{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.7877116975187082, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 0.00017015520980041642, "loss": 0.5575, "step": 100 }, { "epoch": 0.0, "learning_rate": 0.00017015035788629153, "loss": 0.3065, "step": 200 }, { "epoch": 0.01, "learning_rate": 0.00017014224985982174, "loss": 0.2229, "step": 300 }, { "epoch": 0.01, "learning_rate": 0.0001701308860313402, "loss": 0.1812, "step": 400 }, { "epoch": 0.01, "learning_rate": 0.00017011626683579524, "loss": 0.1572, "step": 500 }, { "epoch": 0.01, "learning_rate": 0.00017009839283273364, "loss": 0.1335, "step": 600 }, { "epoch": 0.01, "learning_rate": 0.00017007726470627936, "loss": 0.1193, "step": 700 }, { "epoch": 0.02, "learning_rate": 0.00017005288326510734, "loss": 0.1075, "step": 800 }, { "epoch": 0.02, "learning_rate": 0.0001700252494424124, "loss": 0.0986, "step": 900 }, { "epoch": 0.02, "learning_rate": 0.00016999436429587366, "loss": 0.0952, "step": 1000 }, { "epoch": 0.02, "learning_rate": 0.00016996022900761407, "loss": 0.0865, "step": 1100 }, { "epoch": 0.02, "learning_rate": 0.000169922844884155, "loss": 0.0816, "step": 1200 }, { "epoch": 0.03, "learning_rate": 0.00016988221335636648, "loss": 0.0775, "step": 1300 }, { "epoch": 0.03, "learning_rate": 0.00016983833597941224, "loss": 0.0744, "step": 1400 }, { "epoch": 0.03, "learning_rate": 0.00016979121443269025, "loss": 0.0694, "step": 1500 }, { "epoch": 0.03, "learning_rate": 0.0001697408505197684, "loss": 0.0641, "step": 1600 }, { "epoch": 0.03, "learning_rate": 0.00016968724616831557, "loss": 0.0629, "step": 1700 }, { "epoch": 0.04, "learning_rate": 0.0001696304034300278, "loss": 0.0696, "step": 1800 }, { "epoch": 0.04, "learning_rate": 0.00016957032448054968, "loss": 0.0591, "step": 1900 }, { "epoch": 0.04, "learning_rate": 0.0001695070116193912, "loss": 0.06, "step": 2000 }, { "epoch": 0.04, "learning_rate": 0.0001694404672698396, "loss": 0.0558, "step": 2100 }, { "epoch": 0.04, "learning_rate": 0.00016937069397886687, "loss": 0.0529, "step": 2200 }, { "epoch": 0.05, "learning_rate": 0.00016929769441703196, "loss": 0.0536, "step": 2300 }, { "epoch": 0.05, "learning_rate": 0.00016922147137837868, "loss": 0.0537, "step": 2400 }, { "epoch": 0.05, "learning_rate": 0.00016914202778032893, "loss": 0.0508, "step": 2500 }, { "epoch": 0.05, "learning_rate": 0.0001690593666635707, "loss": 0.0524, "step": 2600 }, { "epoch": 0.05, "learning_rate": 0.00016897349119194207, "loss": 0.0494, "step": 2700 }, { "epoch": 0.06, "learning_rate": 0.00016888440465230977, "loss": 0.047, "step": 2800 }, { "epoch": 0.06, "learning_rate": 0.00016879211045444354, "loss": 0.0502, "step": 2900 }, { "epoch": 0.06, "learning_rate": 0.00016869661213088575, "loss": 0.046, "step": 3000 }, { "epoch": 0.06, "learning_rate": 0.00016859791333681583, "loss": 0.0419, "step": 3100 }, { "epoch": 0.06, "learning_rate": 0.0001684960178499108, "loss": 0.0456, "step": 3200 }, { "epoch": 0.06, "learning_rate": 0.00016839092957020028, "loss": 0.0432, "step": 3300 }, { "epoch": 0.07, "learning_rate": 0.00016828265251991761, "loss": 0.0423, "step": 3400 }, { "epoch": 0.07, "learning_rate": 0.00016817119084334555, "loss": 0.0407, "step": 3500 }, { "epoch": 0.07, "learning_rate": 0.00016805654880665776, "loss": 0.0409, "step": 3600 }, { "epoch": 0.07, "learning_rate": 0.00016793873079775577, "loss": 0.0397, "step": 3700 }, { "epoch": 0.07, "learning_rate": 0.00016781774132610059, "loss": 0.0419, "step": 3800 }, { "epoch": 0.08, "learning_rate": 0.0001676935850225405, "loss": 0.0399, "step": 3900 }, { "epoch": 0.08, "learning_rate": 0.00016756626663913358, "loss": 0.0389, "step": 4000 }, { "epoch": 0.08, "learning_rate": 0.00016743579104896593, "loss": 0.0385, "step": 4100 }, { "epoch": 0.08, "learning_rate": 0.00016730216324596504, "loss": 0.0357, "step": 4200 }, { "epoch": 0.08, "learning_rate": 0.0001671653883447088, "loss": 0.0373, "step": 4300 }, { "epoch": 0.09, "learning_rate": 0.00016702547158022968, "loss": 0.0375, "step": 4400 }, { "epoch": 0.09, "learning_rate": 0.0001668824183078143, "loss": 0.0393, "step": 4500 }, { "epoch": 0.09, "learning_rate": 0.00016673623400279849, "loss": 0.0358, "step": 4600 }, { "epoch": 0.09, "learning_rate": 0.00016658692426035782, "loss": 0.0334, "step": 4700 }, { "epoch": 0.09, "learning_rate": 0.00016643449479529325, "loss": 0.035, "step": 4800 }, { "epoch": 0.1, "learning_rate": 0.00016627895144181258, "loss": 0.0343, "step": 4900 }, { "epoch": 0.1, "learning_rate": 0.0001661203001533071, "loss": 0.0335, "step": 5000 }, { "epoch": 0.1, "learning_rate": 0.00016595854700212362, "loss": 0.0353, "step": 5100 }, { "epoch": 0.1, "learning_rate": 0.0001657936981793322, "loss": 0.0331, "step": 5200 }, { "epoch": 0.1, "learning_rate": 0.0001656257599944891, "loss": 0.0387, "step": 5300 }, { "epoch": 0.11, "learning_rate": 0.00016545473887539532, "loss": 0.0327, "step": 5400 }, { "epoch": 0.11, "learning_rate": 0.00016528064136785056, "loss": 0.0316, "step": 5500 }, { "epoch": 0.11, "learning_rate": 0.00016510347413540262, "loss": 0.0319, "step": 5600 }, { "epoch": 0.11, "learning_rate": 0.0001649232439590925, "loss": 0.0314, "step": 5700 }, { "epoch": 0.11, "learning_rate": 0.0001647399577371947, "loss": 0.0321, "step": 5800 }, { "epoch": 0.12, "learning_rate": 0.00016455362248495338, "loss": 0.0333, "step": 5900 }, { "epoch": 0.12, "learning_rate": 0.00016436424533431362, "loss": 0.0319, "step": 6000 }, { "epoch": 0.12, "learning_rate": 0.0001641718335336486, "loss": 0.0315, "step": 6100 }, { "epoch": 0.12, "learning_rate": 0.0001639763944474821, "loss": 0.0311, "step": 6200 }, { "epoch": 0.12, "learning_rate": 0.0001637779355562068, "loss": 0.031, "step": 6300 }, { "epoch": 0.13, "learning_rate": 0.00016357646445579763, "loss": 0.0299, "step": 6400 }, { "epoch": 0.13, "learning_rate": 0.00016337198885752133, "loss": 0.0317, "step": 6500 }, { "epoch": 0.13, "learning_rate": 0.00016316451658764122, "loss": 0.0302, "step": 6600 }, { "epoch": 0.13, "learning_rate": 0.0001629540555871176, "loss": 0.0295, "step": 6700 }, { "epoch": 0.13, "learning_rate": 0.00016274061391130388, "loss": 0.03, "step": 6800 }, { "epoch": 0.14, "learning_rate": 0.0001625241997296382, "loss": 0.0292, "step": 6900 }, { "epoch": 0.14, "learning_rate": 0.00016230482132533077, "loss": 0.0289, "step": 7000 }, { "epoch": 0.14, "learning_rate": 0.00016208472504084003, "loss": 0.0318, "step": 7100 }, { "epoch": 0.14, "learning_rate": 0.0001618594729250462, "loss": 0.0301, "step": 7200 }, { "epoch": 0.14, "learning_rate": 0.00016163128202889828, "loss": 0.0295, "step": 7300 }, { "epoch": 0.15, "learning_rate": 0.00016140016108635798, "loss": 0.029, "step": 7400 }, { "epoch": 0.15, "learning_rate": 0.00016116611894353386, "loss": 0.0291, "step": 7500 }, { "epoch": 0.15, "learning_rate": 0.00016092916455834295, "loss": 0.0311, "step": 7600 }, { "epoch": 0.15, "learning_rate": 0.00016068930700016766, "loss": 0.0285, "step": 7700 }, { "epoch": 0.15, "learning_rate": 0.00016044655544950889, "loss": 0.0287, "step": 7800 }, { "epoch": 0.16, "learning_rate": 0.00016020091919763445, "loss": 0.0293, "step": 7900 }, { "epoch": 0.16, "learning_rate": 0.00015995240764622357, "loss": 0.0259, "step": 8000 }, { "epoch": 0.16, "learning_rate": 0.000159701030307007, "loss": 0.0293, "step": 8100 }, { "epoch": 0.16, "learning_rate": 0.00015944679680140295, "loss": 0.0277, "step": 8200 }, { "epoch": 0.16, "learning_rate": 0.0001591897168601488, "loss": 0.0304, "step": 8300 }, { "epoch": 0.17, "learning_rate": 0.00015892980032292876, "loss": 0.026, "step": 8400 }, { "epoch": 0.17, "learning_rate": 0.00015866705713799714, "loss": 0.0294, "step": 8500 }, { "epoch": 0.17, "learning_rate": 0.00015840149736179762, "loss": 0.0321, "step": 8600 }, { "epoch": 0.17, "learning_rate": 0.0001581331311585785, "loss": 0.0279, "step": 8700 }, { "epoch": 0.17, "learning_rate": 0.00015786196880000325, "loss": 0.0277, "step": 8800 }, { "epoch": 0.18, "learning_rate": 0.0001575880206647579, "loss": 0.0268, "step": 8900 }, { "epoch": 0.18, "learning_rate": 0.00015731129723815343, "loss": 0.0281, "step": 9000 }, { "epoch": 0.18, "learning_rate": 0.00015703180911172453, "loss": 0.028, "step": 9100 }, { "epoch": 0.18, "learning_rate": 0.0001567495669828243, "loss": 0.026, "step": 9200 }, { "epoch": 0.18, "learning_rate": 0.0001564645816542146, "loss": 0.0256, "step": 9300 }, { "epoch": 0.19, "learning_rate": 0.0001561768640336529, "loss": 0.027, "step": 9400 }, { "epoch": 0.19, "learning_rate": 0.0001558864251334745, "loss": 0.0249, "step": 9500 }, { "epoch": 0.19, "learning_rate": 0.00015559327607017119, "loss": 0.0256, "step": 9600 }, { "epoch": 0.19, "learning_rate": 0.00015529742806396564, "loss": 0.0251, "step": 9700 }, { "epoch": 0.19, "learning_rate": 0.00015499889243838211, "loss": 0.0257, "step": 9800 }, { "epoch": 0.19, "learning_rate": 0.00015469768061981295, "loss": 0.0264, "step": 9900 }, { "epoch": 0.2, "learning_rate": 0.00015439685605389073, "loss": 0.0247, "step": 10000 }, { "epoch": 0.2, "learning_rate": 0.00015409035301023626, "loss": 0.0256, "step": 10100 }, { "epoch": 0.2, "learning_rate": 0.00015378120854776404, "loss": 0.0247, "step": 10200 }, { "epoch": 0.2, "learning_rate": 0.00015346943449891755, "loss": 0.0254, "step": 10300 }, { "epoch": 0.2, "learning_rate": 0.00015315504279678714, "loss": 0.0246, "step": 10400 }, { "epoch": 0.21, "learning_rate": 0.00015283804547465337, "loss": 0.0249, "step": 10500 }, { "epoch": 0.21, "learning_rate": 0.0001525184546655264, "loss": 0.026, "step": 10600 }, { "epoch": 0.21, "learning_rate": 0.0001521962826016816, "loss": 0.0254, "step": 10700 }, { "epoch": 0.21, "learning_rate": 0.00015187154161419122, "loss": 0.0272, "step": 10800 }, { "epoch": 0.21, "learning_rate": 0.0001515442441324528, "loss": 0.0253, "step": 10900 }, { "epoch": 0.22, "learning_rate": 0.00015121440268371297, "loss": 0.0263, "step": 11000 }, { "epoch": 0.22, "learning_rate": 0.00015088202989258835, "loss": 0.0256, "step": 11100 }, { "epoch": 0.22, "learning_rate": 0.00015054713848058212, "loss": 0.023, "step": 11200 }, { "epoch": 0.22, "learning_rate": 0.00015020974126559716, "loss": 0.0236, "step": 11300 }, { "epoch": 0.22, "learning_rate": 0.00014986985116144554, "loss": 0.0241, "step": 11400 }, { "epoch": 0.23, "learning_rate": 0.00014952748117735409, "loss": 0.0235, "step": 11500 }, { "epoch": 0.23, "learning_rate": 0.0001491826444174666, "loss": 0.0242, "step": 11600 }, { "epoch": 0.23, "learning_rate": 0.00014883535408034227, "loss": 0.0233, "step": 11700 }, { "epoch": 0.23, "learning_rate": 0.00014848562345845032, "loss": 0.0228, "step": 11800 }, { "epoch": 0.23, "learning_rate": 0.00014813346593766164, "loss": 0.023, "step": 11900 }, { "epoch": 0.24, "learning_rate": 0.0001477788949967359, "loss": 0.0236, "step": 12000 }, { "epoch": 0.24, "learning_rate": 0.00014742192420680626, "loss": 0.0237, "step": 12100 }, { "epoch": 0.24, "learning_rate": 0.00014706256723085937, "loss": 0.0237, "step": 12200 }, { "epoch": 0.24, "learning_rate": 0.0001467008378232128, "loss": 0.024, "step": 12300 }, { "epoch": 0.24, "learning_rate": 0.00014633674982898854, "loss": 0.0227, "step": 12400 }, { "epoch": 0.25, "learning_rate": 0.00014597031718358285, "loss": 0.0247, "step": 12500 }, { "epoch": 0.25, "learning_rate": 0.00014560155391213318, "loss": 0.0233, "step": 12600 }, { "epoch": 0.25, "learning_rate": 0.0001452304741289812, "loss": 0.0234, "step": 12700 }, { "epoch": 0.25, "learning_rate": 0.00014485709203713263, "loss": 0.0221, "step": 12800 }, { "epoch": 0.25, "learning_rate": 0.00014448142192771354, "loss": 0.0244, "step": 12900 }, { "epoch": 0.26, "learning_rate": 0.00014410347817942347, "loss": 0.0242, "step": 13000 }, { "epoch": 0.26, "learning_rate": 0.00014372327525798503, "loss": 0.0231, "step": 13100 }, { "epoch": 0.26, "learning_rate": 0.00014334082771559026, "loss": 0.0238, "step": 13200 }, { "epoch": 0.26, "learning_rate": 0.00014295615019034358, "loss": 0.0233, "step": 13300 }, { "epoch": 0.26, "learning_rate": 0.0001425692574057016, "loss": 0.024, "step": 13400 }, { "epoch": 0.27, "learning_rate": 0.00014218016416990954, "loss": 0.0249, "step": 13500 }, { "epoch": 0.27, "learning_rate": 0.0001417928089327897, "loss": 0.0265, "step": 13600 }, { "epoch": 0.27, "learning_rate": 0.00014139938118717103, "loss": 0.0206, "step": 13700 }, { "epoch": 0.27, "learning_rate": 0.00014100379776718443, "loss": 0.0216, "step": 13800 }, { "epoch": 0.27, "learning_rate": 0.00014060607381370744, "loss": 0.0239, "step": 13900 }, { "epoch": 0.28, "learning_rate": 0.00014020622454954606, "loss": 0.024, "step": 14000 }, { "epoch": 0.28, "learning_rate": 0.00013980426527885237, "loss": 0.0234, "step": 14100 }, { "epoch": 0.28, "learning_rate": 0.00013940021138653824, "loss": 0.0228, "step": 14200 }, { "epoch": 0.28, "learning_rate": 0.00013899407833768695, "loss": 0.0228, "step": 14300 }, { "epoch": 0.28, "learning_rate": 0.0001385858816769611, "loss": 0.0234, "step": 14400 }, { "epoch": 0.29, "learning_rate": 0.0001381756370280075, "loss": 0.0233, "step": 14500 }, { "epoch": 0.29, "learning_rate": 0.00013776336009285937, "loss": 0.0216, "step": 14600 }, { "epoch": 0.29, "learning_rate": 0.00013734906665133537, "loss": 0.0268, "step": 14700 }, { "epoch": 0.29, "learning_rate": 0.0001369327725604354, "loss": 0.022, "step": 14800 }, { "epoch": 0.29, "learning_rate": 0.00013651449375373396, "loss": 0.0214, "step": 14900 }, { "epoch": 0.3, "learning_rate": 0.00013609424624077, "loss": 0.0235, "step": 15000 }, { "epoch": 0.3, "learning_rate": 0.0001356720461064345, "loss": 0.0225, "step": 15100 }, { "epoch": 0.3, "learning_rate": 0.00013524790951035441, "loss": 0.0218, "step": 15200 }, { "epoch": 0.3, "learning_rate": 0.00013482185268627465, "loss": 0.0225, "step": 15300 }, { "epoch": 0.3, "learning_rate": 0.00013439389194143625, "loss": 0.0216, "step": 15400 }, { "epoch": 0.31, "learning_rate": 0.00013396404365595253, "loss": 0.0223, "step": 15500 }, { "epoch": 0.31, "learning_rate": 0.00013353232428218212, "loss": 0.0212, "step": 15600 }, { "epoch": 0.31, "learning_rate": 0.00013309875034409903, "loss": 0.023, "step": 15700 }, { "epoch": 0.31, "learning_rate": 0.0001326633384366605, "loss": 0.0214, "step": 15800 }, { "epoch": 0.31, "learning_rate": 0.0001322261052251716, "loss": 0.021, "step": 15900 }, { "epoch": 0.32, "learning_rate": 0.00013178706744464749, "loss": 0.0211, "step": 16000 }, { "epoch": 0.32, "learning_rate": 0.00013134624189917282, "loss": 0.0212, "step": 16100 }, { "epoch": 0.32, "learning_rate": 0.00013090808013598785, "loss": 0.0252, "step": 16200 }, { "epoch": 0.32, "learning_rate": 0.00013046374720137184, "loss": 0.0214, "step": 16300 }, { "epoch": 0.32, "learning_rate": 0.00013001767715162726, "loss": 0.0202, "step": 16400 }, { "epoch": 0.32, "learning_rate": 0.00012956988705999754, "loss": 0.0205, "step": 16500 }, { "epoch": 0.33, "learning_rate": 0.00012912039406556028, "loss": 0.0235, "step": 16600 }, { "epoch": 0.33, "learning_rate": 0.00012866921537257149, "loss": 0.0215, "step": 16700 }, { "epoch": 0.33, "learning_rate": 0.00012821636824980682, "loss": 0.0198, "step": 16800 }, { "epoch": 0.33, "learning_rate": 0.00012776187002990082, "loss": 0.0232, "step": 16900 }, { "epoch": 0.33, "learning_rate": 0.00012730573810868347, "loss": 0.0235, "step": 17000 }, { "epoch": 0.34, "learning_rate": 0.00012684798994451428, "loss": 0.0207, "step": 17100 }, { "epoch": 0.34, "learning_rate": 0.00012638864305761417, "loss": 0.0218, "step": 17200 }, { "epoch": 0.34, "learning_rate": 0.00012592771502939492, "loss": 0.0204, "step": 17300 }, { "epoch": 0.34, "learning_rate": 0.0001254652235017861, "loss": 0.0212, "step": 17400 }, { "epoch": 0.34, "learning_rate": 0.00012500118617655992, "loss": 0.0207, "step": 17500 }, { "epoch": 0.35, "learning_rate": 0.00012453562081465375, "loss": 0.0207, "step": 17600 }, { "epoch": 0.35, "learning_rate": 0.0001240685452354902, "loss": 0.0204, "step": 17700 }, { "epoch": 0.35, "learning_rate": 0.00012359997731629517, "loss": 0.0234, "step": 17800 }, { "epoch": 0.35, "learning_rate": 0.00012312993499141364, "loss": 0.02, "step": 17900 }, { "epoch": 0.35, "learning_rate": 0.0001226584362516231, "loss": 0.0191, "step": 18000 }, { "epoch": 0.36, "learning_rate": 0.00012218549914344497, "loss": 0.0196, "step": 18100 }, { "epoch": 0.36, "learning_rate": 0.00012171114176845411, "loss": 0.0203, "step": 18200 }, { "epoch": 0.36, "learning_rate": 0.0001212353822825857, "loss": 0.0224, "step": 18300 }, { "epoch": 0.36, "learning_rate": 0.00012075823889544048, "loss": 0.0211, "step": 18400 }, { "epoch": 0.36, "learning_rate": 0.00012027972986958772, "loss": 0.0188, "step": 18500 }, { "epoch": 0.37, "learning_rate": 0.0001197998735198662, "loss": 0.0214, "step": 18600 }, { "epoch": 0.37, "learning_rate": 0.00011931868821268337, "loss": 0.0201, "step": 18700 }, { "epoch": 0.37, "learning_rate": 0.00011883619236531218, "loss": 0.0206, "step": 18800 }, { "epoch": 0.37, "learning_rate": 0.00011835240444518623, "loss": 0.0206, "step": 18900 }, { "epoch": 0.37, "learning_rate": 0.000117867342969193, "loss": 0.0197, "step": 19000 }, { "epoch": 0.38, "learning_rate": 0.00011738102650296509, "loss": 0.0198, "step": 19100 }, { "epoch": 0.38, "learning_rate": 0.00011689347366016955, "loss": 0.0211, "step": 19200 }, { "epoch": 0.38, "learning_rate": 0.0001164047031017955, "loss": 0.0204, "step": 19300 }, { "epoch": 0.38, "learning_rate": 0.00011591473353543992, "loss": 0.0206, "step": 19400 }, { "epoch": 0.38, "learning_rate": 0.00011542358371459156, "loss": 0.0192, "step": 19500 }, { "epoch": 0.39, "learning_rate": 0.0001149312724379132, "loss": 0.0202, "step": 19600 }, { "epoch": 0.39, "learning_rate": 0.00011443781854852212, "loss": 0.0197, "step": 19700 }, { "epoch": 0.39, "learning_rate": 0.0001139432409332688, "loss": 0.021, "step": 19800 }, { "epoch": 0.39, "learning_rate": 0.00011344755852201423, "loss": 0.022, "step": 19900 }, { "epoch": 0.39, "learning_rate": 0.00011295079028690513, "loss": 0.0196, "step": 20000 }, { "epoch": 0.4, "learning_rate": 0.00011245295524164799, "loss": 0.0195, "step": 20100 }, { "epoch": 0.4, "learning_rate": 0.00011195407244078124, "loss": 0.0197, "step": 20200 }, { "epoch": 0.4, "learning_rate": 0.00011145916512264182, "loss": 0.0227, "step": 20300 }, { "epoch": 0.4, "learning_rate": 0.0001109582541342795, "loss": 0.0197, "step": 20400 }, { "epoch": 0.4, "learning_rate": 0.00011045635259969838, "loss": 0.0203, "step": 20500 }, { "epoch": 0.41, "learning_rate": 0.0001099585132027323, "loss": 0.0216, "step": 20600 }, { "epoch": 0.41, "learning_rate": 0.00010945469766893083, "loss": 0.0224, "step": 20700 }, { "epoch": 0.41, "learning_rate": 0.00010894994913723849, "loss": 0.0193, "step": 20800 }, { "epoch": 0.41, "learning_rate": 0.00010844428692680601, "loss": 0.0203, "step": 20900 }, { "epoch": 0.41, "learning_rate": 0.000107937730391755, "loss": 0.02, "step": 21000 }, { "epoch": 0.42, "learning_rate": 0.00010743029892043725, "loss": 0.0197, "step": 21100 }, { "epoch": 0.42, "learning_rate": 0.00010692201193469241, "loss": 0.0192, "step": 21200 }, { "epoch": 0.42, "learning_rate": 0.00010641288888910492, "loss": 0.0189, "step": 21300 }, { "epoch": 0.42, "learning_rate": 0.00010590294927025919, "loss": 0.02, "step": 21400 }, { "epoch": 0.42, "learning_rate": 0.00010539221259599377, "loss": 0.0182, "step": 21500 }, { "epoch": 0.43, "learning_rate": 0.00010488069841465444, "loss": 0.0191, "step": 21600 }, { "epoch": 0.43, "learning_rate": 0.00010436842630434579, "loss": 0.0188, "step": 21700 }, { "epoch": 0.43, "learning_rate": 0.0001038554158721821, "loss": 0.023, "step": 21800 }, { "epoch": 0.43, "learning_rate": 0.00010334168675353674, "loss": 0.0206, "step": 21900 }, { "epoch": 0.43, "learning_rate": 0.00010282725861129068, "loss": 0.0187, "step": 22000 }, { "epoch": 0.44, "learning_rate": 0.00010231215113507977, "loss": 0.018, "step": 22100 }, { "epoch": 0.44, "learning_rate": 0.00010179638404054145, "loss": 0.0202, "step": 22200 }, { "epoch": 0.44, "learning_rate": 0.00010127997706855969, "loss": 0.0199, "step": 22300 }, { "epoch": 0.44, "learning_rate": 0.00010076294998450981, "loss": 0.019, "step": 22400 }, { "epoch": 0.44, "learning_rate": 0.0001002453225775017, "loss": 0.0193, "step": 22500 }, { "epoch": 0.45, "learning_rate": 9.972711465962257e-05, "loss": 0.019, "step": 22600 }, { "epoch": 0.45, "learning_rate": 9.920834606517847e-05, "loss": 0.0195, "step": 22700 }, { "epoch": 0.45, "learning_rate": 9.868903664993534e-05, "loss": 0.018, "step": 22800 }, { "epoch": 0.45, "learning_rate": 9.816920629035887e-05, "loss": 0.0202, "step": 22900 }, { "epoch": 0.45, "learning_rate": 9.764887488285381e-05, "loss": 0.0185, "step": 23000 }, { "epoch": 0.45, "learning_rate": 9.71280623430024e-05, "loss": 0.0189, "step": 23100 }, { "epoch": 0.46, "learning_rate": 9.660678860480215e-05, "loss": 0.0186, "step": 23200 }, { "epoch": 0.46, "learning_rate": 9.608507361990283e-05, "loss": 0.0195, "step": 23300 }, { "epoch": 0.46, "learning_rate": 9.556293735684288e-05, "loss": 0.0206, "step": 23400 }, { "epoch": 0.46, "learning_rate": 9.50456270965957e-05, "loss": 0.0204, "step": 23500 }, { "epoch": 0.46, "learning_rate": 9.452271196044817e-05, "loss": 0.0189, "step": 23600 }, { "epoch": 0.47, "learning_rate": 9.399943534522518e-05, "loss": 0.0198, "step": 23700 }, { "epoch": 0.47, "learning_rate": 9.347581727923617e-05, "loss": 0.0197, "step": 23800 }, { "epoch": 0.47, "learning_rate": 9.295187780385963e-05, "loss": 0.0187, "step": 23900 }, { "epoch": 0.47, "learning_rate": 9.242763697277588e-05, "loss": 0.0201, "step": 24000 }, { "epoch": 0.47, "learning_rate": 9.19031148511996e-05, "loss": 0.0194, "step": 24100 }, { "epoch": 0.48, "learning_rate": 9.137833151511182e-05, "loss": 0.0186, "step": 24200 }, { "epoch": 0.48, "learning_rate": 9.08533070504915e-05, "loss": 0.0215, "step": 24300 }, { "epoch": 0.48, "learning_rate": 9.032806155254666e-05, "loss": 0.0205, "step": 24400 }, { "epoch": 0.48, "learning_rate": 8.980261512494549e-05, "loss": 0.0188, "step": 24500 }, { "epoch": 0.48, "learning_rate": 8.927698787904661e-05, "loss": 0.0188, "step": 24600 }, { "epoch": 0.49, "learning_rate": 8.875119993312938e-05, "loss": 0.0204, "step": 24700 }, { "epoch": 0.49, "learning_rate": 8.822527141162407e-05, "loss": 0.0191, "step": 24800 }, { "epoch": 0.49, "learning_rate": 8.769922244434138e-05, "loss": 0.017, "step": 24900 }, { "epoch": 0.49, "learning_rate": 8.717307316570196e-05, "loss": 0.0186, "step": 25000 }, { "epoch": 0.49, "learning_rate": 8.664684371396603e-05, "loss": 0.0195, "step": 25100 }, { "epoch": 0.5, "learning_rate": 8.612055423046226e-05, "loss": 0.0199, "step": 25200 }, { "epoch": 0.5, "learning_rate": 8.559422485881711e-05, "loss": 0.0191, "step": 25300 }, { "epoch": 0.5, "learning_rate": 8.506787574418376e-05, "loss": 0.0191, "step": 25400 }, { "epoch": 0.5, "learning_rate": 8.4541527032471e-05, "loss": 0.0196, "step": 25500 }, { "epoch": 0.5, "learning_rate": 8.401519886957223e-05, "loss": 0.0184, "step": 25600 }, { "epoch": 0.51, "learning_rate": 8.348891140059429e-05, "loss": 0.0182, "step": 25700 }, { "epoch": 0.51, "learning_rate": 8.29626847690866e-05, "loss": 0.0216, "step": 25800 }, { "epoch": 0.51, "learning_rate": 8.243653911626992e-05, "loss": 0.0177, "step": 25900 }, { "epoch": 0.51, "learning_rate": 8.191575445898302e-05, "loss": 0.019, "step": 26000 }, { "epoch": 0.51, "learning_rate": 8.13898298618816e-05, "loss": 0.0187, "step": 26100 }, { "epoch": 0.52, "learning_rate": 8.086404644418334e-05, "loss": 0.0189, "step": 26200 }, { "epoch": 0.52, "learning_rate": 8.033842433014505e-05, "loss": 0.0178, "step": 26300 }, { "epoch": 0.52, "learning_rate": 7.981298363784983e-05, "loss": 0.0182, "step": 26400 }, { "epoch": 0.52, "learning_rate": 7.928774447843676e-05, "loss": 0.0197, "step": 26500 }, { "epoch": 0.52, "learning_rate": 7.876272695533133e-05, "loss": 0.0182, "step": 26600 }, { "epoch": 0.53, "learning_rate": 7.823795116347596e-05, "loss": 0.0203, "step": 26700 }, { "epoch": 0.53, "learning_rate": 7.771343718856085e-05, "loss": 0.0193, "step": 26800 }, { "epoch": 0.53, "learning_rate": 7.718920510625514e-05, "loss": 0.0182, "step": 26900 }, { "epoch": 0.53, "learning_rate": 7.666527498143874e-05, "loss": 0.0202, "step": 27000 }, { "epoch": 0.53, "learning_rate": 7.614166686743405e-05, "loss": 0.019, "step": 27100 }, { "epoch": 0.54, "learning_rate": 7.561840080523864e-05, "loss": 0.0174, "step": 27200 }, { "epoch": 0.54, "learning_rate": 7.509549682275809e-05, "loss": 0.018, "step": 27300 }, { "epoch": 0.54, "learning_rate": 7.457297493403944e-05, "loss": 0.019, "step": 27400 }, { "epoch": 0.54, "learning_rate": 7.405085513850521e-05, "loss": 0.0205, "step": 27500 }, { "epoch": 0.54, "learning_rate": 7.352915742018785e-05, "loss": 0.0169, "step": 27600 }, { "epoch": 0.55, "learning_rate": 7.30079017469649e-05, "loss": 0.0183, "step": 27700 }, { "epoch": 0.55, "learning_rate": 7.248710806979468e-05, "loss": 0.0181, "step": 27800 }, { "epoch": 0.55, "learning_rate": 7.196679632195274e-05, "loss": 0.0171, "step": 27900 }, { "epoch": 0.55, "learning_rate": 7.144698641826882e-05, "loss": 0.0202, "step": 28000 }, { "epoch": 0.55, "learning_rate": 7.092769825436483e-05, "loss": 0.0184, "step": 28100 }, { "epoch": 0.56, "learning_rate": 7.040895170589299e-05, "loss": 0.0191, "step": 28200 }, { "epoch": 0.56, "learning_rate": 6.989594563413708e-05, "loss": 0.0206, "step": 28300 }, { "epoch": 0.56, "learning_rate": 6.937833594866479e-05, "loss": 0.0184, "step": 28400 }, { "epoch": 0.56, "learning_rate": 6.886132718016212e-05, "loss": 0.0213, "step": 28500 }, { "epoch": 0.56, "learning_rate": 6.834493911703791e-05, "loss": 0.0181, "step": 28600 }, { "epoch": 0.57, "learning_rate": 6.78291915239438e-05, "loss": 0.0205, "step": 28700 }, { "epoch": 0.57, "learning_rate": 6.731410414101738e-05, "loss": 0.0184, "step": 28800 }, { "epoch": 0.57, "learning_rate": 6.679969668312693e-05, "loss": 0.0179, "step": 28900 }, { "epoch": 0.57, "learning_rate": 6.628598883911667e-05, "loss": 0.0206, "step": 29000 }, { "epoch": 0.57, "learning_rate": 6.577300027105329e-05, "loss": 0.0172, "step": 29100 }, { "epoch": 0.58, "learning_rate": 6.526075061347325e-05, "loss": 0.0194, "step": 29200 }, { "epoch": 0.58, "learning_rate": 6.474925947263152e-05, "loss": 0.0195, "step": 29300 }, { "epoch": 0.58, "learning_rate": 6.423854642575082e-05, "loss": 0.0193, "step": 29400 }, { "epoch": 0.58, "learning_rate": 6.372863102027257e-05, "loss": 0.0182, "step": 29500 }, { "epoch": 0.58, "learning_rate": 6.321953277310858e-05, "loss": 0.0195, "step": 29600 }, { "epoch": 0.58, "learning_rate": 6.27112711698941e-05, "loss": 0.0174, "step": 29700 }, { "epoch": 0.59, "learning_rate": 6.220386566424204e-05, "loss": 0.0214, "step": 29800 }, { "epoch": 0.59, "learning_rate": 6.169733567699826e-05, "loss": 0.0176, "step": 29900 }, { "epoch": 0.59, "learning_rate": 6.11917005954984e-05, "loss": 0.017, "step": 30000 }, { "epoch": 0.59, "learning_rate": 6.068697977282572e-05, "loss": 0.0185, "step": 30100 }, { "epoch": 0.59, "learning_rate": 6.018319252707041e-05, "loss": 0.0175, "step": 30200 }, { "epoch": 0.6, "learning_rate": 5.968035814059016e-05, "loss": 0.0166, "step": 30300 }, { "epoch": 0.6, "learning_rate": 5.9178495859272244e-05, "loss": 0.0167, "step": 30400 }, { "epoch": 0.6, "learning_rate": 5.8677624891796735e-05, "loss": 0.018, "step": 30500 }, { "epoch": 0.6, "learning_rate": 5.8177764408901366e-05, "loss": 0.0182, "step": 30600 }, { "epoch": 0.6, "learning_rate": 5.767893354264776e-05, "loss": 0.019, "step": 30700 }, { "epoch": 0.61, "learning_rate": 5.7186123953555236e-05, "loss": 0.0198, "step": 30800 }, { "epoch": 0.61, "learning_rate": 5.668939878661137e-05, "loss": 0.0177, "step": 30900 }, { "epoch": 0.61, "learning_rate": 5.619376020321073e-05, "loss": 0.0165, "step": 31000 }, { "epoch": 0.61, "learning_rate": 5.5699227173822495e-05, "loss": 0.0205, "step": 31100 }, { "epoch": 0.61, "learning_rate": 5.5205818626601115e-05, "loss": 0.0181, "step": 31200 }, { "epoch": 0.62, "learning_rate": 5.4713553446661644e-05, "loss": 0.0187, "step": 31300 }, { "epoch": 0.62, "learning_rate": 5.4222450475357026e-05, "loss": 0.0169, "step": 31400 }, { "epoch": 0.62, "learning_rate": 5.373252850955681e-05, "loss": 0.0174, "step": 31500 }, { "epoch": 0.62, "learning_rate": 5.3243806300927926e-05, "loss": 0.0179, "step": 31600 }, { "epoch": 0.62, "learning_rate": 5.2756302555216735e-05, "loss": 0.0175, "step": 31700 }, { "epoch": 0.63, "learning_rate": 5.2270035931533207e-05, "loss": 0.0199, "step": 31800 }, { "epoch": 0.63, "learning_rate": 5.1785025041636665e-05, "loss": 0.0184, "step": 31900 }, { "epoch": 0.63, "learning_rate": 5.130128844922355e-05, "loss": 0.0185, "step": 32000 }, { "epoch": 0.63, "learning_rate": 5.081884466921669e-05, "loss": 0.0174, "step": 32100 }, { "epoch": 0.63, "learning_rate": 5.0337712167056944e-05, "loss": 0.0201, "step": 32200 }, { "epoch": 0.64, "learning_rate": 4.985790935799614e-05, "loss": 0.0178, "step": 32300 }, { "epoch": 0.64, "learning_rate": 4.937945460639236e-05, "loss": 0.0176, "step": 32400 }, { "epoch": 0.64, "learning_rate": 4.890236622500717e-05, "loss": 0.0239, "step": 32500 }, { "epoch": 0.64, "learning_rate": 4.8426662474304465e-05, "loss": 0.0177, "step": 32600 }, { "epoch": 0.64, "learning_rate": 4.795236156175173e-05, "loss": 0.0182, "step": 32700 }, { "epoch": 0.65, "learning_rate": 4.7479481641123066e-05, "loss": 0.016, "step": 32800 }, { "epoch": 0.65, "learning_rate": 4.7008040811804424e-05, "loss": 0.0175, "step": 32900 }, { "epoch": 0.65, "learning_rate": 4.653805711810078e-05, "loss": 0.0173, "step": 33000 }, { "epoch": 0.65, "learning_rate": 4.6069548548545565e-05, "loss": 0.0179, "step": 33100 }, { "epoch": 0.65, "learning_rate": 4.560253303521206e-05, "loss": 0.0186, "step": 33200 }, { "epoch": 0.66, "learning_rate": 4.513702845302723e-05, "loss": 0.019, "step": 33300 }, { "epoch": 0.66, "learning_rate": 4.467305261908736e-05, "loss": 0.0164, "step": 33400 }, { "epoch": 0.66, "learning_rate": 4.42106232919762e-05, "loss": 0.019, "step": 33500 }, { "epoch": 0.66, "learning_rate": 4.374975817108527e-05, "loss": 0.0174, "step": 33600 }, { "epoch": 0.66, "learning_rate": 4.3295059840780387e-05, "loss": 0.0187, "step": 33700 }, { "epoch": 0.67, "learning_rate": 4.283735990928788e-05, "loss": 0.019, "step": 33800 }, { "epoch": 0.67, "learning_rate": 4.2381276745401456e-05, "loss": 0.0175, "step": 33900 }, { "epoch": 0.67, "learning_rate": 4.192682780561444e-05, "loss": 0.0186, "step": 34000 }, { "epoch": 0.67, "learning_rate": 4.147403048387059e-05, "loss": 0.02, "step": 34100 }, { "epoch": 0.67, "learning_rate": 4.1022902110898326e-05, "loss": 0.0169, "step": 34200 }, { "epoch": 0.68, "learning_rate": 4.057345995354738e-05, "loss": 0.0176, "step": 34300 }, { "epoch": 0.68, "learning_rate": 4.0125721214127854e-05, "loss": 0.0167, "step": 34400 }, { "epoch": 0.68, "learning_rate": 3.9679703029752e-05, "loss": 0.0179, "step": 34500 }, { "epoch": 0.68, "learning_rate": 3.923542247167802e-05, "loss": 0.0162, "step": 34600 }, { "epoch": 0.68, "learning_rate": 3.879289654465689e-05, "loss": 0.0194, "step": 34700 }, { "epoch": 0.69, "learning_rate": 3.835214218628141e-05, "loss": 0.0187, "step": 34800 }, { "epoch": 0.69, "learning_rate": 3.7913176266337885e-05, "loss": 0.0165, "step": 34900 }, { "epoch": 0.69, "learning_rate": 3.747601558616062e-05, "loss": 0.0162, "step": 35000 }, { "epoch": 0.69, "learning_rate": 3.70406768779886e-05, "loss": 0.0169, "step": 35100 }, { "epoch": 0.69, "learning_rate": 3.660717680432526e-05, "loss": 0.0173, "step": 35200 }, { "epoch": 0.7, "learning_rate": 3.61755319573006e-05, "loss": 0.016, "step": 35300 }, { "epoch": 0.7, "learning_rate": 3.5745758858036255e-05, "loss": 0.0188, "step": 35400 }, { "epoch": 0.7, "learning_rate": 3.5317873956012995e-05, "loss": 0.0198, "step": 35500 }, { "epoch": 0.7, "learning_rate": 3.4891893628441334e-05, "loss": 0.0171, "step": 35600 }, { "epoch": 0.7, "learning_rate": 3.446783417963448e-05, "loss": 0.0179, "step": 35700 }, { "epoch": 0.71, "learning_rate": 3.4045711840384404e-05, "loss": 0.0158, "step": 35800 }, { "epoch": 0.71, "learning_rate": 3.36255427673406e-05, "loss": 0.0171, "step": 35900 }, { "epoch": 0.71, "learning_rate": 3.320734304239168e-05, "loss": 0.0162, "step": 36000 }, { "epoch": 0.71, "learning_rate": 3.279112867204983e-05, "loss": 0.0172, "step": 36100 }, { "epoch": 0.71, "learning_rate": 3.2376915586838195e-05, "loss": 0.0166, "step": 36200 }, { "epoch": 0.71, "learning_rate": 3.196471964068113e-05, "loss": 0.0171, "step": 36300 }, { "epoch": 0.72, "learning_rate": 3.1554556610297393e-05, "loss": 0.019, "step": 36400 }, { "epoch": 0.72, "learning_rate": 3.1146442194596243e-05, "loss": 0.0167, "step": 36500 }, { "epoch": 0.72, "learning_rate": 3.074039201407669e-05, "loss": 0.0172, "step": 36600 }, { "epoch": 0.72, "learning_rate": 3.0336421610229454e-05, "loss": 0.0182, "step": 36700 }, { "epoch": 0.72, "learning_rate": 2.9934546444942337e-05, "loss": 0.0171, "step": 36800 }, { "epoch": 0.73, "learning_rate": 2.9534781899908195e-05, "loss": 0.0165, "step": 36900 }, { "epoch": 0.73, "learning_rate": 2.9141109088924195e-05, "loss": 0.02, "step": 37000 }, { "epoch": 0.73, "learning_rate": 2.87455901192796e-05, "loss": 0.0174, "step": 37100 }, { "epoch": 0.73, "learning_rate": 2.835222727695746e-05, "loss": 0.0171, "step": 37200 }, { "epoch": 0.73, "learning_rate": 2.7961035617843112e-05, "loss": 0.0172, "step": 37300 }, { "epoch": 0.74, "learning_rate": 2.757203011472043e-05, "loss": 0.0164, "step": 37400 }, { "epoch": 0.74, "learning_rate": 2.718522565669847e-05, "loss": 0.017, "step": 37500 }, { "epoch": 0.74, "learning_rate": 2.6800637048641775e-05, "loss": 0.0167, "step": 37600 }, { "epoch": 0.74, "learning_rate": 2.6418279010603663e-05, "loss": 0.0184, "step": 37700 }, { "epoch": 0.74, "learning_rate": 2.6038166177262784e-05, "loss": 0.0158, "step": 37800 }, { "epoch": 0.75, "learning_rate": 2.566031309736311e-05, "loss": 0.0165, "step": 37900 }, { "epoch": 0.75, "learning_rate": 2.5284734233156903e-05, "loss": 0.0165, "step": 38000 }, { "epoch": 0.75, "learning_rate": 2.4911443959851313e-05, "loss": 0.017, "step": 38100 }, { "epoch": 0.75, "learning_rate": 2.4540456565058115e-05, "loss": 0.0177, "step": 38200 }, { "epoch": 0.75, "learning_rate": 2.417178624824684e-05, "loss": 0.0165, "step": 38300 }, { "epoch": 0.76, "learning_rate": 2.3805447120201323e-05, "loss": 0.0173, "step": 38400 }, { "epoch": 0.76, "learning_rate": 2.344145320247968e-05, "loss": 0.0189, "step": 38500 }, { "epoch": 0.76, "learning_rate": 2.3079818426877448e-05, "loss": 0.0183, "step": 38600 }, { "epoch": 0.76, "learning_rate": 2.2720556634894542e-05, "loss": 0.0161, "step": 38700 }, { "epoch": 0.76, "learning_rate": 2.2363681577205363e-05, "loss": 0.0161, "step": 38800 }, { "epoch": 0.77, "learning_rate": 2.2009206913132545e-05, "loss": 0.017, "step": 38900 }, { "epoch": 0.77, "learning_rate": 2.1657146210124072e-05, "loss": 0.0175, "step": 39000 }, { "epoch": 0.77, "learning_rate": 2.1307512943234157e-05, "loss": 0.0162, "step": 39100 }, { "epoch": 0.77, "learning_rate": 2.0960320494607276e-05, "loss": 0.0161, "step": 39200 }, { "epoch": 0.77, "learning_rate": 2.0615582152966087e-05, "loss": 0.0174, "step": 39300 }, { "epoch": 0.78, "learning_rate": 2.027672156726409e-05, "loss": 0.0194, "step": 39400 }, { "epoch": 0.78, "learning_rate": 1.9936906060978636e-05, "loss": 0.0185, "step": 39500 }, { "epoch": 0.78, "learning_rate": 1.9599583832664966e-05, "loss": 0.0169, "step": 39600 }, { "epoch": 0.78, "learning_rate": 1.9264767793265046e-05, "loss": 0.0174, "step": 39700 }, { "epoch": 0.78, "learning_rate": 1.8932470757796908e-05, "loss": 0.0172, "step": 39800 }, { "epoch": 0.79, "learning_rate": 1.86027054448642e-05, "loss": 0.0157, "step": 39900 }, { "epoch": 0.79, "learning_rate": 1.8275484476169405e-05, "loss": 0.0166, "step": 40000 } ], "logging_steps": 100, "max_steps": 50780, "num_train_epochs": 1, "save_steps": 10000, "total_flos": 1.5971349454848e+19, "trial_name": null, "trial_params": null }