{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "global_step": 134942, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.981473521957582e-05, "loss": 2.1777, "step": 500 }, { "epoch": 0.01, "learning_rate": 4.962947043915164e-05, "loss": 2.1697, "step": 1000 }, { "epoch": 0.02, "learning_rate": 4.944420565872746e-05, "loss": 2.208, "step": 1500 }, { "epoch": 0.03, "learning_rate": 4.9258940878303276e-05, "loss": 2.206, "step": 2000 }, { "epoch": 0.04, "learning_rate": 4.907367609787909e-05, "loss": 2.2025, "step": 2500 }, { "epoch": 0.04, "learning_rate": 4.8888411317454905e-05, "loss": 2.2037, "step": 3000 }, { "epoch": 0.05, "learning_rate": 4.8703146537030724e-05, "loss": 2.1928, "step": 3500 }, { "epoch": 0.06, "learning_rate": 4.851788175660654e-05, "loss": 2.2034, "step": 4000 }, { "epoch": 0.07, "learning_rate": 4.833261697618236e-05, "loss": 2.1958, "step": 4500 }, { "epoch": 0.07, "learning_rate": 4.814735219575818e-05, "loss": 2.221, "step": 5000 }, { "epoch": 0.08, "learning_rate": 4.7962087415334e-05, "loss": 2.2158, "step": 5500 }, { "epoch": 0.09, "learning_rate": 4.7776822634909815e-05, "loss": 2.206, "step": 6000 }, { "epoch": 0.1, "learning_rate": 4.7591557854485634e-05, "loss": 2.1996, "step": 6500 }, { "epoch": 0.1, "learning_rate": 4.740629307406145e-05, "loss": 2.21, "step": 7000 }, { "epoch": 0.11, "learning_rate": 4.722102829363727e-05, "loss": 2.2156, "step": 7500 }, { "epoch": 0.12, "learning_rate": 4.703576351321309e-05, "loss": 2.2285, "step": 8000 }, { "epoch": 0.13, "learning_rate": 4.685049873278891e-05, "loss": 2.2124, "step": 8500 }, { "epoch": 0.13, "learning_rate": 4.6665233952364725e-05, "loss": 2.2073, "step": 9000 }, { "epoch": 0.14, "learning_rate": 4.6479969171940543e-05, "loss": 2.1891, "step": 9500 }, { "epoch": 0.15, "learning_rate": 4.6294704391516355e-05, "loss": 2.1992, "step": 10000 }, { "epoch": 0.16, "learning_rate": 4.610943961109217e-05, "loss": 2.1831, "step": 10500 }, { "epoch": 0.16, "learning_rate": 4.592417483066799e-05, "loss": 2.2132, "step": 11000 }, { "epoch": 0.17, "learning_rate": 4.573891005024381e-05, "loss": 2.2117, "step": 11500 }, { "epoch": 0.18, "learning_rate": 4.555364526981963e-05, "loss": 2.1857, "step": 12000 }, { "epoch": 0.19, "learning_rate": 4.5368380489395447e-05, "loss": 2.2008, "step": 12500 }, { "epoch": 0.19, "learning_rate": 4.5183115708971265e-05, "loss": 2.1957, "step": 13000 }, { "epoch": 0.2, "learning_rate": 4.499785092854708e-05, "loss": 2.212, "step": 13500 }, { "epoch": 0.21, "learning_rate": 4.48125861481229e-05, "loss": 2.2022, "step": 14000 }, { "epoch": 0.21, "learning_rate": 4.462732136769871e-05, "loss": 2.1796, "step": 14500 }, { "epoch": 0.22, "learning_rate": 4.444205658727453e-05, "loss": 2.1775, "step": 15000 }, { "epoch": 0.23, "learning_rate": 4.425679180685035e-05, "loss": 2.2016, "step": 15500 }, { "epoch": 0.24, "learning_rate": 4.407152702642617e-05, "loss": 2.1926, "step": 16000 }, { "epoch": 0.24, "learning_rate": 4.3886262246001986e-05, "loss": 2.2146, "step": 16500 }, { "epoch": 0.25, "learning_rate": 4.3700997465577804e-05, "loss": 2.2015, "step": 17000 }, { "epoch": 0.26, "learning_rate": 4.351573268515363e-05, "loss": 2.1924, "step": 17500 }, { "epoch": 0.27, "learning_rate": 4.333046790472944e-05, "loss": 2.2027, "step": 18000 }, { "epoch": 0.27, "learning_rate": 4.314520312430526e-05, "loss": 2.1899, "step": 18500 }, { "epoch": 0.28, "learning_rate": 4.295993834388108e-05, "loss": 2.1938, "step": 19000 }, { "epoch": 0.29, "learning_rate": 4.2774673563456896e-05, "loss": 2.1898, "step": 19500 }, { "epoch": 0.3, "learning_rate": 4.2589408783032714e-05, "loss": 2.1971, "step": 20000 }, { "epoch": 0.3, "learning_rate": 4.240414400260853e-05, "loss": 2.1934, "step": 20500 }, { "epoch": 0.31, "learning_rate": 4.221887922218435e-05, "loss": 2.1759, "step": 21000 }, { "epoch": 0.32, "learning_rate": 4.203361444176017e-05, "loss": 2.1806, "step": 21500 }, { "epoch": 0.33, "learning_rate": 4.184834966133598e-05, "loss": 2.1956, "step": 22000 }, { "epoch": 0.33, "learning_rate": 4.16630848809118e-05, "loss": 2.2144, "step": 22500 }, { "epoch": 0.34, "learning_rate": 4.147782010048762e-05, "loss": 2.2047, "step": 23000 }, { "epoch": 0.35, "learning_rate": 4.1292555320063436e-05, "loss": 2.1988, "step": 23500 }, { "epoch": 0.36, "learning_rate": 4.1107290539639254e-05, "loss": 2.209, "step": 24000 }, { "epoch": 0.36, "learning_rate": 4.092202575921507e-05, "loss": 2.1929, "step": 24500 }, { "epoch": 0.37, "learning_rate": 4.073676097879089e-05, "loss": 2.1703, "step": 25000 }, { "epoch": 0.38, "learning_rate": 4.055149619836671e-05, "loss": 2.2071, "step": 25500 }, { "epoch": 0.39, "learning_rate": 4.036623141794252e-05, "loss": 2.1894, "step": 26000 }, { "epoch": 0.39, "learning_rate": 4.018096663751834e-05, "loss": 2.2136, "step": 26500 }, { "epoch": 0.4, "learning_rate": 3.999570185709416e-05, "loss": 2.2125, "step": 27000 }, { "epoch": 0.41, "learning_rate": 3.9810437076669975e-05, "loss": 2.1986, "step": 27500 }, { "epoch": 0.41, "learning_rate": 3.9625172296245794e-05, "loss": 2.203, "step": 28000 }, { "epoch": 0.42, "learning_rate": 3.943990751582161e-05, "loss": 2.1836, "step": 28500 }, { "epoch": 0.43, "learning_rate": 3.925464273539744e-05, "loss": 2.1815, "step": 29000 }, { "epoch": 0.44, "learning_rate": 3.906937795497325e-05, "loss": 2.2034, "step": 29500 }, { "epoch": 0.44, "learning_rate": 3.888411317454907e-05, "loss": 2.1769, "step": 30000 }, { "epoch": 0.45, "learning_rate": 3.8698848394124885e-05, "loss": 2.1994, "step": 30500 }, { "epoch": 0.46, "learning_rate": 3.8513583613700704e-05, "loss": 2.1856, "step": 31000 }, { "epoch": 0.47, "learning_rate": 3.832831883327652e-05, "loss": 2.1853, "step": 31500 }, { "epoch": 0.47, "learning_rate": 3.814305405285234e-05, "loss": 2.1941, "step": 32000 }, { "epoch": 0.48, "learning_rate": 3.795778927242816e-05, "loss": 2.2018, "step": 32500 }, { "epoch": 0.49, "learning_rate": 3.777252449200398e-05, "loss": 2.2126, "step": 33000 }, { "epoch": 0.5, "learning_rate": 3.7587259711579795e-05, "loss": 2.2037, "step": 33500 }, { "epoch": 0.5, "learning_rate": 3.7401994931155607e-05, "loss": 2.1956, "step": 34000 }, { "epoch": 0.51, "learning_rate": 3.7216730150731425e-05, "loss": 2.1852, "step": 34500 }, { "epoch": 0.52, "learning_rate": 3.703146537030724e-05, "loss": 2.1795, "step": 35000 }, { "epoch": 0.53, "learning_rate": 3.684620058988306e-05, "loss": 2.2005, "step": 35500 }, { "epoch": 0.53, "learning_rate": 3.666093580945888e-05, "loss": 2.1901, "step": 36000 }, { "epoch": 0.54, "learning_rate": 3.64756710290347e-05, "loss": 2.1844, "step": 36500 }, { "epoch": 0.55, "learning_rate": 3.6290406248610516e-05, "loss": 2.1723, "step": 37000 }, { "epoch": 0.56, "learning_rate": 3.6105141468186335e-05, "loss": 2.1801, "step": 37500 }, { "epoch": 0.56, "learning_rate": 3.5919876687762146e-05, "loss": 2.1857, "step": 38000 }, { "epoch": 0.57, "learning_rate": 3.5734611907337965e-05, "loss": 2.1781, "step": 38500 }, { "epoch": 0.58, "learning_rate": 3.554934712691378e-05, "loss": 2.1711, "step": 39000 }, { "epoch": 0.59, "learning_rate": 3.53640823464896e-05, "loss": 2.1726, "step": 39500 }, { "epoch": 0.59, "learning_rate": 3.5178817566065426e-05, "loss": 2.2039, "step": 40000 }, { "epoch": 0.6, "learning_rate": 3.4993552785641245e-05, "loss": 2.1792, "step": 40500 }, { "epoch": 0.61, "learning_rate": 3.480828800521706e-05, "loss": 2.1813, "step": 41000 }, { "epoch": 0.62, "learning_rate": 3.4623023224792874e-05, "loss": 2.1731, "step": 41500 }, { "epoch": 0.62, "learning_rate": 3.443775844436869e-05, "loss": 2.1821, "step": 42000 }, { "epoch": 0.63, "learning_rate": 3.425249366394451e-05, "loss": 2.1767, "step": 42500 }, { "epoch": 0.64, "learning_rate": 3.406722888352033e-05, "loss": 2.1693, "step": 43000 }, { "epoch": 0.64, "learning_rate": 3.388196410309615e-05, "loss": 2.1945, "step": 43500 }, { "epoch": 0.65, "learning_rate": 3.3696699322671966e-05, "loss": 2.1872, "step": 44000 }, { "epoch": 0.66, "learning_rate": 3.3511434542247784e-05, "loss": 2.168, "step": 44500 }, { "epoch": 0.67, "learning_rate": 3.33261697618236e-05, "loss": 2.1672, "step": 45000 }, { "epoch": 0.67, "learning_rate": 3.314090498139942e-05, "loss": 2.1779, "step": 45500 }, { "epoch": 0.68, "learning_rate": 3.295564020097523e-05, "loss": 2.1873, "step": 46000 }, { "epoch": 0.69, "learning_rate": 3.277037542055105e-05, "loss": 2.1972, "step": 46500 }, { "epoch": 0.7, "learning_rate": 3.258511064012687e-05, "loss": 2.1923, "step": 47000 }, { "epoch": 0.7, "learning_rate": 3.239984585970269e-05, "loss": 2.1494, "step": 47500 }, { "epoch": 0.71, "learning_rate": 3.2214581079278506e-05, "loss": 2.1825, "step": 48000 }, { "epoch": 0.72, "learning_rate": 3.2029316298854324e-05, "loss": 2.1538, "step": 48500 }, { "epoch": 0.73, "learning_rate": 3.184405151843014e-05, "loss": 2.1737, "step": 49000 }, { "epoch": 0.73, "learning_rate": 3.165878673800596e-05, "loss": 2.194, "step": 49500 }, { "epoch": 0.74, "learning_rate": 3.147352195758177e-05, "loss": 2.1813, "step": 50000 }, { "epoch": 0.75, "learning_rate": 3.128825717715759e-05, "loss": 2.1702, "step": 50500 }, { "epoch": 0.76, "learning_rate": 3.110299239673341e-05, "loss": 2.1708, "step": 51000 }, { "epoch": 0.76, "learning_rate": 3.0917727616309234e-05, "loss": 2.1997, "step": 51500 }, { "epoch": 0.77, "learning_rate": 3.073246283588505e-05, "loss": 2.1953, "step": 52000 }, { "epoch": 0.78, "learning_rate": 3.054719805546087e-05, "loss": 2.1743, "step": 52500 }, { "epoch": 0.79, "learning_rate": 3.0361933275036685e-05, "loss": 2.1933, "step": 53000 }, { "epoch": 0.79, "learning_rate": 3.0176668494612504e-05, "loss": 2.1624, "step": 53500 }, { "epoch": 0.8, "learning_rate": 2.9991403714188322e-05, "loss": 2.1986, "step": 54000 }, { "epoch": 0.81, "learning_rate": 2.9806138933764137e-05, "loss": 2.159, "step": 54500 }, { "epoch": 0.82, "learning_rate": 2.9620874153339955e-05, "loss": 2.187, "step": 55000 }, { "epoch": 0.82, "learning_rate": 2.9435609372915773e-05, "loss": 2.1734, "step": 55500 }, { "epoch": 0.83, "learning_rate": 2.9250344592491592e-05, "loss": 2.1653, "step": 56000 }, { "epoch": 0.84, "learning_rate": 2.9065079812067407e-05, "loss": 2.176, "step": 56500 }, { "epoch": 0.84, "learning_rate": 2.8879815031643225e-05, "loss": 2.1778, "step": 57000 }, { "epoch": 0.85, "learning_rate": 2.8694550251219043e-05, "loss": 2.1855, "step": 57500 }, { "epoch": 0.86, "learning_rate": 2.850928547079486e-05, "loss": 2.1771, "step": 58000 }, { "epoch": 0.87, "learning_rate": 2.832402069037068e-05, "loss": 2.1719, "step": 58500 }, { "epoch": 0.87, "learning_rate": 2.8138755909946495e-05, "loss": 2.1632, "step": 59000 }, { "epoch": 0.88, "learning_rate": 2.7953491129522313e-05, "loss": 2.1678, "step": 59500 }, { "epoch": 0.89, "learning_rate": 2.776822634909813e-05, "loss": 2.1671, "step": 60000 }, { "epoch": 0.9, "learning_rate": 2.758296156867395e-05, "loss": 2.1724, "step": 60500 }, { "epoch": 0.9, "learning_rate": 2.7397696788249765e-05, "loss": 2.1716, "step": 61000 }, { "epoch": 0.91, "learning_rate": 2.7212432007825583e-05, "loss": 2.1606, "step": 61500 }, { "epoch": 0.92, "learning_rate": 2.70271672274014e-05, "loss": 2.1842, "step": 62000 }, { "epoch": 0.93, "learning_rate": 2.684190244697722e-05, "loss": 2.1652, "step": 62500 }, { "epoch": 0.93, "learning_rate": 2.665663766655304e-05, "loss": 2.1792, "step": 63000 }, { "epoch": 0.94, "learning_rate": 2.647137288612886e-05, "loss": 2.1641, "step": 63500 }, { "epoch": 0.95, "learning_rate": 2.6286108105704678e-05, "loss": 2.1609, "step": 64000 }, { "epoch": 0.96, "learning_rate": 2.6100843325280493e-05, "loss": 2.1632, "step": 64500 }, { "epoch": 0.96, "learning_rate": 2.591557854485631e-05, "loss": 2.1831, "step": 65000 }, { "epoch": 0.97, "learning_rate": 2.573031376443213e-05, "loss": 2.1679, "step": 65500 }, { "epoch": 0.98, "learning_rate": 2.5545048984007948e-05, "loss": 2.1602, "step": 66000 }, { "epoch": 0.99, "learning_rate": 2.5359784203583763e-05, "loss": 2.1663, "step": 66500 }, { "epoch": 0.99, "learning_rate": 2.517451942315958e-05, "loss": 2.1854, "step": 67000 }, { "epoch": 1.0, "learning_rate": 2.49892546427354e-05, "loss": 2.165, "step": 67500 }, { "epoch": 1.01, "learning_rate": 2.4803989862311218e-05, "loss": 2.007, "step": 68000 }, { "epoch": 1.02, "learning_rate": 2.4618725081887032e-05, "loss": 2.021, "step": 68500 }, { "epoch": 1.02, "learning_rate": 2.443346030146285e-05, "loss": 2.0256, "step": 69000 }, { "epoch": 1.03, "learning_rate": 2.424819552103867e-05, "loss": 2.0075, "step": 69500 }, { "epoch": 1.04, "learning_rate": 2.4062930740614487e-05, "loss": 2.0186, "step": 70000 }, { "epoch": 1.04, "learning_rate": 2.3877665960190306e-05, "loss": 2.0247, "step": 70500 }, { "epoch": 1.05, "learning_rate": 2.369240117976612e-05, "loss": 2.0036, "step": 71000 }, { "epoch": 1.06, "learning_rate": 2.3507136399341942e-05, "loss": 2.0254, "step": 71500 }, { "epoch": 1.07, "learning_rate": 2.332187161891776e-05, "loss": 2.0216, "step": 72000 }, { "epoch": 1.07, "learning_rate": 2.3136606838493576e-05, "loss": 2.0236, "step": 72500 }, { "epoch": 1.08, "learning_rate": 2.2951342058069394e-05, "loss": 2.0131, "step": 73000 }, { "epoch": 1.09, "learning_rate": 2.2766077277645212e-05, "loss": 2.0136, "step": 73500 }, { "epoch": 1.1, "learning_rate": 2.258081249722103e-05, "loss": 2.0133, "step": 74000 }, { "epoch": 1.1, "learning_rate": 2.2395547716796845e-05, "loss": 2.017, "step": 74500 }, { "epoch": 1.11, "learning_rate": 2.2210282936372664e-05, "loss": 2.0366, "step": 75000 }, { "epoch": 1.12, "learning_rate": 2.2025018155948482e-05, "loss": 2.0203, "step": 75500 }, { "epoch": 1.13, "learning_rate": 2.18397533755243e-05, "loss": 2.0226, "step": 76000 }, { "epoch": 1.13, "learning_rate": 2.165448859510012e-05, "loss": 2.0357, "step": 76500 }, { "epoch": 1.14, "learning_rate": 2.1469223814675937e-05, "loss": 2.0249, "step": 77000 }, { "epoch": 1.15, "learning_rate": 2.1283959034251755e-05, "loss": 2.0487, "step": 77500 }, { "epoch": 1.16, "learning_rate": 2.1098694253827574e-05, "loss": 2.0277, "step": 78000 }, { "epoch": 1.16, "learning_rate": 2.091342947340339e-05, "loss": 2.0468, "step": 78500 }, { "epoch": 1.17, "learning_rate": 2.0728164692979207e-05, "loss": 2.035, "step": 79000 }, { "epoch": 1.18, "learning_rate": 2.0542899912555025e-05, "loss": 2.0286, "step": 79500 }, { "epoch": 1.19, "learning_rate": 2.0357635132130843e-05, "loss": 2.0409, "step": 80000 }, { "epoch": 1.19, "learning_rate": 2.0172370351706658e-05, "loss": 2.024, "step": 80500 }, { "epoch": 1.2, "learning_rate": 1.9987105571282477e-05, "loss": 2.0153, "step": 81000 }, { "epoch": 1.21, "learning_rate": 1.9801840790858295e-05, "loss": 2.0234, "step": 81500 }, { "epoch": 1.22, "learning_rate": 1.9616576010434113e-05, "loss": 2.0282, "step": 82000 }, { "epoch": 1.22, "learning_rate": 1.9431311230009928e-05, "loss": 2.025, "step": 82500 }, { "epoch": 1.23, "learning_rate": 1.924604644958575e-05, "loss": 2.0451, "step": 83000 }, { "epoch": 1.24, "learning_rate": 1.9060781669161568e-05, "loss": 2.0347, "step": 83500 }, { "epoch": 1.24, "learning_rate": 1.8875516888737386e-05, "loss": 2.0412, "step": 84000 }, { "epoch": 1.25, "learning_rate": 1.86902521083132e-05, "loss": 2.0236, "step": 84500 }, { "epoch": 1.26, "learning_rate": 1.850498732788902e-05, "loss": 2.0277, "step": 85000 }, { "epoch": 1.27, "learning_rate": 1.8319722547464838e-05, "loss": 2.0431, "step": 85500 }, { "epoch": 1.27, "learning_rate": 1.8134457767040656e-05, "loss": 2.0354, "step": 86000 }, { "epoch": 1.28, "learning_rate": 1.794919298661647e-05, "loss": 2.0388, "step": 86500 }, { "epoch": 1.29, "learning_rate": 1.776392820619229e-05, "loss": 2.0355, "step": 87000 }, { "epoch": 1.3, "learning_rate": 1.7578663425768108e-05, "loss": 2.0239, "step": 87500 }, { "epoch": 1.3, "learning_rate": 1.7393398645343926e-05, "loss": 2.0209, "step": 88000 }, { "epoch": 1.31, "learning_rate": 1.7208133864919744e-05, "loss": 2.0399, "step": 88500 }, { "epoch": 1.32, "learning_rate": 1.7022869084495563e-05, "loss": 2.0427, "step": 89000 }, { "epoch": 1.33, "learning_rate": 1.683760430407138e-05, "loss": 2.0448, "step": 89500 }, { "epoch": 1.33, "learning_rate": 1.66523395236472e-05, "loss": 2.0358, "step": 90000 }, { "epoch": 1.34, "learning_rate": 1.6467074743223014e-05, "loss": 2.0449, "step": 90500 }, { "epoch": 1.35, "learning_rate": 1.6281809962798833e-05, "loss": 2.0267, "step": 91000 }, { "epoch": 1.36, "learning_rate": 1.609654518237465e-05, "loss": 2.0233, "step": 91500 }, { "epoch": 1.36, "learning_rate": 1.591128040195047e-05, "loss": 2.0241, "step": 92000 }, { "epoch": 1.37, "learning_rate": 1.5726015621526284e-05, "loss": 2.0344, "step": 92500 }, { "epoch": 1.38, "learning_rate": 1.5540750841102102e-05, "loss": 2.0414, "step": 93000 }, { "epoch": 1.39, "learning_rate": 1.535548606067792e-05, "loss": 2.0423, "step": 93500 }, { "epoch": 1.39, "learning_rate": 1.5170221280253737e-05, "loss": 2.0231, "step": 94000 }, { "epoch": 1.4, "learning_rate": 1.4984956499829559e-05, "loss": 2.0444, "step": 94500 }, { "epoch": 1.41, "learning_rate": 1.4799691719405376e-05, "loss": 2.0327, "step": 95000 }, { "epoch": 1.42, "learning_rate": 1.4614426938981194e-05, "loss": 2.0347, "step": 95500 }, { "epoch": 1.42, "learning_rate": 1.442916215855701e-05, "loss": 2.0287, "step": 96000 }, { "epoch": 1.43, "learning_rate": 1.4243897378132829e-05, "loss": 2.0359, "step": 96500 }, { "epoch": 1.44, "learning_rate": 1.4058632597708645e-05, "loss": 2.0329, "step": 97000 }, { "epoch": 1.45, "learning_rate": 1.3873367817284464e-05, "loss": 2.0221, "step": 97500 }, { "epoch": 1.45, "learning_rate": 1.368810303686028e-05, "loss": 2.0234, "step": 98000 }, { "epoch": 1.46, "learning_rate": 1.3502838256436099e-05, "loss": 2.0478, "step": 98500 }, { "epoch": 1.47, "learning_rate": 1.3317573476011915e-05, "loss": 2.0401, "step": 99000 }, { "epoch": 1.47, "learning_rate": 1.3132308695587734e-05, "loss": 2.0402, "step": 99500 }, { "epoch": 1.48, "learning_rate": 1.2947043915163554e-05, "loss": 2.0335, "step": 100000 }, { "epoch": 1.49, "learning_rate": 1.2761779134739372e-05, "loss": 2.0453, "step": 100500 }, { "epoch": 1.5, "learning_rate": 1.2576514354315189e-05, "loss": 2.0324, "step": 101000 }, { "epoch": 1.5, "learning_rate": 1.2391249573891007e-05, "loss": 2.0247, "step": 101500 }, { "epoch": 1.51, "learning_rate": 1.2205984793466823e-05, "loss": 2.0301, "step": 102000 }, { "epoch": 1.52, "learning_rate": 1.2020720013042642e-05, "loss": 2.0293, "step": 102500 }, { "epoch": 1.53, "learning_rate": 1.1835455232618458e-05, "loss": 2.0333, "step": 103000 }, { "epoch": 1.53, "learning_rate": 1.1650190452194277e-05, "loss": 2.0599, "step": 103500 }, { "epoch": 1.54, "learning_rate": 1.1464925671770093e-05, "loss": 2.0278, "step": 104000 }, { "epoch": 1.55, "learning_rate": 1.1279660891345913e-05, "loss": 2.039, "step": 104500 }, { "epoch": 1.56, "learning_rate": 1.109439611092173e-05, "loss": 2.042, "step": 105000 }, { "epoch": 1.56, "learning_rate": 1.0909131330497548e-05, "loss": 2.0592, "step": 105500 }, { "epoch": 1.57, "learning_rate": 1.0723866550073365e-05, "loss": 2.0512, "step": 106000 }, { "epoch": 1.58, "learning_rate": 1.0538601769649183e-05, "loss": 2.0496, "step": 106500 }, { "epoch": 1.59, "learning_rate": 1.0353336989225001e-05, "loss": 2.0359, "step": 107000 }, { "epoch": 1.59, "learning_rate": 1.016807220880082e-05, "loss": 2.0427, "step": 107500 }, { "epoch": 1.6, "learning_rate": 9.982807428376636e-06, "loss": 2.0655, "step": 108000 }, { "epoch": 1.61, "learning_rate": 9.797542647952455e-06, "loss": 2.0554, "step": 108500 }, { "epoch": 1.62, "learning_rate": 9.612277867528271e-06, "loss": 2.0359, "step": 109000 }, { "epoch": 1.62, "learning_rate": 9.42701308710409e-06, "loss": 2.056, "step": 109500 }, { "epoch": 1.63, "learning_rate": 9.241748306679908e-06, "loss": 2.0273, "step": 110000 }, { "epoch": 1.64, "learning_rate": 9.056483526255726e-06, "loss": 2.0323, "step": 110500 }, { "epoch": 1.65, "learning_rate": 8.871218745831543e-06, "loss": 2.0288, "step": 111000 }, { "epoch": 1.65, "learning_rate": 8.685953965407361e-06, "loss": 2.0188, "step": 111500 }, { "epoch": 1.66, "learning_rate": 8.500689184983178e-06, "loss": 2.0162, "step": 112000 }, { "epoch": 1.67, "learning_rate": 8.315424404558996e-06, "loss": 2.0407, "step": 112500 }, { "epoch": 1.67, "learning_rate": 8.130159624134814e-06, "loss": 2.059, "step": 113000 }, { "epoch": 1.68, "learning_rate": 7.944894843710633e-06, "loss": 2.0614, "step": 113500 }, { "epoch": 1.69, "learning_rate": 7.75963006328645e-06, "loss": 2.0158, "step": 114000 }, { "epoch": 1.7, "learning_rate": 7.574365282862267e-06, "loss": 2.0197, "step": 114500 }, { "epoch": 1.7, "learning_rate": 7.389100502438084e-06, "loss": 2.0359, "step": 115000 }, { "epoch": 1.71, "learning_rate": 7.203835722013903e-06, "loss": 2.0233, "step": 115500 }, { "epoch": 1.72, "learning_rate": 7.018570941589721e-06, "loss": 2.0306, "step": 116000 }, { "epoch": 1.73, "learning_rate": 6.833306161165538e-06, "loss": 2.0297, "step": 116500 }, { "epoch": 1.73, "learning_rate": 6.648041380741356e-06, "loss": 2.0465, "step": 117000 }, { "epoch": 1.74, "learning_rate": 6.462776600317173e-06, "loss": 2.0386, "step": 117500 }, { "epoch": 1.75, "learning_rate": 6.277511819892991e-06, "loss": 2.0534, "step": 118000 }, { "epoch": 1.76, "learning_rate": 6.092247039468809e-06, "loss": 2.0351, "step": 118500 }, { "epoch": 1.76, "learning_rate": 5.906982259044627e-06, "loss": 2.0361, "step": 119000 }, { "epoch": 1.77, "learning_rate": 5.721717478620445e-06, "loss": 2.0462, "step": 119500 }, { "epoch": 1.78, "learning_rate": 5.536452698196262e-06, "loss": 2.0551, "step": 120000 }, { "epoch": 1.79, "learning_rate": 5.3511879177720805e-06, "loss": 2.0461, "step": 120500 }, { "epoch": 1.79, "learning_rate": 5.165923137347898e-06, "loss": 2.0384, "step": 121000 }, { "epoch": 1.8, "learning_rate": 4.980658356923715e-06, "loss": 2.0189, "step": 121500 }, { "epoch": 1.81, "learning_rate": 4.795393576499534e-06, "loss": 2.0275, "step": 122000 }, { "epoch": 1.82, "learning_rate": 4.610128796075351e-06, "loss": 2.0477, "step": 122500 }, { "epoch": 1.82, "learning_rate": 4.424864015651169e-06, "loss": 2.0411, "step": 123000 }, { "epoch": 1.83, "learning_rate": 4.239599235226987e-06, "loss": 2.0358, "step": 123500 }, { "epoch": 1.84, "learning_rate": 4.054334454802804e-06, "loss": 2.0485, "step": 124000 }, { "epoch": 1.85, "learning_rate": 3.869069674378622e-06, "loss": 2.0406, "step": 124500 }, { "epoch": 1.85, "learning_rate": 3.68380489395444e-06, "loss": 2.038, "step": 125000 }, { "epoch": 1.86, "learning_rate": 3.4985401135302576e-06, "loss": 2.0437, "step": 125500 }, { "epoch": 1.87, "learning_rate": 3.313275333106075e-06, "loss": 2.026, "step": 126000 }, { "epoch": 1.87, "learning_rate": 3.1280105526818934e-06, "loss": 2.054, "step": 126500 }, { "epoch": 1.88, "learning_rate": 2.942745772257711e-06, "loss": 2.0416, "step": 127000 }, { "epoch": 1.89, "learning_rate": 2.7574809918335287e-06, "loss": 2.0198, "step": 127500 }, { "epoch": 1.9, "learning_rate": 2.5722162114093466e-06, "loss": 2.0431, "step": 128000 }, { "epoch": 1.9, "learning_rate": 2.386951430985164e-06, "loss": 2.0247, "step": 128500 }, { "epoch": 1.91, "learning_rate": 2.201686650560982e-06, "loss": 2.0406, "step": 129000 }, { "epoch": 1.92, "learning_rate": 2.0164218701368e-06, "loss": 2.0438, "step": 129500 }, { "epoch": 1.93, "learning_rate": 1.8311570897126173e-06, "loss": 2.073, "step": 130000 }, { "epoch": 1.93, "learning_rate": 1.645892309288435e-06, "loss": 2.0388, "step": 130500 }, { "epoch": 1.94, "learning_rate": 1.4606275288642529e-06, "loss": 2.0336, "step": 131000 }, { "epoch": 1.95, "learning_rate": 1.2753627484400705e-06, "loss": 2.0644, "step": 131500 }, { "epoch": 1.96, "learning_rate": 1.0900979680158882e-06, "loss": 2.0585, "step": 132000 }, { "epoch": 1.96, "learning_rate": 9.048331875917061e-07, "loss": 2.0549, "step": 132500 }, { "epoch": 1.97, "learning_rate": 7.195684071675239e-07, "loss": 2.0455, "step": 133000 }, { "epoch": 1.98, "learning_rate": 5.343036267433415e-07, "loss": 2.0656, "step": 133500 }, { "epoch": 1.99, "learning_rate": 3.490388463191593e-07, "loss": 2.0592, "step": 134000 }, { "epoch": 1.99, "learning_rate": 1.637740658949771e-07, "loss": 2.0737, "step": 134500 }, { "epoch": 2.0, "step": 134942, "total_flos": 2.5064145824199475e+17, "train_loss": 2.1117283579871065, "train_runtime": 41123.1656, "train_samples_per_second": 6.563, "train_steps_per_second": 3.281 } ], "max_steps": 134942, "num_train_epochs": 2, "total_flos": 2.5064145824199475e+17, "trial_name": null, "trial_params": null }