.5110483977825444,
"learning_rate": 9.968540458782952e-06,
"loss": 0.4635943412780762,
"step": 450
},
{
"epoch": 0.135663201356632,
"grad_norm": 0.5126440917199716,
"learning_rate": 9.96251083195586e-06,
"loss": 0.4557194709777832,
"step": 460
},
{
"epoch": 0.13861240138612402,
"grad_norm": 0.4678349449098876,
"learning_rate": 9.95595505102676e-06,
"loss": 0.4488111734390259,
"step": 470
},
{
"epoch": 0.14156160141561602,
"grad_norm": 0.5000466024203911,
"learning_rate": 9.94887381107747e-06,
"loss": 0.45911078453063964,
"step": 480
},
{
"epoch": 0.14451080144510803,
"grad_norm": 0.4932634694381353,
"learning_rate": 9.941267862902003e-06,
"loss": 0.44985857605934143,
"step": 490
},
{
"epoch": 0.1474600014746,
"grad_norm": 0.4564322133280632,
"learning_rate": 9.933138012926982e-06,
"loss": 0.4541158199310303,
"step": 500
},
{
"epoch": 0.150409201504092,
"grad_norm": 0.5036486290390941,
"learning_rate": 9.924485123126124e-06,
"loss": 0.45795326232910155,
"step": 510
},
{
"epoch": 0.15335840153358402,
"grad_norm": 0.399805589982632,
"learning_rate": 9.915310110928855e-06,
"loss": 0.45582425594329834,
"step": 520
},
{
"epoch": 0.15630760156307602,
"grad_norm": 0.40969373127916736,
"learning_rate": 9.905613949123036e-06,
"loss": 0.46780738830566404,
"step": 530
},
{
"epoch": 0.15925680159256803,
"grad_norm": 0.4289897010610082,
"learning_rate": 9.895397665751827e-06,
"loss": 0.4596126556396484,
"step": 540
},
{
"epoch": 0.16220600162206,
"grad_norm": 0.44681181591945346,
"learning_rate": 9.884662344004681e-06,
"loss": 0.4619132876396179,
"step": 550
},
{
"epoch": 0.165155201651552,
"grad_norm": 0.4800596902106988,
"learning_rate": 9.873409122102505e-06,
"loss": 0.46259474754333496,
"step": 560
},
{
"epoch": 0.168104401681044,
"grad_norm": 0.46407908994594294,
"learning_rate": 9.861639193176976e-06,
"loss": 0.4620525360107422,
"step": 570
},
{
"epoch": 0.17105360171053602,
"grad_norm": 0.49093572505096134,
"learning_rate": 9.849353805144033e-06,
"loss": 0.45901032856532503,
"step": 580
},
{
"epoch": 0.17400280174002802,
"grad_norm": 0.8726756489944044,
"learning_rate": 9.836554260571577e-06,
"loss": 0.45867166519165037,
"step": 590
},
{
"epoch": 0.17695200176952003,
"grad_norm": 0.7341156370839694,
"learning_rate": 9.82324191654136e-06,
"loss": 0.4561937252680461,
"step": 600
},
{
"epoch": 0.179901201799012,
"grad_norm": 0.4058632919452235,
"learning_rate": 9.809418184505091e-06,
"loss": 0.4548793792724609,
"step": 610
},
{
"epoch": 0.182850401828504,
"grad_norm": 0.4511042225488971,
"learning_rate": 9.795084530134801e-06,
"loss": 0.453752326965332,
"step": 620
},
{
"epoch": 0.18579960185799602,
"grad_norm": 0.45173742731920835,
"learning_rate": 9.780242473167434e-06,
"loss": 0.45865148968166775,
"step": 630
},
{
"epoch": 0.18874880188748802,
"grad_norm": 0.47189314459800014,
"learning_rate": 9.764893587243717e-06,
"loss": 0.4552666664123535,
"step": 640
},
{
"epoch": 0.19169800191698003,
"grad_norm": 0.6214044854130906,
"learning_rate": 9.749039499741313e-06,
"loss": 0.46270686388015747,
"step": 650
},
{
"epoch": 0.19464720194647203,
"grad_norm": 0.4624015500769992,
"learning_rate": 9.732681891602283e-06,
"loss": 0.4482471466064453,
"step": 660
},
{
"epoch": 0.197596401975964,
"grad_norm": 0.4989793290332055,
"learning_rate": 9.715822497154855e-06,
"loss": 0.4525990145547049,
"step": 670
},
{
"epoch": 0.20054560200545601,
"grad_norm": 0.47559063203929464,
"learning_rate": 9.698463103929542e-06,
"loss": 0.4550197601318359,
"step": 680
},
{
"epoch": 0.20349480203494802,
"grad_norm": 0.4454621718199833,
"learning_rate": 9.680605552469625e-06,
"loss": 0.4487155437469482,
"step": 690
},
{
"epoch": 0.20644400206444002,
"grad_norm": 0.5915797439826518,
"learning_rate": 9.662251736136e-06,
"loss": 0.45226011276245115,
"step": 700
},
{
"epoch": 0.20939320209393203,
"grad_norm": 0.46721506797188894,
"learning_rate": 9.643403600906433e-06,
"loss": 0.4483489513397217,
"step": 710
},
{
"epoch": 0.21234240212342403,
"grad_norm": 0.4654852693012983,
"learning_rate": 9.624063145169242e-06,
"loss": 0.44864265124003094,
"step": 720
},
{
"epoch": 0.215291602152916,
"grad_norm": 0.6323687473017191,
"learning_rate": 9.604232419511408e-06,
"loss": 0.4535382270812988,
"step": 730
},
{
"epoch": 0.21824080218240802,
"grad_norm": 0.5697010040942148,
"learning_rate": 9.58391352650117e-06,
"loss": 0.44826154708862304,
"step": 740
},
{
"epoch": 0.22119000221190002,
"grad_norm": 0.7362634384867952,
"learning_rate": 9.563108620465092e-06,
"loss": 0.434189240137736,
"step": 750
},
{
"epoch": 0.22413920224139203,
"grad_norm": 0.4839592091153744,
"learning_rate": 9.541819907259649e-06,
"loss": 0.4535334587097168,
"step": 760
},
{
"epoch": 0.22708840227088403,
"grad_norm": 0.49056139316525293,
"learning_rate": 9.520049644037349e-06,
"loss": 0.4511109352111816,
"step": 770
},
{
"epoch": 0.230037602300376,
"grad_norm": 0.5330837147313883,
"learning_rate": 9.49780013900742e-06,
"loss": 0.45012299219767254,
"step": 780
},
{
"epoch": 0.23298680232986801,
"grad_norm": 0.422780974383286,
"learning_rate": 9.475073751191082e-06,
"loss": 0.45609140396118164,
"step": 790
},
{
"epoch": 0.23593600235936002,
"grad_norm": 0.4442784347012951,
"learning_rate": 9.451872890171419e-06,
"loss": 0.4553239345550537,
"step": 800
},
{
"epoch": 0.23888520238885202,
"grad_norm": 0.4375022755271826,
"learning_rate": 9.428200015837914e-06,
"loss": 0.44987063937717015,
"step": 810
},
{
"epoch": 0.24183440241834403,
"grad_norm": 0.45699354225933403,
"learning_rate": 9.40405763812563e-06,
"loss": 0.44800944328308107,
"step": 820
},
{
"epoch": 0.24478360244783604,
"grad_norm": 0.444393405911582,
"learning_rate": 9.379448316749092e-06,
"loss": 0.44734683632850647,
"step": 830
},
{
"epoch": 0.247732802477328,
"grad_norm": 0.4624536549377123,
"learning_rate": 9.354374660930897e-06,
"loss": 0.45065832138061523,
"step": 840
},
{
"epoch": 0.25068200250682005,
"grad_norm": 0.5448404056143255,
"learning_rate": 9.32883932912506e-06,
"loss": 0.45664215087890625,
"step": 850
},
{
"epoch": 0.25363120253631205,
"grad_norm": 0.40963527831908814,
"learning_rate": 9.30284502873516e-06,
"loss": 0.44174957275390625,
"step": 860
},
{
"epoch": 0.256580402565804,
"grad_norm": 0.5893521896841538,
"learning_rate": 9.276394515827278e-06,
"loss": 0.4503978729248047,
"step": 870
},
{
"epoch": 0.259529602595296,
"grad_norm": 0.6759445609146097,
"learning_rate": 9.249490594837784e-06,
"loss": 0.44724044799804685,
"step": 880
},
{
"epoch": 0.262478802624788,
"grad_norm": 0.4770857199170203,
"learning_rate": 9.222136118275996e-06,
"loss": 0.4475719928741455,
"step": 890
},
{
"epoch": 0.26542800265428,
"grad_norm": 0.5407655237678989,
"learning_rate": 9.194333986421739e-06,
"loss": 0.44543004035949707,
"step": 900
},
{
"epoch": 0.268377202683772,
"grad_norm": 0.4313836536721019,
"learning_rate": 9.166087147017843e-06,
"loss": 0.4491748809814453,
"step": 910
},
{
"epoch": 0.271326402713264,
"grad_norm": 0.514027363001843,
"learning_rate": 9.137398594957605e-06,
"loss": 0.44903249740600587,
"step": 920
},
{
"epoch": 0.27427560274275603,
"grad_norm": 0.5570660212650201,
"learning_rate": 9.108271371967247e-06,
"loss": 0.44883127212524415,
"step": 930
},
{
"epoch": 0.27722480277224804,
"grad_norm": 0.49414270655924086,
"learning_rate": 9.078708566283432e-06,
"loss": 0.4476127624511719,
"step": 940
},
{
"epoch": 0.28017400280174004,
"grad_norm": 0.5435708465253872,
"learning_rate": 9.048713312325806e-06,
"loss": 0.43948898315429685,
"step": 950
},
{
"epoch": 0.28312320283123205,
"grad_norm": 0.4147716445408524,
"learning_rate": 9.018288790364692e-06,
"loss": 0.44974651336669924,
"step": 960
},
{
"epoch": 0.28607240286072405,
"grad_norm": 0.4419046760769447,
"learning_rate": 8.987438226183886e-06,
"loss": 0.44513511657714844,
"step": 970
},
{
"epoch": 0.28902160289021606,
"grad_norm": 0.46008010730858606,
"learning_rate": 8.956164890738643e-06,
"loss": 0.4487351417541504,
"step": 980
},
{
"epoch": 0.291970802919708,
"grad_norm": 0.4122144023210693,
"learning_rate": 8.924472099808878e-06,
"loss": 0.44594388008117675,
"step": 990
},
{
"epoch": 0.2949200029492,
"grad_norm": 0.4283365251996404,
"learning_rate": 8.892363213647597e-06,
"loss": 0.44366854429244995,
"step": 1000
},
{
"epoch": 0.297869202978692,
"grad_norm": 0.43330313156753963,
"learning_rate": 8.859841636624632e-06,
"loss": 0.4339793026447296,
"step": 1010
},
{
"epoch": 0.300818403008184,
"grad_norm": 0.5965674016711119,
"learning_rate": 8.826910816865687e-06,
"loss": 0.4508774757385254,
"step": 1020
},
{
"epoch": 0.303767603037676,
"grad_norm": 0.40748634997353456,
"learning_rate": 8.793574245886747e-06,
"loss": 0.44038143157958987,
"step": 1030
},
{
"epoch": 0.30671680306716803,
"grad_norm": 0.42190516326680844,
"learning_rate": 8.759835458223889e-06,
"loss": 0.44063931703567505,
"step": 1040
},
{
"epoch": 0.30966600309666004,
"grad_norm": 0.46164218622009234,
"learning_rate": 8.72569803105853e-06,
"loss": 0.44178342819213867,
"step": 1050
},
{
"epoch": 0.31261520312615204,
"grad_norm": 0.4929848096318575,
"learning_rate": 8.69116558383815e-06,
"loss": 0.4523301124572754,
"step": 1060
},
{
"epoch": 0.31556440315564405,
"grad_norm": 0.4924653229379322,
"learning_rate": 8.656241777892544e-06,
"loss": 0.43601031303405763,
"step": 1070
},
{
"epoch": 0.31851360318513605,
"grad_norm": 0.4226604314473932,
"learning_rate": 8.620930316045619e-06,
"loss": 0.44405059814453124,
"step": 1080
},
{
"epoch": 0.32146280321462806,
"grad_norm": 0.4236547546134905,
"learning_rate": 8.585234942222806e-06,
"loss": 0.43758702278137207,
"step": 1090
},
{
"epoch": 0.32441200324412,
"grad_norm": 0.4805889787560502,
"learning_rate": 8.549159441054105e-06,
"loss": 0.4501505851745605,
"step": 1100
},
{
"epoch": 0.327361203273612,
"grad_norm": 0.4743601561407433,
"learning_rate": 8.512707637472808e-06,
"loss": 0.43948731422424314,
"step": 1110
},
{
"epoch": 0.330310403303104,
"grad_norm": 0.4171265101033706,
"learning_rate": 8.475883396309977e-06,
"loss": 0.44372354234967915,
"step": 1120
},
{
"epoch": 0.333259603332596,
"grad_norm": 0.447416162788826,
"learning_rate": 8.43869062188465e-06,
"loss": 0.43704824447631835,
"step": 1130
},
{
"epoch": 0.336208803362088,
"grad_norm": 0.45418718637069994,
"learning_rate": 8.401133257589902e-06,
"loss": 0.4376950740814209,
"step": 1140
},
{
"epoch": 0.33915800339158003,
"grad_norm": 0.38725385605408097,
"learning_rate": 8.363215285474727e-06,
"loss": 0.44152375629970003,
"step": 1150
},
{
"epoch": 0.34210720342107204,
"grad_norm": 0.5016629711168561,
"learning_rate": 8.324940725821853e-06,
"loss": 0.4489162445068359,
"step": 1160
},
{
"epoch": 0.34505640345056404,
"grad_norm": 0.5328079856793903,
"learning_rate": 8.286313636721474e-06,
"loss": 0.446546745300293,
"step": 1170
},
{
"epoch": 0.34800560348005605,
"grad_norm": 0.5248220390584734,
"learning_rate": 8.247338113641003e-06,
"loss": 0.4443833827972412,
"step": 1180
},
{
"epoch": 0.35095480350954805,
"grad_norm": 0.47953642120467227,
"learning_rate": 8.208018288990832e-06,
"loss": 0.43544740676879884,
"step": 1190
},
{
"epoch": 0.35390400353904006,
"grad_norm": 0.5048123170861494,
"learning_rate": 8.168358331686201e-06,
"loss": 0.4377775192260742,
"step": 1200
},
{
"epoch": 0.35685320356853206,
"grad_norm": 0.4323962795333619,
"learning_rate": 8.12836244670518e-06,
"loss": 0.4346789360046387,
"step": 1210
},
{
"epoch": 0.359802403598024,
"grad_norm": 0.43275038314567504,
"learning_rate": 8.088034874642834e-06,
"loss": 0.438964319229126,
"step": 1220
},
{
"epoch": 0.362751603627516,
"grad_norm": 0.4723419704201015,
"learning_rate": 8.04737989126161e-06,
"loss": 0.43390822410583496,
"step": 1230
},
{
"epoch": 0.365700803657008,
"grad_norm": 0.5261345573735546,
"learning_rate": 8.006401807038003e-06,
"loss": 0.44441862106323243,
"step": 1240
},
{
"epoch": 0.3686500036865,
"grad_norm": 0.46122411208354347,
"learning_rate": 7.965104966705518e-06,
"loss": 0.441965389251709,
"step": 1250
},
{
"epoch": 0.37159920371599203,
"grad_norm": 0.509468856231947,
"learning_rate": 7.923493748794033e-06,
"loss": 0.44285959667629665,
"step": 1260
},
{
"epoch": 0.37454840374548404,
"grad_norm": 0.43985121690800727,
"learning_rate": 7.881572565165556e-06,
"loss": 0.4370553016662598,
"step": 1270
},
{
"epoch": 0.37749760377497604,
"grad_norm": 0.4193419303408407,
"learning_rate": 7.839345860546448e-06,
"loss": 0.44532525539398193,
"step": 1280
},
{
"epoch": 0.38044680380446805,
"grad_norm": 0.43580986209337313,
"learning_rate": 7.796818112056176e-06,
"loss": 0.43182549476623533,
"step": 1290
},
{
"epoch": 0.38339600383396005,
"grad_norm": 0.4767234521883533,
"learning_rate": 7.753993828732624e-06,
"loss": 0.44052562713623045,
"step": 1300
},
{
"epoch": 0.38634520386345206,
"grad_norm": 0.45215748877092227,
"learning_rate": 7.710877551054004e-06,
"loss": 0.43268642425537107,
"step": 1310
},
{
"epoch": 0.38929440389294406,
"grad_norm": 0.45322506466283746,
"learning_rate": 7.667473850457466e-06,
"loss": 0.44087610244750974,
"step": 1320
},
{
"epoch": 0.392243603922436,
"grad_norm": 0.48165493215175825,
"learning_rate": 7.623787328854397e-06,
"loss": 0.4389915466308594,
"step": 1330
},
{
"epoch": 0.395192803951928,
"grad_norm": 0.47990402640999297,
"learning_rate": 7.579822618142505e-06,
"loss": 0.43350309133529663,
"step": 1340
},
{
"epoch": 0.39814200398142,
"grad_norm": 0.44359105052104236,
"learning_rate": 7.535584379714717e-06,
"loss": 0.43778276443481445,
"step": 1350
},
{
"epoch": 0.40109120401091203,
"grad_norm": 0.4126099256240836,
"learning_rate": 7.491077303964946e-06,
"loss": 0.42278945446014404,
"step": 1360
},
{
"epoch": 0.40404040404040403,
"grad_norm": 0.6115035984776925,
"learning_rate": 7.446306109790798e-06,
"loss": 0.4358266830444336,
"step": 1370
},
{
"epoch": 0.40698960406989604,
"grad_norm": 0.45211683003520525,
"learning_rate": 7.401275544093238e-06,
"loss": 0.42867231369018555,
"step": 1380
},
{
"epoch": 0.40993880409938804,
"grad_norm": 0.5058543077254085,
"learning_rate": 7.355990381273302e-06,
"loss": 0.43923482298851013,
"step": 1390
},
{
"epoch": 0.41288800412888005,
"grad_norm": 0.47681056444954306,
"learning_rate": 7.3104554227258895e-06,
"loss": 0.43735685348510744,
"step": 1400
},
{
"epoch": 0.41583720415837205,
"grad_norm": 0.47907345262988904,
"learning_rate": 7.26467549633068e-06,
"loss": 0.4366196632385254,
"step": 1410
},
{
"epoch": 0.41878640418786406,
"grad_norm": 1.498448882465359,
"learning_rate": 7.218655455940266e-06,
"loss": 0.43742281198501587,
"step": 1420
},
{
"epoch": 0.42173560421735606,
"grad_norm": 0.5267975509057776,
"learning_rate": 7.172400180865514e-06,
"loss": 0.43933544158935545,
"step": 1430
},
{
"epoch": 0.42468480424684807,
"grad_norm": 0.5529750180943933,
"learning_rate": 7.125914575358231e-06,
"loss": 0.4414161205291748,
"step": 1440
},
{
"epoch": 0.42763400427634,
"grad_norm": 0.4213668929430919,
"learning_rate": 7.079203568091192e-06,
"loss": 0.43733177866254536,
"step": 1450
},
{
"epoch": 0.430583204305832,
"grad_norm": 0.5481726528287952,
"learning_rate": 7.032272111635565e-06,
"loss": 0.43358192443847654,
"step": 1460
},
{
"epoch": 0.43353240433532403,
"grad_norm": 0.46667586699579716,
"learning_rate": 6.9851251819358215e-06,
"loss": 0.44203882217407225,
"step": 1470
},
{
"epoch": 0.43648160436481603,
"grad_norm": 0.6402953577068929,
"learning_rate": 6.9377677777821504e-06,
"loss": 0.43342137336730957,
"step": 1480
},
{
"epoch": 0.43943080439430804,
"grad_norm": 0.5358248162028951,
"learning_rate": 6.8902049202804574e-06,
"loss": 0.44051499366760255,
"step": 1490
},
{
"epoch": 0.44238000442380004,
"grad_norm": 0.4741782148014334,
"learning_rate": 6.842441652320005e-06,
"loss": 0.44628814856211346,
"step": 1500
},
{
"epoch": 0.44532920445329205,
"grad_norm": 0.4968077776577462,
"learning_rate": 6.794483038038726e-06,
"loss": 0.43202877044677734,
"step": 1510
},
{
"epoch": 0.44827840448278405,
"grad_norm": 0.44198655309225987,
"learning_rate": 6.7463341622863074e-06,
"loss": 0.4334902763366699,
"step": 1520
},
{
"epoch": 0.45122760451227606,
"grad_norm": 0.48757609224205706,
"learning_rate": 6.698000130085052e-06,
"loss": 0.43942566712697345,
"step": 1530
},
{
"epoch": 0.45417680454176806,
"grad_norm": 0.5543149635616764,
"learning_rate": 6.649486066088622e-06,
"loss": 0.43965444564819334,
"step": 1540
},
{
"epoch": 0.45712600457126007,
"grad_norm": 0.43824689185514526,
"learning_rate": 6.6007971140386915e-06,
"loss": 0.4329670906066895,
"step": 1550
},
{
"epoch": 0.460075204600752,
"grad_norm": 0.5465274476694729,
"learning_rate": 6.551938436219578e-06,
"loss": 0.4395641326904297,
"step": 1560
},
{
"epoch": 0.463024404630244,
"grad_norm": 0.5537509643631231,
"learning_rate": 6.502915212910908e-06,
"loss": 0.43081774030412945,
"step": 1570
},
{
"epoch": 0.46597360465973603,
"grad_norm": 0.6891055017422132,
"learning_rate": 6.453732641838372e-06,
"loss": 0.436126184463501,
"step": 1580
},
{
"epoch": 0.46892280468922803,
"grad_norm": 0.4727794064469008,
"learning_rate": 6.4043959376226285e-06,
"loss": 0.43718743324279785,
"step": 1590
},
{
"epoch": 0.47187200471872004,
"grad_norm": 0.5174849373071831,
"learning_rate": 6.354910331226427e-06,
"loss": 0.43861846923828124,
"step": 1600
},
{
"epoch": 0.47482120474821204,
"grad_norm": 0.5451182940012586,
"learning_rate": 6.305281069399989e-06,
"loss": 0.4318662643432617,
"step": 1610
},
{
"epoch": 0.47777040477770405,
"grad_norm": 0.49243422723637303,
"learning_rate": 6.255513414124712e-06,
"loss": 0.43900103039211696,
"step": 1620
},
{
"epoch": 0.48071960480719605,
"grad_norm": 0.7191775498372631,
"learning_rate": 6.20561264205527e-06,
"loss": 0.42850513458251954,
"step": 1630
},
{
"epoch": 0.48366880483668806,
"grad_norm": 0.5338547106433559,
"learning_rate": 6.155584043960145e-06,
"loss": 0.41286659240722656,
"step": 1640
},
{
"epoch": 0.48661800486618007,
"grad_norm": 0.4559088327055419,
"learning_rate": 6.105432924160679e-06,
"loss": 0.43925085067749026,
"step": 1650
},
{
"epoch": 0.48956720489567207,
"grad_norm": 0.44070286117587115,
"learning_rate": 6.055164599968674e-06,
"loss": 0.43514204025268555,
"step": 1660
},
{
"epoch": 0.4925164049251641,
"grad_norm": 0.4153353893595977,
"learning_rate": 6.004784401122613e-06,
"loss": 0.43522300720214846,
"step": 1670
},
{
"epoch": 0.495465604954656,
"grad_norm": 0.5375605831166275,
"learning_rate": 5.954297669222584e-06,
"loss": 0.4359435081481934,
"step": 1680
},
{
"epoch": 0.49841480498414803,
"grad_norm": 0.6454662789331427,
"learning_rate": 5.903709757163922e-06,
"loss": 0.42930173873901367,
"step": 1690
},
{
"epoch": 0.5013640050136401,
"grad_norm": 0.5500985765215042,
"learning_rate": 5.8530260285696674e-06,
"loss": 0.4256927490234375,
"step": 1700
},
{
"epoch": 0.504313205043132,
"grad_norm": 0.6208010483461102,
"learning_rate": 5.802251857221886e-06,
"loss": 0.4257539749145508,
"step": 1710
},
{
"epoch": 0.5072624050726241,
"grad_norm": 0.47566478958679537,
"learning_rate": 5.751392626491907e-06,
"loss": 0.42954330444335936,
"step": 1720
},
{
"epoch": 0.510211605102116,
"grad_norm": 0.42787071882123745,
"learning_rate": 5.700453728769545e-06,
"loss": 0.4258143901824951,
"step": 1730
},
{
"epoch": 0.513160805131608,
"grad_norm": 0.4629048423358173,
"learning_rate": 5.649440564891373e-06,
"loss": 0.42579212188720705,
"step": 1740
},
{
"epoch": 0.5161100051611001,
"grad_norm": 0.4593224146242963,
"learning_rate": 5.5983585435680894e-06,
"loss": 0.4214206337928772,
"step": 1750
},
{
"epoch": 0.519059205190592,
"grad_norm": 0.5983325684323427,
"learning_rate": 5.5472130808110595e-06,
"loss": 0.4341601371765137,
"step": 1760
},
{
"epoch": 0.5220084052200841,
"grad_norm": 0.474667229133082,
"learning_rate": 5.496009599358072e-06,
"loss": 0.4302968978881836,
"step": 1770
},
{
"epoch": 0.524957605249576,
"grad_norm": 0.46177780626394316,
"learning_rate": 5.444753528098396e-06,
"loss": 0.43253841400146487,
"step": 1780
},
{
"epoch": 0.5279068052790681,
"grad_norm": 0.44669661110471875,
"learning_rate": 5.39345030149718e-06,
"loss": 0.42998533248901366,
"step": 1790
},
{
"epoch": 0.53085600530856,
"grad_norm": 0.49526829989644183,
"learning_rate": 5.342105359019244e-06,
"loss": 0.4352678457895915,
"step": 1800
},
{
"epoch": 0.5338052053380521,
"grad_norm": 0.475814297372346,
"learning_rate": 5.290724144552379e-06,
"loss": 0.4338411331176758,
"step": 1810
},
{
"epoch": 0.536754405367544,
"grad_norm": 0.4942643058141533,
"learning_rate": 5.239312105830135e-06,
"loss": 0.43322124481201174,
"step": 1820
},
{
"epoch": 0.5397036053970361,
"grad_norm": 0.4608256469051304,
"learning_rate": 5.187874693854228e-06,
"loss": 0.42023173967997235,
"step": 1830
},
{
"epoch": 0.542652805426528,
"grad_norm": 0.4795252576988978,
"learning_rate": 5.136417362316594e-06,
"loss": 0.42961969375610354,
"step": 1840
},
{
"epoch": 0.54560200545602,
"grad_norm": 0.5189901674665132,
"learning_rate": 5.084945567021159e-06,
"loss": 0.4320012092590332,
"step": 1850
},
{
"epoch": 0.5485512054855121,
"grad_norm": 0.48997129382028376,
"learning_rate": 5.033464765305375e-06,
"loss": 0.43167175187004936,
"step": 1860
},
{
"epoch": 0.551500405515004,
"grad_norm": 0.46295253554316024,
"learning_rate": 4.981980415461611e-06,
"loss": 0.431025641305106,
"step": 1870
},
{
"epoch": 0.5544496055444961,
"grad_norm": 0.4378611722025286,
"learning_rate": 4.9304979761584256e-06,
"loss": 0.4259642601013184,
"step": 1880
},
{
"epoch": 0.557398805573988,
"grad_norm": 0.47013994944311244,
"learning_rate": 4.8790229058618134e-06,
"loss": 0.43539981842041015,
"step": 1890
},
{
"epoch": 0.5603480056034801,
"grad_norm": 0.4643181066969001,
"learning_rate": 4.827560662256465e-06,
"loss": 0.4295026779174805,
"step": 1900
},
{
"epoch": 0.563297205632972,
"grad_norm": 0.4421560226823286,
"learning_rate": 4.7761167016671064e-06,
"loss": 0.43505077362060546,
"step": 1910
},
{
"epoch": 0.5662464056624641,
"grad_norm": 0.512252383210586,
"learning_rate": 4.724696478479999e-06,
"loss": 0.4244855244954427,
"step": 1920
},
{
"epoch": 0.569195605691956,
"grad_norm": 0.4338777525477151,
"learning_rate": 4.673305444564623e-06,
"loss": 0.42738046646118166,
"step": 1930
},
{
"epoch": 0.5721448057214481,
"grad_norm": 0.4951562183022821,
"learning_rate": 4.621949048695646e-06,
"loss": 0.4277921676635742,
"step": 1940
},
{
"epoch": 0.57509400575094,
"grad_norm": 0.4406700161724765,
"learning_rate": 4.570632735975205e-06,
"loss": 0.4272439479827881,
"step": 1950
},
{
"epoch": 0.5780432057804321,
"grad_norm": 0.4135316597546389,
"learning_rate": 4.519361947255595e-06,
"loss": 0.42667450904846194,
"step": 1960
},
{
"epoch": 0.5809924058099241,
"grad_norm": 0.4079029470615887,
"learning_rate": 4.468142118562389e-06,
"loss": 0.4228860855102539,
"step": 1970
},
{
"epoch": 0.583941605839416,
"grad_norm": 0.4333532180482263,
"learning_rate": 4.416978680518087e-06,
"loss": 0.4278755982716878,
"step": 1980
},
{
"epoch": 0.5868908058689081,
"grad_norm": 0.4274178984768739,
"learning_rate": 4.365877057766323e-06,
"loss": 0.4284843444824219,
"step": 1990
},
{
"epoch": 0.5898400058984,
"grad_norm": 0.4374330569823035,
"learning_rate": 4.314842668396716e-06,
"loss": 0.41731327772140503,
"step": 2000
},
{
"epoch": 0.5927892059278921,
"grad_norm": 0.4495624037630845,
"learning_rate": 4.263880923370418e-06,
"loss": 0.42450475692749023,
"step": 2010
},
{
"epoch": 0.595738405957384,
"grad_norm": 0.45091936894990575,
"learning_rate": 4.212997225946402e-06,
"loss": 0.4308480262756348,
"step": 2020
},
{
"epoch": 0.5986876059868761,
"grad_norm": 0.44501674575899014,
"learning_rate": 4.162196971108584e-06,
"loss": 0.4261492729187012,
"step": 2030
},
{
"epoch": 0.601636806016368,
"grad_norm": 0.41951268457332463,
"learning_rate": 4.111485544993817e-06,
"loss": 0.42758755683898925,
"step": 2040
},
{
"epoch": 0.6045860060458601,
"grad_norm": 0.5142796824437533,
"learning_rate": 4.0608683243208165e-06,
"loss": 0.42105774879455565,
"step": 2050
},
{
"epoch": 0.607535206075352,
"grad_norm": 0.4818313886940324,
"learning_rate": 4.010350675820091e-06,
"loss": 0.42716293334960936,
"step": 2060
},
{
"epoch": 0.6104844061048441,
"grad_norm": 0.46864624486768397,
"learning_rate": 3.959937955664935e-06,
"loss": 0.4280477099948459,
"step": 2070
},
{
"epoch": 0.6134336061343361,
"grad_norm": 0.43290161198243887,
"learning_rate": 3.909635508903528e-06,
"loss": 0.42287391424179077,
"step": 2080
},
{
"epoch": 0.616382806163828,
"grad_norm": 0.6372283409230178,
"learning_rate": 3.859448668892233e-06,
"loss": 0.4306667327880859,
"step": 2090
},
{
"epoch": 0.6193320061933201,
"grad_norm": 0.4452103223236772,
"learning_rate": 3.8093827567301124e-06,
"loss": 0.4375183582305908,
"step": 2100
},
{
"epoch": 0.622281206222812,
"grad_norm": 0.4779849175516226,
"learning_rate": 3.7594430806947584e-06,
"loss": 0.4284537315368652,
"step": 2110
},
{
"epoch": 0.6252304062523041,
"grad_norm": 0.5228639444081449,
"learning_rate": 3.7096349356794803e-06,
"loss": 0.4255107641220093,
"step": 2120
},
{
"epoch": 0.628179606281796,
"grad_norm": 0.486553349845169,
"learning_rate": 3.6599636026319073e-06,
"loss": 0.42174453735351564,
"step": 2130
},
{
"epoch": 0.6311288063112881,
"grad_norm": 0.7457013177519835,
"learning_rate": 3.610434347994071e-06,
"loss": 0.41699570417404175,
"step": 2140
},
{
"epoch": 0.63407800634078,
"grad_norm": 0.4473716326553472,
"learning_rate": 3.5610524231440324e-06,
"loss": 0.42831082344055177,
"step": 2150
},
{
"epoch": 0.6370272063702721,
"grad_norm": 0.4182714960508355,
"learning_rate": 3.5118230638390915e-06,
"loss": 0.42322230339050293,
"step": 2160
},
{
"epoch": 0.639976406399764,
"grad_norm": 0.4466605654685248,
"learning_rate": 3.4627514896606707e-06,
"loss": 0.4218415915966034,
"step": 2170
},
{
"epoch": 0.6429256064292561,
"grad_norm": 0.41267104203160443,
"learning_rate": 3.413842903460896e-06,
"loss": 0.4298990249633789,
"step": 2180
},
{
"epoch": 0.6458748064587481,
"grad_norm": 1.1820562807879373,
"learning_rate": 3.3651024908109698e-06,
"loss": 0.43236680030822755,
"step": 2190
},
{
"epoch": 0.64882400648824,
"grad_norm": 0.4004799801119464,
"learning_rate": 3.31653541945136e-06,
"loss": 0.4253828525543213,
"step": 2200
},
{
"epoch": 0.6517732065177321,
"grad_norm": 0.6488056302248444,
"learning_rate": 3.2681468387438876e-06,
"loss": 0.430925989151001,
"step": 2210
},
{
"epoch": 0.654722406547224,
"grad_norm": 0.5319457156936115,
"learning_rate": 3.2199418791257635e-06,
"loss": 0.4280077934265137,
"step": 2220
},
{
"epoch": 0.6576716065767161,
"grad_norm": 0.5805743070958141,
"learning_rate": 3.171925651565627e-06,
"loss": 0.42512601613998413,
"step": 2230
},
{
"epoch": 0.660620806606208,
"grad_norm": 0.4786720826427019,
"learning_rate": 3.1241032470216564e-06,
"loss": 0.42143568992614744,
"step": 2240
},
{
"epoch": 0.6635700066357001,
"grad_norm": 0.4638984311110436,
"learning_rate": 3.0764797359017894e-06,
"loss": 0.4212714672088623,
"step": 2250
},
{
"epoch": 0.666519206665192,
"grad_norm": 0.45634717378048567,
"learning_rate": 3.029060167526133e-06,
"loss": 0.4226480892726353,
"step": 2260
},
{
"epoch": 0.6694684066946841,
"grad_norm": 0.5185236096195818,
"learning_rate": 2.981849569591606e-06,
"loss": 0.42209339141845703,
"step": 2270
},
{
"epoch": 0.672417606724176,
"grad_norm": 0.5142168440328658,
"learning_rate": 2.9348529476388722e-06,
"loss": 0.4218320846557617,
"step": 2280
},
{
"epoch": 0.6753668067536681,
"grad_norm": 0.491078700341383,
"learning_rate": 2.888075284521622e-06,
"loss": 0.43361196517944334,
"step": 2290
},
{
"epoch": 0.6783160067831601,
"grad_norm": 0.4452481113098441,
"learning_rate": 2.8415215398782657e-06,
"loss": 0.422395133972168,
"step": 2300
},
{
"epoch": 0.681265206812652,
"grad_norm": 0.41323136227240675,
"learning_rate": 2.7951966496060834e-06,
"loss": 0.4265224536259969,
"step": 2310
},
{
"epoch": 0.6842144068421441,
"grad_norm": 0.5199847126406363,
"learning_rate": 2.749105525337886e-06,
"loss": 0.4258094787597656,
"step": 2320
},
{
"epoch": 0.687163606871636,
"grad_norm": 0.4770218877441737,
"learning_rate": 2.703253053921266e-06,
"loss": 0.4214162349700928,
"step": 2330
},
{
"epoch": 0.6901128069011281,
"grad_norm": 0.5240527237231183,
"learning_rate": 2.6576440969004562e-06,
"loss": 0.41685261726379397,
"step": 2340
},
{
"epoch": 0.69306200693062,
"grad_norm": 0.8307379407302324,
"learning_rate": 2.612283490000887e-06,
"loss": 0.42280903458595276,
"step": 2350
},
{
"epoch": 0.6960112069601121,
"grad_norm": 0.46917160863642615,
"learning_rate": 2.567176042616471e-06,
"loss": 0.4258403778076172,
"step": 2360
},
{
"epoch": 0.698960406989604,
"grad_norm": 0.8170575013309473,
"learning_rate": 2.522326537299684e-06,
"loss": 0.42054362297058107,
"step": 2370
},
{
"epoch": 0.7019096070190961,
"grad_norm": 0.4551781718665858,
"learning_rate": 2.477739729254497e-06,
"loss": 0.4166623183659145,
"step": 2380
},
{
"epoch": 0.704858807048588,
"grad_norm": 0.44128590419576835,
"learning_rate": 2.433420345832191e-06,
"loss": 0.427097225189209,
"step": 2390
},
{
"epoch": 0.7078080070780801,
"grad_norm": 0.5090921441586733,
"learning_rate": 2.3893730860301433e-06,
"loss": 0.43317718505859376,
"step": 2400
},
{
"epoch": 0.7107572071075721,
"grad_norm": 0.433061012122296,
"learning_rate": 2.3456026199936123e-06,
"loss": 0.42459826469421386,
"step": 2410
},
{
"epoch": 0.7137064071370641,
"grad_norm": 0.4230000543889016,
"learning_rate": 2.302113588520578e-06,
"loss": 0.4215096473693848,
"step": 2420
},
{
"epoch": 0.7166556071665561,
"grad_norm": 0.5019341190829107,
"learning_rate": 2.2589106025697027e-06,
"loss": 0.4279549916585286,
"step": 2430
},
{
"epoch": 0.719604807196048,
"grad_norm": 0.4509552796846411,
"learning_rate": 2.215998242771445e-06,
"loss": 0.4007607102394104,
"step": 2440
},
{
"epoch": 0.7225540072255401,
"grad_norm": 0.4067492226746083,
"learning_rate": 2.173381058942402e-06,
"loss": 0.4229306221008301,
"step": 2450
},
{
"epoch": 0.725503207255032,
"grad_norm": 0.4302732221102772,
"learning_rate": 2.1310635696029046e-06,
"loss": 0.424178409576416,
"step": 2460
},
{
"epoch": 0.7284524072845241,
"grad_norm": 0.4204556283750899,
"learning_rate": 2.0890502614979397e-06,
"loss": 0.4222692966461182,
"step": 2470
},
{
"epoch": 0.731401607314016,
"grad_norm": 0.9459508044859516,
"learning_rate": 2.0473455891214416e-06,
"loss": 0.4112105965614319,
"step": 2480
},
{
"epoch": 0.7343508073435081,
"grad_norm": 0.4905240594294945,
"learning_rate": 2.0059539742439993e-06,
"loss": 0.42390217781066897,
"step": 2490
},
{
"epoch": 0.737300007373,
"grad_norm": 0.4897228845154844,
"learning_rate": 1.9648798054440345e-06,
"loss": 0.4177046298980713,
"step": 2500
},
{
"epoch": 0.7402492074024921,
"grad_norm": 0.47141631810303875,
"learning_rate": 1.9241274376425e-06,
"loss": 0.4214902877807617,
"step": 2510
},
{
"epoch": 0.7431984074319841,
"grad_norm": 0.4311234608800366,
"learning_rate": 1.8837011916411458e-06,
"loss": 0.42144293785095216,
"step": 2520
},
{
"epoch": 0.7461476074614761,
"grad_norm": 0.4154980039716799,
"learning_rate": 1.8436053536644017e-06,
"loss": 0.4247028350830078,
"step": 2530
},
{
"epoch": 0.7490968074909681,
"grad_norm": 0.483115715597217,
"learning_rate": 1.803844174904928e-06,
"loss": 0.4199508726596832,
"step": 2540
},
{
"epoch": 0.75204600752046,
"grad_norm": 0.49963260739817345,
"learning_rate": 1.764421871072879e-06,
"loss": 0.41815831926133895,
"step": 2550
},
{
"epoch": 0.7549952075499521,
"grad_norm": 0.49998836248739126,
"learning_rate": 1.7253426219489284e-06,
"loss": 0.4235694885253906,
"step": 2560
},
{
"epoch": 0.757944407579444,
"grad_norm": 0.4002836885548947,
"learning_rate": 1.6866105709411069e-06,
"loss": 0.4307806968688965,
"step": 2570
},
{
"epoch": 0.7608936076089361,
"grad_norm": 0.4316111808556622,
"learning_rate": 1.648229824645492e-06,
"loss": 0.42072014808654784,
"step": 2580
},
{
"epoch": 0.763842807638428,
"grad_norm": 0.48884144907161353,
"learning_rate": 1.6102044524108014e-06,
"loss": 0.42240183694022043,
"step": 2590
},
{
"epoch": 0.7667920076679201,
"grad_norm": 0.46988962920527444,
"learning_rate": 1.5725384859069454e-06,
"loss": 0.42186822891235354,
"step": 2600
},
{
"epoch": 0.769741207697412,
"grad_norm": 0.46649424069974776,
"learning_rate": 1.5352359186975534e-06,
"loss": 0.42093173662821454,
"step": 2610
},
{
"epoch": 0.7726904077269041,
"grad_norm": 0.6385606884920392,
"learning_rate": 1.498300705816564e-06,
"loss": 0.4217710018157959,
"step": 2620
},
{
"epoch": 0.7756396077563961,
"grad_norm": 0.42991362684616335,
"learning_rate": 1.4617367633488816e-06,
"loss": 0.4215093612670898,
"step": 2630
},
{
"epoch": 0.7785888077858881,
"grad_norm": 0.5154955484211314,
"learning_rate": 1.4255479680151768e-06,
"loss": 0.4325077136357625,
"step": 2640
},
{
"epoch": 0.7815380078153801,
"grad_norm": 0.4705190403100384,
"learning_rate": 1.3897381567608498e-06,
"loss": 0.42404894828796386,
"step": 2650
},
{
"epoch": 0.784487207844872,
"grad_norm": 0.5878037530065117,
"learning_rate": 1.3543111263492165e-06,
"loss": 0.42200336456298826,
"step": 2660
},
{
"epoch": 0.7874364078743641,
"grad_norm": 0.5078063275913126,
"learning_rate": 1.3192706329589544e-06,
"loss": 0.4225406249364217,
"step": 2670
},
{
"epoch": 0.790385607903856,
"grad_norm": 0.5678496323247609,
"learning_rate": 1.2846203917858508e-06,
"loss": 0.4167942523956299,
"step": 2680
},
{
"epoch": 0.7933348079333481,
"grad_norm": 0.43824377809371,
"learning_rate": 1.250364076648894e-06,
"loss": 0.41694241762161255,
"step": 2690
},
{
"epoch": 0.79628400796284,
"grad_norm": 0.49382977556600033,
"learning_rate": 1.2165053196007565e-06,
"loss": 0.4200151443481445,
"step": 2700
},
{
"epoch": 0.7992332079923321,
"grad_norm": 0.4640803232672988,
"learning_rate": 1.183047710542703e-06,
"loss": 0.43220198154449463,
"step": 2710
},
{
"epoch": 0.8021824080218241,
"grad_norm": 0.5429123288187059,
"learning_rate": 1.1499947968439673e-06,
"loss": 0.42777442932128906,
"step": 2720
},
{
"epoch": 0.8051316080513161,
"grad_norm": 0.44445203079455214,
"learning_rate": 1.117350082965641e-06,
"loss": 0.4181947708129883,
"step": 2730
},
{
"epoch": 0.8080808080808081,
"grad_norm": 0.5608624463335975,
"learning_rate": 1.0851170300891089e-06,
"loss": 0.42819342017173767,
"step": 2740
},
{
"epoch": 0.8110300081103001,
"grad_norm": 0.5073433312971042,
"learning_rate": 1.0532990557490768e-06,
"loss": 0.42534780502319336,
"step": 2750
},
{
"epoch": 0.8139792081397921,
"grad_norm": 0.43482432065064375,
"learning_rate": 1.0218995334712205e-06,
"loss": 0.4315471649169922,
"step": 2760
},
{
"epoch": 0.8169284081692841,
"grad_norm": 0.4355732036652417,
"learning_rate": 9.909217924145114e-07,
"loss": 0.417811529976981,
"step": 2770
},
{
"epoch": 0.8198776081987761,
"grad_norm": 0.4621367304882082,
"learning_rate": 9.603691170182316e-07,
"loss": 0.41324310302734374,
"step": 2780
},
{
"epoch": 0.822826808228268,
"grad_norm": 0.4049121392981768,
"learning_rate": 9.302447466537496e-07,
"loss": 0.4109320243199666,
"step": 2790
},
{
"epoch": 0.8257760082577601,
"grad_norm": 0.5334245178977874,
"learning_rate": 9.005518752810516e-07,
"loss": 0.42446060180664064,
"step": 2800
},
{
"epoch": 0.828725208287252,
"grad_norm": 0.48771063215043997,
"learning_rate": 8.712936511101056e-07,
"loss": 0.4142922401428223,
"step": 2810
},
{
"epoch": 0.8316744083167441,
"grad_norm": 0.46166981422310766,
"learning_rate": 8.424731762670723e-07,
"loss": 0.41686129570007324,
"step": 2820
},
{
"epoch": 0.8346236083462361,
"grad_norm": 0.8079406321561716,
"learning_rate": 8.140935064653949e-07,
"loss": 0.42085041999816897,
"step": 2830
},
{
"epoch": 0.8375728083757281,
"grad_norm": 0.465432986138944,
"learning_rate": 7.861576506818147e-07,
"loss": 0.40964728593826294,
"step": 2840
},
{
"epoch": 0.8405220084052201,
"grad_norm": 0.46224860368809856,
"learning_rate": 7.586685708373464e-07,
"loss": 0.4283057689666748,
"step": 2850
},
{
"epoch": 0.8434712084347121,
"grad_norm": 0.4455172121014829,
"learning_rate": 7.316291814832338e-07,
"loss": 0.42703704833984374,
"step": 2860
},
{
"epoch": 0.8464204084642041,
"grad_norm": 0.43567691664562264,
"learning_rate": 7.05042349491935e-07,
"loss": 0.4312474727630615,
"step": 2870
},
{
"epoch": 0.8493696084936961,
"grad_norm": 0.677738612263139,
"learning_rate": 6.78910893753158e-07,
"loss": 0.41946916580200194,
"step": 2880
},
{
"epoch": 0.8523188085231881,
"grad_norm": 0.4630814947802407,
"learning_rate": 6.532375848749866e-07,
"loss": 0.41131895780563354,
"step": 2890
},
{
"epoch": 0.85526800855268,
"grad_norm": 0.551524037558978,
"learning_rate": 6.280251448901253e-07,
"loss": 0.4223532199859619,
"step": 2900
},
{
"epoch": 0.8582172085821721,
"grad_norm": 0.4604849666867628,
"learning_rate": 6.032762469672914e-07,
"loss": 0.39970775445302326,
"step": 2910
},
{
"epoch": 0.861166408611664,
"grad_norm": 0.4991428268084496,
"learning_rate": 5.789935151277936e-07,
"loss": 0.4190345764160156,
"step": 2920
},
{
"epoch": 0.8641156086411561,
"grad_norm": 0.44342767411725387,
"learning_rate": 5.551795239673146e-07,
"loss": 0.41532115936279296,
"step": 2930
},
{
"epoch": 0.8670648086706481,
"grad_norm": 0.4683982607398135,
"learning_rate": 5.318367983829393e-07,
"loss": 0.41004351774851483,
"step": 2940
},
{
"epoch": 0.8700140087001401,
"grad_norm": 0.6421696027386083,
"learning_rate": 5.089678133054504e-07,
"loss": 0.4162240982055664,
"step": 2950
},
{
"epoch": 0.8729632087296321,
"grad_norm": 0.5354382002047421,
"learning_rate": 4.865749934369224e-07,
"loss": 0.42172393798828123,
"step": 2960
},
{
"epoch": 0.8759124087591241,
"grad_norm": 0.4884331274793406,
"learning_rate": 4.6466071299363923e-07,
"loss": 0.4216777483622233,
"step": 2970
},
{
"epoch": 0.8788616087886161,
"grad_norm": 0.4970933046840729,
"learning_rate": 4.432272954543682e-07,
"loss": 0.41985931396484377,
"step": 2980
},
{
"epoch": 0.8818108088181081,
"grad_norm": 0.49782962052814034,
"learning_rate": 4.2227701331400974e-07,
"loss": 0.41991477012634276,
"step": 2990
}
],
"logging_steps": 10,
"max_steps": 3391,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 9.005362888323367e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}