{ "best_metric": null, "best_model_checkpoint": null, "epoch": 12.886142983230362, "global_step": 73000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.09, "learning_rate": 4.975482985191724e-05, "loss": 5.0818, "step": 500 }, { "epoch": 0.18, "learning_rate": 4.9509659703834464e-05, "loss": 4.0909, "step": 1000 }, { "epoch": 0.26, "learning_rate": 4.926448955575169e-05, "loss": 3.4691, "step": 1500 }, { "epoch": 0.35, "learning_rate": 4.9019319407668926e-05, "loss": 2.6149, "step": 2000 }, { "epoch": 0.44, "learning_rate": 4.8774149259586154e-05, "loss": 1.9173, "step": 2500 }, { "epoch": 0.53, "learning_rate": 4.852897911150339e-05, "loss": 1.5632, "step": 3000 }, { "epoch": 0.62, "learning_rate": 4.8283808963420616e-05, "loss": 1.4404, "step": 3500 }, { "epoch": 0.71, "learning_rate": 4.8038638815337844e-05, "loss": 1.3445, "step": 4000 }, { "epoch": 0.79, "learning_rate": 4.779346866725508e-05, "loss": 1.2212, "step": 4500 }, { "epoch": 0.88, "learning_rate": 4.7548298519172305e-05, "loss": 1.1768, "step": 5000 }, { "epoch": 0.97, "learning_rate": 4.730312837108954e-05, "loss": 1.1509, "step": 5500 }, { "epoch": 1.06, "learning_rate": 4.705795822300677e-05, "loss": 0.9679, "step": 6000 }, { "epoch": 1.15, "learning_rate": 4.6812788074924e-05, "loss": 0.8976, "step": 6500 }, { "epoch": 1.24, "learning_rate": 4.656761792684123e-05, "loss": 0.8739, "step": 7000 }, { "epoch": 1.32, "learning_rate": 4.632244777875846e-05, "loss": 0.8976, "step": 7500 }, { "epoch": 1.41, "learning_rate": 4.607727763067569e-05, "loss": 0.8516, "step": 8000 }, { "epoch": 1.5, "learning_rate": 4.5832107482592926e-05, "loss": 0.8655, "step": 8500 }, { "epoch": 1.59, "learning_rate": 4.558693733451015e-05, "loss": 0.8692, "step": 9000 }, { "epoch": 1.68, "learning_rate": 4.534176718642738e-05, "loss": 0.8289, "step": 9500 }, { "epoch": 1.77, "learning_rate": 4.5096597038344615e-05, "loss": 0.8132, "step": 10000 }, { "epoch": 1.85, "learning_rate": 4.485142689026184e-05, "loss": 0.7894, "step": 10500 }, { "epoch": 1.94, "learning_rate": 4.460625674217908e-05, "loss": 0.796, "step": 11000 }, { "epoch": 2.03, "learning_rate": 4.4361086594096305e-05, "loss": 0.7169, "step": 11500 }, { "epoch": 2.12, "learning_rate": 4.411591644601353e-05, "loss": 0.5873, "step": 12000 }, { "epoch": 2.21, "learning_rate": 4.387074629793077e-05, "loss": 0.5975, "step": 12500 }, { "epoch": 2.29, "learning_rate": 4.3625576149847994e-05, "loss": 0.6024, "step": 13000 }, { "epoch": 2.38, "learning_rate": 4.338040600176523e-05, "loss": 0.5982, "step": 13500 }, { "epoch": 2.47, "learning_rate": 4.3135235853682456e-05, "loss": 0.6028, "step": 14000 }, { "epoch": 2.56, "learning_rate": 4.2890065705599684e-05, "loss": 0.6075, "step": 14500 }, { "epoch": 2.65, "learning_rate": 4.264489555751692e-05, "loss": 0.5916, "step": 15000 }, { "epoch": 2.74, "learning_rate": 4.2399725409434146e-05, "loss": 0.6201, "step": 15500 }, { "epoch": 2.82, "learning_rate": 4.215455526135138e-05, "loss": 0.5912, "step": 16000 }, { "epoch": 2.91, "learning_rate": 4.1909385113268615e-05, "loss": 0.586, "step": 16500 }, { "epoch": 3.0, "learning_rate": 4.1664214965185836e-05, "loss": 0.6191, "step": 17000 }, { "epoch": 3.09, "learning_rate": 4.141904481710307e-05, "loss": 0.4129, "step": 17500 }, { "epoch": 3.18, "learning_rate": 4.1173874669020304e-05, "loss": 0.4322, "step": 18000 }, { "epoch": 3.27, "learning_rate": 4.092870452093753e-05, "loss": 0.439, "step": 18500 }, { "epoch": 3.35, "learning_rate": 4.0683534372854766e-05, "loss": 0.4423, "step": 19000 }, { "epoch": 3.44, "learning_rate": 4.0438364224771994e-05, "loss": 0.4428, "step": 19500 }, { "epoch": 3.53, "learning_rate": 4.019319407668922e-05, "loss": 0.4441, "step": 20000 }, { "epoch": 3.62, "learning_rate": 3.9948023928606456e-05, "loss": 0.4499, "step": 20500 }, { "epoch": 3.71, "learning_rate": 3.9702853780523683e-05, "loss": 0.4561, "step": 21000 }, { "epoch": 3.8, "learning_rate": 3.945768363244092e-05, "loss": 0.4601, "step": 21500 }, { "epoch": 3.88, "learning_rate": 3.9212513484358145e-05, "loss": 0.455, "step": 22000 }, { "epoch": 3.97, "learning_rate": 3.896734333627537e-05, "loss": 0.4718, "step": 22500 }, { "epoch": 4.06, "learning_rate": 3.872217318819261e-05, "loss": 0.3575, "step": 23000 }, { "epoch": 4.15, "learning_rate": 3.8477003040109835e-05, "loss": 0.3211, "step": 23500 }, { "epoch": 4.24, "learning_rate": 3.823183289202707e-05, "loss": 0.3413, "step": 24000 }, { "epoch": 4.32, "learning_rate": 3.7986662743944304e-05, "loss": 0.3325, "step": 24500 }, { "epoch": 4.41, "learning_rate": 3.7741492595861525e-05, "loss": 0.3552, "step": 25000 }, { "epoch": 4.5, "learning_rate": 3.749632244777876e-05, "loss": 0.3475, "step": 25500 }, { "epoch": 4.59, "learning_rate": 3.725115229969599e-05, "loss": 0.3617, "step": 26000 }, { "epoch": 4.68, "learning_rate": 3.700598215161322e-05, "loss": 0.3457, "step": 26500 }, { "epoch": 4.77, "learning_rate": 3.6760812003530455e-05, "loss": 0.3658, "step": 27000 }, { "epoch": 4.85, "learning_rate": 3.651564185544768e-05, "loss": 0.348, "step": 27500 }, { "epoch": 4.94, "learning_rate": 3.627047170736491e-05, "loss": 0.3547, "step": 28000 }, { "epoch": 5.03, "learning_rate": 3.6025301559282145e-05, "loss": 0.3103, "step": 28500 }, { "epoch": 5.12, "learning_rate": 3.578013141119937e-05, "loss": 0.2441, "step": 29000 }, { "epoch": 5.21, "learning_rate": 3.553496126311661e-05, "loss": 0.2516, "step": 29500 }, { "epoch": 5.3, "learning_rate": 3.528979111503384e-05, "loss": 0.261, "step": 30000 }, { "epoch": 5.38, "learning_rate": 3.504462096695106e-05, "loss": 0.2665, "step": 30500 }, { "epoch": 5.47, "learning_rate": 3.4799450818868296e-05, "loss": 0.2585, "step": 31000 }, { "epoch": 5.56, "learning_rate": 3.4554280670785524e-05, "loss": 0.2794, "step": 31500 }, { "epoch": 5.65, "learning_rate": 3.430911052270276e-05, "loss": 0.2757, "step": 32000 }, { "epoch": 5.74, "learning_rate": 3.406394037461999e-05, "loss": 0.289, "step": 32500 }, { "epoch": 5.83, "learning_rate": 3.3818770226537214e-05, "loss": 0.2775, "step": 33000 }, { "epoch": 5.91, "learning_rate": 3.357360007845445e-05, "loss": 0.282, "step": 33500 }, { "epoch": 6.0, "learning_rate": 3.332842993037168e-05, "loss": 0.28, "step": 34000 }, { "epoch": 6.09, "learning_rate": 3.308325978228891e-05, "loss": 0.193, "step": 34500 }, { "epoch": 6.18, "learning_rate": 3.2838089634206144e-05, "loss": 0.1936, "step": 35000 }, { "epoch": 6.27, "learning_rate": 3.259291948612337e-05, "loss": 0.2037, "step": 35500 }, { "epoch": 6.35, "learning_rate": 3.23477493380406e-05, "loss": 0.2154, "step": 36000 }, { "epoch": 6.44, "learning_rate": 3.2102579189957834e-05, "loss": 0.2103, "step": 36500 }, { "epoch": 6.53, "learning_rate": 3.185740904187506e-05, "loss": 0.216, "step": 37000 }, { "epoch": 6.62, "learning_rate": 3.1612238893792296e-05, "loss": 0.2115, "step": 37500 }, { "epoch": 6.71, "learning_rate": 3.1367068745709523e-05, "loss": 0.226, "step": 38000 }, { "epoch": 6.8, "learning_rate": 3.112189859762675e-05, "loss": 0.2232, "step": 38500 }, { "epoch": 6.88, "learning_rate": 3.0876728449543985e-05, "loss": 0.2272, "step": 39000 }, { "epoch": 6.97, "learning_rate": 3.063155830146121e-05, "loss": 0.2344, "step": 39500 }, { "epoch": 7.06, "learning_rate": 3.0386388153378447e-05, "loss": 0.1742, "step": 40000 }, { "epoch": 7.15, "learning_rate": 3.014121800529568e-05, "loss": 0.1587, "step": 40500 }, { "epoch": 7.24, "learning_rate": 2.9896047857212906e-05, "loss": 0.1653, "step": 41000 }, { "epoch": 7.33, "learning_rate": 2.9650877709130137e-05, "loss": 0.1679, "step": 41500 }, { "epoch": 7.41, "learning_rate": 2.9405707561047368e-05, "loss": 0.1728, "step": 42000 }, { "epoch": 7.5, "learning_rate": 2.91605374129646e-05, "loss": 0.1729, "step": 42500 }, { "epoch": 7.59, "learning_rate": 2.8915367264881833e-05, "loss": 0.1737, "step": 43000 }, { "epoch": 7.68, "learning_rate": 2.8670197116799058e-05, "loss": 0.181, "step": 43500 }, { "epoch": 7.77, "learning_rate": 2.842502696871629e-05, "loss": 0.1771, "step": 44000 }, { "epoch": 7.86, "learning_rate": 2.817985682063352e-05, "loss": 0.1818, "step": 44500 }, { "epoch": 7.94, "learning_rate": 2.7934686672550754e-05, "loss": 0.1847, "step": 45000 }, { "epoch": 8.03, "learning_rate": 2.7689516524467985e-05, "loss": 0.1643, "step": 45500 }, { "epoch": 8.12, "learning_rate": 2.744434637638521e-05, "loss": 0.1249, "step": 46000 }, { "epoch": 8.21, "learning_rate": 2.7199176228302443e-05, "loss": 0.1315, "step": 46500 }, { "epoch": 8.3, "learning_rate": 2.6954006080219674e-05, "loss": 0.1372, "step": 47000 }, { "epoch": 8.38, "learning_rate": 2.6708835932136905e-05, "loss": 0.1437, "step": 47500 }, { "epoch": 8.47, "learning_rate": 2.6463665784054136e-05, "loss": 0.1447, "step": 48000 }, { "epoch": 8.56, "learning_rate": 2.6218495635971364e-05, "loss": 0.1417, "step": 48500 }, { "epoch": 8.65, "learning_rate": 2.5973325487888595e-05, "loss": 0.1483, "step": 49000 }, { "epoch": 8.74, "learning_rate": 2.5728155339805826e-05, "loss": 0.1392, "step": 49500 }, { "epoch": 8.83, "learning_rate": 2.5482985191723057e-05, "loss": 0.1479, "step": 50000 }, { "epoch": 8.91, "learning_rate": 2.5237815043640288e-05, "loss": 0.1509, "step": 50500 }, { "epoch": 9.0, "learning_rate": 2.499264489555752e-05, "loss": 0.1498, "step": 51000 }, { "epoch": 9.09, "learning_rate": 2.474747474747475e-05, "loss": 0.1, "step": 51500 }, { "epoch": 9.18, "learning_rate": 2.4502304599391977e-05, "loss": 0.1103, "step": 52000 }, { "epoch": 9.27, "learning_rate": 2.4257134451309212e-05, "loss": 0.1057, "step": 52500 }, { "epoch": 9.36, "learning_rate": 2.401196430322644e-05, "loss": 0.1201, "step": 53000 }, { "epoch": 9.44, "learning_rate": 2.376679415514367e-05, "loss": 0.122, "step": 53500 }, { "epoch": 9.53, "learning_rate": 2.35216240070609e-05, "loss": 0.1172, "step": 54000 }, { "epoch": 9.62, "learning_rate": 2.3276453858978132e-05, "loss": 0.1173, "step": 54500 }, { "epoch": 9.71, "learning_rate": 2.3031283710895363e-05, "loss": 0.1149, "step": 55000 }, { "epoch": 9.8, "learning_rate": 2.2786113562812594e-05, "loss": 0.125, "step": 55500 }, { "epoch": 9.89, "learning_rate": 2.2540943414729822e-05, "loss": 0.1214, "step": 56000 }, { "epoch": 9.97, "learning_rate": 2.2295773266647056e-05, "loss": 0.1233, "step": 56500 }, { "epoch": 10.06, "learning_rate": 2.2050603118564284e-05, "loss": 0.0946, "step": 57000 }, { "epoch": 10.15, "learning_rate": 2.1805432970481515e-05, "loss": 0.0892, "step": 57500 }, { "epoch": 10.24, "learning_rate": 2.1560262822398746e-05, "loss": 0.0911, "step": 58000 }, { "epoch": 10.33, "learning_rate": 2.1315092674315977e-05, "loss": 0.0912, "step": 58500 }, { "epoch": 10.41, "learning_rate": 2.1069922526233208e-05, "loss": 0.0917, "step": 59000 }, { "epoch": 10.5, "learning_rate": 2.0824752378150436e-05, "loss": 0.0957, "step": 59500 }, { "epoch": 10.59, "learning_rate": 2.0579582230067666e-05, "loss": 0.0992, "step": 60000 }, { "epoch": 10.68, "learning_rate": 2.03344120819849e-05, "loss": 0.0973, "step": 60500 }, { "epoch": 10.77, "learning_rate": 2.008924193390213e-05, "loss": 0.1042, "step": 61000 }, { "epoch": 10.86, "learning_rate": 1.984407178581936e-05, "loss": 0.1027, "step": 61500 }, { "epoch": 10.94, "learning_rate": 1.959890163773659e-05, "loss": 0.1047, "step": 62000 }, { "epoch": 11.03, "learning_rate": 1.935373148965382e-05, "loss": 0.088, "step": 62500 }, { "epoch": 11.12, "learning_rate": 1.9108561341571052e-05, "loss": 0.0695, "step": 63000 }, { "epoch": 11.21, "learning_rate": 1.886339119348828e-05, "loss": 0.0733, "step": 63500 }, { "epoch": 11.3, "learning_rate": 1.861822104540551e-05, "loss": 0.0727, "step": 64000 }, { "epoch": 11.39, "learning_rate": 1.8373050897322745e-05, "loss": 0.0784, "step": 64500 }, { "epoch": 11.47, "learning_rate": 1.8127880749239973e-05, "loss": 0.076, "step": 65000 }, { "epoch": 11.56, "learning_rate": 1.7882710601157204e-05, "loss": 0.0814, "step": 65500 }, { "epoch": 11.65, "learning_rate": 1.763754045307443e-05, "loss": 0.0816, "step": 66000 }, { "epoch": 11.74, "learning_rate": 1.7392370304991666e-05, "loss": 0.0848, "step": 66500 }, { "epoch": 11.83, "learning_rate": 1.7147200156908897e-05, "loss": 0.0827, "step": 67000 }, { "epoch": 11.92, "learning_rate": 1.6902030008826124e-05, "loss": 0.0839, "step": 67500 }, { "epoch": 12.0, "learning_rate": 1.6656859860743355e-05, "loss": 0.0808, "step": 68000 }, { "epoch": 12.09, "learning_rate": 1.641168971266059e-05, "loss": 0.0545, "step": 68500 }, { "epoch": 12.18, "learning_rate": 1.6166519564577817e-05, "loss": 0.0593, "step": 69000 }, { "epoch": 12.27, "learning_rate": 1.592134941649505e-05, "loss": 0.0626, "step": 69500 }, { "epoch": 12.36, "learning_rate": 1.5676179268412276e-05, "loss": 0.0633, "step": 70000 }, { "epoch": 12.44, "learning_rate": 1.543100912032951e-05, "loss": 0.0673, "step": 70500 }, { "epoch": 12.53, "learning_rate": 1.5185838972246741e-05, "loss": 0.0661, "step": 71000 }, { "epoch": 12.62, "learning_rate": 1.4940668824163969e-05, "loss": 0.067, "step": 71500 }, { "epoch": 12.71, "learning_rate": 1.4695498676081202e-05, "loss": 0.0691, "step": 72000 }, { "epoch": 12.8, "learning_rate": 1.4450328527998433e-05, "loss": 0.0674, "step": 72500 }, { "epoch": 12.89, "learning_rate": 1.4205158379915662e-05, "loss": 0.067, "step": 73000 } ], "max_steps": 101970, "num_train_epochs": 18, "total_flos": 1.2452822711842406e+17, "trial_name": null, "trial_params": null }