{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.3938558487593541, "eval_steps": 500, "global_step": 20000, "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 } ], "logging_steps": 100, "max_steps": 50780, "num_train_epochs": 1, "save_steps": 10000, "total_flos": 7.985674727424e+18, "trial_name": null, "trial_params": null }