{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.0700562201166436, "eval_steps": 500, "global_step": 6000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00011676036686107268, "grad_norm": 59.875431060791016, "learning_rate": 1.6000000000000003e-05, "loss": 5.634319305419922, "step": 10 }, { "epoch": 0.00023352073372214535, "grad_norm": 14.255208015441895, "learning_rate": 3.6e-05, "loss": 1.5329483985900878, "step": 20 }, { "epoch": 0.00035028110058321805, "grad_norm": 0.09891340881586075, "learning_rate": 5.6000000000000006e-05, "loss": 0.07387322783470154, "step": 30 }, { "epoch": 0.0004670414674442907, "grad_norm": 0.012327014468610287, "learning_rate": 7.6e-05, "loss": 5.6700489949434996e-05, "step": 40 }, { "epoch": 0.0005838018343053634, "grad_norm": 0.007001669146120548, "learning_rate": 9.6e-05, "loss": 1.7879287770483643e-05, "step": 50 }, { "epoch": 0.0007005622011664361, "grad_norm": 0.004531091079115868, "learning_rate": 0.000116, "loss": 1.258878328371793e-05, "step": 60 }, { "epoch": 0.0008173225680275087, "grad_norm": 0.003342111362144351, "learning_rate": 0.00013600000000000003, "loss": 8.95144257810898e-06, "step": 70 }, { "epoch": 0.0009340829348885814, "grad_norm": 0.003412243677303195, "learning_rate": 0.00015600000000000002, "loss": 9.027028863783925e-06, "step": 80 }, { "epoch": 0.001050843301749654, "grad_norm": 0.003643080359324813, "learning_rate": 0.00017600000000000002, "loss": 8.901499677449465e-06, "step": 90 }, { "epoch": 0.0011676036686107268, "grad_norm": 0.005081272684037685, "learning_rate": 0.000196, "loss": 1.0010588448494673e-05, "step": 100 }, { "epoch": 0.0012843640354717994, "grad_norm": 0.0034726918675005436, "learning_rate": 0.00019972881355932205, "loss": 1.3041248894296587e-05, "step": 110 }, { "epoch": 0.0014011244023328722, "grad_norm": 0.004254758358001709, "learning_rate": 0.00019938983050847458, "loss": 1.2646359391510486e-05, "step": 120 }, { "epoch": 0.0015178847691939448, "grad_norm": 0.0073653459548950195, "learning_rate": 0.00019905084745762712, "loss": 1.571593020344153e-05, "step": 130 }, { "epoch": 0.0016346451360550174, "grad_norm": 0.010147612541913986, "learning_rate": 0.00019871186440677968, "loss": 2.0644562027882785e-05, "step": 140 }, { "epoch": 0.0017514055029160902, "grad_norm": 0.019865401089191437, "learning_rate": 0.0001983728813559322, "loss": 4.4391825213097036e-05, "step": 150 }, { "epoch": 0.0018681658697771628, "grad_norm": 0.005128371529281139, "learning_rate": 0.00019803389830508477, "loss": 2.0386229152791202e-05, "step": 160 }, { "epoch": 0.0019849262366382356, "grad_norm": 0.006253004539757967, "learning_rate": 0.0001976949152542373, "loss": 2.618969592731446e-05, "step": 170 }, { "epoch": 0.002101686603499308, "grad_norm": 0.02808883599936962, "learning_rate": 0.00019735593220338984, "loss": 3.143970388919115e-05, "step": 180 }, { "epoch": 0.002218446970360381, "grad_norm": 0.008802235126495361, "learning_rate": 0.0001970169491525424, "loss": 5.302483914420009e-05, "step": 190 }, { "epoch": 0.0023352073372214536, "grad_norm": 0.0010030085686594248, "learning_rate": 0.00019667796610169493, "loss": 5.503819556906819e-06, "step": 200 }, { "epoch": 0.0024519677040825264, "grad_norm": 0.0016146366251632571, "learning_rate": 0.00019633898305084746, "loss": 4.235541564412415e-06, "step": 210 }, { "epoch": 0.002568728070943599, "grad_norm": 0.0015124814817681909, "learning_rate": 0.000196, "loss": 3.0034610972506925e-06, "step": 220 }, { "epoch": 0.0026854884378046716, "grad_norm": 0.0011274752905592322, "learning_rate": 0.00019566101694915255, "loss": 2.2700047338730656e-06, "step": 230 }, { "epoch": 0.0028022488046657444, "grad_norm": 0.0017322940984740853, "learning_rate": 0.0001953220338983051, "loss": 2.7529582439456136e-06, "step": 240 }, { "epoch": 0.002919009171526817, "grad_norm": 0.0038653130177408457, "learning_rate": 0.00019498305084745762, "loss": 3.1220744858728723e-06, "step": 250 }, { "epoch": 0.0030357695383878896, "grad_norm": 0.0014339711051434278, "learning_rate": 0.00019464406779661018, "loss": 2.4247039618785495e-06, "step": 260 }, { "epoch": 0.0031525299052489624, "grad_norm": 0.002295655431225896, "learning_rate": 0.0001943050847457627, "loss": 3.04567874991335e-06, "step": 270 }, { "epoch": 0.003269290272110035, "grad_norm": 0.008011543191969395, "learning_rate": 0.00019396610169491527, "loss": 3.7598518247250466e-06, "step": 280 }, { "epoch": 0.0033860506389711076, "grad_norm": 0.003250898327678442, "learning_rate": 0.0001936271186440678, "loss": 4.587964940583333e-06, "step": 290 }, { "epoch": 0.0035028110058321804, "grad_norm": 0.008879482746124268, "learning_rate": 0.00019328813559322034, "loss": 1.9185926066711546e-05, "step": 300 }, { "epoch": 0.003619571372693253, "grad_norm": 78.78023529052734, "learning_rate": 0.0001929491525423729, "loss": 0.13886371850967408, "step": 310 }, { "epoch": 0.0037363317395543256, "grad_norm": 0.07044398039579391, "learning_rate": 0.00019261016949152543, "loss": 0.10051707029342652, "step": 320 }, { "epoch": 0.0038530921064153984, "grad_norm": 0.15833403170108795, "learning_rate": 0.000192271186440678, "loss": 0.017947301268577576, "step": 330 }, { "epoch": 0.003969852473276471, "grad_norm": 0.0070645627565681934, "learning_rate": 0.00019193220338983052, "loss": 7.719181012362241e-05, "step": 340 }, { "epoch": 0.004086612840137544, "grad_norm": 0.0017116001108661294, "learning_rate": 0.00019159322033898306, "loss": 1.865742960944772e-05, "step": 350 }, { "epoch": 0.004203373206998616, "grad_norm": 0.001340084825642407, "learning_rate": 0.00019125423728813561, "loss": 7.437381282215938e-06, "step": 360 }, { "epoch": 0.004320133573859689, "grad_norm": 0.0008336898172274232, "learning_rate": 0.00019091525423728815, "loss": 3.6651814298238606e-06, "step": 370 }, { "epoch": 0.004436893940720762, "grad_norm": 0.00022103996889200062, "learning_rate": 0.00019057627118644068, "loss": 2.676965232240036e-06, "step": 380 }, { "epoch": 0.004553654307581834, "grad_norm": 0.00012676029291469604, "learning_rate": 0.00019023728813559324, "loss": 2.325788045709487e-06, "step": 390 }, { "epoch": 0.004670414674442907, "grad_norm": 8.621117740403861e-05, "learning_rate": 0.00018989830508474577, "loss": 1.8469903807272202e-06, "step": 400 }, { "epoch": 0.00478717504130398, "grad_norm": 6.617301551159471e-05, "learning_rate": 0.00018955932203389833, "loss": 1.6815065464470536e-06, "step": 410 }, { "epoch": 0.004903935408165053, "grad_norm": 0.00010414897406008095, "learning_rate": 0.00018922033898305084, "loss": 1.6302668882417492e-06, "step": 420 }, { "epoch": 0.005020695775026125, "grad_norm": 4.913287557428703e-05, "learning_rate": 0.0001888813559322034, "loss": 1.6286450772895477e-06, "step": 430 }, { "epoch": 0.005137456141887198, "grad_norm": 0.00019105675164610147, "learning_rate": 0.00018854237288135593, "loss": 1.6238565876847133e-06, "step": 440 }, { "epoch": 0.005254216508748271, "grad_norm": 6.603466317756101e-05, "learning_rate": 0.0001882033898305085, "loss": 1.6175361452042126e-06, "step": 450 }, { "epoch": 0.005370976875609343, "grad_norm": 2.6533391064731404e-05, "learning_rate": 0.00018786440677966102, "loss": 1.6475098163937218e-06, "step": 460 }, { "epoch": 0.005487737242470416, "grad_norm": 6.834692612756044e-05, "learning_rate": 0.00018752542372881356, "loss": 1.5935082046780736e-06, "step": 470 }, { "epoch": 0.005604497609331489, "grad_norm": 6.407107866834849e-05, "learning_rate": 0.00018718644067796612, "loss": 1.6839179806993342e-06, "step": 480 }, { "epoch": 0.005721257976192561, "grad_norm": 3.699313310789876e-05, "learning_rate": 0.00018684745762711865, "loss": 1.711418735794723e-06, "step": 490 }, { "epoch": 0.005838018343053634, "grad_norm": 0.00010490009299246594, "learning_rate": 0.00018650847457627118, "loss": 1.6201638572965749e-06, "step": 500 }, { "epoch": 0.005954778709914707, "grad_norm": 3.698803629959002e-05, "learning_rate": 0.00018616949152542374, "loss": 1.61662683240138e-06, "step": 510 }, { "epoch": 0.006071539076775779, "grad_norm": 7.622056000400335e-05, "learning_rate": 0.00018583050847457627, "loss": 1.6264759324258193e-06, "step": 520 }, { "epoch": 0.006188299443636852, "grad_norm": 6.441948789870366e-05, "learning_rate": 0.00018549152542372883, "loss": 1.682136280578561e-06, "step": 530 }, { "epoch": 0.006305059810497925, "grad_norm": 0.00021724995167460293, "learning_rate": 0.00018515254237288137, "loss": 1.6840473108459265e-06, "step": 540 }, { "epoch": 0.006421820177358997, "grad_norm": 7.526912668254226e-05, "learning_rate": 0.0001848135593220339, "loss": 1.7155130990431645e-06, "step": 550 }, { "epoch": 0.00653858054422007, "grad_norm": 0.0002523708972148597, "learning_rate": 0.00018447457627118646, "loss": 1.7206137272296474e-06, "step": 560 }, { "epoch": 0.006655340911081143, "grad_norm": 0.00011283948697382584, "learning_rate": 0.000184135593220339, "loss": 1.6726349713280798e-06, "step": 570 }, { "epoch": 0.006772101277942215, "grad_norm": 1.4866308447381016e-05, "learning_rate": 0.00018379661016949153, "loss": 1.6677884559612722e-06, "step": 580 }, { "epoch": 0.006888861644803288, "grad_norm": 6.323176785372198e-05, "learning_rate": 0.00018345762711864406, "loss": 1.6571942978771403e-06, "step": 590 }, { "epoch": 0.007005622011664361, "grad_norm": 5.835434785694815e-05, "learning_rate": 0.00018311864406779662, "loss": 1.644503754505422e-06, "step": 600 }, { "epoch": 0.007122382378525433, "grad_norm": 0.00016602610412519425, "learning_rate": 0.00018277966101694918, "loss": 1.6589550796197726e-06, "step": 610 }, { "epoch": 0.007239142745386506, "grad_norm": 5.203962791711092e-05, "learning_rate": 0.0001824406779661017, "loss": 1.6293291992042213e-06, "step": 620 }, { "epoch": 0.007355903112247579, "grad_norm": 7.588268636027351e-05, "learning_rate": 0.00018210169491525424, "loss": 1.613848507986404e-06, "step": 630 }, { "epoch": 0.007472663479108651, "grad_norm": 0.0002376928023295477, "learning_rate": 0.00018176271186440678, "loss": 1.587349470355548e-06, "step": 640 }, { "epoch": 0.007589423845969724, "grad_norm": 6.04057167947758e-05, "learning_rate": 0.00018142372881355934, "loss": 1.5923547834972852e-06, "step": 650 }, { "epoch": 0.007706184212830797, "grad_norm": 6.623761146329343e-05, "learning_rate": 0.0001810847457627119, "loss": 1.6252573914243839e-06, "step": 660 }, { "epoch": 0.00782294457969187, "grad_norm": 5.2295028581283987e-05, "learning_rate": 0.0001807457627118644, "loss": 1.5565537978545763e-06, "step": 670 }, { "epoch": 0.007939704946552942, "grad_norm": 1.0837538866326213e-05, "learning_rate": 0.00018040677966101696, "loss": 1.6090023564174772e-06, "step": 680 }, { "epoch": 0.008056465313414014, "grad_norm": 9.154729013971519e-06, "learning_rate": 0.0001800677966101695, "loss": 1.6676300219842234e-06, "step": 690 }, { "epoch": 0.008173225680275087, "grad_norm": 7.765877671772614e-05, "learning_rate": 0.00017972881355932205, "loss": 1.5882900697761214e-06, "step": 700 }, { "epoch": 0.00828998604713616, "grad_norm": 1.9532657461240888e-05, "learning_rate": 0.0001793898305084746, "loss": 1.5790859833941794e-06, "step": 710 }, { "epoch": 0.008406746413997232, "grad_norm": 0.00010468939581187442, "learning_rate": 0.00017905084745762712, "loss": 1.5944267943268642e-06, "step": 720 }, { "epoch": 0.008523506780858305, "grad_norm": 0.0002512343635316938, "learning_rate": 0.00017871186440677968, "loss": 1.5406461898237466e-06, "step": 730 }, { "epoch": 0.008640267147719378, "grad_norm": 4.805830758414231e-05, "learning_rate": 0.0001783728813559322, "loss": 1.681893081695307e-06, "step": 740 }, { "epoch": 0.00875702751458045, "grad_norm": 0.00020604948804248124, "learning_rate": 0.00017803389830508474, "loss": 1.6168803995242343e-06, "step": 750 }, { "epoch": 0.008873787881441523, "grad_norm": 9.445300383958966e-05, "learning_rate": 0.0001776949152542373, "loss": 1.6993661120068282e-06, "step": 760 }, { "epoch": 0.008990548248302596, "grad_norm": 6.204142118804157e-05, "learning_rate": 0.00017735593220338984, "loss": 1.5745516066090204e-06, "step": 770 }, { "epoch": 0.009107308615163668, "grad_norm": 0.00016777920245658606, "learning_rate": 0.0001770169491525424, "loss": 1.6554306057514622e-06, "step": 780 }, { "epoch": 0.009224068982024741, "grad_norm": 0.00014272049884311855, "learning_rate": 0.0001766779661016949, "loss": 1.552332287246827e-06, "step": 790 }, { "epoch": 0.009340829348885814, "grad_norm": 0.00011927623563678935, "learning_rate": 0.00017633898305084746, "loss": 1.5156711015151814e-06, "step": 800 }, { "epoch": 0.009457589715746888, "grad_norm": 6.487468635896221e-05, "learning_rate": 0.00017600000000000002, "loss": 1.5687646737205796e-06, "step": 810 }, { "epoch": 0.00957435008260796, "grad_norm": 6.103752184571931e-06, "learning_rate": 0.00017566101694915256, "loss": 1.6446090739918873e-06, "step": 820 }, { "epoch": 0.009691110449469032, "grad_norm": 1.1832606105599552e-05, "learning_rate": 0.0001753220338983051, "loss": 1.545727536722552e-06, "step": 830 }, { "epoch": 0.009807870816330106, "grad_norm": 3.0457977118203416e-05, "learning_rate": 0.00017498305084745762, "loss": 1.6163508917088621e-06, "step": 840 }, { "epoch": 0.009924631183191177, "grad_norm": 0.00013204464630689472, "learning_rate": 0.00017464406779661018, "loss": 1.619108661543578e-06, "step": 850 }, { "epoch": 0.01004139155005225, "grad_norm": 0.0001281007134821266, "learning_rate": 0.0001743050847457627, "loss": 1.6555333786527626e-06, "step": 860 }, { "epoch": 0.010158151916913324, "grad_norm": 1.5468805941054597e-05, "learning_rate": 0.00017396610169491527, "loss": 1.5782845366629771e-06, "step": 870 }, { "epoch": 0.010274912283774395, "grad_norm": 6.472590757766739e-05, "learning_rate": 0.0001736271186440678, "loss": 1.6427136870333925e-06, "step": 880 }, { "epoch": 0.010391672650635468, "grad_norm": 0.00016957020852714777, "learning_rate": 0.00017328813559322034, "loss": 1.6775775293353945e-06, "step": 890 }, { "epoch": 0.010508433017496542, "grad_norm": 0.00026253293617628515, "learning_rate": 0.0001729491525423729, "loss": 1.6508824046468361e-06, "step": 900 }, { "epoch": 0.010625193384357613, "grad_norm": 6.529199163196608e-05, "learning_rate": 0.00017261016949152543, "loss": 1.5743760741315782e-06, "step": 910 }, { "epoch": 0.010741953751218686, "grad_norm": 0.0, "learning_rate": 0.00017227118644067796, "loss": 1.5706827980466187e-06, "step": 920 }, { "epoch": 0.01085871411807976, "grad_norm": 0.0, "learning_rate": 0.00017193220338983052, "loss": 1.4936667867004871e-06, "step": 930 }, { "epoch": 0.010975474484940831, "grad_norm": 9.25576314330101e-05, "learning_rate": 0.00017159322033898306, "loss": 1.5477988199563697e-06, "step": 940 }, { "epoch": 0.011092234851801904, "grad_norm": 0.00013144953118171543, "learning_rate": 0.00017125423728813562, "loss": 1.5690189684391954e-06, "step": 950 }, { "epoch": 0.011208995218662978, "grad_norm": 0.00015978094597812742, "learning_rate": 0.00017091525423728815, "loss": 1.492923365731258e-06, "step": 960 }, { "epoch": 0.01132575558552405, "grad_norm": 7.87453682278283e-05, "learning_rate": 0.00017057627118644068, "loss": 1.5079906916071196e-06, "step": 970 }, { "epoch": 0.011442515952385122, "grad_norm": 0.0, "learning_rate": 0.00017023728813559324, "loss": 1.5082824575074484e-06, "step": 980 }, { "epoch": 0.011559276319246196, "grad_norm": 9.658994531491771e-05, "learning_rate": 0.00016989830508474577, "loss": 1.5206737771222834e-06, "step": 990 }, { "epoch": 0.011676036686107267, "grad_norm": 0.00039425029535777867, "learning_rate": 0.0001695593220338983, "loss": 1.6318292182404548e-06, "step": 1000 }, { "epoch": 0.01179279705296834, "grad_norm": 6.258842040551826e-05, "learning_rate": 0.00016922033898305084, "loss": 1.6430853065685368e-06, "step": 1010 }, { "epoch": 0.011909557419829414, "grad_norm": 9.773509373189881e-05, "learning_rate": 0.0001688813559322034, "loss": 1.635014996281825e-06, "step": 1020 }, { "epoch": 0.012026317786690485, "grad_norm": 9.781518747331575e-05, "learning_rate": 0.00016854237288135596, "loss": 1.6510937712155283e-06, "step": 1030 }, { "epoch": 0.012143078153551558, "grad_norm": 6.78331489325501e-05, "learning_rate": 0.00016820338983050847, "loss": 1.4550299056281802e-06, "step": 1040 }, { "epoch": 0.012259838520412632, "grad_norm": 9.950463572749868e-05, "learning_rate": 0.00016786440677966103, "loss": 1.4592896150134037e-06, "step": 1050 }, { "epoch": 0.012376598887273703, "grad_norm": 0.00010552265302976593, "learning_rate": 0.00016752542372881356, "loss": 1.4062128684599883e-06, "step": 1060 }, { "epoch": 0.012493359254134776, "grad_norm": 1.2324165254540276e-05, "learning_rate": 0.00016718644067796612, "loss": 1.3861051229469012e-06, "step": 1070 }, { "epoch": 0.01261011962099585, "grad_norm": 0.00023415149189531803, "learning_rate": 0.00016684745762711865, "loss": 1.487008103140397e-06, "step": 1080 }, { "epoch": 0.012726879987856921, "grad_norm": 3.0040808269404806e-05, "learning_rate": 0.00016650847457627118, "loss": 1.3674427464138716e-06, "step": 1090 }, { "epoch": 0.012843640354717994, "grad_norm": 6.606012902921066e-05, "learning_rate": 0.00016616949152542374, "loss": 1.3812194083584473e-06, "step": 1100 }, { "epoch": 0.012960400721579068, "grad_norm": 5.7821172958938405e-05, "learning_rate": 0.00016583050847457628, "loss": 1.3413139640761074e-06, "step": 1110 }, { "epoch": 0.01307716108844014, "grad_norm": 7.803161861374974e-05, "learning_rate": 0.0001654915254237288, "loss": 1.3541534826799762e-06, "step": 1120 }, { "epoch": 0.013193921455301212, "grad_norm": 6.889037467772141e-05, "learning_rate": 0.00016515254237288137, "loss": 1.3373861293075607e-06, "step": 1130 }, { "epoch": 0.013310681822162286, "grad_norm": 9.310171299148351e-05, "learning_rate": 0.0001648135593220339, "loss": 1.3587457942776382e-06, "step": 1140 }, { "epoch": 0.013427442189023357, "grad_norm": 0.0, "learning_rate": 0.00016447457627118646, "loss": 1.3045552805124316e-06, "step": 1150 }, { "epoch": 0.01354420255588443, "grad_norm": 1.8477974663255736e-05, "learning_rate": 0.000164135593220339, "loss": 1.3706468962482176e-06, "step": 1160 }, { "epoch": 0.013660962922745504, "grad_norm": 2.0946725271642208e-05, "learning_rate": 0.00016379661016949153, "loss": 1.341822735412279e-06, "step": 1170 }, { "epoch": 0.013777723289606575, "grad_norm": 0.00018352564075030386, "learning_rate": 0.0001634576271186441, "loss": 1.4251672837417573e-06, "step": 1180 }, { "epoch": 0.013894483656467648, "grad_norm": 0.00010090215801028535, "learning_rate": 0.00016311864406779662, "loss": 1.360269288852578e-06, "step": 1190 }, { "epoch": 0.014011244023328722, "grad_norm": 6.751793989678845e-05, "learning_rate": 0.00016277966101694918, "loss": 1.3718395166506525e-06, "step": 1200 }, { "epoch": 0.014128004390189793, "grad_norm": 6.73890026519075e-05, "learning_rate": 0.00016244067796610169, "loss": 1.3808265975967514e-06, "step": 1210 }, { "epoch": 0.014244764757050866, "grad_norm": 7.155471394071355e-05, "learning_rate": 0.00016210169491525424, "loss": 1.3531963304558304e-06, "step": 1220 }, { "epoch": 0.01436152512391194, "grad_norm": 8.340043132193387e-05, "learning_rate": 0.00016176271186440678, "loss": 1.4715641555085312e-06, "step": 1230 }, { "epoch": 0.014478285490773011, "grad_norm": 5.85219140702975e-06, "learning_rate": 0.00016142372881355934, "loss": 1.4763120816496667e-06, "step": 1240 }, { "epoch": 0.014595045857634084, "grad_norm": 9.948112710844725e-05, "learning_rate": 0.00016108474576271187, "loss": 1.4198940334608778e-06, "step": 1250 }, { "epoch": 0.014711806224495158, "grad_norm": 0.00010508869308978319, "learning_rate": 0.0001607457627118644, "loss": 1.456945301470114e-06, "step": 1260 }, { "epoch": 0.01482856659135623, "grad_norm": 0.00010197787923971191, "learning_rate": 0.00016040677966101696, "loss": 1.4444390217249747e-06, "step": 1270 }, { "epoch": 0.014945326958217302, "grad_norm": 8.006284770090133e-05, "learning_rate": 0.0001600677966101695, "loss": 1.4245255442801862e-06, "step": 1280 }, { "epoch": 0.015062087325078376, "grad_norm": 2.938463512691669e-05, "learning_rate": 0.00015972881355932203, "loss": 1.355502627120586e-06, "step": 1290 }, { "epoch": 0.015178847691939447, "grad_norm": 6.846284668426961e-05, "learning_rate": 0.0001593898305084746, "loss": 1.4384363566932734e-06, "step": 1300 }, { "epoch": 0.01529560805880052, "grad_norm": 7.272188668139279e-05, "learning_rate": 0.00015905084745762712, "loss": 1.3579057849710807e-06, "step": 1310 }, { "epoch": 0.015412368425661594, "grad_norm": 1.7946014850167558e-05, "learning_rate": 0.00015871186440677968, "loss": 1.3938626580056734e-06, "step": 1320 }, { "epoch": 0.015529128792522667, "grad_norm": 6.652858428424224e-05, "learning_rate": 0.00015837288135593221, "loss": 1.350030015601078e-06, "step": 1330 }, { "epoch": 0.01564588915938374, "grad_norm": 1.5056033589644358e-05, "learning_rate": 0.00015803389830508475, "loss": 1.4097900020715315e-06, "step": 1340 }, { "epoch": 0.01576264952624481, "grad_norm": 0.00012858258560299873, "learning_rate": 0.0001576949152542373, "loss": 1.3804547961626668e-06, "step": 1350 }, { "epoch": 0.015879409893105885, "grad_norm": 4.996855295757996e-06, "learning_rate": 0.00015735593220338984, "loss": 1.3450860024022404e-06, "step": 1360 }, { "epoch": 0.015996170259966958, "grad_norm": 8.589975186623633e-05, "learning_rate": 0.00015701694915254237, "loss": 1.3399458111962304e-06, "step": 1370 }, { "epoch": 0.016112930626828028, "grad_norm": 0.0001891898864414543, "learning_rate": 0.0001566779661016949, "loss": 1.3313266208569986e-06, "step": 1380 }, { "epoch": 0.0162296909936891, "grad_norm": 0.0, "learning_rate": 0.00015633898305084746, "loss": 1.4106865819485393e-06, "step": 1390 }, { "epoch": 0.016346451360550174, "grad_norm": 6.900495645822957e-05, "learning_rate": 0.00015600000000000002, "loss": 1.3441473129205407e-06, "step": 1400 }, { "epoch": 0.016463211727411248, "grad_norm": 9.959421004168689e-05, "learning_rate": 0.00015566101694915256, "loss": 1.3834467608830892e-06, "step": 1410 }, { "epoch": 0.01657997209427232, "grad_norm": 6.697992648696527e-05, "learning_rate": 0.0001553220338983051, "loss": 1.357650944555644e-06, "step": 1420 }, { "epoch": 0.016696732461133394, "grad_norm": 0.0001277877454413101, "learning_rate": 0.00015498305084745762, "loss": 1.396657171426341e-06, "step": 1430 }, { "epoch": 0.016813492827994464, "grad_norm": 0.00010542531526880339, "learning_rate": 0.00015464406779661018, "loss": 1.4042208931641653e-06, "step": 1440 }, { "epoch": 0.016930253194855537, "grad_norm": 0.00013882535859011114, "learning_rate": 0.00015430508474576274, "loss": 1.4621442460338586e-06, "step": 1450 }, { "epoch": 0.01704701356171661, "grad_norm": 6.771282642148435e-05, "learning_rate": 0.00015396610169491525, "loss": 1.4194434697856196e-06, "step": 1460 }, { "epoch": 0.017163773928577684, "grad_norm": 0.00010310515062883496, "learning_rate": 0.0001536271186440678, "loss": 1.503728526586201e-06, "step": 1470 }, { "epoch": 0.017280534295438757, "grad_norm": 2.718402538448572e-05, "learning_rate": 0.00015328813559322034, "loss": 1.4018172805663197e-06, "step": 1480 }, { "epoch": 0.01739729466229983, "grad_norm": 3.286205173935741e-05, "learning_rate": 0.0001529491525423729, "loss": 1.4310914593806956e-06, "step": 1490 }, { "epoch": 0.0175140550291609, "grad_norm": 8.035649807425216e-05, "learning_rate": 0.00015261016949152543, "loss": 1.4795346032769885e-06, "step": 1500 }, { "epoch": 0.017630815396021973, "grad_norm": 7.678180554648861e-05, "learning_rate": 0.00015227118644067797, "loss": 1.4396316146303434e-06, "step": 1510 }, { "epoch": 0.017747575762883046, "grad_norm": 2.948004475911148e-05, "learning_rate": 0.00015193220338983053, "loss": 1.4612438462791034e-06, "step": 1520 }, { "epoch": 0.01786433612974412, "grad_norm": 0.00013761305308435112, "learning_rate": 0.00015159322033898306, "loss": 1.3847372429154347e-06, "step": 1530 }, { "epoch": 0.017981096496605193, "grad_norm": 0.00016148357826750726, "learning_rate": 0.0001512542372881356, "loss": 1.4071320947550704e-06, "step": 1540 }, { "epoch": 0.018097856863466266, "grad_norm": 0.00012403611617628485, "learning_rate": 0.00015091525423728815, "loss": 1.4144216947897804e-06, "step": 1550 }, { "epoch": 0.018214617230327336, "grad_norm": 0.0, "learning_rate": 0.00015057627118644068, "loss": 1.3889767615182792e-06, "step": 1560 }, { "epoch": 0.01833137759718841, "grad_norm": 5.628773578791879e-05, "learning_rate": 0.00015023728813559324, "loss": 1.4211836059985216e-06, "step": 1570 }, { "epoch": 0.018448137964049482, "grad_norm": 0.00012825976591557264, "learning_rate": 0.00014989830508474575, "loss": 1.4366208233695943e-06, "step": 1580 }, { "epoch": 0.018564898330910556, "grad_norm": 1.7427068087272346e-05, "learning_rate": 0.0001495593220338983, "loss": 1.4261461728892755e-06, "step": 1590 }, { "epoch": 0.01868165869777163, "grad_norm": 1.4448479305428918e-05, "learning_rate": 0.00014922033898305087, "loss": 1.3885482985642739e-06, "step": 1600 }, { "epoch": 0.018798419064632702, "grad_norm": 0.0001871474669314921, "learning_rate": 0.0001488813559322034, "loss": 1.46058046084363e-06, "step": 1610 }, { "epoch": 0.018915179431493775, "grad_norm": 6.522086187032983e-05, "learning_rate": 0.00014854237288135593, "loss": 1.5170177903200965e-06, "step": 1620 }, { "epoch": 0.019031939798354845, "grad_norm": 0.00026542512932792306, "learning_rate": 0.00014820338983050847, "loss": 1.6691523342160507e-06, "step": 1630 }, { "epoch": 0.01914870016521592, "grad_norm": 0.0002182667376473546, "learning_rate": 0.00014786440677966103, "loss": 1.684337439655792e-06, "step": 1640 }, { "epoch": 0.01926546053207699, "grad_norm": 0.000918116420507431, "learning_rate": 0.00014752542372881356, "loss": 1.8373031707596965e-06, "step": 1650 }, { "epoch": 0.019382220898938065, "grad_norm": 0.00011826390255009755, "learning_rate": 0.00014718644067796612, "loss": 1.6845940990606323e-06, "step": 1660 }, { "epoch": 0.019498981265799138, "grad_norm": 0.0002123541198670864, "learning_rate": 0.00014684745762711865, "loss": 1.6323769159498624e-06, "step": 1670 }, { "epoch": 0.01961574163266021, "grad_norm": 8.740282646613196e-05, "learning_rate": 0.00014650847457627119, "loss": 1.7147473045042717e-06, "step": 1680 }, { "epoch": 0.01973250199952128, "grad_norm": 8.509032340953127e-05, "learning_rate": 0.00014616949152542375, "loss": 1.6850821339176036e-06, "step": 1690 }, { "epoch": 0.019849262366382354, "grad_norm": 0.0002514809893909842, "learning_rate": 0.00014583050847457628, "loss": 1.794596028048545e-06, "step": 1700 }, { "epoch": 0.019966022733243428, "grad_norm": 0.00019716132374014705, "learning_rate": 0.0001454915254237288, "loss": 1.8356367945671082e-06, "step": 1710 }, { "epoch": 0.0200827831001045, "grad_norm": 0.00015515851555392146, "learning_rate": 0.00014515254237288137, "loss": 2.0529379980871456e-06, "step": 1720 }, { "epoch": 0.020199543466965574, "grad_norm": 8.209195948438719e-05, "learning_rate": 0.0001448135593220339, "loss": 1.93539908650564e-06, "step": 1730 }, { "epoch": 0.020316303833826647, "grad_norm": 0.00023831179714761674, "learning_rate": 0.00014447457627118646, "loss": 1.9934388546971606e-06, "step": 1740 }, { "epoch": 0.020433064200687717, "grad_norm": 7.010973058640957e-05, "learning_rate": 0.000144135593220339, "loss": 2.090843736368697e-06, "step": 1750 }, { "epoch": 0.02054982456754879, "grad_norm": 9.501949534751475e-05, "learning_rate": 0.00014379661016949153, "loss": 1.8209009795100429e-06, "step": 1760 }, { "epoch": 0.020666584934409864, "grad_norm": 6.99472293490544e-05, "learning_rate": 0.0001434576271186441, "loss": 1.8174021533923224e-06, "step": 1770 }, { "epoch": 0.020783345301270937, "grad_norm": 7.495882164221257e-05, "learning_rate": 0.00014311864406779662, "loss": 1.781230457709171e-06, "step": 1780 }, { "epoch": 0.02090010566813201, "grad_norm": 0.00020178320119157434, "learning_rate": 0.00014277966101694915, "loss": 1.7899070371640847e-06, "step": 1790 }, { "epoch": 0.021016866034993083, "grad_norm": 0.00014918761735316366, "learning_rate": 0.0001424406779661017, "loss": 1.7568227121955715e-06, "step": 1800 }, { "epoch": 0.021133626401854153, "grad_norm": 1.9064373191213235e-05, "learning_rate": 0.00014210169491525425, "loss": 1.7985246813623235e-06, "step": 1810 }, { "epoch": 0.021250386768715226, "grad_norm": 0.00012646097457036376, "learning_rate": 0.0001417627118644068, "loss": 1.8512891983846202e-06, "step": 1820 }, { "epoch": 0.0213671471355763, "grad_norm": 0.00017406328697688878, "learning_rate": 0.0001414237288135593, "loss": 1.8382943380856887e-06, "step": 1830 }, { "epoch": 0.021483907502437373, "grad_norm": 1.7785694581107236e-05, "learning_rate": 0.00014108474576271187, "loss": 1.9178693037247286e-06, "step": 1840 }, { "epoch": 0.021600667869298446, "grad_norm": 0.0002715683658607304, "learning_rate": 0.0001407457627118644, "loss": 1.962191890925169e-06, "step": 1850 }, { "epoch": 0.02171742823615952, "grad_norm": 0.00039442337583750486, "learning_rate": 0.00014040677966101696, "loss": 2.1291600205586294e-06, "step": 1860 }, { "epoch": 0.02183418860302059, "grad_norm": 0.00044150848407298326, "learning_rate": 0.0001400677966101695, "loss": 2.294996193086263e-06, "step": 1870 }, { "epoch": 0.021950948969881662, "grad_norm": 0.0009703988907858729, "learning_rate": 0.00013972881355932203, "loss": 3.0259834602475166e-06, "step": 1880 }, { "epoch": 0.022067709336742736, "grad_norm": 0.0004824126372113824, "learning_rate": 0.0001393898305084746, "loss": 3.0716255423612893e-06, "step": 1890 }, { "epoch": 0.02218446970360381, "grad_norm": 0.0009920555166900158, "learning_rate": 0.00013905084745762712, "loss": 3.824635132332332e-06, "step": 1900 }, { "epoch": 0.022301230070464882, "grad_norm": 0.00033314264146611094, "learning_rate": 0.00013871186440677966, "loss": 2.861424582079053e-06, "step": 1910 }, { "epoch": 0.022417990437325955, "grad_norm": 0.0003477221471257508, "learning_rate": 0.00013837288135593222, "loss": 2.431200300634373e-06, "step": 1920 }, { "epoch": 0.022534750804187025, "grad_norm": 0.00010379795276094228, "learning_rate": 0.00013803389830508475, "loss": 2.044691973424051e-06, "step": 1930 }, { "epoch": 0.0226515111710481, "grad_norm": 2.0585646780091338e-05, "learning_rate": 0.0001376949152542373, "loss": 1.7322965504718013e-06, "step": 1940 }, { "epoch": 0.02276827153790917, "grad_norm": 0.00014282431220635772, "learning_rate": 0.00013735593220338984, "loss": 1.7086696971091441e-06, "step": 1950 }, { "epoch": 0.022885031904770245, "grad_norm": 0.00010931365250144154, "learning_rate": 0.00013701694915254237, "loss": 1.687408257566858e-06, "step": 1960 }, { "epoch": 0.023001792271631318, "grad_norm": 6.643973028985783e-05, "learning_rate": 0.00013667796610169493, "loss": 1.6163905456778593e-06, "step": 1970 }, { "epoch": 0.02311855263849239, "grad_norm": 9.641289943829179e-05, "learning_rate": 0.00013633898305084747, "loss": 1.5819381587789393e-06, "step": 1980 }, { "epoch": 0.02323531300535346, "grad_norm": 6.208644481375813e-05, "learning_rate": 0.00013600000000000003, "loss": 1.5344317944254727e-06, "step": 1990 }, { "epoch": 0.023352073372214534, "grad_norm": 5.661721388605656e-06, "learning_rate": 0.00013566101694915253, "loss": 1.54183799168095e-06, "step": 2000 }, { "epoch": 0.023468833739075608, "grad_norm": 6.737840158166364e-05, "learning_rate": 0.00013535593220338985, "loss": 1.5509236618527212e-06, "step": 2010 }, { "epoch": 0.02358559410593668, "grad_norm": 8.187992352759466e-05, "learning_rate": 0.00013501694915254238, "loss": 1.5634514056728221e-06, "step": 2020 }, { "epoch": 0.023702354472797754, "grad_norm": 2.138152194675058e-05, "learning_rate": 0.0001346779661016949, "loss": 1.5208880540740211e-06, "step": 2030 }, { "epoch": 0.023819114839658827, "grad_norm": 5.823801075166557e-06, "learning_rate": 0.00013433898305084747, "loss": 1.5027135304990225e-06, "step": 2040 }, { "epoch": 0.023935875206519897, "grad_norm": 0.00018770445603877306, "learning_rate": 0.000134, "loss": 1.523056926089339e-06, "step": 2050 }, { "epoch": 0.02405263557338097, "grad_norm": 0.0001309715589741245, "learning_rate": 0.00013366101694915254, "loss": 1.5320663806051016e-06, "step": 2060 }, { "epoch": 0.024169395940242044, "grad_norm": 8.546516619389877e-05, "learning_rate": 0.0001333220338983051, "loss": 1.6208254237426444e-06, "step": 2070 }, { "epoch": 0.024286156307103117, "grad_norm": 0.00023168772167991847, "learning_rate": 0.00013298305084745763, "loss": 1.6410729585913942e-06, "step": 2080 }, { "epoch": 0.02440291667396419, "grad_norm": 0.0003386707976460457, "learning_rate": 0.0001326440677966102, "loss": 1.819970020733308e-06, "step": 2090 }, { "epoch": 0.024519677040825263, "grad_norm": 1.7884540284285322e-05, "learning_rate": 0.0001323050847457627, "loss": 1.6884436263353563e-06, "step": 2100 }, { "epoch": 0.024636437407686333, "grad_norm": 9.627863619243726e-05, "learning_rate": 0.00013196610169491526, "loss": 1.740954212436918e-06, "step": 2110 }, { "epoch": 0.024753197774547406, "grad_norm": 0.00016727342153899372, "learning_rate": 0.00013162711864406782, "loss": 1.822072226786986e-06, "step": 2120 }, { "epoch": 0.02486995814140848, "grad_norm": 0.0001092248858185485, "learning_rate": 0.00013128813559322035, "loss": 1.8791550246533007e-06, "step": 2130 }, { "epoch": 0.024986718508269553, "grad_norm": 0.00042687152745202184, "learning_rate": 0.00013094915254237288, "loss": 2.346935980312992e-06, "step": 2140 }, { "epoch": 0.025103478875130626, "grad_norm": 0.005975034087896347, "learning_rate": 0.0001306101694915254, "loss": 3.751758049475029e-06, "step": 2150 }, { "epoch": 0.0252202392419917, "grad_norm": 0.0002233933482784778, "learning_rate": 0.00013027118644067797, "loss": 2.2646194338449276e-06, "step": 2160 }, { "epoch": 0.025336999608852773, "grad_norm": 3.1968527764547616e-05, "learning_rate": 0.00012993220338983053, "loss": 1.8630344129633158e-06, "step": 2170 }, { "epoch": 0.025453759975713842, "grad_norm": 0.00013106416736263782, "learning_rate": 0.00012959322033898307, "loss": 1.6936999600147829e-06, "step": 2180 }, { "epoch": 0.025570520342574916, "grad_norm": 0.00012995705765206367, "learning_rate": 0.0001292542372881356, "loss": 1.5745910786790773e-06, "step": 2190 }, { "epoch": 0.02568728070943599, "grad_norm": 2.0045872588525526e-05, "learning_rate": 0.00012891525423728813, "loss": 1.5972793335095048e-06, "step": 2200 }, { "epoch": 0.025804041076297062, "grad_norm": 3.028578430530615e-05, "learning_rate": 0.0001285762711864407, "loss": 1.6016567315091379e-06, "step": 2210 }, { "epoch": 0.025920801443158135, "grad_norm": 0.0001439986372133717, "learning_rate": 0.00012823728813559322, "loss": 1.5990770407370292e-06, "step": 2220 }, { "epoch": 0.02603756181001921, "grad_norm": 2.008549199672416e-05, "learning_rate": 0.00012789830508474576, "loss": 1.6530904758838005e-06, "step": 2230 }, { "epoch": 0.02615432217688028, "grad_norm": 0.00010255297820549458, "learning_rate": 0.00012755932203389832, "loss": 1.629211510589812e-06, "step": 2240 }, { "epoch": 0.02627108254374135, "grad_norm": 0.00012852097279392183, "learning_rate": 0.00012722033898305085, "loss": 1.636833076190669e-06, "step": 2250 }, { "epoch": 0.026387842910602425, "grad_norm": 2.7599591703619808e-05, "learning_rate": 0.0001268813559322034, "loss": 1.6537753253942355e-06, "step": 2260 }, { "epoch": 0.026504603277463498, "grad_norm": 0.0001625096338102594, "learning_rate": 0.00012654237288135594, "loss": 1.6621592294541187e-06, "step": 2270 }, { "epoch": 0.02662136364432457, "grad_norm": 8.66148984641768e-05, "learning_rate": 0.00012620338983050847, "loss": 1.6177982615772635e-06, "step": 2280 }, { "epoch": 0.026738124011185645, "grad_norm": 9.50808243942447e-05, "learning_rate": 0.00012586440677966103, "loss": 1.6551437511225231e-06, "step": 2290 }, { "epoch": 0.026854884378046714, "grad_norm": 0.00011600083962548524, "learning_rate": 0.00012552542372881357, "loss": 1.6295243767672218e-06, "step": 2300 }, { "epoch": 0.026971644744907788, "grad_norm": 0.00030933969537727535, "learning_rate": 0.0001251864406779661, "loss": 1.7880485756904818e-06, "step": 2310 }, { "epoch": 0.02708840511176886, "grad_norm": 0.0001261148281628266, "learning_rate": 0.00012484745762711866, "loss": 1.8836488379747607e-06, "step": 2320 }, { "epoch": 0.027205165478629934, "grad_norm": 0.0007181658875197172, "learning_rate": 0.0001245084745762712, "loss": 2.163295903301332e-06, "step": 2330 }, { "epoch": 0.027321925845491007, "grad_norm": 0.00026710331439971924, "learning_rate": 0.00012416949152542375, "loss": 2.493400825187564e-06, "step": 2340 }, { "epoch": 0.02743868621235208, "grad_norm": 6.985353684285656e-05, "learning_rate": 0.00012383050847457626, "loss": 2.357956691412255e-06, "step": 2350 }, { "epoch": 0.02755544657921315, "grad_norm": 0.0004108547291252762, "learning_rate": 0.00012349152542372882, "loss": 2.3179190975497477e-06, "step": 2360 }, { "epoch": 0.027672206946074224, "grad_norm": 0.0004732685047201812, "learning_rate": 0.00012315254237288135, "loss": 2.3334543584496716e-06, "step": 2370 }, { "epoch": 0.027788967312935297, "grad_norm": 0.00011048079613829032, "learning_rate": 0.0001228135593220339, "loss": 2.450296233291738e-06, "step": 2380 }, { "epoch": 0.02790572767979637, "grad_norm": 0.00017211842350661755, "learning_rate": 0.00012247457627118644, "loss": 2.3725997380097397e-06, "step": 2390 }, { "epoch": 0.028022488046657443, "grad_norm": 0.00044060120126232505, "learning_rate": 0.00012213559322033898, "loss": 2.3969145331648178e-06, "step": 2400 }, { "epoch": 0.028139248413518517, "grad_norm": 5.874623457202688e-05, "learning_rate": 0.00012179661016949154, "loss": 1.7556896636961028e-06, "step": 2410 }, { "epoch": 0.028256008780379586, "grad_norm": 0.00011153460218338296, "learning_rate": 0.00012145762711864408, "loss": 1.7029842638294213e-06, "step": 2420 }, { "epoch": 0.02837276914724066, "grad_norm": 0.00011762795475078747, "learning_rate": 0.0001211186440677966, "loss": 1.7962009223992937e-06, "step": 2430 }, { "epoch": 0.028489529514101733, "grad_norm": 0.0004375619755592197, "learning_rate": 0.00012077966101694915, "loss": 2.118217526003718e-06, "step": 2440 }, { "epoch": 0.028606289880962806, "grad_norm": 0.0005378103232942522, "learning_rate": 0.0001204406779661017, "loss": 3.5935878258896992e-06, "step": 2450 }, { "epoch": 0.02872305024782388, "grad_norm": 0.0005620762822218239, "learning_rate": 0.00012010169491525425, "loss": 4.4984088162891565e-06, "step": 2460 }, { "epoch": 0.028839810614684953, "grad_norm": 0.0021721464581787586, "learning_rate": 0.0001197627118644068, "loss": 5.6281347497133535e-06, "step": 2470 }, { "epoch": 0.028956570981546022, "grad_norm": 0.0015677689807489514, "learning_rate": 0.00011942372881355932, "loss": 5.305397280608304e-06, "step": 2480 }, { "epoch": 0.029073331348407096, "grad_norm": 0.001409961492754519, "learning_rate": 0.00011908474576271187, "loss": 7.273031951626763e-06, "step": 2490 }, { "epoch": 0.02919009171526817, "grad_norm": 0.0038958522491157055, "learning_rate": 0.00011874576271186441, "loss": 1.9833706028293817e-05, "step": 2500 }, { "epoch": 0.029306852082129242, "grad_norm": 0.0010552983731031418, "learning_rate": 0.00011840677966101696, "loss": 8.511309715686366e-06, "step": 2510 }, { "epoch": 0.029423612448990315, "grad_norm": 0.0018230744171887636, "learning_rate": 0.00011806779661016949, "loss": 5.045130819780752e-06, "step": 2520 }, { "epoch": 0.02954037281585139, "grad_norm": 0.0023112548515200615, "learning_rate": 0.00011772881355932204, "loss": 2.0276603754609825e-05, "step": 2530 }, { "epoch": 0.02965713318271246, "grad_norm": 0.619185745716095, "learning_rate": 0.00011738983050847458, "loss": 7.167774019762873e-05, "step": 2540 }, { "epoch": 0.02977389354957353, "grad_norm": 2.633053779602051, "learning_rate": 0.00011708474576271187, "loss": 0.3269044399261475, "step": 2550 }, { "epoch": 0.029890653916434605, "grad_norm": 78.88597869873047, "learning_rate": 0.00011677966101694915, "loss": 0.1233400583267212, "step": 2560 }, { "epoch": 0.030007414283295678, "grad_norm": 70.32923889160156, "learning_rate": 0.00011644067796610169, "loss": 0.18804351091384888, "step": 2570 }, { "epoch": 0.03012417465015675, "grad_norm": 0.061514776200056076, "learning_rate": 0.00011610169491525424, "loss": 0.5769382953643799, "step": 2580 }, { "epoch": 0.030240935017017825, "grad_norm": 0.0027894896920770407, "learning_rate": 0.0001157627118644068, "loss": 0.0004052900243550539, "step": 2590 }, { "epoch": 0.030357695383878894, "grad_norm": 0.002143974881619215, "learning_rate": 0.00011542372881355932, "loss": 0.0006741383578628302, "step": 2600 }, { "epoch": 0.030474455750739968, "grad_norm": 0.00237264740280807, "learning_rate": 0.00011508474576271186, "loss": 2.506431774236262e-05, "step": 2610 }, { "epoch": 0.03059121611760104, "grad_norm": 0.00042826461140066385, "learning_rate": 0.00011474576271186441, "loss": 8.783912198850885e-06, "step": 2620 }, { "epoch": 0.030707976484462114, "grad_norm": 0.0003991922130808234, "learning_rate": 0.00011440677966101696, "loss": 5.77748978685122e-06, "step": 2630 }, { "epoch": 0.030824736851323187, "grad_norm": 0.00013551986194215715, "learning_rate": 0.00011406779661016949, "loss": 4.757464193971828e-06, "step": 2640 }, { "epoch": 0.03094149721818426, "grad_norm": 0.0001976866042241454, "learning_rate": 0.00011372881355932204, "loss": 3.982192720286548e-06, "step": 2650 }, { "epoch": 0.031058257585045334, "grad_norm": 0.00016013045387808233, "learning_rate": 0.00011338983050847458, "loss": 3.4830805816454814e-06, "step": 2660 }, { "epoch": 0.031175017951906404, "grad_norm": 0.00014570701750926673, "learning_rate": 0.00011305084745762713, "loss": 3.5921588278142735e-06, "step": 2670 }, { "epoch": 0.03129177831876748, "grad_norm": 0.00011156608525197953, "learning_rate": 0.00011271186440677967, "loss": 3.2941421522991733e-06, "step": 2680 }, { "epoch": 0.031408538685628554, "grad_norm": 0.00021060730796307325, "learning_rate": 0.00011237288135593221, "loss": 3.206400288036093e-06, "step": 2690 }, { "epoch": 0.03152529905248962, "grad_norm": 0.00014615162217523903, "learning_rate": 0.00011203389830508475, "loss": 3.0735405744053424e-06, "step": 2700 }, { "epoch": 0.03164205941935069, "grad_norm": 8.24721428216435e-06, "learning_rate": 0.0001116949152542373, "loss": 2.964516716019716e-06, "step": 2710 }, { "epoch": 0.03175881978621177, "grad_norm": 9.397340181749314e-05, "learning_rate": 0.00011135593220338985, "loss": 2.933578798547387e-06, "step": 2720 }, { "epoch": 0.03187558015307284, "grad_norm": 8.81102605490014e-05, "learning_rate": 0.00011101694915254236, "loss": 2.7795455025625416e-06, "step": 2730 }, { "epoch": 0.031992340519933916, "grad_norm": 0.00010666410526027903, "learning_rate": 0.00011067796610169492, "loss": 2.8226597351022066e-06, "step": 2740 }, { "epoch": 0.032109100886794986, "grad_norm": 7.704042945988476e-05, "learning_rate": 0.00011033898305084747, "loss": 2.9511858883779498e-06, "step": 2750 }, { "epoch": 0.032225861253656056, "grad_norm": 2.1477530026459135e-05, "learning_rate": 0.00011000000000000002, "loss": 2.6188161427853627e-06, "step": 2760 }, { "epoch": 0.03234262162051713, "grad_norm": 1.7250757082365453e-05, "learning_rate": 0.00010966101694915254, "loss": 2.622844112920575e-06, "step": 2770 }, { "epoch": 0.0324593819873782, "grad_norm": 0.00013947318075224757, "learning_rate": 0.00010932203389830508, "loss": 2.5740868295542895e-06, "step": 2780 }, { "epoch": 0.03257614235423928, "grad_norm": 6.799829134251922e-05, "learning_rate": 0.00010898305084745763, "loss": 2.5692570488899946e-06, "step": 2790 }, { "epoch": 0.03269290272110035, "grad_norm": 8.138755219988525e-06, "learning_rate": 0.00010864406779661019, "loss": 2.6186939066974445e-06, "step": 2800 }, { "epoch": 0.032809663087961426, "grad_norm": 4.546531727100955e-06, "learning_rate": 0.00010830508474576271, "loss": 2.671362381079234e-06, "step": 2810 }, { "epoch": 0.032926423454822495, "grad_norm": 2.267390300403349e-05, "learning_rate": 0.00010796610169491525, "loss": 2.7489344574860296e-06, "step": 2820 }, { "epoch": 0.033043183821683565, "grad_norm": 0.00013649604807142168, "learning_rate": 0.0001076271186440678, "loss": 2.5407276552869008e-06, "step": 2830 }, { "epoch": 0.03315994418854464, "grad_norm": 7.409998215734959e-05, "learning_rate": 0.00010728813559322035, "loss": 2.3465197955374604e-06, "step": 2840 }, { "epoch": 0.03327670455540571, "grad_norm": 1.3193653103371616e-05, "learning_rate": 0.00010694915254237288, "loss": 2.431076609354932e-06, "step": 2850 }, { "epoch": 0.03339346492226679, "grad_norm": 7.322274177568033e-05, "learning_rate": 0.00010661016949152543, "loss": 2.3664873879170044e-06, "step": 2860 }, { "epoch": 0.03351022528912786, "grad_norm": 1.4847065358480904e-05, "learning_rate": 0.00010627118644067797, "loss": 2.424820922897197e-06, "step": 2870 }, { "epoch": 0.03362698565598893, "grad_norm": 1.4212756468623411e-05, "learning_rate": 0.00010593220338983052, "loss": 2.2911957785254343e-06, "step": 2880 }, { "epoch": 0.033743746022850005, "grad_norm": 6.81228857501992e-06, "learning_rate": 0.00010559322033898305, "loss": 2.3263673938345164e-06, "step": 2890 }, { "epoch": 0.033860506389711074, "grad_norm": 2.190048235206632e-06, "learning_rate": 0.0001052542372881356, "loss": 2.2670785256195815e-06, "step": 2900 }, { "epoch": 0.03397726675657215, "grad_norm": 6.678269073745469e-06, "learning_rate": 0.00010491525423728814, "loss": 2.264988870592788e-06, "step": 2910 }, { "epoch": 0.03409402712343322, "grad_norm": 7.456969615304843e-05, "learning_rate": 0.00010457627118644069, "loss": 2.175485496991314e-06, "step": 2920 }, { "epoch": 0.0342107874902943, "grad_norm": 1.9885351321136113e-06, "learning_rate": 0.00010423728813559324, "loss": 2.287617462570779e-06, "step": 2930 }, { "epoch": 0.03432754785715537, "grad_norm": 2.060169890683028e-06, "learning_rate": 0.00010389830508474576, "loss": 2.315424535481725e-06, "step": 2940 }, { "epoch": 0.03444430822401644, "grad_norm": 7.341393211390823e-05, "learning_rate": 0.00010355932203389832, "loss": 2.377605596848298e-06, "step": 2950 }, { "epoch": 0.034561068590877514, "grad_norm": 2.2036572318029357e-06, "learning_rate": 0.00010322033898305086, "loss": 2.2663543859380297e-06, "step": 2960 }, { "epoch": 0.034677828957738584, "grad_norm": 9.463338938076049e-05, "learning_rate": 0.00010288135593220341, "loss": 2.26012161874678e-06, "step": 2970 }, { "epoch": 0.03479458932459966, "grad_norm": 1.5640855053788982e-05, "learning_rate": 0.00010254237288135593, "loss": 2.0918838345096447e-06, "step": 2980 }, { "epoch": 0.03491134969146073, "grad_norm": 9.752279765962157e-06, "learning_rate": 0.00010220338983050847, "loss": 2.331857831450179e-06, "step": 2990 }, { "epoch": 0.0350281100583218, "grad_norm": 0.00010294884123140946, "learning_rate": 0.00010186440677966102, "loss": 2.1787878722534516e-06, "step": 3000 }, { "epoch": 0.03514487042518288, "grad_norm": 0.00010099232167704031, "learning_rate": 0.00010152542372881357, "loss": 2.174664223275613e-06, "step": 3010 }, { "epoch": 0.035261630792043946, "grad_norm": 7.355651177931577e-05, "learning_rate": 0.0001011864406779661, "loss": 2.1843752620043233e-06, "step": 3020 }, { "epoch": 0.03537839115890502, "grad_norm": 1.63390432135202e-05, "learning_rate": 0.00010084745762711865, "loss": 2.182403295591939e-06, "step": 3030 }, { "epoch": 0.03549515152576609, "grad_norm": 7.151766476454213e-05, "learning_rate": 0.00010050847457627119, "loss": 2.1563122572842985e-06, "step": 3040 }, { "epoch": 0.03561191189262717, "grad_norm": 2.100063056786894e-06, "learning_rate": 0.00010016949152542374, "loss": 2.141538425348699e-06, "step": 3050 }, { "epoch": 0.03572867225948824, "grad_norm": 0.00016249011969193816, "learning_rate": 9.983050847457628e-05, "loss": 2.136301736754831e-06, "step": 3060 }, { "epoch": 0.03584543262634931, "grad_norm": 6.550679972860962e-05, "learning_rate": 9.949152542372882e-05, "loss": 2.0307767044869252e-06, "step": 3070 }, { "epoch": 0.035962192993210386, "grad_norm": 1.0432325325382408e-05, "learning_rate": 9.915254237288136e-05, "loss": 2.1234629457467237e-06, "step": 3080 }, { "epoch": 0.036078953360071456, "grad_norm": 0.00010071093856822699, "learning_rate": 9.88135593220339e-05, "loss": 2.0928424419253134e-06, "step": 3090 }, { "epoch": 0.03619571372693253, "grad_norm": 5.0666458264458925e-05, "learning_rate": 9.847457627118644e-05, "loss": 2.1181087504373865e-06, "step": 3100 }, { "epoch": 0.0363124740937936, "grad_norm": 0.00012666780094150454, "learning_rate": 9.813559322033899e-05, "loss": 1.997594335989561e-06, "step": 3110 }, { "epoch": 0.03642923446065467, "grad_norm": 5.238919402472675e-05, "learning_rate": 9.779661016949154e-05, "loss": 2.063276042463258e-06, "step": 3120 }, { "epoch": 0.03654599482751575, "grad_norm": 7.314996764762327e-05, "learning_rate": 9.745762711864407e-05, "loss": 1.9893661374226214e-06, "step": 3130 }, { "epoch": 0.03666275519437682, "grad_norm": 6.618754014198203e-06, "learning_rate": 9.711864406779661e-05, "loss": 2.0250694433343596e-06, "step": 3140 }, { "epoch": 0.036779515561237895, "grad_norm": 5.7765271776588634e-05, "learning_rate": 9.677966101694915e-05, "loss": 2.039803803199902e-06, "step": 3150 }, { "epoch": 0.036896275928098965, "grad_norm": 5.030070315115154e-05, "learning_rate": 9.64406779661017e-05, "loss": 2.049379145319108e-06, "step": 3160 }, { "epoch": 0.03701303629496004, "grad_norm": 0.00014426227426156402, "learning_rate": 9.610169491525424e-05, "loss": 2.065716398647055e-06, "step": 3170 }, { "epoch": 0.03712979666182111, "grad_norm": 0.0001042195362970233, "learning_rate": 9.576271186440679e-05, "loss": 2.0669653167715295e-06, "step": 3180 }, { "epoch": 0.03724655702868218, "grad_norm": 3.6196935980115086e-05, "learning_rate": 9.542372881355933e-05, "loss": 2.0056630091858095e-06, "step": 3190 }, { "epoch": 0.03736331739554326, "grad_norm": 2.1439184365590336e-06, "learning_rate": 9.508474576271187e-05, "loss": 1.960424742719624e-06, "step": 3200 }, { "epoch": 0.03748007776240433, "grad_norm": 2.0883240722469054e-05, "learning_rate": 9.474576271186441e-05, "loss": 1.931091901496984e-06, "step": 3210 }, { "epoch": 0.037596838129265404, "grad_norm": 0.0001155544669018127, "learning_rate": 9.440677966101696e-05, "loss": 1.9896011508535595e-06, "step": 3220 }, { "epoch": 0.037713598496126474, "grad_norm": 6.560485871887067e-06, "learning_rate": 9.40677966101695e-05, "loss": 1.947508644661866e-06, "step": 3230 }, { "epoch": 0.03783035886298755, "grad_norm": 9.319369382865261e-06, "learning_rate": 9.372881355932204e-05, "loss": 2.071146809612401e-06, "step": 3240 }, { "epoch": 0.03794711922984862, "grad_norm": 8.688485831953585e-05, "learning_rate": 9.338983050847458e-05, "loss": 1.9745541067095475e-06, "step": 3250 }, { "epoch": 0.03806387959670969, "grad_norm": 2.1601556454697857e-06, "learning_rate": 9.305084745762712e-05, "loss": 1.968435935850721e-06, "step": 3260 }, { "epoch": 0.03818063996357077, "grad_norm": 2.1765763449366204e-06, "learning_rate": 9.271186440677968e-05, "loss": 1.980983688554261e-06, "step": 3270 }, { "epoch": 0.03829740033043184, "grad_norm": 1.0872706297959667e-05, "learning_rate": 9.237288135593221e-05, "loss": 1.938048990268726e-06, "step": 3280 }, { "epoch": 0.038414160697292914, "grad_norm": 4.855420411331579e-05, "learning_rate": 9.203389830508475e-05, "loss": 2.0312449123593978e-06, "step": 3290 }, { "epoch": 0.03853092106415398, "grad_norm": 0.0, "learning_rate": 9.169491525423729e-05, "loss": 1.9035969671676867e-06, "step": 3300 }, { "epoch": 0.03864768143101505, "grad_norm": 5.165005859453231e-05, "learning_rate": 9.135593220338983e-05, "loss": 1.897206311696209e-06, "step": 3310 }, { "epoch": 0.03876444179787613, "grad_norm": 5.7258494052803144e-05, "learning_rate": 9.101694915254238e-05, "loss": 1.961050838872325e-06, "step": 3320 }, { "epoch": 0.0388812021647372, "grad_norm": 1.8185855878982693e-05, "learning_rate": 9.067796610169493e-05, "loss": 1.9104940292891114e-06, "step": 3330 }, { "epoch": 0.038997962531598276, "grad_norm": 5.590146429312881e-06, "learning_rate": 9.033898305084746e-05, "loss": 1.927184530359227e-06, "step": 3340 }, { "epoch": 0.039114722898459346, "grad_norm": 7.191507756942883e-05, "learning_rate": 9e-05, "loss": 1.9486598830553705e-06, "step": 3350 }, { "epoch": 0.03923148326532042, "grad_norm": 6.619741179747507e-05, "learning_rate": 8.966101694915254e-05, "loss": 2.0193221644149164e-06, "step": 3360 }, { "epoch": 0.03934824363218149, "grad_norm": 5.084141321276547e-06, "learning_rate": 8.932203389830508e-05, "loss": 1.9432471162872387e-06, "step": 3370 }, { "epoch": 0.03946500399904256, "grad_norm": 5.648166552418843e-05, "learning_rate": 8.898305084745763e-05, "loss": 1.8662518414203078e-06, "step": 3380 }, { "epoch": 0.03958176436590364, "grad_norm": 7.099672802723944e-05, "learning_rate": 8.864406779661018e-05, "loss": 1.926773620652966e-06, "step": 3390 }, { "epoch": 0.03969852473276471, "grad_norm": 5.490633157023694e-06, "learning_rate": 8.830508474576271e-05, "loss": 2.003668669203762e-06, "step": 3400 }, { "epoch": 0.039815285099625786, "grad_norm": 4.8940302804112434e-05, "learning_rate": 8.796610169491526e-05, "loss": 1.951141894096509e-06, "step": 3410 }, { "epoch": 0.039932045466486855, "grad_norm": 5.092696665087715e-05, "learning_rate": 8.76271186440678e-05, "loss": 1.8971470126416534e-06, "step": 3420 }, { "epoch": 0.040048805833347925, "grad_norm": 4.840248220716603e-05, "learning_rate": 8.728813559322035e-05, "loss": 1.7974445654544979e-06, "step": 3430 }, { "epoch": 0.040165566200209, "grad_norm": 4.424721737450454e-06, "learning_rate": 8.694915254237288e-05, "loss": 1.9609513401519505e-06, "step": 3440 }, { "epoch": 0.04028232656707007, "grad_norm": 2.1310127067408757e-06, "learning_rate": 8.661016949152543e-05, "loss": 2.0018365830765107e-06, "step": 3450 }, { "epoch": 0.04039908693393115, "grad_norm": 7.513844320783392e-05, "learning_rate": 8.627118644067797e-05, "loss": 1.8380125766270794e-06, "step": 3460 }, { "epoch": 0.04051584730079222, "grad_norm": 4.8595178668620065e-05, "learning_rate": 8.593220338983051e-05, "loss": 1.8887243641074746e-06, "step": 3470 }, { "epoch": 0.040632607667653295, "grad_norm": 5.385456097428687e-05, "learning_rate": 8.559322033898305e-05, "loss": 1.940351830853615e-06, "step": 3480 }, { "epoch": 0.040749368034514365, "grad_norm": 0.00014665677736047655, "learning_rate": 8.52542372881356e-05, "loss": 1.8199347323388792e-06, "step": 3490 }, { "epoch": 0.040866128401375434, "grad_norm": 9.344716818304732e-05, "learning_rate": 8.491525423728815e-05, "loss": 1.8320717572350986e-06, "step": 3500 }, { "epoch": 0.04098288876823651, "grad_norm": 2.1979137727612397e-06, "learning_rate": 8.457627118644068e-05, "loss": 1.8974973500007763e-06, "step": 3510 }, { "epoch": 0.04109964913509758, "grad_norm": 5.3389358072308823e-05, "learning_rate": 8.423728813559323e-05, "loss": 1.8122955225408077e-06, "step": 3520 }, { "epoch": 0.04121640950195866, "grad_norm": 7.409727550111711e-05, "learning_rate": 8.389830508474577e-05, "loss": 1.8840732082026078e-06, "step": 3530 }, { "epoch": 0.04133316986881973, "grad_norm": 5.480295294546522e-05, "learning_rate": 8.355932203389832e-05, "loss": 1.8401231500320137e-06, "step": 3540 }, { "epoch": 0.0414499302356808, "grad_norm": 0.0, "learning_rate": 8.322033898305085e-05, "loss": 1.764593253028579e-06, "step": 3550 }, { "epoch": 0.041566690602541874, "grad_norm": 2.2396238819055725e-06, "learning_rate": 8.28813559322034e-05, "loss": 1.8283782992511987e-06, "step": 3560 }, { "epoch": 0.041683450969402944, "grad_norm": 9.503438195679337e-05, "learning_rate": 8.254237288135593e-05, "loss": 1.913698542921338e-06, "step": 3570 }, { "epoch": 0.04180021133626402, "grad_norm": 4.277124389773235e-06, "learning_rate": 8.220338983050848e-05, "loss": 1.8454575183568522e-06, "step": 3580 }, { "epoch": 0.04191697170312509, "grad_norm": 0.00010249170736642554, "learning_rate": 8.186440677966102e-05, "loss": 1.791014256014023e-06, "step": 3590 }, { "epoch": 0.04203373206998617, "grad_norm": 2.1688040305889444e-06, "learning_rate": 8.152542372881357e-05, "loss": 1.8011172869591973e-06, "step": 3600 }, { "epoch": 0.04215049243684724, "grad_norm": 6.839696288807318e-05, "learning_rate": 8.11864406779661e-05, "loss": 1.7736805602908134e-06, "step": 3610 }, { "epoch": 0.042267252803708306, "grad_norm": 7.180737156886607e-06, "learning_rate": 8.084745762711865e-05, "loss": 1.7980677512241527e-06, "step": 3620 }, { "epoch": 0.04238401317056938, "grad_norm": 6.958799076528521e-06, "learning_rate": 8.050847457627118e-05, "loss": 1.7815744286053814e-06, "step": 3630 }, { "epoch": 0.04250077353743045, "grad_norm": 6.6404591052560136e-06, "learning_rate": 8.016949152542374e-05, "loss": 1.8455340978107415e-06, "step": 3640 }, { "epoch": 0.04261753390429153, "grad_norm": 6.840089918114245e-05, "learning_rate": 7.983050847457627e-05, "loss": 1.7297694284934551e-06, "step": 3650 }, { "epoch": 0.0427342942711526, "grad_norm": 0.0, "learning_rate": 7.949152542372882e-05, "loss": 1.774441079760436e-06, "step": 3660 }, { "epoch": 0.04285105463801367, "grad_norm": 4.816091313841753e-05, "learning_rate": 7.915254237288135e-05, "loss": 1.8213042494608089e-06, "step": 3670 }, { "epoch": 0.042967815004874746, "grad_norm": 6.61901940475218e-05, "learning_rate": 7.88135593220339e-05, "loss": 1.796543801901862e-06, "step": 3680 }, { "epoch": 0.043084575371735816, "grad_norm": 2.207036914114724e-06, "learning_rate": 7.847457627118644e-05, "loss": 1.727639028104022e-06, "step": 3690 }, { "epoch": 0.04320133573859689, "grad_norm": 4.7766596253495663e-05, "learning_rate": 7.813559322033899e-05, "loss": 1.7181819202960469e-06, "step": 3700 }, { "epoch": 0.04331809610545796, "grad_norm": 9.358324314234778e-05, "learning_rate": 7.779661016949152e-05, "loss": 1.8226328393211588e-06, "step": 3710 }, { "epoch": 0.04343485647231904, "grad_norm": 4.626426743925549e-05, "learning_rate": 7.745762711864407e-05, "loss": 1.7751455743564292e-06, "step": 3720 }, { "epoch": 0.04355161683918011, "grad_norm": 0.0, "learning_rate": 7.711864406779662e-05, "loss": 1.760644772730302e-06, "step": 3730 }, { "epoch": 0.04366837720604118, "grad_norm": 5.273880105960416e-06, "learning_rate": 7.677966101694916e-05, "loss": 1.7392259906046092e-06, "step": 3740 }, { "epoch": 0.043785137572902255, "grad_norm": 0.0, "learning_rate": 7.644067796610171e-05, "loss": 1.8046715922537259e-06, "step": 3750 }, { "epoch": 0.043901897939763325, "grad_norm": 5.032142871641554e-05, "learning_rate": 7.610169491525424e-05, "loss": 1.821439764171373e-06, "step": 3760 }, { "epoch": 0.0440186583066244, "grad_norm": 2.19600315176649e-06, "learning_rate": 7.576271186440679e-05, "loss": 1.7603120795683936e-06, "step": 3770 }, { "epoch": 0.04413541867348547, "grad_norm": 0.0, "learning_rate": 7.542372881355932e-05, "loss": 1.7147014659713022e-06, "step": 3780 }, { "epoch": 0.04425217904034655, "grad_norm": 9.782256529433653e-05, "learning_rate": 7.508474576271187e-05, "loss": 1.7871834643301554e-06, "step": 3790 }, { "epoch": 0.04436893940720762, "grad_norm": 9.953362314263359e-05, "learning_rate": 7.474576271186441e-05, "loss": 1.7436228517908604e-06, "step": 3800 }, { "epoch": 0.04448569977406869, "grad_norm": 7.75048101786524e-05, "learning_rate": 7.440677966101696e-05, "loss": 1.7289874449488707e-06, "step": 3810 }, { "epoch": 0.044602460140929764, "grad_norm": 0.00010488975385669619, "learning_rate": 7.406779661016949e-05, "loss": 1.7794043742469512e-06, "step": 3820 }, { "epoch": 0.044719220507790834, "grad_norm": 2.131126166204922e-06, "learning_rate": 7.372881355932204e-05, "loss": 1.784406958904583e-06, "step": 3830 }, { "epoch": 0.04483598087465191, "grad_norm": 6.641692016273737e-05, "learning_rate": 7.338983050847457e-05, "loss": 1.8203621948487124e-06, "step": 3840 }, { "epoch": 0.04495274124151298, "grad_norm": 9.44149142014794e-05, "learning_rate": 7.305084745762713e-05, "loss": 1.7521047993795947e-06, "step": 3850 }, { "epoch": 0.04506950160837405, "grad_norm": 0.0, "learning_rate": 7.271186440677966e-05, "loss": 1.65750188898528e-06, "step": 3860 }, { "epoch": 0.04518626197523513, "grad_norm": 0.0, "learning_rate": 7.237288135593221e-05, "loss": 1.7596286852494813e-06, "step": 3870 }, { "epoch": 0.0453030223420962, "grad_norm": 5.394325853558257e-05, "learning_rate": 7.203389830508474e-05, "loss": 1.6870504623511807e-06, "step": 3880 }, { "epoch": 0.045419782708957274, "grad_norm": 2.284836000399082e-06, "learning_rate": 7.169491525423729e-05, "loss": 1.689726377662737e-06, "step": 3890 }, { "epoch": 0.04553654307581834, "grad_norm": 4.559060471365228e-05, "learning_rate": 7.135593220338984e-05, "loss": 1.7255473721888847e-06, "step": 3900 }, { "epoch": 0.04565330344267942, "grad_norm": 4.5206386857898906e-05, "learning_rate": 7.101694915254238e-05, "loss": 1.740222614898812e-06, "step": 3910 }, { "epoch": 0.04577006380954049, "grad_norm": 2.1293201371008763e-06, "learning_rate": 7.067796610169491e-05, "loss": 1.7261534594581463e-06, "step": 3920 }, { "epoch": 0.04588682417640156, "grad_norm": 1.0025795745605137e-05, "learning_rate": 7.033898305084746e-05, "loss": 1.8548162188380957e-06, "step": 3930 }, { "epoch": 0.046003584543262636, "grad_norm": 8.444023842457682e-05, "learning_rate": 7e-05, "loss": 1.645698466745671e-06, "step": 3940 }, { "epoch": 0.046120344910123706, "grad_norm": 6.808809121139348e-05, "learning_rate": 6.966101694915254e-05, "loss": 1.6807753127068281e-06, "step": 3950 }, { "epoch": 0.04623710527698478, "grad_norm": 6.510812818305567e-05, "learning_rate": 6.932203389830509e-05, "loss": 1.7703574485494756e-06, "step": 3960 }, { "epoch": 0.04635386564384585, "grad_norm": 1.5896932382020168e-05, "learning_rate": 6.898305084745763e-05, "loss": 1.7425098121748306e-06, "step": 3970 }, { "epoch": 0.04647062601070692, "grad_norm": 5.356021938496269e-05, "learning_rate": 6.864406779661017e-05, "loss": 1.7232412574230694e-06, "step": 3980 }, { "epoch": 0.046587386377568, "grad_norm": 4.5757722546113655e-05, "learning_rate": 6.830508474576271e-05, "loss": 1.657305983826518e-06, "step": 3990 }, { "epoch": 0.04670414674442907, "grad_norm": 0.00017296714941039681, "learning_rate": 6.796610169491526e-05, "loss": 1.7679916709312237e-06, "step": 4000 }, { "epoch": 0.046820907111290146, "grad_norm": 4.453680230653845e-05, "learning_rate": 6.76271186440678e-05, "loss": 1.641594644752331e-06, "step": 4010 }, { "epoch": 0.046937667478151215, "grad_norm": 8.189085747289937e-06, "learning_rate": 6.728813559322035e-05, "loss": 1.6620160749880597e-06, "step": 4020 }, { "epoch": 0.04705442784501229, "grad_norm": 2.0963232145732036e-06, "learning_rate": 6.694915254237288e-05, "loss": 1.6667652744217776e-06, "step": 4030 }, { "epoch": 0.04717118821187336, "grad_norm": 1.5700910807936452e-05, "learning_rate": 6.661016949152543e-05, "loss": 1.7349459085380658e-06, "step": 4040 }, { "epoch": 0.04728794857873443, "grad_norm": 0.0001244883897015825, "learning_rate": 6.627118644067796e-05, "loss": 1.6490987036377191e-06, "step": 4050 }, { "epoch": 0.04740470894559551, "grad_norm": 0.0, "learning_rate": 6.593220338983052e-05, "loss": 1.6804051483632065e-06, "step": 4060 }, { "epoch": 0.04752146931245658, "grad_norm": 0.00012983976921532303, "learning_rate": 6.559322033898306e-05, "loss": 1.6431975382147357e-06, "step": 4070 }, { "epoch": 0.047638229679317655, "grad_norm": 0.0, "learning_rate": 6.52542372881356e-05, "loss": 1.6927364413277245e-06, "step": 4080 }, { "epoch": 0.047754990046178725, "grad_norm": 9.585384395904839e-05, "learning_rate": 6.491525423728813e-05, "loss": 1.7203876268467865e-06, "step": 4090 }, { "epoch": 0.047871750413039794, "grad_norm": 8.841085946187377e-05, "learning_rate": 6.457627118644068e-05, "loss": 1.7156202375190332e-06, "step": 4100 }, { "epoch": 0.04798851077990087, "grad_norm": 9.684242832008749e-05, "learning_rate": 6.423728813559323e-05, "loss": 1.7227321222890168e-06, "step": 4110 }, { "epoch": 0.04810527114676194, "grad_norm": 0.0, "learning_rate": 6.389830508474577e-05, "loss": 1.6056968888733535e-06, "step": 4120 }, { "epoch": 0.04822203151362302, "grad_norm": 4.84180100102094e-06, "learning_rate": 6.35593220338983e-05, "loss": 1.6854257410159334e-06, "step": 4130 }, { "epoch": 0.04833879188048409, "grad_norm": 0.0, "learning_rate": 6.322033898305085e-05, "loss": 1.6812249668873846e-06, "step": 4140 }, { "epoch": 0.048455552247345164, "grad_norm": 0.0, "learning_rate": 6.288135593220339e-05, "loss": 1.6822612451505848e-06, "step": 4150 }, { "epoch": 0.048572312614206234, "grad_norm": 5.203986347623868e-06, "learning_rate": 6.254237288135593e-05, "loss": 1.697816878731828e-06, "step": 4160 }, { "epoch": 0.048689072981067304, "grad_norm": 9.379423136124387e-05, "learning_rate": 6.220338983050848e-05, "loss": 1.6019837858038955e-06, "step": 4170 }, { "epoch": 0.04880583334792838, "grad_norm": 2.1230739548627753e-06, "learning_rate": 6.186440677966102e-05, "loss": 1.6596117347944527e-06, "step": 4180 }, { "epoch": 0.04892259371478945, "grad_norm": 0.0, "learning_rate": 6.152542372881356e-05, "loss": 1.618065834918525e-06, "step": 4190 }, { "epoch": 0.04903935408165053, "grad_norm": 9.350182517664507e-05, "learning_rate": 6.11864406779661e-05, "loss": 1.6505833627888934e-06, "step": 4200 }, { "epoch": 0.0491561144485116, "grad_norm": 4.2700281483121216e-05, "learning_rate": 6.084745762711864e-05, "loss": 1.6859339666552842e-06, "step": 4210 }, { "epoch": 0.049272874815372666, "grad_norm": 0.00019017797603737563, "learning_rate": 6.050847457627119e-05, "loss": 1.6512869478901848e-06, "step": 4220 }, { "epoch": 0.04938963518223374, "grad_norm": 6.502112228190526e-05, "learning_rate": 6.016949152542373e-05, "loss": 1.5450377759407274e-06, "step": 4230 }, { "epoch": 0.04950639554909481, "grad_norm": 4.248663390171714e-05, "learning_rate": 5.9830508474576275e-05, "loss": 1.6007126760086976e-06, "step": 4240 }, { "epoch": 0.04962315591595589, "grad_norm": 0.0001021979987854138, "learning_rate": 5.9491525423728814e-05, "loss": 1.5740963135613128e-06, "step": 4250 }, { "epoch": 0.04973991628281696, "grad_norm": 0.0, "learning_rate": 5.915254237288136e-05, "loss": 1.7099528122344054e-06, "step": 4260 }, { "epoch": 0.049856676649678036, "grad_norm": 4.2370869778096676e-05, "learning_rate": 5.881355932203391e-05, "loss": 1.6406565919169225e-06, "step": 4270 }, { "epoch": 0.049973437016539106, "grad_norm": 4.322805398260243e-05, "learning_rate": 5.8474576271186446e-05, "loss": 1.5791374607942998e-06, "step": 4280 }, { "epoch": 0.050090197383400176, "grad_norm": 2.1999978798703523e-06, "learning_rate": 5.813559322033899e-05, "loss": 1.626175617275294e-06, "step": 4290 }, { "epoch": 0.05020695775026125, "grad_norm": 4.189540049992502e-05, "learning_rate": 5.7796610169491525e-05, "loss": 1.5821848137420603e-06, "step": 4300 }, { "epoch": 0.05032371811712232, "grad_norm": 6.657272024312988e-05, "learning_rate": 5.745762711864407e-05, "loss": 1.658791006775573e-06, "step": 4310 }, { "epoch": 0.0504404784839834, "grad_norm": 6.365696754073724e-05, "learning_rate": 5.711864406779661e-05, "loss": 1.6204114217543974e-06, "step": 4320 }, { "epoch": 0.05055723885084447, "grad_norm": 0.00016623760166112334, "learning_rate": 5.677966101694916e-05, "loss": 1.714504833216779e-06, "step": 4330 }, { "epoch": 0.050673999217705545, "grad_norm": 4.169995736447163e-05, "learning_rate": 5.64406779661017e-05, "loss": 1.6089798009488732e-06, "step": 4340 }, { "epoch": 0.050790759584566615, "grad_norm": 2.1216033928794786e-05, "learning_rate": 5.6101694915254243e-05, "loss": 1.643040195631329e-06, "step": 4350 }, { "epoch": 0.050907519951427685, "grad_norm": 2.1762675714853685e-06, "learning_rate": 5.576271186440678e-05, "loss": 1.6593365216976963e-06, "step": 4360 }, { "epoch": 0.05102428031828876, "grad_norm": 0.000107581727206707, "learning_rate": 5.542372881355933e-05, "loss": 1.6118299754452892e-06, "step": 4370 }, { "epoch": 0.05114104068514983, "grad_norm": 5.259860336082056e-05, "learning_rate": 5.508474576271186e-05, "loss": 1.5759527741465717e-06, "step": 4380 }, { "epoch": 0.05125780105201091, "grad_norm": 0.00010669691255316138, "learning_rate": 5.4745762711864415e-05, "loss": 1.5893943782430143e-06, "step": 4390 }, { "epoch": 0.05137456141887198, "grad_norm": 0.000116826115117874, "learning_rate": 5.440677966101695e-05, "loss": 1.5961008102749473e-06, "step": 4400 }, { "epoch": 0.05149132178573305, "grad_norm": 0.00015807825548108667, "learning_rate": 5.4067796610169494e-05, "loss": 1.6570907973800787e-06, "step": 4410 }, { "epoch": 0.051608082152594124, "grad_norm": 0.0, "learning_rate": 5.3728813559322034e-05, "loss": 1.5699330106144772e-06, "step": 4420 }, { "epoch": 0.051724842519455194, "grad_norm": 4.22196535510011e-06, "learning_rate": 5.338983050847458e-05, "loss": 1.5922509192023425e-06, "step": 4430 }, { "epoch": 0.05184160288631627, "grad_norm": 4.248529876349494e-05, "learning_rate": 5.305084745762712e-05, "loss": 1.5383542631752788e-06, "step": 4440 }, { "epoch": 0.05195836325317734, "grad_norm": 3.999047112301923e-05, "learning_rate": 5.2711864406779666e-05, "loss": 1.5818350220797583e-06, "step": 4450 }, { "epoch": 0.05207512362003842, "grad_norm": 5.459059320855886e-05, "learning_rate": 5.23728813559322e-05, "loss": 1.5423203876707702e-06, "step": 4460 }, { "epoch": 0.05219188398689949, "grad_norm": 5.246176806394942e-05, "learning_rate": 5.203389830508475e-05, "loss": 1.603524106030818e-06, "step": 4470 }, { "epoch": 0.05230864435376056, "grad_norm": 5.46584851690568e-06, "learning_rate": 5.1694915254237284e-05, "loss": 1.5377483578049577e-06, "step": 4480 }, { "epoch": 0.052425404720621634, "grad_norm": 4.1143452108372e-05, "learning_rate": 5.135593220338983e-05, "loss": 1.541577876196243e-06, "step": 4490 }, { "epoch": 0.0525421650874827, "grad_norm": 0.00018789821479003876, "learning_rate": 5.101694915254237e-05, "loss": 1.5388426618301309e-06, "step": 4500 }, { "epoch": 0.05265892545434378, "grad_norm": 0.00010700323764467612, "learning_rate": 5.0677966101694917e-05, "loss": 1.5483585229958407e-06, "step": 4510 }, { "epoch": 0.05277568582120485, "grad_norm": 5.43225251021795e-05, "learning_rate": 5.0338983050847456e-05, "loss": 1.5612771676387637e-06, "step": 4520 }, { "epoch": 0.05289244618806592, "grad_norm": 6.245359691092744e-05, "learning_rate": 5e-05, "loss": 1.6059664631029591e-06, "step": 4530 }, { "epoch": 0.053009206554926996, "grad_norm": 8.126306784106418e-05, "learning_rate": 4.966101694915254e-05, "loss": 1.5144147255341523e-06, "step": 4540 }, { "epoch": 0.053125966921788066, "grad_norm": 0.0, "learning_rate": 4.932203389830509e-05, "loss": 1.506343778601149e-06, "step": 4550 }, { "epoch": 0.05324272728864914, "grad_norm": 8.041210094233975e-05, "learning_rate": 4.898305084745763e-05, "loss": 1.554262235003989e-06, "step": 4560 }, { "epoch": 0.05335948765551021, "grad_norm": 3.988751268479973e-05, "learning_rate": 4.8644067796610174e-05, "loss": 1.54417721205391e-06, "step": 4570 }, { "epoch": 0.05347624802237129, "grad_norm": 5.1112729124724865e-06, "learning_rate": 4.8305084745762714e-05, "loss": 1.5159579561441205e-06, "step": 4580 }, { "epoch": 0.05359300838923236, "grad_norm": 0.0, "learning_rate": 4.796610169491525e-05, "loss": 1.5396437447634525e-06, "step": 4590 }, { "epoch": 0.05370976875609343, "grad_norm": 0.00011359211202943698, "learning_rate": 4.76271186440678e-05, "loss": 1.5359306416939945e-06, "step": 4600 }, { "epoch": 0.053826529122954506, "grad_norm": 8.150914072757587e-05, "learning_rate": 4.728813559322034e-05, "loss": 1.5157635971263516e-06, "step": 4610 }, { "epoch": 0.053943289489815575, "grad_norm": 3.945328717236407e-05, "learning_rate": 4.694915254237288e-05, "loss": 1.473337397328578e-06, "step": 4620 }, { "epoch": 0.05406004985667665, "grad_norm": 5.413919552665902e-06, "learning_rate": 4.6610169491525425e-05, "loss": 1.524340950709302e-06, "step": 4630 }, { "epoch": 0.05417681022353772, "grad_norm": 0.00017984810983762145, "learning_rate": 4.6271186440677964e-05, "loss": 1.469545986765297e-06, "step": 4640 }, { "epoch": 0.05429357059039879, "grad_norm": 0.0, "learning_rate": 4.593220338983051e-05, "loss": 1.4799027667322662e-06, "step": 4650 }, { "epoch": 0.05441033095725987, "grad_norm": 6.315819337032735e-05, "learning_rate": 4.559322033898305e-05, "loss": 1.495204651291715e-06, "step": 4660 }, { "epoch": 0.05452709132412094, "grad_norm": 4.7498702770099044e-05, "learning_rate": 4.5254237288135596e-05, "loss": 1.5366334991995244e-06, "step": 4670 }, { "epoch": 0.054643851690982015, "grad_norm": 8.457310468656942e-05, "learning_rate": 4.491525423728814e-05, "loss": 1.534151851956267e-06, "step": 4680 }, { "epoch": 0.054760612057843085, "grad_norm": 0.00012268357386346906, "learning_rate": 4.457627118644068e-05, "loss": 1.4569609447789844e-06, "step": 4690 }, { "epoch": 0.05487737242470416, "grad_norm": 0.00010966585978167132, "learning_rate": 4.423728813559322e-05, "loss": 1.5641469872207381e-06, "step": 4700 }, { "epoch": 0.05499413279156523, "grad_norm": 3.980071414844133e-05, "learning_rate": 4.389830508474577e-05, "loss": 1.464406796003459e-06, "step": 4710 }, { "epoch": 0.0551108931584263, "grad_norm": 5.441743633127771e-05, "learning_rate": 4.355932203389831e-05, "loss": 1.446505484636873e-06, "step": 4720 }, { "epoch": 0.05522765352528738, "grad_norm": 2.770342234725831e-06, "learning_rate": 4.3220338983050854e-05, "loss": 1.4363635273184626e-06, "step": 4730 }, { "epoch": 0.05534441389214845, "grad_norm": 8.978364348877221e-06, "learning_rate": 4.2881355932203394e-05, "loss": 1.482755487813847e-06, "step": 4740 }, { "epoch": 0.055461174259009524, "grad_norm": 0.00010691554052755237, "learning_rate": 4.254237288135593e-05, "loss": 1.4542250937665812e-06, "step": 4750 }, { "epoch": 0.055577934625870594, "grad_norm": 1.002875706035411e-05, "learning_rate": 4.220338983050848e-05, "loss": 1.4257126167649404e-06, "step": 4760 }, { "epoch": 0.05569469499273167, "grad_norm": 5.3997257055016235e-05, "learning_rate": 4.186440677966102e-05, "loss": 1.4622179151047022e-06, "step": 4770 }, { "epoch": 0.05581145535959274, "grad_norm": 8.037842053454369e-05, "learning_rate": 4.152542372881356e-05, "loss": 1.4390406249731313e-06, "step": 4780 }, { "epoch": 0.05592821572645381, "grad_norm": 5.411546953837387e-05, "learning_rate": 4.1186440677966105e-05, "loss": 1.5547482689726167e-06, "step": 4790 }, { "epoch": 0.05604497609331489, "grad_norm": 5.027368388255127e-06, "learning_rate": 4.0847457627118644e-05, "loss": 1.4949664546293207e-06, "step": 4800 }, { "epoch": 0.05616173646017596, "grad_norm": 0.0, "learning_rate": 4.050847457627119e-05, "loss": 1.4146327885100618e-06, "step": 4810 }, { "epoch": 0.05627849682703703, "grad_norm": 0.0, "learning_rate": 4.016949152542373e-05, "loss": 1.4047062904865014e-06, "step": 4820 }, { "epoch": 0.0563952571938981, "grad_norm": 0.0, "learning_rate": 3.983050847457627e-05, "loss": 1.378285014652647e-06, "step": 4830 }, { "epoch": 0.05651201756075917, "grad_norm": 6.339018727885559e-05, "learning_rate": 3.9491525423728816e-05, "loss": 1.4362078218255192e-06, "step": 4840 }, { "epoch": 0.05662877792762025, "grad_norm": 0.0, "learning_rate": 3.9152542372881355e-05, "loss": 1.3823691006109583e-06, "step": 4850 }, { "epoch": 0.05674553829448132, "grad_norm": 0.00011063187412219122, "learning_rate": 3.88135593220339e-05, "loss": 1.4159812963043804e-06, "step": 4860 }, { "epoch": 0.056862298661342396, "grad_norm": 3.853217276628129e-05, "learning_rate": 3.847457627118644e-05, "loss": 1.443984638171969e-06, "step": 4870 }, { "epoch": 0.056979059028203466, "grad_norm": 0.00016084953676909208, "learning_rate": 3.813559322033898e-05, "loss": 1.4580738024960739e-06, "step": 4880 }, { "epoch": 0.05709581939506454, "grad_norm": 0.0, "learning_rate": 3.779661016949153e-05, "loss": 1.385260657116305e-06, "step": 4890 }, { "epoch": 0.05721257976192561, "grad_norm": 3.9170579839264974e-05, "learning_rate": 3.745762711864407e-05, "loss": 1.444784902560059e-06, "step": 4900 }, { "epoch": 0.05732934012878668, "grad_norm": 0.00013803069305140525, "learning_rate": 3.711864406779661e-05, "loss": 1.3933115951658693e-06, "step": 4910 }, { "epoch": 0.05744610049564776, "grad_norm": 2.30526734412706e-06, "learning_rate": 3.677966101694915e-05, "loss": 1.3226872397353873e-06, "step": 4920 }, { "epoch": 0.05756286086250883, "grad_norm": 8.447455911664292e-05, "learning_rate": 3.644067796610169e-05, "loss": 1.3623964150610845e-06, "step": 4930 }, { "epoch": 0.057679621229369905, "grad_norm": 0.0, "learning_rate": 3.610169491525424e-05, "loss": 1.4465447748079896e-06, "step": 4940 }, { "epoch": 0.057796381596230975, "grad_norm": 0.0002463430864736438, "learning_rate": 3.5762711864406785e-05, "loss": 1.4583848496840802e-06, "step": 4950 }, { "epoch": 0.057913141963092045, "grad_norm": 8.406710549024865e-05, "learning_rate": 3.5423728813559324e-05, "loss": 1.3667741768585985e-06, "step": 4960 }, { "epoch": 0.05802990232995312, "grad_norm": 4.017229002784006e-05, "learning_rate": 3.508474576271187e-05, "loss": 1.3913191651226954e-06, "step": 4970 }, { "epoch": 0.05814666269681419, "grad_norm": 3.8721009332221e-05, "learning_rate": 3.474576271186441e-05, "loss": 1.4052521692065057e-06, "step": 4980 }, { "epoch": 0.05826342306367527, "grad_norm": 5.4408383221016265e-06, "learning_rate": 3.440677966101695e-05, "loss": 1.399311895511346e-06, "step": 4990 }, { "epoch": 0.05838018343053634, "grad_norm": 8.536592940799892e-05, "learning_rate": 3.4067796610169496e-05, "loss": 1.473233260185225e-06, "step": 5000 }, { "epoch": 0.058496943797397415, "grad_norm": 0.0, "learning_rate": 3.3728813559322035e-05, "loss": 1.3745911019213962e-06, "step": 5010 }, { "epoch": 0.058613704164258484, "grad_norm": 4.98253484693123e-06, "learning_rate": 3.338983050847458e-05, "loss": 1.400658493366791e-06, "step": 5020 }, { "epoch": 0.058730464531119554, "grad_norm": 4.87361648993101e-06, "learning_rate": 3.305084745762712e-05, "loss": 1.4142950931272934e-06, "step": 5030 }, { "epoch": 0.05884722489798063, "grad_norm": 0.0001477235637139529, "learning_rate": 3.271186440677966e-05, "loss": 1.3812918041367083e-06, "step": 5040 }, { "epoch": 0.0589639852648417, "grad_norm": 5.3765750635648146e-05, "learning_rate": 3.237288135593221e-05, "loss": 1.3543650311476086e-06, "step": 5050 }, { "epoch": 0.05908074563170278, "grad_norm": 0.0, "learning_rate": 3.203389830508475e-05, "loss": 1.378987781208707e-06, "step": 5060 }, { "epoch": 0.05919750599856385, "grad_norm": 0.0, "learning_rate": 3.169491525423729e-05, "loss": 1.304357010667445e-06, "step": 5070 }, { "epoch": 0.05931426636542492, "grad_norm": 0.00012571342813316733, "learning_rate": 3.135593220338983e-05, "loss": 1.4625479707319755e-06, "step": 5080 }, { "epoch": 0.059431026732285994, "grad_norm": 0.0, "learning_rate": 3.101694915254237e-05, "loss": 1.4507661944662687e-06, "step": 5090 }, { "epoch": 0.05954778709914706, "grad_norm": 0.0001415027363691479, "learning_rate": 3.067796610169492e-05, "loss": 1.4688024748465977e-06, "step": 5100 }, { "epoch": 0.05966454746600814, "grad_norm": 1.5932922906358726e-05, "learning_rate": 3.0338983050847458e-05, "loss": 1.3704868251807056e-06, "step": 5110 }, { "epoch": 0.05978130783286921, "grad_norm": 0.00010421362094348297, "learning_rate": 3e-05, "loss": 1.3911039786762558e-06, "step": 5120 }, { "epoch": 0.05989806819973029, "grad_norm": 0.00010565033153397962, "learning_rate": 2.9661016949152544e-05, "loss": 1.392373724229401e-06, "step": 5130 }, { "epoch": 0.060014828566591356, "grad_norm": 0.0, "learning_rate": 2.9322033898305083e-05, "loss": 1.3754704923485406e-06, "step": 5140 }, { "epoch": 0.060131588933452426, "grad_norm": 8.762565994402394e-05, "learning_rate": 2.8983050847457626e-05, "loss": 1.326165693171788e-06, "step": 5150 }, { "epoch": 0.0602483493003135, "grad_norm": 3.855280374409631e-05, "learning_rate": 2.864406779661017e-05, "loss": 1.4523289792123252e-06, "step": 5160 }, { "epoch": 0.06036510966717457, "grad_norm": 0.00011001485836459324, "learning_rate": 2.8305084745762712e-05, "loss": 1.3749999197898432e-06, "step": 5170 }, { "epoch": 0.06048187003403565, "grad_norm": 8.036902727326378e-05, "learning_rate": 2.7966101694915255e-05, "loss": 1.4094521247898228e-06, "step": 5180 }, { "epoch": 0.06059863040089672, "grad_norm": 3.893943357979879e-05, "learning_rate": 2.7627118644067794e-05, "loss": 1.419243199052289e-06, "step": 5190 }, { "epoch": 0.06071539076775779, "grad_norm": 0.0, "learning_rate": 2.7288135593220337e-05, "loss": 1.3803946785628796e-06, "step": 5200 }, { "epoch": 0.060832151134618866, "grad_norm": 0.0, "learning_rate": 2.6949152542372884e-05, "loss": 1.3023631254327483e-06, "step": 5210 }, { "epoch": 0.060948911501479935, "grad_norm": 0.0, "learning_rate": 2.6610169491525427e-05, "loss": 1.3296630640979856e-06, "step": 5220 }, { "epoch": 0.06106567186834101, "grad_norm": 3.859197749989107e-05, "learning_rate": 2.627118644067797e-05, "loss": 1.3933507034380455e-06, "step": 5230 }, { "epoch": 0.06118243223520208, "grad_norm": 9.67649684753269e-05, "learning_rate": 2.5932203389830512e-05, "loss": 1.361144859401975e-06, "step": 5240 }, { "epoch": 0.06129919260206316, "grad_norm": 5.280922050587833e-05, "learning_rate": 2.5593220338983055e-05, "loss": 1.3954211681266316e-06, "step": 5250 }, { "epoch": 0.06141595296892423, "grad_norm": 1.5948218788253143e-05, "learning_rate": 2.5254237288135595e-05, "loss": 1.3136381312506273e-06, "step": 5260 }, { "epoch": 0.0615327133357853, "grad_norm": 0.00010590072633931413, "learning_rate": 2.4915254237288138e-05, "loss": 1.389070075674681e-06, "step": 5270 }, { "epoch": 0.061649473702646375, "grad_norm": 9.394597145728767e-05, "learning_rate": 2.457627118644068e-05, "loss": 1.3415250577963888e-06, "step": 5280 }, { "epoch": 0.061766234069507445, "grad_norm": 5.027367024013074e-06, "learning_rate": 2.4237288135593224e-05, "loss": 1.3301514627528377e-06, "step": 5290 }, { "epoch": 0.06188299443636852, "grad_norm": 0.0, "learning_rate": 2.3898305084745763e-05, "loss": 1.372791848552879e-06, "step": 5300 }, { "epoch": 0.06199975480322959, "grad_norm": 2.136297098331852e-06, "learning_rate": 2.3559322033898306e-05, "loss": 1.3066026440355927e-06, "step": 5310 }, { "epoch": 0.06211651517009067, "grad_norm": 8.22246220195666e-05, "learning_rate": 2.322033898305085e-05, "loss": 1.354950018139789e-06, "step": 5320 }, { "epoch": 0.06223327553695174, "grad_norm": 0.0, "learning_rate": 2.2881355932203392e-05, "loss": 1.3250906704342925e-06, "step": 5330 }, { "epoch": 0.06235003590381281, "grad_norm": 3.910596569767222e-05, "learning_rate": 2.2542372881355935e-05, "loss": 1.329311908193631e-06, "step": 5340 }, { "epoch": 0.062466796270673884, "grad_norm": 8.18129992694594e-05, "learning_rate": 2.2203389830508474e-05, "loss": 1.3537390259443783e-06, "step": 5350 }, { "epoch": 0.06258355663753495, "grad_norm": 4.9395052883483e-06, "learning_rate": 2.1864406779661017e-05, "loss": 1.3296433280629572e-06, "step": 5360 }, { "epoch": 0.06270031700439603, "grad_norm": 0.0, "learning_rate": 2.152542372881356e-05, "loss": 1.347232137050014e-06, "step": 5370 }, { "epoch": 0.06281707737125711, "grad_norm": 0.00010895914601860568, "learning_rate": 2.1186440677966103e-05, "loss": 1.3500055501936004e-06, "step": 5380 }, { "epoch": 0.06293383773811817, "grad_norm": 3.876193659380078e-05, "learning_rate": 2.0847457627118643e-05, "loss": 1.332556075794855e-06, "step": 5390 }, { "epoch": 0.06305059810497925, "grad_norm": 0.00017675173876341432, "learning_rate": 2.0508474576271186e-05, "loss": 1.340800736215897e-06, "step": 5400 }, { "epoch": 0.06316735847184032, "grad_norm": 8.157679258147255e-05, "learning_rate": 2.0169491525423732e-05, "loss": 1.3102765478834045e-06, "step": 5410 }, { "epoch": 0.06328411883870139, "grad_norm": 6.882925663376227e-05, "learning_rate": 1.9830508474576275e-05, "loss": 1.3631774891109671e-06, "step": 5420 }, { "epoch": 0.06340087920556246, "grad_norm": 4.993570655642543e-06, "learning_rate": 1.9491525423728814e-05, "loss": 1.3343914361030328e-06, "step": 5430 }, { "epoch": 0.06351763957242354, "grad_norm": 0.00017276369908358902, "learning_rate": 1.9152542372881357e-05, "loss": 1.2962072105437982e-06, "step": 5440 }, { "epoch": 0.0636343999392846, "grad_norm": 3.778476457227953e-05, "learning_rate": 1.88135593220339e-05, "loss": 1.3215922081144527e-06, "step": 5450 }, { "epoch": 0.06375116030614568, "grad_norm": 4.939194695907645e-05, "learning_rate": 1.8474576271186443e-05, "loss": 1.2814136425731703e-06, "step": 5460 }, { "epoch": 0.06386792067300676, "grad_norm": 0.00018234833260066807, "learning_rate": 1.8135593220338983e-05, "loss": 1.4039794223208446e-06, "step": 5470 }, { "epoch": 0.06398468103986783, "grad_norm": 5.087385943625122e-06, "learning_rate": 1.7796610169491526e-05, "loss": 1.3452759958454409e-06, "step": 5480 }, { "epoch": 0.0641014414067289, "grad_norm": 0.00018187820387538522, "learning_rate": 1.745762711864407e-05, "loss": 1.3168809346097986e-06, "step": 5490 }, { "epoch": 0.06421820177358997, "grad_norm": 8.086241723503917e-05, "learning_rate": 1.711864406779661e-05, "loss": 1.2804553080059122e-06, "step": 5500 }, { "epoch": 0.06433496214045105, "grad_norm": 1.03999000202748e-05, "learning_rate": 1.677966101694915e-05, "loss": 1.288898420170881e-06, "step": 5510 }, { "epoch": 0.06445172250731211, "grad_norm": 4.8271394916810095e-06, "learning_rate": 1.6440677966101694e-05, "loss": 1.3110979125485755e-06, "step": 5520 }, { "epoch": 0.06456848287417319, "grad_norm": 0.0, "learning_rate": 1.6101694915254237e-05, "loss": 1.2629267075681128e-06, "step": 5530 }, { "epoch": 0.06468524324103427, "grad_norm": 5.420911475084722e-05, "learning_rate": 1.5762711864406783e-05, "loss": 1.267206243937835e-06, "step": 5540 }, { "epoch": 0.06480200360789534, "grad_norm": 3.802979699685238e-05, "learning_rate": 1.5423728813559323e-05, "loss": 1.3049223525740672e-06, "step": 5550 }, { "epoch": 0.0649187639747564, "grad_norm": 8.084896398941055e-05, "learning_rate": 1.5084745762711865e-05, "loss": 1.3566611414717044e-06, "step": 5560 }, { "epoch": 0.06503552434161748, "grad_norm": 0.00024158322776202112, "learning_rate": 1.4745762711864408e-05, "loss": 1.2927865100209602e-06, "step": 5570 }, { "epoch": 0.06515228470847856, "grad_norm": 0.0, "learning_rate": 1.440677966101695e-05, "loss": 1.3083025805826764e-06, "step": 5580 }, { "epoch": 0.06526904507533962, "grad_norm": 0.0, "learning_rate": 1.4067796610169493e-05, "loss": 1.3237998246040661e-06, "step": 5590 }, { "epoch": 0.0653858054422007, "grad_norm": 5.126770247443346e-06, "learning_rate": 1.3728813559322034e-05, "loss": 1.2794596841558814e-06, "step": 5600 }, { "epoch": 0.06550256580906177, "grad_norm": 0.0, "learning_rate": 1.3389830508474577e-05, "loss": 1.3023432074987795e-06, "step": 5610 }, { "epoch": 0.06561932617592285, "grad_norm": 9.926976781571284e-05, "learning_rate": 1.305084745762712e-05, "loss": 1.3155726264812983e-06, "step": 5620 }, { "epoch": 0.06573608654278391, "grad_norm": 0.0, "learning_rate": 1.2711864406779661e-05, "loss": 1.2964418601768556e-06, "step": 5630 }, { "epoch": 0.06585284690964499, "grad_norm": 0.000179946786374785, "learning_rate": 1.2372881355932204e-05, "loss": 1.3129935723554808e-06, "step": 5640 }, { "epoch": 0.06596960727650607, "grad_norm": 5.199072620598599e-05, "learning_rate": 1.2033898305084747e-05, "loss": 1.214970416185679e-06, "step": 5650 }, { "epoch": 0.06608636764336713, "grad_norm": 0.0, "learning_rate": 1.169491525423729e-05, "loss": 1.273596444661962e-06, "step": 5660 }, { "epoch": 0.06620312801022821, "grad_norm": 0.0, "learning_rate": 1.135593220338983e-05, "loss": 1.2817466085834895e-06, "step": 5670 }, { "epoch": 0.06631988837708928, "grad_norm": 0.0, "learning_rate": 1.1016949152542374e-05, "loss": 1.3487366231856869e-06, "step": 5680 }, { "epoch": 0.06643664874395035, "grad_norm": 0.000116236180474516, "learning_rate": 1.0677966101694915e-05, "loss": 1.2826248166675213e-06, "step": 5690 }, { "epoch": 0.06655340911081142, "grad_norm": 3.861305958707817e-05, "learning_rate": 1.0338983050847458e-05, "loss": 1.3293899428390432e-06, "step": 5700 }, { "epoch": 0.0666701694776725, "grad_norm": 7.040800846880302e-05, "learning_rate": 1e-05, "loss": 1.3549687537306453e-06, "step": 5710 }, { "epoch": 0.06678692984453358, "grad_norm": 8.289927063742653e-05, "learning_rate": 9.661016949152544e-06, "loss": 1.2531360880529973e-06, "step": 5720 }, { "epoch": 0.06690369021139464, "grad_norm": 3.7534187868004665e-05, "learning_rate": 9.322033898305085e-06, "loss": 1.3300335922394878e-06, "step": 5730 }, { "epoch": 0.06702045057825572, "grad_norm": 3.810706766671501e-05, "learning_rate": 8.983050847457628e-06, "loss": 1.26019103845465e-06, "step": 5740 }, { "epoch": 0.0671372109451168, "grad_norm": 3.717451909324154e-05, "learning_rate": 8.644067796610169e-06, "loss": 1.2743178558594082e-06, "step": 5750 }, { "epoch": 0.06725397131197786, "grad_norm": 0.0, "learning_rate": 8.305084745762712e-06, "loss": 1.266815797862364e-06, "step": 5760 }, { "epoch": 0.06737073167883893, "grad_norm": 3.764754728763364e-05, "learning_rate": 7.966101694915253e-06, "loss": 1.2933539437653962e-06, "step": 5770 }, { "epoch": 0.06748749204570001, "grad_norm": 6.362178100971505e-05, "learning_rate": 7.627118644067798e-06, "loss": 1.2529993909993208e-06, "step": 5780 }, { "epoch": 0.06760425241256109, "grad_norm": 4.811870894627646e-06, "learning_rate": 7.28813559322034e-06, "loss": 1.2315996173128952e-06, "step": 5790 }, { "epoch": 0.06772101277942215, "grad_norm": 0.0, "learning_rate": 6.949152542372882e-06, "loss": 1.2378348401398398e-06, "step": 5800 }, { "epoch": 0.06783777314628323, "grad_norm": 3.929694867110811e-05, "learning_rate": 6.610169491525424e-06, "loss": 1.2902270100312308e-06, "step": 5810 }, { "epoch": 0.0679545335131443, "grad_norm": 0.0, "learning_rate": 6.271186440677966e-06, "loss": 1.3066031897324137e-06, "step": 5820 }, { "epoch": 0.06807129388000537, "grad_norm": 3.9525595639133826e-05, "learning_rate": 5.932203389830509e-06, "loss": 1.2427204637788237e-06, "step": 5830 }, { "epoch": 0.06818805424686644, "grad_norm": 6.335924990708008e-05, "learning_rate": 5.593220338983051e-06, "loss": 1.3962799130240455e-06, "step": 5840 }, { "epoch": 0.06830481461372752, "grad_norm": 0.00014715190627612174, "learning_rate": 5.254237288135593e-06, "loss": 1.3100194337312133e-06, "step": 5850 }, { "epoch": 0.0684215749805886, "grad_norm": 3.666737029561773e-05, "learning_rate": 4.915254237288136e-06, "loss": 1.2644506568904034e-06, "step": 5860 }, { "epoch": 0.06853833534744966, "grad_norm": 8.057400555117056e-05, "learning_rate": 4.576271186440678e-06, "loss": 1.1796381841122638e-06, "step": 5870 }, { "epoch": 0.06865509571431073, "grad_norm": 3.7795114621985704e-05, "learning_rate": 4.23728813559322e-06, "loss": 1.2283956493774895e-06, "step": 5880 }, { "epoch": 0.06877185608117181, "grad_norm": 0.0, "learning_rate": 3.898305084745763e-06, "loss": 1.1908361557289026e-06, "step": 5890 }, { "epoch": 0.06888861644803287, "grad_norm": 0.00015865970635786653, "learning_rate": 3.5593220338983053e-06, "loss": 1.2719737242150585e-06, "step": 5900 }, { "epoch": 0.06900537681489395, "grad_norm": 0.0, "learning_rate": 3.2203389830508473e-06, "loss": 1.3039256373303942e-06, "step": 5910 }, { "epoch": 0.06912213718175503, "grad_norm": 0.0, "learning_rate": 2.88135593220339e-06, "loss": 1.2635128769034055e-06, "step": 5920 }, { "epoch": 0.0692388975486161, "grad_norm": 3.7257887015584856e-05, "learning_rate": 2.5423728813559323e-06, "loss": 1.2475281437218654e-06, "step": 5930 }, { "epoch": 0.06935565791547717, "grad_norm": 6.361248961184174e-05, "learning_rate": 2.203389830508475e-06, "loss": 1.2045544281136245e-06, "step": 5940 }, { "epoch": 0.06947241828233824, "grad_norm": 6.462608143920079e-05, "learning_rate": 1.8644067796610171e-06, "loss": 1.2439124475349672e-06, "step": 5950 }, { "epoch": 0.06958917864919932, "grad_norm": 1.5946430721669458e-05, "learning_rate": 1.5254237288135594e-06, "loss": 1.260463977814652e-06, "step": 5960 }, { "epoch": 0.06970593901606038, "grad_norm": 3.700413799379021e-05, "learning_rate": 1.1864406779661017e-06, "loss": 1.2278292160772253e-06, "step": 5970 }, { "epoch": 0.06982269938292146, "grad_norm": 3.6791527236346155e-05, "learning_rate": 8.474576271186441e-07, "loss": 1.3399623639998026e-06, "step": 5980 }, { "epoch": 0.06993945974978254, "grad_norm": 0.0, "learning_rate": 5.084745762711865e-07, "loss": 1.3250516531115863e-06, "step": 5990 }, { "epoch": 0.0700562201166436, "grad_norm": 4.887090199190425e-06, "learning_rate": 1.694915254237288e-07, "loss": 1.2329284800216555e-06, "step": 6000 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7.442698862592e+16, "train_batch_size": 8, "trial_name": null, "trial_params": null }