{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 123732, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.979795040894837e-05, "loss": 2.8372, "step": 500 }, { "epoch": 0.02, "learning_rate": 4.9595900817896746e-05, "loss": 2.4789, "step": 1000 }, { "epoch": 0.04, "learning_rate": 4.9393851226845114e-05, "loss": 2.2893, "step": 1500 }, { "epoch": 0.05, "learning_rate": 4.9191801635793496e-05, "loss": 2.1965, "step": 2000 }, { "epoch": 0.06, "learning_rate": 4.8989752044741864e-05, "loss": 2.1373, "step": 2500 }, { "epoch": 0.07, "learning_rate": 4.878770245369024e-05, "loss": 2.0767, "step": 3000 }, { "epoch": 0.08, "learning_rate": 4.858565286263861e-05, "loss": 2.0443, "step": 3500 }, { "epoch": 0.1, "learning_rate": 4.838360327158698e-05, "loss": 2.024, "step": 4000 }, { "epoch": 0.11, "learning_rate": 4.818155368053535e-05, "loss": 1.9501, "step": 4500 }, { "epoch": 0.12, "learning_rate": 4.7979504089483726e-05, "loss": 1.9562, "step": 5000 }, { "epoch": 0.13, "learning_rate": 4.7777454498432095e-05, "loss": 1.9222, "step": 5500 }, { "epoch": 0.15, "learning_rate": 4.757540490738047e-05, "loss": 1.8926, "step": 6000 }, { "epoch": 0.16, "learning_rate": 4.7373355316328845e-05, "loss": 1.8646, "step": 6500 }, { "epoch": 0.17, "learning_rate": 4.717130572527721e-05, "loss": 1.8655, "step": 7000 }, { "epoch": 0.18, "learning_rate": 4.696925613422559e-05, "loss": 1.8596, "step": 7500 }, { "epoch": 0.19, "learning_rate": 4.676720654317396e-05, "loss": 1.8408, "step": 8000 }, { "epoch": 0.21, "learning_rate": 4.656515695212233e-05, "loss": 1.8267, "step": 8500 }, { "epoch": 0.22, "learning_rate": 4.63631073610707e-05, "loss": 1.7901, "step": 9000 }, { "epoch": 0.23, "learning_rate": 4.6161057770019075e-05, "loss": 1.7815, "step": 9500 }, { "epoch": 0.24, "learning_rate": 4.595900817896745e-05, "loss": 1.785, "step": 10000 }, { "epoch": 0.25, "learning_rate": 4.575695858791582e-05, "loss": 1.7634, "step": 10500 }, { "epoch": 0.27, "learning_rate": 4.5554908996864194e-05, "loss": 1.7426, "step": 11000 }, { "epoch": 0.28, "learning_rate": 4.535285940581256e-05, "loss": 1.7472, "step": 11500 }, { "epoch": 0.29, "learning_rate": 4.515080981476094e-05, "loss": 1.7534, "step": 12000 }, { "epoch": 0.3, "learning_rate": 4.4948760223709305e-05, "loss": 1.7196, "step": 12500 }, { "epoch": 0.32, "learning_rate": 4.474671063265768e-05, "loss": 1.7174, "step": 13000 }, { "epoch": 0.33, "learning_rate": 4.4544661041606056e-05, "loss": 1.7015, "step": 13500 }, { "epoch": 0.34, "learning_rate": 4.434261145055443e-05, "loss": 1.6568, "step": 14000 }, { "epoch": 0.35, "learning_rate": 4.41405618595028e-05, "loss": 1.6854, "step": 14500 }, { "epoch": 0.36, "learning_rate": 4.393851226845117e-05, "loss": 1.7124, "step": 15000 }, { "epoch": 0.38, "learning_rate": 4.373646267739954e-05, "loss": 1.6685, "step": 15500 }, { "epoch": 0.39, "learning_rate": 4.353441308634791e-05, "loss": 1.6657, "step": 16000 }, { "epoch": 0.4, "learning_rate": 4.3332363495296286e-05, "loss": 1.6733, "step": 16500 }, { "epoch": 0.41, "learning_rate": 4.313031390424466e-05, "loss": 1.6467, "step": 17000 }, { "epoch": 0.42, "learning_rate": 4.2928264313193036e-05, "loss": 1.6647, "step": 17500 }, { "epoch": 0.44, "learning_rate": 4.2726214722141404e-05, "loss": 1.6391, "step": 18000 }, { "epoch": 0.45, "learning_rate": 4.252416513108978e-05, "loss": 1.6388, "step": 18500 }, { "epoch": 0.46, "learning_rate": 4.232211554003815e-05, "loss": 1.6363, "step": 19000 }, { "epoch": 0.47, "learning_rate": 4.212006594898652e-05, "loss": 1.6375, "step": 19500 }, { "epoch": 0.48, "learning_rate": 4.191801635793489e-05, "loss": 1.6179, "step": 20000 }, { "epoch": 0.5, "learning_rate": 4.1715966766883266e-05, "loss": 1.6001, "step": 20500 }, { "epoch": 0.51, "learning_rate": 4.151391717583164e-05, "loss": 1.6116, "step": 21000 }, { "epoch": 0.52, "learning_rate": 4.131186758478001e-05, "loss": 1.592, "step": 21500 }, { "epoch": 0.53, "learning_rate": 4.1109817993728385e-05, "loss": 1.5869, "step": 22000 }, { "epoch": 0.55, "learning_rate": 4.090776840267675e-05, "loss": 1.5869, "step": 22500 }, { "epoch": 0.56, "learning_rate": 4.070571881162513e-05, "loss": 1.6034, "step": 23000 }, { "epoch": 0.57, "learning_rate": 4.05036692205735e-05, "loss": 1.5646, "step": 23500 }, { "epoch": 0.58, "learning_rate": 4.030161962952187e-05, "loss": 1.5619, "step": 24000 }, { "epoch": 0.59, "learning_rate": 4.009957003847025e-05, "loss": 1.5718, "step": 24500 }, { "epoch": 0.61, "learning_rate": 3.9897520447418615e-05, "loss": 1.5903, "step": 25000 }, { "epoch": 0.62, "learning_rate": 3.969547085636699e-05, "loss": 1.5563, "step": 25500 }, { "epoch": 0.63, "learning_rate": 3.949342126531536e-05, "loss": 1.5549, "step": 26000 }, { "epoch": 0.64, "learning_rate": 3.9291371674263734e-05, "loss": 1.5373, "step": 26500 }, { "epoch": 0.65, "learning_rate": 3.90893220832121e-05, "loss": 1.5607, "step": 27000 }, { "epoch": 0.67, "learning_rate": 3.888727249216048e-05, "loss": 1.5519, "step": 27500 }, { "epoch": 0.68, "learning_rate": 3.8685222901108846e-05, "loss": 1.5209, "step": 28000 }, { "epoch": 0.69, "learning_rate": 3.848317331005723e-05, "loss": 1.5354, "step": 28500 }, { "epoch": 0.7, "learning_rate": 3.8281123719005596e-05, "loss": 1.5082, "step": 29000 }, { "epoch": 0.72, "learning_rate": 3.807907412795397e-05, "loss": 1.506, "step": 29500 }, { "epoch": 0.73, "learning_rate": 3.787702453690234e-05, "loss": 1.5307, "step": 30000 }, { "epoch": 0.74, "learning_rate": 3.767497494585071e-05, "loss": 1.501, "step": 30500 }, { "epoch": 0.75, "learning_rate": 3.747292535479908e-05, "loss": 1.5334, "step": 31000 }, { "epoch": 0.76, "learning_rate": 3.727087576374745e-05, "loss": 1.513, "step": 31500 }, { "epoch": 0.78, "learning_rate": 3.706882617269583e-05, "loss": 1.5083, "step": 32000 }, { "epoch": 0.79, "learning_rate": 3.68667765816442e-05, "loss": 1.4999, "step": 32500 }, { "epoch": 0.8, "learning_rate": 3.6664726990592576e-05, "loss": 1.4755, "step": 33000 }, { "epoch": 0.81, "learning_rate": 3.6462677399540945e-05, "loss": 1.5159, "step": 33500 }, { "epoch": 0.82, "learning_rate": 3.626062780848932e-05, "loss": 1.5126, "step": 34000 }, { "epoch": 0.84, "learning_rate": 3.605857821743769e-05, "loss": 1.4948, "step": 34500 }, { "epoch": 0.85, "learning_rate": 3.585652862638606e-05, "loss": 1.4975, "step": 35000 }, { "epoch": 0.86, "learning_rate": 3.565447903533443e-05, "loss": 1.5006, "step": 35500 }, { "epoch": 0.87, "learning_rate": 3.5452429444282807e-05, "loss": 1.4772, "step": 36000 }, { "epoch": 0.88, "learning_rate": 3.525037985323118e-05, "loss": 1.4721, "step": 36500 }, { "epoch": 0.9, "learning_rate": 3.504833026217955e-05, "loss": 1.4934, "step": 37000 }, { "epoch": 0.91, "learning_rate": 3.4846280671127925e-05, "loss": 1.47, "step": 37500 }, { "epoch": 0.92, "learning_rate": 3.4644231080076293e-05, "loss": 1.471, "step": 38000 }, { "epoch": 0.93, "learning_rate": 3.444218148902467e-05, "loss": 1.4491, "step": 38500 }, { "epoch": 0.95, "learning_rate": 3.424013189797304e-05, "loss": 1.4718, "step": 39000 }, { "epoch": 0.96, "learning_rate": 3.403808230692141e-05, "loss": 1.4481, "step": 39500 }, { "epoch": 0.97, "learning_rate": 3.383603271586979e-05, "loss": 1.4609, "step": 40000 }, { "epoch": 0.98, "learning_rate": 3.3633983124818155e-05, "loss": 1.4501, "step": 40500 }, { "epoch": 0.99, "learning_rate": 3.343193353376653e-05, "loss": 1.4285, "step": 41000 }, { "epoch": 1.01, "learning_rate": 3.32298839427149e-05, "loss": 1.349, "step": 41500 }, { "epoch": 1.02, "learning_rate": 3.3027834351663274e-05, "loss": 1.3115, "step": 42000 }, { "epoch": 1.03, "learning_rate": 3.282578476061164e-05, "loss": 1.3064, "step": 42500 }, { "epoch": 1.04, "learning_rate": 3.262373516956002e-05, "loss": 1.2838, "step": 43000 }, { "epoch": 1.05, "learning_rate": 3.242168557850839e-05, "loss": 1.2855, "step": 43500 }, { "epoch": 1.07, "learning_rate": 3.221963598745677e-05, "loss": 1.2931, "step": 44000 }, { "epoch": 1.08, "learning_rate": 3.2017586396405136e-05, "loss": 1.2874, "step": 44500 }, { "epoch": 1.09, "learning_rate": 3.1815536805353504e-05, "loss": 1.2677, "step": 45000 }, { "epoch": 1.1, "learning_rate": 3.161348721430188e-05, "loss": 1.2898, "step": 45500 }, { "epoch": 1.12, "learning_rate": 3.141143762325025e-05, "loss": 1.2906, "step": 46000 }, { "epoch": 1.13, "learning_rate": 3.120938803219862e-05, "loss": 1.293, "step": 46500 }, { "epoch": 1.14, "learning_rate": 3.1007338441147e-05, "loss": 1.2844, "step": 47000 }, { "epoch": 1.15, "learning_rate": 3.080528885009537e-05, "loss": 1.295, "step": 47500 }, { "epoch": 1.16, "learning_rate": 3.060323925904374e-05, "loss": 1.3018, "step": 48000 }, { "epoch": 1.18, "learning_rate": 3.0401189667992113e-05, "loss": 1.29, "step": 48500 }, { "epoch": 1.19, "learning_rate": 3.0199140076940485e-05, "loss": 1.2879, "step": 49000 }, { "epoch": 1.2, "learning_rate": 2.9997090485888856e-05, "loss": 1.3044, "step": 49500 }, { "epoch": 1.21, "learning_rate": 2.9795040894837228e-05, "loss": 1.3007, "step": 50000 }, { "epoch": 1.22, "learning_rate": 2.95929913037856e-05, "loss": 1.2731, "step": 50500 }, { "epoch": 1.24, "learning_rate": 2.939094171273398e-05, "loss": 1.2825, "step": 51000 }, { "epoch": 1.25, "learning_rate": 2.918889212168235e-05, "loss": 1.2862, "step": 51500 }, { "epoch": 1.26, "learning_rate": 2.898684253063072e-05, "loss": 1.2784, "step": 52000 }, { "epoch": 1.27, "learning_rate": 2.878479293957909e-05, "loss": 1.2938, "step": 52500 }, { "epoch": 1.29, "learning_rate": 2.8582743348527462e-05, "loss": 1.2863, "step": 53000 }, { "epoch": 1.3, "learning_rate": 2.8380693757475834e-05, "loss": 1.2642, "step": 53500 }, { "epoch": 1.31, "learning_rate": 2.8178644166424205e-05, "loss": 1.2798, "step": 54000 }, { "epoch": 1.32, "learning_rate": 2.7976594575372584e-05, "loss": 1.2822, "step": 54500 }, { "epoch": 1.33, "learning_rate": 2.7774544984320956e-05, "loss": 1.2814, "step": 55000 }, { "epoch": 1.35, "learning_rate": 2.7572495393269327e-05, "loss": 1.2556, "step": 55500 }, { "epoch": 1.36, "learning_rate": 2.73704458022177e-05, "loss": 1.2496, "step": 56000 }, { "epoch": 1.37, "learning_rate": 2.716839621116607e-05, "loss": 1.2556, "step": 56500 }, { "epoch": 1.38, "learning_rate": 2.6966346620114442e-05, "loss": 1.267, "step": 57000 }, { "epoch": 1.39, "learning_rate": 2.6764297029062814e-05, "loss": 1.2647, "step": 57500 }, { "epoch": 1.41, "learning_rate": 2.6562247438011182e-05, "loss": 1.2617, "step": 58000 }, { "epoch": 1.42, "learning_rate": 2.636019784695956e-05, "loss": 1.2625, "step": 58500 }, { "epoch": 1.43, "learning_rate": 2.6158148255907933e-05, "loss": 1.267, "step": 59000 }, { "epoch": 1.44, "learning_rate": 2.5956098664856304e-05, "loss": 1.2708, "step": 59500 }, { "epoch": 1.45, "learning_rate": 2.5754049073804676e-05, "loss": 1.2623, "step": 60000 }, { "epoch": 1.47, "learning_rate": 2.5551999482753048e-05, "loss": 1.2569, "step": 60500 }, { "epoch": 1.48, "learning_rate": 2.534994989170142e-05, "loss": 1.2438, "step": 61000 }, { "epoch": 1.49, "learning_rate": 2.514790030064979e-05, "loss": 1.2695, "step": 61500 }, { "epoch": 1.5, "learning_rate": 2.4945850709598166e-05, "loss": 1.2551, "step": 62000 }, { "epoch": 1.52, "learning_rate": 2.4743801118546535e-05, "loss": 1.2633, "step": 62500 }, { "epoch": 1.53, "learning_rate": 2.454175152749491e-05, "loss": 1.2424, "step": 63000 }, { "epoch": 1.54, "learning_rate": 2.433970193644328e-05, "loss": 1.2659, "step": 63500 }, { "epoch": 1.55, "learning_rate": 2.4137652345391653e-05, "loss": 1.2639, "step": 64000 }, { "epoch": 1.56, "learning_rate": 2.3935602754340025e-05, "loss": 1.2428, "step": 64500 }, { "epoch": 1.58, "learning_rate": 2.37335531632884e-05, "loss": 1.2475, "step": 65000 }, { "epoch": 1.59, "learning_rate": 2.3531503572236772e-05, "loss": 1.2645, "step": 65500 }, { "epoch": 1.6, "learning_rate": 2.3329453981185143e-05, "loss": 1.2685, "step": 66000 }, { "epoch": 1.61, "learning_rate": 2.3127404390133515e-05, "loss": 1.2612, "step": 66500 }, { "epoch": 1.62, "learning_rate": 2.2925354799081887e-05, "loss": 1.2448, "step": 67000 }, { "epoch": 1.64, "learning_rate": 2.272330520803026e-05, "loss": 1.2503, "step": 67500 }, { "epoch": 1.65, "learning_rate": 2.252125561697863e-05, "loss": 1.2255, "step": 68000 }, { "epoch": 1.66, "learning_rate": 2.2319206025927005e-05, "loss": 1.2437, "step": 68500 }, { "epoch": 1.67, "learning_rate": 2.2117156434875377e-05, "loss": 1.2258, "step": 69000 }, { "epoch": 1.69, "learning_rate": 2.191510684382375e-05, "loss": 1.2203, "step": 69500 }, { "epoch": 1.7, "learning_rate": 2.171305725277212e-05, "loss": 1.2359, "step": 70000 }, { "epoch": 1.71, "learning_rate": 2.1511007661720496e-05, "loss": 1.2359, "step": 70500 }, { "epoch": 1.72, "learning_rate": 2.1308958070668867e-05, "loss": 1.239, "step": 71000 }, { "epoch": 1.73, "learning_rate": 2.110690847961724e-05, "loss": 1.21, "step": 71500 }, { "epoch": 1.75, "learning_rate": 2.090485888856561e-05, "loss": 1.1985, "step": 72000 }, { "epoch": 1.76, "learning_rate": 2.0702809297513983e-05, "loss": 1.2627, "step": 72500 }, { "epoch": 1.77, "learning_rate": 2.0500759706462354e-05, "loss": 1.2186, "step": 73000 }, { "epoch": 1.78, "learning_rate": 2.0298710115410726e-05, "loss": 1.2453, "step": 73500 }, { "epoch": 1.79, "learning_rate": 2.0096660524359098e-05, "loss": 1.227, "step": 74000 }, { "epoch": 1.81, "learning_rate": 1.9894610933307473e-05, "loss": 1.2383, "step": 74500 }, { "epoch": 1.82, "learning_rate": 1.9692561342255844e-05, "loss": 1.2272, "step": 75000 }, { "epoch": 1.83, "learning_rate": 1.9490511751204216e-05, "loss": 1.2031, "step": 75500 }, { "epoch": 1.84, "learning_rate": 1.928846216015259e-05, "loss": 1.2024, "step": 76000 }, { "epoch": 1.85, "learning_rate": 1.9086412569100963e-05, "loss": 1.2257, "step": 76500 }, { "epoch": 1.87, "learning_rate": 1.8884362978049335e-05, "loss": 1.2075, "step": 77000 }, { "epoch": 1.88, "learning_rate": 1.8682313386997703e-05, "loss": 1.2449, "step": 77500 }, { "epoch": 1.89, "learning_rate": 1.8480263795946078e-05, "loss": 1.2141, "step": 78000 }, { "epoch": 1.9, "learning_rate": 1.827821420489445e-05, "loss": 1.2161, "step": 78500 }, { "epoch": 1.92, "learning_rate": 1.807616461384282e-05, "loss": 1.2141, "step": 79000 }, { "epoch": 1.93, "learning_rate": 1.7874115022791193e-05, "loss": 1.1951, "step": 79500 }, { "epoch": 1.94, "learning_rate": 1.767206543173957e-05, "loss": 1.2105, "step": 80000 }, { "epoch": 1.95, "learning_rate": 1.747001584068794e-05, "loss": 1.187, "step": 80500 }, { "epoch": 1.96, "learning_rate": 1.7267966249636312e-05, "loss": 1.2099, "step": 81000 }, { "epoch": 1.98, "learning_rate": 1.7065916658584684e-05, "loss": 1.2091, "step": 81500 }, { "epoch": 1.99, "learning_rate": 1.6863867067533055e-05, "loss": 1.2034, "step": 82000 }, { "epoch": 2.0, "learning_rate": 1.6661817476481427e-05, "loss": 1.1969, "step": 82500 }, { "epoch": 2.01, "learning_rate": 1.64597678854298e-05, "loss": 1.0541, "step": 83000 }, { "epoch": 2.02, "learning_rate": 1.6257718294378174e-05, "loss": 1.0428, "step": 83500 }, { "epoch": 2.04, "learning_rate": 1.6055668703326546e-05, "loss": 1.0576, "step": 84000 }, { "epoch": 2.05, "learning_rate": 1.5853619112274917e-05, "loss": 1.0312, "step": 84500 }, { "epoch": 2.06, "learning_rate": 1.565156952122329e-05, "loss": 1.0415, "step": 85000 }, { "epoch": 2.07, "learning_rate": 1.5449519930171664e-05, "loss": 1.0685, "step": 85500 }, { "epoch": 2.09, "learning_rate": 1.5247470339120034e-05, "loss": 1.0584, "step": 86000 }, { "epoch": 2.1, "learning_rate": 1.5045420748068406e-05, "loss": 1.0444, "step": 86500 }, { "epoch": 2.11, "learning_rate": 1.4843371157016778e-05, "loss": 1.0456, "step": 87000 }, { "epoch": 2.12, "learning_rate": 1.4641321565965153e-05, "loss": 1.0597, "step": 87500 }, { "epoch": 2.13, "learning_rate": 1.4439271974913524e-05, "loss": 1.0505, "step": 88000 }, { "epoch": 2.15, "learning_rate": 1.4237222383861894e-05, "loss": 1.0605, "step": 88500 }, { "epoch": 2.16, "learning_rate": 1.4035172792810266e-05, "loss": 1.0581, "step": 89000 }, { "epoch": 2.17, "learning_rate": 1.3833123201758641e-05, "loss": 1.0357, "step": 89500 }, { "epoch": 2.18, "learning_rate": 1.3631073610707013e-05, "loss": 1.0408, "step": 90000 }, { "epoch": 2.19, "learning_rate": 1.3429024019655385e-05, "loss": 1.0404, "step": 90500 }, { "epoch": 2.21, "learning_rate": 1.3226974428603758e-05, "loss": 1.039, "step": 91000 }, { "epoch": 2.22, "learning_rate": 1.302492483755213e-05, "loss": 1.0434, "step": 91500 }, { "epoch": 2.23, "learning_rate": 1.2822875246500501e-05, "loss": 1.0348, "step": 92000 }, { "epoch": 2.24, "learning_rate": 1.2620825655448873e-05, "loss": 1.0526, "step": 92500 }, { "epoch": 2.25, "learning_rate": 1.2418776064397247e-05, "loss": 1.0395, "step": 93000 }, { "epoch": 2.27, "learning_rate": 1.2216726473345618e-05, "loss": 1.0409, "step": 93500 }, { "epoch": 2.28, "learning_rate": 1.201467688229399e-05, "loss": 1.0396, "step": 94000 }, { "epoch": 2.29, "learning_rate": 1.1812627291242363e-05, "loss": 1.055, "step": 94500 }, { "epoch": 2.3, "learning_rate": 1.1610577700190735e-05, "loss": 1.0594, "step": 95000 }, { "epoch": 2.32, "learning_rate": 1.1408528109139109e-05, "loss": 1.04, "step": 95500 }, { "epoch": 2.33, "learning_rate": 1.1206478518087479e-05, "loss": 1.0421, "step": 96000 }, { "epoch": 2.34, "learning_rate": 1.1004428927035852e-05, "loss": 1.0507, "step": 96500 }, { "epoch": 2.35, "learning_rate": 1.0802379335984225e-05, "loss": 1.0275, "step": 97000 }, { "epoch": 2.36, "learning_rate": 1.0600329744932597e-05, "loss": 1.0356, "step": 97500 }, { "epoch": 2.38, "learning_rate": 1.0398280153880969e-05, "loss": 1.0379, "step": 98000 }, { "epoch": 2.39, "learning_rate": 1.019623056282934e-05, "loss": 1.0256, "step": 98500 }, { "epoch": 2.4, "learning_rate": 9.994180971777714e-06, "loss": 1.0462, "step": 99000 }, { "epoch": 2.41, "learning_rate": 9.792131380726086e-06, "loss": 1.0387, "step": 99500 }, { "epoch": 2.42, "learning_rate": 9.590081789674459e-06, "loss": 1.0175, "step": 100000 }, { "epoch": 2.44, "learning_rate": 9.38803219862283e-06, "loss": 1.0262, "step": 100500 }, { "epoch": 2.45, "learning_rate": 9.185982607571203e-06, "loss": 1.0545, "step": 101000 }, { "epoch": 2.46, "learning_rate": 8.983933016519574e-06, "loss": 1.0339, "step": 101500 }, { "epoch": 2.47, "learning_rate": 8.781883425467948e-06, "loss": 1.0378, "step": 102000 }, { "epoch": 2.49, "learning_rate": 8.57983383441632e-06, "loss": 1.034, "step": 102500 }, { "epoch": 2.5, "learning_rate": 8.377784243364691e-06, "loss": 1.0412, "step": 103000 }, { "epoch": 2.51, "learning_rate": 8.175734652313063e-06, "loss": 1.0321, "step": 103500 }, { "epoch": 2.52, "learning_rate": 7.973685061261436e-06, "loss": 1.0341, "step": 104000 }, { "epoch": 2.53, "learning_rate": 7.77163547020981e-06, "loss": 1.0207, "step": 104500 }, { "epoch": 2.55, "learning_rate": 7.5695858791581805e-06, "loss": 1.0017, "step": 105000 }, { "epoch": 2.56, "learning_rate": 7.367536288106554e-06, "loss": 1.0238, "step": 105500 }, { "epoch": 2.57, "learning_rate": 7.165486697054926e-06, "loss": 1.0219, "step": 106000 }, { "epoch": 2.58, "learning_rate": 6.963437106003298e-06, "loss": 1.0153, "step": 106500 }, { "epoch": 2.59, "learning_rate": 6.76138751495167e-06, "loss": 1.0296, "step": 107000 }, { "epoch": 2.61, "learning_rate": 6.5593379239000424e-06, "loss": 1.0198, "step": 107500 }, { "epoch": 2.62, "learning_rate": 6.357288332848414e-06, "loss": 1.0137, "step": 108000 }, { "epoch": 2.63, "learning_rate": 6.1552387417967876e-06, "loss": 1.0256, "step": 108500 }, { "epoch": 2.64, "learning_rate": 5.953189150745159e-06, "loss": 1.0119, "step": 109000 }, { "epoch": 2.65, "learning_rate": 5.751139559693532e-06, "loss": 1.0177, "step": 109500 }, { "epoch": 2.67, "learning_rate": 5.5490899686419036e-06, "loss": 1.0131, "step": 110000 }, { "epoch": 2.68, "learning_rate": 5.347040377590276e-06, "loss": 0.9982, "step": 110500 }, { "epoch": 2.69, "learning_rate": 5.144990786538649e-06, "loss": 1.0195, "step": 111000 }, { "epoch": 2.7, "learning_rate": 4.94294119548702e-06, "loss": 1.0126, "step": 111500 }, { "epoch": 2.72, "learning_rate": 4.740891604435393e-06, "loss": 1.0218, "step": 112000 }, { "epoch": 2.73, "learning_rate": 4.538842013383765e-06, "loss": 1.0432, "step": 112500 }, { "epoch": 2.74, "learning_rate": 4.336792422332137e-06, "loss": 1.0058, "step": 113000 }, { "epoch": 2.75, "learning_rate": 4.13474283128051e-06, "loss": 1.0041, "step": 113500 }, { "epoch": 2.76, "learning_rate": 3.9326932402288815e-06, "loss": 1.0208, "step": 114000 }, { "epoch": 2.78, "learning_rate": 3.730643649177254e-06, "loss": 0.9971, "step": 114500 }, { "epoch": 2.79, "learning_rate": 3.5285940581256262e-06, "loss": 1.0126, "step": 115000 }, { "epoch": 2.8, "learning_rate": 3.3265444670739992e-06, "loss": 1.0259, "step": 115500 }, { "epoch": 2.81, "learning_rate": 3.124494876022371e-06, "loss": 1.0374, "step": 116000 }, { "epoch": 2.82, "learning_rate": 2.922445284970743e-06, "loss": 1.025, "step": 116500 }, { "epoch": 2.84, "learning_rate": 2.7203956939191156e-06, "loss": 1.0199, "step": 117000 }, { "epoch": 2.85, "learning_rate": 2.518346102867488e-06, "loss": 1.0116, "step": 117500 }, { "epoch": 2.86, "learning_rate": 2.3162965118158603e-06, "loss": 1.0321, "step": 118000 }, { "epoch": 2.87, "learning_rate": 2.1142469207642325e-06, "loss": 1.0162, "step": 118500 }, { "epoch": 2.89, "learning_rate": 1.9121973297126046e-06, "loss": 1.0218, "step": 119000 }, { "epoch": 2.9, "learning_rate": 1.710147738660977e-06, "loss": 0.9819, "step": 119500 }, { "epoch": 2.91, "learning_rate": 1.5080981476093493e-06, "loss": 1.0366, "step": 120000 }, { "epoch": 2.92, "learning_rate": 1.3060485565577217e-06, "loss": 0.9988, "step": 120500 }, { "epoch": 2.93, "learning_rate": 1.1039989655060938e-06, "loss": 1.0048, "step": 121000 }, { "epoch": 2.95, "learning_rate": 9.019493744544661e-07, "loss": 1.0015, "step": 121500 }, { "epoch": 2.96, "learning_rate": 6.998997834028384e-07, "loss": 0.9998, "step": 122000 }, { "epoch": 2.97, "learning_rate": 4.978501923512107e-07, "loss": 1.0045, "step": 122500 }, { "epoch": 2.98, "learning_rate": 2.95800601299583e-07, "loss": 1.0024, "step": 123000 }, { "epoch": 2.99, "learning_rate": 9.375101024795526e-08, "loss": 0.995, "step": 123500 }, { "epoch": 3.0, "step": 123732, "total_flos": 1.8552133900202803e+17, "train_loss": 1.3215980781360575, "train_runtime": 31676.1684, "train_samples_per_second": 39.061, "train_steps_per_second": 3.906 } ], "max_steps": 123732, "num_train_epochs": 3, "total_flos": 1.8552133900202803e+17, "trial_name": null, "trial_params": null }