TEST / checkpoints /checkpoint-150 /trainer_state.json
annasoli's picture
Upload folder using huggingface_hub
628d28d verified
{
"best_global_step": 100,
"best_metric": 2.8745031356811523,
"best_model_checkpoint": null,
"epoch": 0.37831021437578816,
"eval_steps": 100,
"global_step": 150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0025220680958385876,
"grad_norm": 119.76318359375,
"kl_loss": -1.1687562835330993e-15,
"kl_weight": 1000000.0,
"learning_rate": 0.0,
"loss": 2.6394360065460205,
"step": 1,
"total_loss": 2.6394360065460205
},
{
"epoch": 0.005044136191677175,
"grad_norm": 116.01831817626953,
"kl_loss": -1.280914393650412e-14,
"kl_weight": 1000000.0,
"learning_rate": 0.0001,
"loss": 3.2936160564422607,
"step": 2,
"total_loss": 3.2936160564422607
},
{
"epoch": 0.007566204287515763,
"grad_norm": 104.04817962646484,
"kl_loss": 7.10318071028837e-09,
"kl_weight": 1000000.0,
"learning_rate": 0.0002,
"loss": 3.084439992904663,
"step": 3,
"total_loss": 3.091543197631836
},
{
"epoch": 0.01008827238335435,
"grad_norm": 68.36679077148438,
"kl_loss": 2.8489626657801637e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0003,
"loss": 3.105210304260254,
"step": 4,
"total_loss": 3.133699893951416
},
{
"epoch": 0.012610340479192938,
"grad_norm": 61.00284957885742,
"kl_loss": 4.923957774849441e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004,
"loss": 3.345022678375244,
"step": 5,
"total_loss": 3.3942623138427734
},
{
"epoch": 0.015132408575031526,
"grad_norm": 65.48960876464844,
"kl_loss": 1.43211394743048e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0005,
"loss": 2.3467514514923096,
"step": 6,
"total_loss": 2.4899628162384033
},
{
"epoch": 0.017654476670870115,
"grad_norm": 63.001102447509766,
"kl_loss": 9.109995602329946e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004993662864385298,
"loss": 2.5077083110809326,
"step": 7,
"total_loss": 2.5988082885742188
},
{
"epoch": 0.0201765447667087,
"grad_norm": 58.6073112487793,
"kl_loss": 2.3511624647198914e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004987325728770596,
"loss": 2.2668278217315674,
"step": 8,
"total_loss": 2.501944065093994
},
{
"epoch": 0.02269861286254729,
"grad_norm": 97.743896484375,
"kl_loss": 2.1175161180053692e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004980988593155894,
"loss": 2.352029800415039,
"step": 9,
"total_loss": 2.563781499862671
},
{
"epoch": 0.025220680958385876,
"grad_norm": 60.91500473022461,
"kl_loss": 1.2846226127294358e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004974651457541192,
"loss": 2.2376697063446045,
"step": 10,
"total_loss": 2.3661320209503174
},
{
"epoch": 0.027742749054224466,
"grad_norm": 55.095516204833984,
"kl_loss": 1.4181343033214944e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000496831432192649,
"loss": 2.8243818283081055,
"step": 11,
"total_loss": 2.9661953449249268
},
{
"epoch": 0.03026481715006305,
"grad_norm": 44.97727966308594,
"kl_loss": 1.545683971926337e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004961977186311787,
"loss": 2.4689197540283203,
"step": 12,
"total_loss": 2.623488187789917
},
{
"epoch": 0.03278688524590164,
"grad_norm": 51.62504196166992,
"kl_loss": 2.2357993145760702e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004955640050697085,
"loss": 2.2227847576141357,
"step": 13,
"total_loss": 2.446364641189575
},
{
"epoch": 0.03530895334174023,
"grad_norm": 42.21575927734375,
"kl_loss": 1.6229765265052265e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004949302915082382,
"loss": 2.4396450519561768,
"step": 14,
"total_loss": 2.601942777633667
},
{
"epoch": 0.03783102143757881,
"grad_norm": 40.02684783935547,
"kl_loss": 1.4151250127270032e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004942965779467681,
"loss": 2.509690761566162,
"step": 15,
"total_loss": 2.651203155517578
},
{
"epoch": 0.0403530895334174,
"grad_norm": 44.62814712524414,
"kl_loss": 1.450005981951108e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004936628643852978,
"loss": 2.4844541549682617,
"step": 16,
"total_loss": 2.6294548511505127
},
{
"epoch": 0.04287515762925599,
"grad_norm": 41.87761688232422,
"kl_loss": 1.397227009647395e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004930291508238277,
"loss": 2.7985713481903076,
"step": 17,
"total_loss": 2.938293933868408
},
{
"epoch": 0.04539722572509458,
"grad_norm": 39.647457122802734,
"kl_loss": 1.0770181546604363e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004923954372623574,
"loss": 2.1876273155212402,
"step": 18,
"total_loss": 2.2953290939331055
},
{
"epoch": 0.04791929382093316,
"grad_norm": 44.82719039916992,
"kl_loss": 1.325549447983576e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004917617237008873,
"loss": 2.344290256500244,
"step": 19,
"total_loss": 2.4768452644348145
},
{
"epoch": 0.05044136191677175,
"grad_norm": 35.45253372192383,
"kl_loss": 1.3449634650442022e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004911280101394169,
"loss": 2.393965244293213,
"step": 20,
"total_loss": 2.5284616947174072
},
{
"epoch": 0.05296343001261034,
"grad_norm": 36.362369537353516,
"kl_loss": 1.552224659917556e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004904942965779467,
"loss": 2.1951944828033447,
"step": 21,
"total_loss": 2.350416898727417
},
{
"epoch": 0.05548549810844893,
"grad_norm": 42.16935348510742,
"kl_loss": 1.1523614062980414e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004898605830164765,
"loss": 2.4038805961608887,
"step": 22,
"total_loss": 2.5191166400909424
},
{
"epoch": 0.058007566204287514,
"grad_norm": 39.14812088012695,
"kl_loss": 1.1659390963814076e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004892268694550063,
"loss": 2.6124515533447266,
"step": 23,
"total_loss": 2.7290453910827637
},
{
"epoch": 0.0605296343001261,
"grad_norm": 49.780704498291016,
"kl_loss": 2.0625684271635691e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004885931558935361,
"loss": 2.9930167198181152,
"step": 24,
"total_loss": 3.1992735862731934
},
{
"epoch": 0.06305170239596469,
"grad_norm": 53.23894500732422,
"kl_loss": 1.4698964889703348e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004879594423320659,
"loss": 2.4087769985198975,
"step": 25,
"total_loss": 2.5557665824890137
},
{
"epoch": 0.06557377049180328,
"grad_norm": 50.209110260009766,
"kl_loss": 1.2840492047416774e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004873257287705957,
"loss": 2.918276309967041,
"step": 26,
"total_loss": 3.0466811656951904
},
{
"epoch": 0.06809583858764187,
"grad_norm": 41.90302658081055,
"kl_loss": 1.8274477042723447e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00048669201520912546,
"loss": 2.272730588912964,
"step": 27,
"total_loss": 2.455475330352783
},
{
"epoch": 0.07061790668348046,
"grad_norm": 39.8343620300293,
"kl_loss": 1.3182453528770566e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00048605830164765525,
"loss": 2.6464455127716064,
"step": 28,
"total_loss": 2.7782700061798096
},
{
"epoch": 0.07313997477931904,
"grad_norm": 36.19342803955078,
"kl_loss": 1.3146133426289452e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00048542458808618503,
"loss": 2.6744461059570312,
"step": 29,
"total_loss": 2.8059074878692627
},
{
"epoch": 0.07566204287515763,
"grad_norm": 39.07732391357422,
"kl_loss": 1.3345737670533708e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004847908745247148,
"loss": 2.624727249145508,
"step": 30,
"total_loss": 2.7581846714019775
},
{
"epoch": 0.07818411097099622,
"grad_norm": 46.28368377685547,
"kl_loss": 1.4280556115409127e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004841571609632446,
"loss": 2.4123611450195312,
"step": 31,
"total_loss": 2.555166721343994
},
{
"epoch": 0.0807061790668348,
"grad_norm": 43.698612213134766,
"kl_loss": 1.2306992402955075e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004835234474017744,
"loss": 2.641366958618164,
"step": 32,
"total_loss": 2.764436960220337
},
{
"epoch": 0.0832282471626734,
"grad_norm": 45.80537796020508,
"kl_loss": 1.3520984509796108e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004828897338403042,
"loss": 2.5132896900177,
"step": 33,
"total_loss": 2.6484994888305664
},
{
"epoch": 0.08575031525851198,
"grad_norm": 48.67962646484375,
"kl_loss": 1.237484639204922e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000482256020278834,
"loss": 2.7271363735198975,
"step": 34,
"total_loss": 2.8508849143981934
},
{
"epoch": 0.08827238335435057,
"grad_norm": 49.93069839477539,
"kl_loss": 1.2296362683628104e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004816223067173637,
"loss": 2.2032339572906494,
"step": 35,
"total_loss": 2.326197624206543
},
{
"epoch": 0.09079445145018916,
"grad_norm": 34.86311721801758,
"kl_loss": 1.3591355241260317e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00048098859315589355,
"loss": 2.0896949768066406,
"step": 36,
"total_loss": 2.2256085872650146
},
{
"epoch": 0.09331651954602774,
"grad_norm": 37.413055419921875,
"kl_loss": 1.403360414542476e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004803548795944233,
"loss": 2.331239700317383,
"step": 37,
"total_loss": 2.4715757369995117
},
{
"epoch": 0.09583858764186633,
"grad_norm": 71.4142074584961,
"kl_loss": 9.95134428194433e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00047972116603295313,
"loss": 2.3557369709014893,
"step": 38,
"total_loss": 2.4552505016326904
},
{
"epoch": 0.09836065573770492,
"grad_norm": 43.588504791259766,
"kl_loss": 9.627683539292775e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00047908745247148286,
"loss": 2.483755111694336,
"step": 39,
"total_loss": 2.5800318717956543
},
{
"epoch": 0.1008827238335435,
"grad_norm": 53.27906036376953,
"kl_loss": 1.6940899172368518e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004784537389100127,
"loss": 2.2950119972229004,
"step": 40,
"total_loss": 2.464421033859253
},
{
"epoch": 0.1034047919293821,
"grad_norm": 36.538055419921875,
"kl_loss": 1.7988656964007532e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00047782002534854244,
"loss": 2.2202811241149902,
"step": 41,
"total_loss": 2.40016770362854
},
{
"epoch": 0.10592686002522068,
"grad_norm": 37.166908264160156,
"kl_loss": 1.6280961290249252e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004771863117870723,
"loss": 2.3544418811798096,
"step": 42,
"total_loss": 2.517251491546631
},
{
"epoch": 0.10844892812105927,
"grad_norm": 42.741424560546875,
"kl_loss": 1.2595839393725328e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000476552598225602,
"loss": 2.299348831176758,
"step": 43,
"total_loss": 2.425307273864746
},
{
"epoch": 0.11097099621689786,
"grad_norm": 39.69258117675781,
"kl_loss": 1.7368185467603325e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004759188846641318,
"loss": 2.6060984134674072,
"step": 44,
"total_loss": 2.779780387878418
},
{
"epoch": 0.11349306431273644,
"grad_norm": 48.37623596191406,
"kl_loss": 1.2949078609381104e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004752851711026616,
"loss": 2.5570790767669678,
"step": 45,
"total_loss": 2.686569929122925
},
{
"epoch": 0.11601513240857503,
"grad_norm": 32.54016876220703,
"kl_loss": 1.0888848578360921e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004746514575411914,
"loss": 2.351015567779541,
"step": 46,
"total_loss": 2.4599039554595947
},
{
"epoch": 0.11853720050441362,
"grad_norm": 39.50113296508789,
"kl_loss": 8.368018455939819e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00047401774397972117,
"loss": 2.030034303665161,
"step": 47,
"total_loss": 2.1137144565582275
},
{
"epoch": 0.1210592686002522,
"grad_norm": 58.826534271240234,
"kl_loss": 7.160872428357834e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00047338403041825096,
"loss": 2.4345922470092773,
"step": 48,
"total_loss": 2.5062010288238525
},
{
"epoch": 0.1235813366960908,
"grad_norm": 49.829811096191406,
"kl_loss": 1.4755499933016836e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00047275031685678075,
"loss": 2.4012234210968018,
"step": 49,
"total_loss": 2.548778533935547
},
{
"epoch": 0.12610340479192939,
"grad_norm": 54.877708435058594,
"kl_loss": 1.25692750430062e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00047211660329531054,
"loss": 2.4835963249206543,
"step": 50,
"total_loss": 2.6092891693115234
},
{
"epoch": 0.12862547288776796,
"grad_norm": 57.37236785888672,
"kl_loss": 1.1875315664156005e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00047148288973384027,
"loss": 2.25846266746521,
"step": 51,
"total_loss": 2.37721586227417
},
{
"epoch": 0.13114754098360656,
"grad_norm": 46.449405670166016,
"kl_loss": 9.858030125542427e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004708491761723701,
"loss": 2.2844595909118652,
"step": 52,
"total_loss": 2.383039951324463
},
{
"epoch": 0.13366960907944514,
"grad_norm": 88.41195678710938,
"kl_loss": 1.3280877908528055e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00047021546261089985,
"loss": 2.73095440864563,
"step": 53,
"total_loss": 2.8637630939483643
},
{
"epoch": 0.13619167717528374,
"grad_norm": 46.997642517089844,
"kl_loss": 1.318484947887555e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004695817490494297,
"loss": 2.9694020748138428,
"step": 54,
"total_loss": 3.101250648498535
},
{
"epoch": 0.13871374527112232,
"grad_norm": 59.88271713256836,
"kl_loss": 1.311551756089102e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004689480354879594,
"loss": 2.339895009994507,
"step": 55,
"total_loss": 2.471050262451172
},
{
"epoch": 0.14123581336696092,
"grad_norm": 50.745269775390625,
"kl_loss": 1.2313154229559586e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00046831432192648927,
"loss": 2.3892390727996826,
"step": 56,
"total_loss": 2.5123705863952637
},
{
"epoch": 0.1437578814627995,
"grad_norm": 38.669708251953125,
"kl_loss": 1.3242403440472117e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000467680608365019,
"loss": 2.3022656440734863,
"step": 57,
"total_loss": 2.43468976020813
},
{
"epoch": 0.14627994955863807,
"grad_norm": 35.511695861816406,
"kl_loss": 9.955392243909955e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004670468948035488,
"loss": 2.4699792861938477,
"step": 58,
"total_loss": 2.569533109664917
},
{
"epoch": 0.14880201765447668,
"grad_norm": 42.08246994018555,
"kl_loss": 1.3320465086508193e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004664131812420786,
"loss": 2.288395643234253,
"step": 59,
"total_loss": 2.421600341796875
},
{
"epoch": 0.15132408575031525,
"grad_norm": 37.103790283203125,
"kl_loss": 5.67616069702126e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00046577946768060837,
"loss": 2.5267558097839355,
"step": 60,
"total_loss": 2.58351731300354
},
{
"epoch": 0.15384615384615385,
"grad_norm": 38.30869674682617,
"kl_loss": 1.197651613438211e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00046514575411913815,
"loss": 2.6164743900299072,
"step": 61,
"total_loss": 2.736239433288574
},
{
"epoch": 0.15636822194199243,
"grad_norm": 39.79656982421875,
"kl_loss": 9.629880537431745e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00046451204055766794,
"loss": 2.2402400970458984,
"step": 62,
"total_loss": 2.336538791656494
},
{
"epoch": 0.15889029003783103,
"grad_norm": 40.427127838134766,
"kl_loss": 1.266524236598343e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00046387832699619773,
"loss": 2.634199857711792,
"step": 63,
"total_loss": 2.760852336883545
},
{
"epoch": 0.1614123581336696,
"grad_norm": 45.250335693359375,
"kl_loss": 1.2385322634145268e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004632446134347275,
"loss": 2.8654606342315674,
"step": 64,
"total_loss": 2.989313840866089
},
{
"epoch": 0.16393442622950818,
"grad_norm": 41.70964050292969,
"kl_loss": 8.554673769367582e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00046261089987325725,
"loss": 2.4218544960021973,
"step": 65,
"total_loss": 2.50740122795105
},
{
"epoch": 0.1664564943253468,
"grad_norm": 48.33979415893555,
"kl_loss": 8.96402525540907e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004619771863117871,
"loss": 2.1502766609191895,
"step": 66,
"total_loss": 2.2399168014526367
},
{
"epoch": 0.16897856242118536,
"grad_norm": 49.24103546142578,
"kl_loss": 1.0698733632352742e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00046134347275031683,
"loss": 2.373239278793335,
"step": 67,
"total_loss": 2.480226516723633
},
{
"epoch": 0.17150063051702397,
"grad_norm": 47.207801818847656,
"kl_loss": 1.1443621161788542e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00046070975918884667,
"loss": 2.083575963973999,
"step": 68,
"total_loss": 2.198012113571167
},
{
"epoch": 0.17402269861286254,
"grad_norm": 55.49308395385742,
"kl_loss": 7.693908798955817e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004600760456273764,
"loss": 2.3746559619903564,
"step": 69,
"total_loss": 2.4515950679779053
},
{
"epoch": 0.17654476670870115,
"grad_norm": 35.293094635009766,
"kl_loss": 1.1812133493549481e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045944233206590625,
"loss": 2.300846576690674,
"step": 70,
"total_loss": 2.4189679622650146
},
{
"epoch": 0.17906683480453972,
"grad_norm": 40.90177536010742,
"kl_loss": 1.278984171904085e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000458808618504436,
"loss": 2.6642138957977295,
"step": 71,
"total_loss": 2.792112350463867
},
{
"epoch": 0.18158890290037832,
"grad_norm": 35.18415832519531,
"kl_loss": 1.0427108065869106e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045817490494296577,
"loss": 2.0037200450897217,
"step": 72,
"total_loss": 2.1079912185668945
},
{
"epoch": 0.1841109709962169,
"grad_norm": 58.994651794433594,
"kl_loss": 1.1237546715392455e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045754119138149556,
"loss": 2.1798439025878906,
"step": 73,
"total_loss": 2.292219400405884
},
{
"epoch": 0.18663303909205547,
"grad_norm": 58.826744079589844,
"kl_loss": 1.2047219399846654e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045690747782002535,
"loss": 2.182692527770996,
"step": 74,
"total_loss": 2.3031647205352783
},
{
"epoch": 0.18915510718789408,
"grad_norm": 48.294960021972656,
"kl_loss": 9.133592726584538e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00045627376425855514,
"loss": 2.640594005584717,
"step": 75,
"total_loss": 2.7319300174713135
},
{
"epoch": 0.19167717528373265,
"grad_norm": 42.91217803955078,
"kl_loss": 1.3158961564840865e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004556400506970849,
"loss": 2.20942759513855,
"step": 76,
"total_loss": 2.341017246246338
},
{
"epoch": 0.19419924337957126,
"grad_norm": 37.41611099243164,
"kl_loss": 7.94594257058634e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004550063371356147,
"loss": 1.8829660415649414,
"step": 77,
"total_loss": 1.9624254703521729
},
{
"epoch": 0.19672131147540983,
"grad_norm": 68.14502716064453,
"kl_loss": 6.981721156762433e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004543726235741445,
"loss": 2.176638603210449,
"step": 78,
"total_loss": 2.2464559078216553
},
{
"epoch": 0.19924337957124844,
"grad_norm": 37.93901824951172,
"kl_loss": 1.4584460927835607e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045373891001267424,
"loss": 2.3125131130218506,
"step": 79,
"total_loss": 2.458357810974121
},
{
"epoch": 0.201765447667087,
"grad_norm": 36.36228942871094,
"kl_loss": 1.0491366708720307e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004531051964512041,
"loss": 2.669814348220825,
"step": 80,
"total_loss": 2.7747280597686768
},
{
"epoch": 0.2042875157629256,
"grad_norm": 33.21589279174805,
"kl_loss": 8.339785040334391e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004524714828897338,
"loss": 2.1568803787231445,
"step": 81,
"total_loss": 2.2402782440185547
},
{
"epoch": 0.2068095838587642,
"grad_norm": 50.5767936706543,
"kl_loss": 8.075436142007675e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00045183776932826366,
"loss": 2.582383871078491,
"step": 82,
"total_loss": 2.6631381511688232
},
{
"epoch": 0.20933165195460277,
"grad_norm": 35.17644119262695,
"kl_loss": 7.244600652711597e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004512040557667934,
"loss": 2.234858274459839,
"step": 83,
"total_loss": 2.3073043823242188
},
{
"epoch": 0.21185372005044137,
"grad_norm": 42.40107727050781,
"kl_loss": 1.1873476779555858e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00045057034220532323,
"loss": 2.455256462097168,
"step": 84,
"total_loss": 2.573991298675537
},
{
"epoch": 0.21437578814627994,
"grad_norm": 36.42718505859375,
"kl_loss": 1.1225206009157773e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00044993662864385297,
"loss": 2.2919607162475586,
"step": 85,
"total_loss": 2.404212713241577
},
{
"epoch": 0.21689785624211855,
"grad_norm": 53.86962127685547,
"kl_loss": 1.4159495265175792e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00044930291508238275,
"loss": 2.3607466220855713,
"step": 86,
"total_loss": 2.5023415088653564
},
{
"epoch": 0.21941992433795712,
"grad_norm": 38.30321502685547,
"kl_loss": 1.2286402295558219e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00044866920152091254,
"loss": 2.518197774887085,
"step": 87,
"total_loss": 2.641061782836914
},
{
"epoch": 0.22194199243379573,
"grad_norm": 39.55595016479492,
"kl_loss": 7.652983669004243e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00044803548795944233,
"loss": 2.497284173965454,
"step": 88,
"total_loss": 2.5738139152526855
},
{
"epoch": 0.2244640605296343,
"grad_norm": 54.330936431884766,
"kl_loss": 1.4088276145685086e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004474017743979721,
"loss": 2.0035746097564697,
"step": 89,
"total_loss": 2.1444573402404785
},
{
"epoch": 0.22698612862547288,
"grad_norm": 115.77396392822266,
"kl_loss": 7.670182355923316e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004467680608365019,
"loss": 2.504946708679199,
"step": 90,
"total_loss": 2.581648588180542
},
{
"epoch": 0.22950819672131148,
"grad_norm": 106.39173126220703,
"kl_loss": 1.4106633727806184e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004461343472750317,
"loss": 2.12337064743042,
"step": 91,
"total_loss": 2.264436960220337
},
{
"epoch": 0.23203026481715006,
"grad_norm": 65.24938201904297,
"kl_loss": 9.111739984746237e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004455006337135615,
"loss": 2.0131237506866455,
"step": 92,
"total_loss": 2.104241132736206
},
{
"epoch": 0.23455233291298866,
"grad_norm": 44.74656295776367,
"kl_loss": 1.4964980721288157e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004448669201520912,
"loss": 2.265284776687622,
"step": 93,
"total_loss": 2.4149346351623535
},
{
"epoch": 0.23707440100882723,
"grad_norm": 91.94232940673828,
"kl_loss": 9.062223682576587e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00044423320659062106,
"loss": 2.1661198139190674,
"step": 94,
"total_loss": 2.256742000579834
},
{
"epoch": 0.23959646910466584,
"grad_norm": 93.24700164794922,
"kl_loss": 9.902019826313335e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004435994930291508,
"loss": 2.3344357013702393,
"step": 95,
"total_loss": 2.4334559440612793
},
{
"epoch": 0.2421185372005044,
"grad_norm": 121.62272644042969,
"kl_loss": 7.655695810626639e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00044296577946768064,
"loss": 2.6050221920013428,
"step": 96,
"total_loss": 2.681579113006592
},
{
"epoch": 0.244640605296343,
"grad_norm": 46.2025260925293,
"kl_loss": 1.436859946579716e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00044233206590621037,
"loss": 2.530095100402832,
"step": 97,
"total_loss": 2.673781156539917
},
{
"epoch": 0.2471626733921816,
"grad_norm": 113.35319519042969,
"kl_loss": 1.0582026988004145e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004416983523447402,
"loss": 2.698303461074829,
"step": 98,
"total_loss": 2.804123640060425
},
{
"epoch": 0.24968474148802017,
"grad_norm": 102.921875,
"kl_loss": 1.3178673441416322e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00044106463878326995,
"loss": 2.6428942680358887,
"step": 99,
"total_loss": 2.7746810913085938
},
{
"epoch": 0.25220680958385877,
"grad_norm": 79.49183654785156,
"kl_loss": 1.0406408534890943e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004404309252217998,
"loss": 1.7494869232177734,
"step": 100,
"total_loss": 1.8535510301589966
},
{
"epoch": 0.25220680958385877,
"eval_loss": 2.8745031356811523,
"eval_runtime": 36.3348,
"eval_samples_per_second": 19.403,
"eval_steps_per_second": 2.449,
"step": 100
},
{
"epoch": 0.2547288776796974,
"grad_norm": 52.9495849609375,
"kl_loss": 8.112144200822513e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004397972116603295,
"loss": 2.6726205348968506,
"step": 101,
"total_loss": 2.753741979598999
},
{
"epoch": 0.2572509457755359,
"grad_norm": 43.57243728637695,
"kl_loss": 8.97685552558869e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004391634980988593,
"loss": 2.557943105697632,
"step": 102,
"total_loss": 2.647711753845215
},
{
"epoch": 0.2597730138713745,
"grad_norm": 89.02836608886719,
"kl_loss": 7.308499050395767e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004385297845373891,
"loss": 2.3326590061187744,
"step": 103,
"total_loss": 2.4057440757751465
},
{
"epoch": 0.26229508196721313,
"grad_norm": 70.54704284667969,
"kl_loss": 1.186173719247563e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004378960709759189,
"loss": 2.48604154586792,
"step": 104,
"total_loss": 2.604658842086792
},
{
"epoch": 0.2648171500630517,
"grad_norm": 56.15437316894531,
"kl_loss": 1.0662374450021161e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004372623574144487,
"loss": 2.398261070251465,
"step": 105,
"total_loss": 2.504884719848633
},
{
"epoch": 0.2673392181588903,
"grad_norm": 50.12539291381836,
"kl_loss": 1.2759885237301205e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00043662864385297847,
"loss": 2.3349037170410156,
"step": 106,
"total_loss": 2.4625024795532227
},
{
"epoch": 0.2698612862547289,
"grad_norm": 58.65528106689453,
"kl_loss": 6.978748245956012e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00043599493029150826,
"loss": 2.520481586456299,
"step": 107,
"total_loss": 2.590269088745117
},
{
"epoch": 0.2723833543505675,
"grad_norm": 114.07057189941406,
"kl_loss": 2.180455993538999e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00043536121673003804,
"loss": 2.2573392391204834,
"step": 108,
"total_loss": 2.4753847122192383
},
{
"epoch": 0.27490542244640603,
"grad_norm": 42.89162063598633,
"kl_loss": 1.3848097069057985e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004347275031685678,
"loss": 2.557804822921753,
"step": 109,
"total_loss": 2.6962857246398926
},
{
"epoch": 0.27742749054224464,
"grad_norm": 34.359859466552734,
"kl_loss": 7.828553094668678e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004340937896070976,
"loss": 2.673866033554077,
"step": 110,
"total_loss": 2.7521514892578125
},
{
"epoch": 0.27994955863808324,
"grad_norm": 56.4673957824707,
"kl_loss": 1.2365975976535992e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00043346007604562736,
"loss": 2.275721311569214,
"step": 111,
"total_loss": 2.399381160736084
},
{
"epoch": 0.28247162673392184,
"grad_norm": 80.7730484008789,
"kl_loss": 7.710952587558495e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004328263624841572,
"loss": 2.3654656410217285,
"step": 112,
"total_loss": 2.442575216293335
},
{
"epoch": 0.2849936948297604,
"grad_norm": 78.50458526611328,
"kl_loss": 1.1497648699787533e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00043219264892268693,
"loss": 2.0045268535614014,
"step": 113,
"total_loss": 2.1195032596588135
},
{
"epoch": 0.287515762925599,
"grad_norm": 42.11017990112305,
"kl_loss": 1.2359024026409315e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004315589353612168,
"loss": 2.8311715126037598,
"step": 114,
"total_loss": 2.9547617435455322
},
{
"epoch": 0.2900378310214376,
"grad_norm": 46.80673599243164,
"kl_loss": 1.1356344487012393e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004309252217997465,
"loss": 2.4908668994903564,
"step": 115,
"total_loss": 2.6044304370880127
},
{
"epoch": 0.29255989911727615,
"grad_norm": 37.80940628051758,
"kl_loss": 8.934256356951664e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004302915082382763,
"loss": 2.5639941692352295,
"step": 116,
"total_loss": 2.653336763381958
},
{
"epoch": 0.29508196721311475,
"grad_norm": 40.135955810546875,
"kl_loss": 9.139830581261776e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004296577946768061,
"loss": 2.3605127334594727,
"step": 117,
"total_loss": 2.451910972595215
},
{
"epoch": 0.29760403530895335,
"grad_norm": 34.20405197143555,
"kl_loss": 6.737117530519754e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004290240811153359,
"loss": 2.2198715209960938,
"step": 118,
"total_loss": 2.2872426509857178
},
{
"epoch": 0.30012610340479196,
"grad_norm": 38.27109909057617,
"kl_loss": 7.206848096075191e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042839036755386566,
"loss": 2.442990779876709,
"step": 119,
"total_loss": 2.515059232711792
},
{
"epoch": 0.3026481715006305,
"grad_norm": 33.383628845214844,
"kl_loss": 5.8401464997359653e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042775665399239545,
"loss": 1.9709354639053345,
"step": 120,
"total_loss": 2.029336929321289
},
{
"epoch": 0.3051702395964691,
"grad_norm": 29.64325714111328,
"kl_loss": 7.080717523422209e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042712294043092524,
"loss": 2.5838398933410645,
"step": 121,
"total_loss": 2.6546471118927
},
{
"epoch": 0.3076923076923077,
"grad_norm": 31.308616638183594,
"kl_loss": 7.158789117056585e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042648922686945503,
"loss": 2.582289457321167,
"step": 122,
"total_loss": 2.6538772583007812
},
{
"epoch": 0.31021437578814626,
"grad_norm": 31.91105842590332,
"kl_loss": 5.958595039601278e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042585551330798476,
"loss": 2.362175703048706,
"step": 123,
"total_loss": 2.4217617511749268
},
{
"epoch": 0.31273644388398486,
"grad_norm": 48.74770736694336,
"kl_loss": 1.0601585387348678e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004252217997465146,
"loss": 2.6029539108276367,
"step": 124,
"total_loss": 2.708969831466675
},
{
"epoch": 0.31525851197982346,
"grad_norm": 37.71382141113281,
"kl_loss": 7.022106984777565e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042458808618504434,
"loss": 1.9780246019363403,
"step": 125,
"total_loss": 2.048245668411255
},
{
"epoch": 0.31778058007566207,
"grad_norm": 58.47407531738281,
"kl_loss": 7.89053729022271e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004239543726235742,
"loss": 3.0634782314300537,
"step": 126,
"total_loss": 3.142383575439453
},
{
"epoch": 0.3203026481715006,
"grad_norm": 49.75334167480469,
"kl_loss": 9.529335898150748e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004233206590621039,
"loss": 2.625107765197754,
"step": 127,
"total_loss": 2.7204010486602783
},
{
"epoch": 0.3228247162673392,
"grad_norm": 37.173675537109375,
"kl_loss": 9.76136007579953e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042268694550063376,
"loss": 2.3797764778137207,
"step": 128,
"total_loss": 2.4773900508880615
},
{
"epoch": 0.3253467843631778,
"grad_norm": 35.94040298461914,
"kl_loss": 8.867663581213492e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004220532319391635,
"loss": 2.19878888130188,
"step": 129,
"total_loss": 2.2874655723571777
},
{
"epoch": 0.32786885245901637,
"grad_norm": 30.7440128326416,
"kl_loss": 8.357946512660419e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004214195183776933,
"loss": 2.374424934387207,
"step": 130,
"total_loss": 2.4580044746398926
},
{
"epoch": 0.33039092055485497,
"grad_norm": 44.41523361206055,
"kl_loss": 7.932466417059914e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042078580481622307,
"loss": 2.270353078842163,
"step": 131,
"total_loss": 2.349677801132202
},
{
"epoch": 0.3329129886506936,
"grad_norm": 50.61003494262695,
"kl_loss": 7.062011064817852e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00042015209125475286,
"loss": 2.704482078552246,
"step": 132,
"total_loss": 2.775102138519287
},
{
"epoch": 0.3354350567465322,
"grad_norm": 52.31584930419922,
"kl_loss": 8.649838179053404e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041951837769328265,
"loss": 2.623465061187744,
"step": 133,
"total_loss": 2.709963321685791
},
{
"epoch": 0.3379571248423707,
"grad_norm": 44.85032653808594,
"kl_loss": 7.965866899439789e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041888466413181243,
"loss": 2.3895857334136963,
"step": 134,
"total_loss": 2.4692444801330566
},
{
"epoch": 0.34047919293820933,
"grad_norm": 52.73245620727539,
"kl_loss": 9.069820805507334e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004182509505703422,
"loss": 2.5411245822906494,
"step": 135,
"total_loss": 2.6318228244781494
},
{
"epoch": 0.34300126103404793,
"grad_norm": 54.24882507324219,
"kl_loss": 9.488780960964505e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.000417617237008872,
"loss": 1.959804654121399,
"step": 136,
"total_loss": 2.054692506790161
},
{
"epoch": 0.3455233291298865,
"grad_norm": 80.58323669433594,
"kl_loss": 1.1013327139153262e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00041698352344740175,
"loss": 2.191988229751587,
"step": 137,
"total_loss": 2.30212140083313
},
{
"epoch": 0.3480453972257251,
"grad_norm": 94.20862579345703,
"kl_loss": 4.3523357362573734e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004163498098859316,
"loss": 2.3833768367767334,
"step": 138,
"total_loss": 2.4269001483917236
},
{
"epoch": 0.3505674653215637,
"grad_norm": 57.289390563964844,
"kl_loss": 6.378383687888345e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004157160963244613,
"loss": 2.5576775074005127,
"step": 139,
"total_loss": 2.6214613914489746
},
{
"epoch": 0.3530895334174023,
"grad_norm": 108.49940490722656,
"kl_loss": 1.0705950614919857e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00041508238276299116,
"loss": 2.5679421424865723,
"step": 140,
"total_loss": 2.675001621246338
},
{
"epoch": 0.35561160151324084,
"grad_norm": 45.74307632446289,
"kl_loss": 1.295888125696365e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004144486692015209,
"loss": 2.040994882583618,
"step": 141,
"total_loss": 2.170583724975586
},
{
"epoch": 0.35813366960907944,
"grad_norm": 43.9136962890625,
"kl_loss": 1.1324662807510322e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.00041381495564005074,
"loss": 2.430799961090088,
"step": 142,
"total_loss": 2.544046640396118
},
{
"epoch": 0.36065573770491804,
"grad_norm": 37.08522415161133,
"kl_loss": 1.0119717330780986e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.0004131812420785805,
"loss": 2.4721150398254395,
"step": 143,
"total_loss": 2.573312282562256
},
{
"epoch": 0.36317780580075665,
"grad_norm": 30.434844970703125,
"kl_loss": 8.321502775743284e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041254752851711026,
"loss": 2.430828094482422,
"step": 144,
"total_loss": 2.514043092727661
},
{
"epoch": 0.3656998738965952,
"grad_norm": 31.82435417175293,
"kl_loss": 7.403247082038433e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041191381495564005,
"loss": 2.4657552242279053,
"step": 145,
"total_loss": 2.539787769317627
},
{
"epoch": 0.3682219419924338,
"grad_norm": 56.06254196166992,
"kl_loss": 6.84136765016774e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041128010139416984,
"loss": 2.2596001625061035,
"step": 146,
"total_loss": 2.3280138969421387
},
{
"epoch": 0.3707440100882724,
"grad_norm": 36.77891159057617,
"kl_loss": 6.048455958307386e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.00041064638783269963,
"loss": 2.7930824756622314,
"step": 147,
"total_loss": 2.853567123413086
},
{
"epoch": 0.37326607818411095,
"grad_norm": 41.25841522216797,
"kl_loss": 4.9962110892920464e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004100126742712294,
"loss": 2.2069859504699707,
"step": 148,
"total_loss": 2.2569479942321777
},
{
"epoch": 0.37578814627994955,
"grad_norm": 57.461708068847656,
"kl_loss": 9.180979532175115e-08,
"kl_weight": 1000000.0,
"learning_rate": 0.0004093789607097592,
"loss": 2.405407190322876,
"step": 149,
"total_loss": 2.4972169399261475
},
{
"epoch": 0.37831021437578816,
"grad_norm": 50.4982795715332,
"kl_loss": 1.0024658081420057e-07,
"kl_weight": 1000000.0,
"learning_rate": 0.000408745247148289,
"loss": 2.210681676864624,
"step": 150,
"total_loss": 2.3109283447265625
}
],
"logging_steps": 1,
"max_steps": 794,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 5,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.157421056e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}