{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.4, "eval_steps": 500, "global_step": 2000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9251985549926758, "learning_rate": 3.6e-06, "loss": 1.1283, "step": 10 }, { "grad_norm": 0.31233087182044983, "learning_rate": 7.6e-06, "loss": 1.1005, "step": 20 }, { "grad_norm": 0.17401772737503052, "learning_rate": 1.16e-05, "loss": 1.0621, "step": 30 }, { "grad_norm": 0.08489914983510971, "learning_rate": 1.56e-05, "loss": 1.0479, "step": 40 }, { "grad_norm": 0.06278981268405914, "learning_rate": 1.9600000000000002e-05, "loss": 1.043, "step": 50 }, { "grad_norm": 0.07312581688165665, "learning_rate": 2.36e-05, "loss": 1.0422, "step": 60 }, { "grad_norm": 0.07355771958827972, "learning_rate": 2.7600000000000003e-05, "loss": 1.0359, "step": 70 }, { "grad_norm": 0.09430158138275146, "learning_rate": 3.16e-05, "loss": 1.0311, "step": 80 }, { "grad_norm": 0.1752423793077469, "learning_rate": 3.56e-05, "loss": 1.0166, "step": 90 }, { "grad_norm": 0.34610697627067566, "learning_rate": 3.960000000000001e-05, "loss": 0.985, "step": 100 }, { "grad_norm": 0.9313993453979492, "learning_rate": 4.36e-05, "loss": 0.9318, "step": 110 }, { "grad_norm": 0.6117237210273743, "learning_rate": 4.76e-05, "loss": 0.8528, "step": 120 }, { "grad_norm": 0.7044277191162109, "learning_rate": 5.16e-05, "loss": 0.7745, "step": 130 }, { "grad_norm": 0.7221372127532959, "learning_rate": 5.560000000000001e-05, "loss": 0.6958, "step": 140 }, { "grad_norm": 0.8104686737060547, "learning_rate": 5.96e-05, "loss": 0.6216, "step": 150 }, { "grad_norm": 0.7707035541534424, "learning_rate": 6.36e-05, "loss": 0.5569, "step": 160 }, { "grad_norm": 0.7717885971069336, "learning_rate": 6.76e-05, "loss": 0.5008, "step": 170 }, { "grad_norm": 0.8074904680252075, "learning_rate": 7.16e-05, "loss": 0.4472, "step": 180 }, { "grad_norm": 0.8878680467605591, "learning_rate": 7.560000000000001e-05, "loss": 0.4035, "step": 190 }, { "grad_norm": 0.7049605846405029, "learning_rate": 7.960000000000001e-05, "loss": 0.3654, "step": 200 }, { "grad_norm": 0.9025763869285583, "learning_rate": 8.36e-05, "loss": 0.3315, "step": 210 }, { "grad_norm": 0.976150631904602, "learning_rate": 8.76e-05, "loss": 0.2955, "step": 220 }, { "grad_norm": 0.9648725390434265, "learning_rate": 9.16e-05, "loss": 0.2746, "step": 230 }, { "grad_norm": 0.9989051222801208, "learning_rate": 9.56e-05, "loss": 0.255, "step": 240 }, { "grad_norm": 1.0400114059448242, "learning_rate": 9.960000000000001e-05, "loss": 0.2217, "step": 250 }, { "grad_norm": 1.0539939403533936, "learning_rate": 9.999911419878559e-05, "loss": 0.1905, "step": 260 }, { "grad_norm": 1.063599944114685, "learning_rate": 9.999605221019081e-05, "loss": 0.1741, "step": 270 }, { "grad_norm": 0.9228838086128235, "learning_rate": 9.999080323230761e-05, "loss": 0.1558, "step": 280 }, { "grad_norm": 0.7603300213813782, "learning_rate": 9.998336749474329e-05, "loss": 0.1494, "step": 290 }, { "grad_norm": 0.8506730794906616, "learning_rate": 9.997374532276107e-05, "loss": 0.1387, "step": 300 }, { "grad_norm": 0.8497244715690613, "learning_rate": 9.996193713726596e-05, "loss": 0.1312, "step": 310 }, { "grad_norm": 0.864569365978241, "learning_rate": 9.994794345478624e-05, "loss": 0.1272, "step": 320 }, { "grad_norm": 0.7728807330131531, "learning_rate": 9.99317648874509e-05, "loss": 0.1269, "step": 330 }, { "grad_norm": 0.8701847195625305, "learning_rate": 9.991340214296292e-05, "loss": 0.1205, "step": 340 }, { "grad_norm": 1.436145305633545, "learning_rate": 9.989285602456819e-05, "loss": 0.1121, "step": 350 }, { "grad_norm": 0.7210439443588257, "learning_rate": 9.98701274310205e-05, "loss": 0.11, "step": 360 }, { "grad_norm": 0.8955262303352356, "learning_rate": 9.984521735654218e-05, "loss": 0.0957, "step": 370 }, { "grad_norm": 0.9444258213043213, "learning_rate": 9.981812689078057e-05, "loss": 0.1009, "step": 380 }, { "grad_norm": 0.8208094835281372, "learning_rate": 9.978885721876041e-05, "loss": 0.0949, "step": 390 }, { "grad_norm": 0.7844081521034241, "learning_rate": 9.975740962083198e-05, "loss": 0.0944, "step": 400 }, { "grad_norm": 0.7617616057395935, "learning_rate": 9.972378547261504e-05, "loss": 0.091, "step": 410 }, { "grad_norm": 1.0140830278396606, "learning_rate": 9.968798624493885e-05, "loss": 0.0848, "step": 420 }, { "grad_norm": 0.8056594133377075, "learning_rate": 9.965001350377753e-05, "loss": 0.0831, "step": 430 }, { "grad_norm": 0.8687112927436829, "learning_rate": 9.960986891018183e-05, "loss": 0.0717, "step": 440 }, { "grad_norm": 0.7181530594825745, "learning_rate": 9.95675542202063e-05, "loss": 0.0646, "step": 450 }, { "grad_norm": 0.7142108082771301, "learning_rate": 9.952307128483256e-05, "loss": 0.0598, "step": 460 }, { "grad_norm": 0.757612943649292, "learning_rate": 9.947642204988835e-05, "loss": 0.0599, "step": 470 }, { "grad_norm": 0.7987963557243347, "learning_rate": 9.942760855596226e-05, "loss": 0.0639, "step": 480 }, { "grad_norm": 0.8699095845222473, "learning_rate": 9.937663293831471e-05, "loss": 0.0633, "step": 490 }, { "grad_norm": 0.6249838471412659, "learning_rate": 9.932349742678433e-05, "loss": 0.0636, "step": 500 }, { "grad_norm": 0.756233274936676, "learning_rate": 9.926820434569051e-05, "loss": 0.0603, "step": 510 }, { "grad_norm": 0.7866028547286987, "learning_rate": 9.921075611373179e-05, "loss": 0.0634, "step": 520 }, { "grad_norm": 0.6185405850410461, "learning_rate": 9.915115524387988e-05, "loss": 0.0578, "step": 530 }, { "grad_norm": 0.7116502523422241, "learning_rate": 9.908940434326997e-05, "loss": 0.0541, "step": 540 }, { "grad_norm": 0.7327669262886047, "learning_rate": 9.902550611308645e-05, "loss": 0.0563, "step": 550 }, { "grad_norm": 0.7472072243690491, "learning_rate": 9.895946334844494e-05, "loss": 0.0546, "step": 560 }, { "grad_norm": 0.8047758340835571, "learning_rate": 9.889127893826989e-05, "loss": 0.0558, "step": 570 }, { "grad_norm": 0.6287261247634888, "learning_rate": 9.882095586516831e-05, "loss": 0.0548, "step": 580 }, { "grad_norm": 0.6708779335021973, "learning_rate": 9.874849720529921e-05, "loss": 0.0561, "step": 590 }, { "grad_norm": 0.8972636461257935, "learning_rate": 9.867390612823914e-05, "loss": 0.0558, "step": 600 }, { "grad_norm": 0.6193103194236755, "learning_rate": 9.859718589684344e-05, "loss": 0.0497, "step": 610 }, { "grad_norm": 0.6898216605186462, "learning_rate": 9.851833986710353e-05, "loss": 0.0525, "step": 620 }, { "grad_norm": 0.7660088539123535, "learning_rate": 9.843737148800023e-05, "loss": 0.0503, "step": 630 }, { "grad_norm": 0.7528529167175293, "learning_rate": 9.835428430135271e-05, "loss": 0.0542, "step": 640 }, { "grad_norm": 0.6518449783325195, "learning_rate": 9.82690819416637e-05, "loss": 0.0525, "step": 650 }, { "grad_norm": 0.7395707964897156, "learning_rate": 9.818176813596041e-05, "loss": 0.0528, "step": 660 }, { "grad_norm": 0.6715362668037415, "learning_rate": 9.809234670363159e-05, "loss": 0.0459, "step": 670 }, { "grad_norm": 0.7790312170982361, "learning_rate": 9.800082155626034e-05, "loss": 0.0526, "step": 680 }, { "grad_norm": 0.7286430597305298, "learning_rate": 9.790719669745312e-05, "loss": 0.0459, "step": 690 }, { "grad_norm": 0.7442739009857178, "learning_rate": 9.781147622266455e-05, "loss": 0.0444, "step": 700 }, { "grad_norm": 0.7166370153427124, "learning_rate": 9.771366431901831e-05, "loss": 0.0444, "step": 710 }, { "grad_norm": 0.7265373468399048, "learning_rate": 9.761376526512394e-05, "loss": 0.0492, "step": 720 }, { "grad_norm": 0.7237659096717834, "learning_rate": 9.751178343088963e-05, "loss": 0.046, "step": 730 }, { "grad_norm": 0.7301507592201233, "learning_rate": 9.740772327733123e-05, "loss": 0.0438, "step": 740 }, { "grad_norm": 0.6654811501502991, "learning_rate": 9.730158935637697e-05, "loss": 0.0457, "step": 750 }, { "grad_norm": 0.7930723428726196, "learning_rate": 9.719338631066834e-05, "loss": 0.0464, "step": 760 }, { "grad_norm": 0.6005651950836182, "learning_rate": 9.708311887335713e-05, "loss": 0.0453, "step": 770 }, { "grad_norm": 0.6690271496772766, "learning_rate": 9.697079186789823e-05, "loss": 0.0444, "step": 780 }, { "grad_norm": 0.6243356466293335, "learning_rate": 9.685641020783876e-05, "loss": 0.045, "step": 790 }, { "grad_norm": 0.7267593741416931, "learning_rate": 9.67399788966031e-05, "loss": 0.0454, "step": 800 }, { "grad_norm": 0.626296877861023, "learning_rate": 9.662150302727395e-05, "loss": 0.0417, "step": 810 }, { "grad_norm": 0.6896906495094299, "learning_rate": 9.650098778236968e-05, "loss": 0.0441, "step": 820 }, { "grad_norm": 0.6030279994010925, "learning_rate": 9.637843843361749e-05, "loss": 0.0399, "step": 830 }, { "grad_norm": 0.7116725444793701, "learning_rate": 9.62538603417229e-05, "loss": 0.0431, "step": 840 }, { "grad_norm": 0.6708303689956665, "learning_rate": 9.612725895613526e-05, "loss": 0.0435, "step": 850 }, { "grad_norm": 0.5857799053192139, "learning_rate": 9.599863981480926e-05, "loss": 0.0422, "step": 860 }, { "grad_norm": 0.7056573033332825, "learning_rate": 9.586800854396283e-05, "loss": 0.0405, "step": 870 }, { "grad_norm": 0.5773367285728455, "learning_rate": 9.573537085783095e-05, "loss": 0.0395, "step": 880 }, { "grad_norm": 0.6533933877944946, "learning_rate": 9.560073255841571e-05, "loss": 0.0453, "step": 890 }, { "grad_norm": 0.6993983387947083, "learning_rate": 9.546409953523247e-05, "loss": 0.0435, "step": 900 }, { "grad_norm": 0.6972476840019226, "learning_rate": 9.532547776505229e-05, "loss": 0.0408, "step": 910 }, { "grad_norm": 0.6296972632408142, "learning_rate": 9.518487331164048e-05, "loss": 0.0388, "step": 920 }, { "grad_norm": 0.5836044549942017, "learning_rate": 9.504229232549134e-05, "loss": 0.0391, "step": 930 }, { "grad_norm": 0.7759990096092224, "learning_rate": 9.489774104355909e-05, "loss": 0.0398, "step": 940 }, { "grad_norm": 0.6509332656860352, "learning_rate": 9.475122578898507e-05, "loss": 0.0391, "step": 950 }, { "grad_norm": 0.5787583589553833, "learning_rate": 9.460275297082119e-05, "loss": 0.0376, "step": 960 }, { "grad_norm": 0.6208513379096985, "learning_rate": 9.445232908374948e-05, "loss": 0.0411, "step": 970 }, { "grad_norm": 0.5396055579185486, "learning_rate": 9.429996070779808e-05, "loss": 0.0372, "step": 980 }, { "grad_norm": 0.6517285704612732, "learning_rate": 9.414565450805333e-05, "loss": 0.0398, "step": 990 }, { "grad_norm": 0.5843371152877808, "learning_rate": 9.398941723436831e-05, "loss": 0.0378, "step": 1000 }, { "grad_norm": 0.7116906642913818, "learning_rate": 9.383125572106752e-05, "loss": 0.0389, "step": 1010 }, { "grad_norm": 0.607302725315094, "learning_rate": 9.367117688664791e-05, "loss": 0.0377, "step": 1020 }, { "grad_norm": 0.8494533896446228, "learning_rate": 9.35091877334763e-05, "loss": 0.0396, "step": 1030 }, { "grad_norm": 0.6156577467918396, "learning_rate": 9.334529534748297e-05, "loss": 0.0392, "step": 1040 }, { "grad_norm": 0.7002044916152954, "learning_rate": 9.317950689785188e-05, "loss": 0.0399, "step": 1050 }, { "grad_norm": 0.6531574726104736, "learning_rate": 9.301182963670688e-05, "loss": 0.0379, "step": 1060 }, { "grad_norm": 0.7230827808380127, "learning_rate": 9.284227089879456e-05, "loss": 0.0408, "step": 1070 }, { "grad_norm": 0.5891832113265991, "learning_rate": 9.26708381011634e-05, "loss": 0.0351, "step": 1080 }, { "grad_norm": 0.604454755783081, "learning_rate": 9.249753874283937e-05, "loss": 0.0381, "step": 1090 }, { "grad_norm": 0.6771432757377625, "learning_rate": 9.232238040449779e-05, "loss": 0.0353, "step": 1100 }, { "grad_norm": 0.5565150380134583, "learning_rate": 9.214537074813181e-05, "loss": 0.0355, "step": 1110 }, { "grad_norm": 0.5481780171394348, "learning_rate": 9.196651751671724e-05, "loss": 0.0361, "step": 1120 }, { "grad_norm": 0.48391294479370117, "learning_rate": 9.178582853387384e-05, "loss": 0.0412, "step": 1130 }, { "grad_norm": 0.5970834493637085, "learning_rate": 9.160331170352304e-05, "loss": 0.0383, "step": 1140 }, { "grad_norm": 0.6498259902000427, "learning_rate": 9.141897500954229e-05, "loss": 0.0357, "step": 1150 }, { "grad_norm": 0.5829015374183655, "learning_rate": 9.123282651541576e-05, "loss": 0.0368, "step": 1160 }, { "grad_norm": 0.6121956706047058, "learning_rate": 9.104487436388161e-05, "loss": 0.0381, "step": 1170 }, { "grad_norm": 0.5612373948097229, "learning_rate": 9.085512677657582e-05, "loss": 0.037, "step": 1180 }, { "grad_norm": 0.5658372044563293, "learning_rate": 9.066359205367258e-05, "loss": 0.0368, "step": 1190 }, { "grad_norm": 0.46027734875679016, "learning_rate": 9.047027857352112e-05, "loss": 0.0361, "step": 1200 }, { "grad_norm": 0.6454007625579834, "learning_rate": 9.027519479227935e-05, "loss": 0.0346, "step": 1210 }, { "grad_norm": 0.5289171934127808, "learning_rate": 9.007834924354383e-05, "loss": 0.0353, "step": 1220 }, { "grad_norm": 0.5146805644035339, "learning_rate": 8.987975053797655e-05, "loss": 0.0364, "step": 1230 }, { "grad_norm": 0.661555290222168, "learning_rate": 8.967940736292825e-05, "loss": 0.0361, "step": 1240 }, { "grad_norm": 0.4500121772289276, "learning_rate": 8.947732848205846e-05, "loss": 0.0362, "step": 1250 }, { "grad_norm": 0.6776179671287537, "learning_rate": 8.927352273495204e-05, "loss": 0.0367, "step": 1260 }, { "grad_norm": 0.6280854344367981, "learning_rate": 8.906799903673265e-05, "loss": 0.0338, "step": 1270 }, { "grad_norm": 0.3347318768501282, "learning_rate": 8.88607663776726e-05, "loss": 0.0361, "step": 1280 }, { "grad_norm": 0.4536568522453308, "learning_rate": 8.865183382279978e-05, "loss": 0.0347, "step": 1290 }, { "grad_norm": 0.537060558795929, "learning_rate": 8.844121051150096e-05, "loss": 0.0334, "step": 1300 }, { "grad_norm": 0.7525205612182617, "learning_rate": 8.822890565712211e-05, "loss": 0.0339, "step": 1310 }, { "grad_norm": 0.6858833432197571, "learning_rate": 8.801492854656536e-05, "loss": 0.0363, "step": 1320 }, { "grad_norm": 0.5998900532722473, "learning_rate": 8.779928853988268e-05, "loss": 0.0343, "step": 1330 }, { "grad_norm": 0.605100154876709, "learning_rate": 8.758199506986655e-05, "loss": 0.0338, "step": 1340 }, { "grad_norm": 0.5373857021331787, "learning_rate": 8.73630576416373e-05, "loss": 0.0341, "step": 1350 }, { "grad_norm": 0.5052387118339539, "learning_rate": 8.714248583222726e-05, "loss": 0.0347, "step": 1360 }, { "grad_norm": 0.36015674471855164, "learning_rate": 8.692028929016196e-05, "loss": 0.034, "step": 1370 }, { "grad_norm": 0.5457674264907837, "learning_rate": 8.669647773503797e-05, "loss": 0.034, "step": 1380 }, { "grad_norm": 0.6281781792640686, "learning_rate": 8.647106095709773e-05, "loss": 0.0327, "step": 1390 }, { "grad_norm": 0.6226788759231567, "learning_rate": 8.624404881680139e-05, "loss": 0.0336, "step": 1400 }, { "grad_norm": 0.4499896764755249, "learning_rate": 8.601545124439535e-05, "loss": 0.0298, "step": 1410 }, { "grad_norm": 0.5874620676040649, "learning_rate": 8.5785278239478e-05, "loss": 0.0346, "step": 1420 }, { "grad_norm": 0.5097549557685852, "learning_rate": 8.555353987056224e-05, "loss": 0.0321, "step": 1430 }, { "grad_norm": 0.5932273864746094, "learning_rate": 8.532024627463505e-05, "loss": 0.0314, "step": 1440 }, { "grad_norm": 0.47906121611595154, "learning_rate": 8.508540765671407e-05, "loss": 0.0307, "step": 1450 }, { "grad_norm": 0.4678618907928467, "learning_rate": 8.484903428940121e-05, "loss": 0.0309, "step": 1460 }, { "grad_norm": 0.5211289525032043, "learning_rate": 8.461113651243334e-05, "loss": 0.0354, "step": 1470 }, { "grad_norm": 0.5473387837409973, "learning_rate": 8.437172473222987e-05, "loss": 0.0361, "step": 1480 }, { "grad_norm": 0.6478421092033386, "learning_rate": 8.413080942143767e-05, "loss": 0.0312, "step": 1490 }, { "grad_norm": 0.5109831690788269, "learning_rate": 8.388840111847288e-05, "loss": 0.032, "step": 1500 }, { "grad_norm": 0.4686678647994995, "learning_rate": 8.364451042705998e-05, "loss": 0.0334, "step": 1510 }, { "grad_norm": 0.5508041977882385, "learning_rate": 8.33991480157679e-05, "loss": 0.0293, "step": 1520 }, { "grad_norm": 0.5323015451431274, "learning_rate": 8.315232461754338e-05, "loss": 0.0314, "step": 1530 }, { "grad_norm": 0.4847467839717865, "learning_rate": 8.290405102924144e-05, "loss": 0.0325, "step": 1540 }, { "grad_norm": 0.4998106360435486, "learning_rate": 8.265433811115316e-05, "loss": 0.0325, "step": 1550 }, { "grad_norm": 0.5103840231895447, "learning_rate": 8.240319678653049e-05, "loss": 0.0327, "step": 1560 }, { "grad_norm": 0.5510079264640808, "learning_rate": 8.215063804110857e-05, "loss": 0.0346, "step": 1570 }, { "grad_norm": 0.5957469344139099, "learning_rate": 8.189667292262512e-05, "loss": 0.0311, "step": 1580 }, { "grad_norm": 0.501615583896637, "learning_rate": 8.164131254033716e-05, "loss": 0.0319, "step": 1590 }, { "grad_norm": 0.4492211639881134, "learning_rate": 8.138456806453503e-05, "loss": 0.0313, "step": 1600 }, { "grad_norm": 0.5348633527755737, "learning_rate": 8.112645072605386e-05, "loss": 0.03, "step": 1610 }, { "grad_norm": 0.5107631683349609, "learning_rate": 8.086697181578222e-05, "loss": 0.0345, "step": 1620 }, { "grad_norm": 0.45068928599357605, "learning_rate": 8.060614268416823e-05, "loss": 0.0321, "step": 1630 }, { "grad_norm": 0.3904348611831665, "learning_rate": 8.034397474072309e-05, "loss": 0.0297, "step": 1640 }, { "grad_norm": 0.5581115484237671, "learning_rate": 8.008047945352193e-05, "loss": 0.031, "step": 1650 }, { "grad_norm": 0.5107088685035706, "learning_rate": 7.981566834870225e-05, "loss": 0.0331, "step": 1660 }, { "grad_norm": 0.44165241718292236, "learning_rate": 7.954955300995961e-05, "loss": 0.0307, "step": 1670 }, { "grad_norm": 0.47160977125167847, "learning_rate": 7.928214507804104e-05, "loss": 0.0304, "step": 1680 }, { "grad_norm": 0.4960729479789734, "learning_rate": 7.901345625023576e-05, "loss": 0.0292, "step": 1690 }, { "grad_norm": 0.528746485710144, "learning_rate": 7.874349827986354e-05, "loss": 0.0324, "step": 1700 }, { "grad_norm": 0.47307735681533813, "learning_rate": 7.847228297576053e-05, "loss": 0.032, "step": 1710 }, { "grad_norm": 0.5367241501808167, "learning_rate": 7.819982220176276e-05, "loss": 0.0296, "step": 1720 }, { "grad_norm": 0.47281432151794434, "learning_rate": 7.792612787618714e-05, "loss": 0.0293, "step": 1730 }, { "grad_norm": 0.5008071064949036, "learning_rate": 7.765121197131009e-05, "loss": 0.0321, "step": 1740 }, { "grad_norm": 0.4207312762737274, "learning_rate": 7.737508651284391e-05, "loss": 0.0288, "step": 1750 }, { "grad_norm": 0.49400952458381653, "learning_rate": 7.709776357941069e-05, "loss": 0.0311, "step": 1760 }, { "grad_norm": 0.574874997138977, "learning_rate": 7.681925530201392e-05, "loss": 0.0301, "step": 1770 }, { "grad_norm": 0.4045153558254242, "learning_rate": 7.65395738635079e-05, "loss": 0.0302, "step": 1780 }, { "grad_norm": 0.4946967363357544, "learning_rate": 7.62587314980648e-05, "loss": 0.0306, "step": 1790 }, { "grad_norm": 0.45120808482170105, "learning_rate": 7.597674049063947e-05, "loss": 0.0286, "step": 1800 }, { "grad_norm": 0.5340564250946045, "learning_rate": 7.569361317643211e-05, "loss": 0.029, "step": 1810 }, { "grad_norm": 0.4389083683490753, "learning_rate": 7.540936194034865e-05, "loss": 0.0299, "step": 1820 }, { "grad_norm": 0.47904035449028015, "learning_rate": 7.512399921645901e-05, "loss": 0.0275, "step": 1830 }, { "grad_norm": 0.41724881529808044, "learning_rate": 7.483753748745317e-05, "loss": 0.0279, "step": 1840 }, { "grad_norm": 0.4357088506221771, "learning_rate": 7.454998928409516e-05, "loss": 0.0294, "step": 1850 }, { "grad_norm": 0.4692774713039398, "learning_rate": 7.426136718467493e-05, "loss": 0.0281, "step": 1860 }, { "grad_norm": 0.5714887380599976, "learning_rate": 7.397168381445812e-05, "loss": 0.0304, "step": 1870 }, { "grad_norm": 0.41615596413612366, "learning_rate": 7.368095184513377e-05, "loss": 0.0306, "step": 1880 }, { "grad_norm": 0.625390887260437, "learning_rate": 7.338918399426005e-05, "loss": 0.0273, "step": 1890 }, { "grad_norm": 0.4221304655075073, "learning_rate": 7.309639302470801e-05, "loss": 0.0283, "step": 1900 }, { "grad_norm": 0.44597235321998596, "learning_rate": 7.280259174410312e-05, "loss": 0.032, "step": 1910 }, { "grad_norm": 0.5311533212661743, "learning_rate": 7.250779300426517e-05, "loss": 0.0275, "step": 1920 }, { "grad_norm": 0.45131003856658936, "learning_rate": 7.22120097006461e-05, "loss": 0.0302, "step": 1930 }, { "grad_norm": 0.4558291435241699, "learning_rate": 7.191525477176577e-05, "loss": 0.0277, "step": 1940 }, { "grad_norm": 0.4313548803329468, "learning_rate": 7.161754119864616e-05, "loss": 0.0323, "step": 1950 }, { "grad_norm": 0.4807124733924866, "learning_rate": 7.131888200424339e-05, "loss": 0.031, "step": 1960 }, { "grad_norm": 0.48752525448799133, "learning_rate": 7.101929025287816e-05, "loss": 0.0287, "step": 1970 }, { "grad_norm": 0.3627080023288727, "learning_rate": 7.071877904966423e-05, "loss": 0.028, "step": 1980 }, { "grad_norm": 0.4705303907394409, "learning_rate": 7.04173615399351e-05, "loss": 0.0257, "step": 1990 }, { "grad_norm": 0.4926755428314209, "learning_rate": 7.011505090866913e-05, "loss": 0.028, "step": 2000 } ], "logging_steps": 10, "max_steps": 5000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 96, "trial_name": null, "trial_params": null }