| { |
| "best_global_step": 3900, |
| "best_metric": 0.3069113492965698, |
| "best_model_checkpoint": "Assignment4_Distilled_ModernBERT/run-2/checkpoint-3000", |
| "epoch": 9.0, |
| "eval_steps": 100, |
| "global_step": 4293, |
| "is_hyper_param_search": true, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.20964360587002095, |
| "grad_norm": 15.277114868164062, |
| "learning_rate": 5.99213045445736e-05, |
| "loss": 6.7857, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.20964360587002095, |
| "eval_accuracy": 0.7109677419354838, |
| "eval_loss": 3.844492197036743, |
| "eval_runtime": 24.4798, |
| "eval_samples_per_second": 126.635, |
| "eval_steps_per_second": 15.85, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.4192872117400419, |
| "grad_norm": 9.93507194519043, |
| "learning_rate": 5.968245320843699e-05, |
| "loss": 2.8413, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.4192872117400419, |
| "eval_accuracy": 0.8938709677419355, |
| "eval_loss": 1.8548765182495117, |
| "eval_runtime": 24.5612, |
| "eval_samples_per_second": 126.215, |
| "eval_steps_per_second": 15.797, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "grad_norm": 3.672398805618286, |
| "learning_rate": 5.928471649139723e-05, |
| "loss": 1.4945, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "eval_accuracy": 0.94, |
| "eval_loss": 1.1517655849456787, |
| "eval_runtime": 24.6593, |
| "eval_samples_per_second": 125.713, |
| "eval_steps_per_second": 15.734, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.8385744234800838, |
| "grad_norm": 9.015824317932129, |
| "learning_rate": 5.87302234139009e-05, |
| "loss": 1.0247, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.8385744234800838, |
| "eval_accuracy": 0.9532258064516129, |
| "eval_loss": 0.9113911390304565, |
| "eval_runtime": 24.4669, |
| "eval_samples_per_second": 126.702, |
| "eval_steps_per_second": 15.858, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.0482180293501049, |
| "grad_norm": 13.825676918029785, |
| "learning_rate": 5.802194208788969e-05, |
| "loss": 0.7584, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0482180293501049, |
| "eval_accuracy": 0.9558064516129032, |
| "eval_loss": 0.7309688329696655, |
| "eval_runtime": 24.9176, |
| "eval_samples_per_second": 124.41, |
| "eval_steps_per_second": 15.571, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.2578616352201257, |
| "grad_norm": 2.16816782951355, |
| "learning_rate": 5.716366382897622e-05, |
| "loss": 0.5289, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.2578616352201257, |
| "eval_accuracy": 0.9616129032258065, |
| "eval_loss": 0.6186437606811523, |
| "eval_runtime": 24.8855, |
| "eval_samples_per_second": 124.571, |
| "eval_steps_per_second": 15.591, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.4675052410901468, |
| "grad_norm": 3.3350861072540283, |
| "learning_rate": 5.6159982862143515e-05, |
| "loss": 0.4857, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.4675052410901468, |
| "eval_accuracy": 0.96, |
| "eval_loss": 0.5846336483955383, |
| "eval_runtime": 24.8682, |
| "eval_samples_per_second": 124.657, |
| "eval_steps_per_second": 15.602, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.6771488469601676, |
| "grad_norm": 1.4833406209945679, |
| "learning_rate": 5.5016271729600304e-05, |
| "loss": 0.4169, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.6771488469601676, |
| "eval_accuracy": 0.9638709677419355, |
| "eval_loss": 0.5170760750770569, |
| "eval_runtime": 31.1132, |
| "eval_samples_per_second": 99.636, |
| "eval_steps_per_second": 12.471, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.8867924528301887, |
| "grad_norm": 8.999539375305176, |
| "learning_rate": 5.3738652532429715e-05, |
| "loss": 0.3953, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.8867924528301887, |
| "eval_accuracy": 0.9683870967741935, |
| "eval_loss": 0.4950183629989624, |
| "eval_runtime": 25.9065, |
| "eval_samples_per_second": 119.661, |
| "eval_steps_per_second": 14.977, |
| "step": 900 |
| }, |
| { |
| "epoch": 2.0964360587002098, |
| "grad_norm": 1.5073931217193604, |
| "learning_rate": 5.2333964159970384e-05, |
| "loss": 0.3393, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.0964360587002098, |
| "eval_accuracy": 0.9674193548387097, |
| "eval_loss": 0.45995765924453735, |
| "eval_runtime": 25.1127, |
| "eval_samples_per_second": 123.444, |
| "eval_steps_per_second": 15.45, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.3060796645702304, |
| "grad_norm": 1.204892873764038, |
| "learning_rate": 5.080972568234569e-05, |
| "loss": 0.2747, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.3060796645702304, |
| "eval_accuracy": 0.9661290322580646, |
| "eval_loss": 0.44675561785697937, |
| "eval_runtime": 24.863, |
| "eval_samples_per_second": 124.683, |
| "eval_steps_per_second": 15.606, |
| "step": 1100 |
| }, |
| { |
| "epoch": 2.5157232704402515, |
| "grad_norm": 1.581334114074707, |
| "learning_rate": 4.9174096102095087e-05, |
| "loss": 0.2703, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.5157232704402515, |
| "eval_accuracy": 0.9664516129032258, |
| "eval_loss": 0.43446871638298035, |
| "eval_runtime": 24.788, |
| "eval_samples_per_second": 125.061, |
| "eval_steps_per_second": 15.653, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.7253668763102725, |
| "grad_norm": 1.4468449354171753, |
| "learning_rate": 4.7435830680350456e-05, |
| "loss": 0.2482, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.7253668763102725, |
| "eval_accuracy": 0.967741935483871, |
| "eval_loss": 0.4083913564682007, |
| "eval_runtime": 30.9276, |
| "eval_samples_per_second": 100.234, |
| "eval_steps_per_second": 12.545, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.9350104821802936, |
| "grad_norm": 1.0635446310043335, |
| "learning_rate": 4.5604234071336463e-05, |
| "loss": 0.2342, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.9350104821802936, |
| "eval_accuracy": 0.9674193548387097, |
| "eval_loss": 0.3973832428455353, |
| "eval_runtime": 25.2744, |
| "eval_samples_per_second": 122.654, |
| "eval_steps_per_second": 15.351, |
| "step": 1400 |
| }, |
| { |
| "epoch": 3.1446540880503147, |
| "grad_norm": 1.0511268377304077, |
| "learning_rate": 4.368911051605842e-05, |
| "loss": 0.2073, |
| "step": 1500 |
| }, |
| { |
| "epoch": 3.1446540880503147, |
| "eval_accuracy": 0.9680645161290322, |
| "eval_loss": 0.3778529465198517, |
| "eval_runtime": 24.3903, |
| "eval_samples_per_second": 127.1, |
| "eval_steps_per_second": 15.908, |
| "step": 1500 |
| }, |
| { |
| "epoch": 3.3542976939203353, |
| "grad_norm": 1.3159152269363403, |
| "learning_rate": 4.1700711361782675e-05, |
| "loss": 0.1973, |
| "step": 1600 |
| }, |
| { |
| "epoch": 3.3542976939203353, |
| "eval_accuracy": 0.9696774193548388, |
| "eval_loss": 0.3735784590244293, |
| "eval_runtime": 24.7557, |
| "eval_samples_per_second": 125.224, |
| "eval_steps_per_second": 15.673, |
| "step": 1600 |
| }, |
| { |
| "epoch": 3.5639412997903563, |
| "grad_norm": 0.8998211622238159, |
| "learning_rate": 3.9649680188229416e-05, |
| "loss": 0.1876, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.5639412997903563, |
| "eval_accuracy": 0.9696774193548388, |
| "eval_loss": 0.3719204366207123, |
| "eval_runtime": 25.0523, |
| "eval_samples_per_second": 123.741, |
| "eval_steps_per_second": 15.488, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.7735849056603774, |
| "grad_norm": 1.0484949350357056, |
| "learning_rate": 3.754699583420843e-05, |
| "loss": 0.1836, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.7735849056603774, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.35943037271499634, |
| "eval_runtime": 29.7226, |
| "eval_samples_per_second": 104.298, |
| "eval_steps_per_second": 13.054, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.9832285115303985, |
| "grad_norm": 1.2159234285354614, |
| "learning_rate": 3.5403913629667045e-05, |
| "loss": 0.184, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.9832285115303985, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.35340631008148193, |
| "eval_runtime": 26.1611, |
| "eval_samples_per_second": 118.496, |
| "eval_steps_per_second": 14.831, |
| "step": 1900 |
| }, |
| { |
| "epoch": 4.1928721174004195, |
| "grad_norm": 0.8908376097679138, |
| "learning_rate": 3.323190514772574e-05, |
| "loss": 0.1558, |
| "step": 2000 |
| }, |
| { |
| "epoch": 4.1928721174004195, |
| "eval_accuracy": 0.97, |
| "eval_loss": 0.34400272369384766, |
| "eval_runtime": 24.8773, |
| "eval_samples_per_second": 124.612, |
| "eval_steps_per_second": 15.597, |
| "step": 2000 |
| }, |
| { |
| "epoch": 4.40251572327044, |
| "grad_norm": 0.9733275175094604, |
| "learning_rate": 3.1042596799198844e-05, |
| "loss": 0.1496, |
| "step": 2100 |
| }, |
| { |
| "epoch": 4.40251572327044, |
| "eval_accuracy": 0.9716129032258064, |
| "eval_loss": 0.33769491314888, |
| "eval_runtime": 24.7885, |
| "eval_samples_per_second": 125.058, |
| "eval_steps_per_second": 15.652, |
| "step": 2100 |
| }, |
| { |
| "epoch": 4.612159329140461, |
| "grad_norm": 0.7378196120262146, |
| "learning_rate": 2.8847707598294083e-05, |
| "loss": 0.1461, |
| "step": 2200 |
| }, |
| { |
| "epoch": 4.612159329140461, |
| "eval_accuracy": 0.9706451612903226, |
| "eval_loss": 0.3369283974170685, |
| "eval_runtime": 24.7525, |
| "eval_samples_per_second": 125.24, |
| "eval_steps_per_second": 15.675, |
| "step": 2200 |
| }, |
| { |
| "epoch": 4.821802935010482, |
| "grad_norm": 1.0811831951141357, |
| "learning_rate": 2.665898643262115e-05, |
| "loss": 0.1457, |
| "step": 2300 |
| }, |
| { |
| "epoch": 4.821802935010482, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.3287930190563202, |
| "eval_runtime": 27.4125, |
| "eval_samples_per_second": 113.087, |
| "eval_steps_per_second": 14.154, |
| "step": 2300 |
| }, |
| { |
| "epoch": 5.031446540880503, |
| "grad_norm": 0.89784175157547, |
| "learning_rate": 2.448814917329275e-05, |
| "loss": 0.1406, |
| "step": 2400 |
| }, |
| { |
| "epoch": 5.031446540880503, |
| "eval_accuracy": 0.97, |
| "eval_loss": 0.3258771300315857, |
| "eval_runtime": 26.8839, |
| "eval_samples_per_second": 115.311, |
| "eval_steps_per_second": 14.432, |
| "step": 2400 |
| }, |
| { |
| "epoch": 5.241090146750524, |
| "grad_norm": 0.6693010330200195, |
| "learning_rate": 2.23468159617576e-05, |
| "loss": 0.1254, |
| "step": 2500 |
| }, |
| { |
| "epoch": 5.241090146750524, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.3203785717487335, |
| "eval_runtime": 24.583, |
| "eval_samples_per_second": 126.104, |
| "eval_steps_per_second": 15.783, |
| "step": 2500 |
| }, |
| { |
| "epoch": 5.450733752620545, |
| "grad_norm": 0.6166190505027771, |
| "learning_rate": 2.024644900905892e-05, |
| "loss": 0.1245, |
| "step": 2600 |
| }, |
| { |
| "epoch": 5.450733752620545, |
| "eval_accuracy": 0.97, |
| "eval_loss": 0.3191240429878235, |
| "eval_runtime": 25.4943, |
| "eval_samples_per_second": 121.596, |
| "eval_steps_per_second": 15.219, |
| "step": 2600 |
| }, |
| { |
| "epoch": 5.660377358490566, |
| "grad_norm": 0.5683078169822693, |
| "learning_rate": 1.8198291240468836e-05, |
| "loss": 0.1221, |
| "step": 2700 |
| }, |
| { |
| "epoch": 5.660377358490566, |
| "eval_accuracy": 0.9719354838709677, |
| "eval_loss": 0.3174494206905365, |
| "eval_runtime": 25.2702, |
| "eval_samples_per_second": 122.674, |
| "eval_steps_per_second": 15.354, |
| "step": 2700 |
| }, |
| { |
| "epoch": 5.870020964360587, |
| "grad_norm": 0.6431116461753845, |
| "learning_rate": 1.621330611392417e-05, |
| "loss": 0.1209, |
| "step": 2800 |
| }, |
| { |
| "epoch": 5.870020964360587, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.3135199248790741, |
| "eval_runtime": 27.0594, |
| "eval_samples_per_second": 114.563, |
| "eval_steps_per_second": 14.339, |
| "step": 2800 |
| }, |
| { |
| "epoch": 6.079664570230608, |
| "grad_norm": 0.5339530110359192, |
| "learning_rate": 1.4302118934405711e-05, |
| "loss": 0.1173, |
| "step": 2900 |
| }, |
| { |
| "epoch": 6.079664570230608, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.31273338198661804, |
| "eval_runtime": 28.6742, |
| "eval_samples_per_second": 108.111, |
| "eval_steps_per_second": 13.531, |
| "step": 2900 |
| }, |
| { |
| "epoch": 6.289308176100629, |
| "grad_norm": 0.4836990237236023, |
| "learning_rate": 1.2474959978395775e-05, |
| "loss": 0.1113, |
| "step": 3000 |
| }, |
| { |
| "epoch": 6.289308176100629, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.31234365701675415, |
| "eval_runtime": 24.6684, |
| "eval_samples_per_second": 125.667, |
| "eval_steps_per_second": 15.729, |
| "step": 3000 |
| }, |
| { |
| "epoch": 6.49895178197065, |
| "grad_norm": 0.5658986568450928, |
| "learning_rate": 1.0741609732859415e-05, |
| "loss": 0.1065, |
| "step": 3100 |
| }, |
| { |
| "epoch": 6.49895178197065, |
| "eval_accuracy": 0.9712903225806452, |
| "eval_loss": 0.30965664982795715, |
| "eval_runtime": 25.2448, |
| "eval_samples_per_second": 122.797, |
| "eval_steps_per_second": 15.369, |
| "step": 3100 |
| }, |
| { |
| "epoch": 6.7085953878406706, |
| "grad_norm": 0.5603812336921692, |
| "learning_rate": 9.111346541876427e-06, |
| "loss": 0.1061, |
| "step": 3200 |
| }, |
| { |
| "epoch": 6.7085953878406706, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.30989164113998413, |
| "eval_runtime": 24.9967, |
| "eval_samples_per_second": 124.017, |
| "eval_steps_per_second": 15.522, |
| "step": 3200 |
| }, |
| { |
| "epoch": 6.918238993710692, |
| "grad_norm": 0.5175463557243347, |
| "learning_rate": 7.5928969411631525e-06, |
| "loss": 0.1057, |
| "step": 3300 |
| }, |
| { |
| "epoch": 6.918238993710692, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.3099421262741089, |
| "eval_runtime": 25.1564, |
| "eval_samples_per_second": 123.229, |
| "eval_steps_per_second": 15.424, |
| "step": 3300 |
| }, |
| { |
| "epoch": 7.127882599580713, |
| "grad_norm": 0.47962552309036255, |
| "learning_rate": 6.194388946335337e-06, |
| "loss": 0.1026, |
| "step": 3400 |
| }, |
| { |
| "epoch": 7.127882599580713, |
| "eval_accuracy": 0.9706451612903226, |
| "eval_loss": 0.3079238533973694, |
| "eval_runtime": 30.478, |
| "eval_samples_per_second": 101.713, |
| "eval_steps_per_second": 12.73, |
| "step": 3400 |
| }, |
| { |
| "epoch": 7.337526205450734, |
| "grad_norm": 0.5649577975273132, |
| "learning_rate": 4.923308544952481e-06, |
| "loss": 0.099, |
| "step": 3500 |
| }, |
| { |
| "epoch": 7.337526205450734, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.3081134557723999, |
| "eval_runtime": 24.6885, |
| "eval_samples_per_second": 125.565, |
| "eval_steps_per_second": 15.716, |
| "step": 3500 |
| }, |
| { |
| "epoch": 7.547169811320755, |
| "grad_norm": 0.4486183226108551, |
| "learning_rate": 3.786459625234828e-06, |
| "loss": 0.0999, |
| "step": 3600 |
| }, |
| { |
| "epoch": 7.547169811320755, |
| "eval_accuracy": 0.9703225806451613, |
| "eval_loss": 0.30738693475723267, |
| "eval_runtime": 24.9276, |
| "eval_samples_per_second": 124.36, |
| "eval_steps_per_second": 15.565, |
| "step": 3600 |
| }, |
| { |
| "epoch": 7.756813417190775, |
| "grad_norm": 0.5032072067260742, |
| "learning_rate": 2.7899275559478822e-06, |
| "loss": 0.0997, |
| "step": 3700 |
| }, |
| { |
| "epoch": 7.756813417190775, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.3071826696395874, |
| "eval_runtime": 24.6968, |
| "eval_samples_per_second": 125.522, |
| "eval_steps_per_second": 15.711, |
| "step": 3700 |
| }, |
| { |
| "epoch": 7.966457023060797, |
| "grad_norm": 0.4989459216594696, |
| "learning_rate": 1.9390466124057314e-06, |
| "loss": 0.0987, |
| "step": 3800 |
| }, |
| { |
| "epoch": 7.966457023060797, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.30702054500579834, |
| "eval_runtime": 24.6975, |
| "eval_samples_per_second": 125.519, |
| "eval_steps_per_second": 15.71, |
| "step": 3800 |
| }, |
| { |
| "epoch": 8.176100628930818, |
| "grad_norm": 0.4764493405818939, |
| "learning_rate": 1.238371422956248e-06, |
| "loss": 0.0952, |
| "step": 3900 |
| }, |
| { |
| "epoch": 8.176100628930818, |
| "eval_accuracy": 0.9706451612903226, |
| "eval_loss": 0.3069113492965698, |
| "eval_runtime": 32.1081, |
| "eval_samples_per_second": 96.549, |
| "eval_steps_per_second": 12.084, |
| "step": 3900 |
| }, |
| { |
| "epoch": 8.385744234800839, |
| "grad_norm": 0.44305703043937683, |
| "learning_rate": 6.916525887907699e-07, |
| "loss": 0.0964, |
| "step": 4000 |
| }, |
| { |
| "epoch": 8.385744234800839, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.3070818781852722, |
| "eval_runtime": 25.2525, |
| "eval_samples_per_second": 122.76, |
| "eval_steps_per_second": 15.365, |
| "step": 4000 |
| }, |
| { |
| "epoch": 8.59538784067086, |
| "grad_norm": 0.42053186893463135, |
| "learning_rate": 3.0181660758127563e-07, |
| "loss": 0.0952, |
| "step": 4100 |
| }, |
| { |
| "epoch": 8.59538784067086, |
| "eval_accuracy": 0.9709677419354839, |
| "eval_loss": 0.306934118270874, |
| "eval_runtime": 25.3837, |
| "eval_samples_per_second": 122.126, |
| "eval_steps_per_second": 15.285, |
| "step": 4100 |
| }, |
| { |
| "epoch": 8.80503144654088, |
| "grad_norm": 0.5232221484184265, |
| "learning_rate": 7.095020841074385e-08, |
| "loss": 0.0977, |
| "step": 4200 |
| }, |
| { |
| "epoch": 8.80503144654088, |
| "eval_accuracy": 0.9706451612903226, |
| "eval_loss": 0.3069760203361511, |
| "eval_runtime": 25.0708, |
| "eval_samples_per_second": 123.65, |
| "eval_steps_per_second": 15.476, |
| "step": 4200 |
| } |
| ], |
| "logging_steps": 100, |
| "max_steps": 4293, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 9, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1878717068327988.0, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": { |
| "alpha": 0.497477897443408, |
| "num_train_epochs": 9, |
| "temperature": 7 |
| } |
| } |
|
|
