| { | |
| "best_global_step": 10000, | |
| "best_metric": 10.865962028503418, | |
| "best_model_checkpoint": "distilbert_rand_100_v1/checkpoint-10000", | |
| "epoch": 25.0, | |
| "eval_steps": 10000, | |
| "global_step": 59550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 2.623448133468628, | |
| "learning_rate": 5e-06, | |
| "loss": 13.935, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.2489383220672607, | |
| "learning_rate": 1e-05, | |
| "loss": 12.2125, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 2.007150173187256, | |
| "learning_rate": 1.5e-05, | |
| "loss": 11.6791, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 2.0411810874938965, | |
| "learning_rate": 2e-05, | |
| "loss": 11.5312, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 2.287830114364624, | |
| "learning_rate": 2.5e-05, | |
| "loss": 11.3996, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 2.2175443172454834, | |
| "learning_rate": 3e-05, | |
| "loss": 11.2859, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.2127342224121094, | |
| "learning_rate": 3.5e-05, | |
| "loss": 11.1939, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 2.384563446044922, | |
| "learning_rate": 4e-05, | |
| "loss": 11.1184, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 2.4358389377593994, | |
| "learning_rate": 4.5e-05, | |
| "loss": 11.0592, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.2787206172943115, | |
| "learning_rate": 5e-05, | |
| "loss": 11.0078, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.558605194091797, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.969, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.6357815265655518, | |
| "learning_rate": 6e-05, | |
| "loss": 10.9397, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.2616066932678223, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.9113, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.3792471885681152, | |
| "learning_rate": 7e-05, | |
| "loss": 10.8861, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.501986026763916, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.8125, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.8892018795013428, | |
| "learning_rate": 8e-05, | |
| "loss": 10.7806, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.484477996826172, | |
| "learning_rate": 8.5e-05, | |
| "loss": 10.7731, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 2.767636775970459, | |
| "learning_rate": 9e-05, | |
| "loss": 10.7591, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 3.018613576889038, | |
| "learning_rate": 9.5e-05, | |
| "loss": 10.7479, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.2642602920532227, | |
| "learning_rate": 0.0001, | |
| "loss": 10.5482, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.1512006455382732, | |
| "eval_loss": 10.865962028503418, | |
| "eval_runtime": 0.9243, | |
| "eval_samples_per_second": 518.203, | |
| "eval_steps_per_second": 5.409, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.994490385055542, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 10.5576, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 3.3885488510131836, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 10.5624, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.3888919353485107, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 10.5753, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.4450531005859375, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 10.5203, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.6954619884490967, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 10.2986, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 3.551410436630249, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 10.3352, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 3.592194080352783, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 10.3653, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.599605083465576, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 10.3831, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 3.4054486751556396, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 10.2463, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 3.6060092449188232, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 10.1034, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 4.198487758636475, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 10.1445, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.7223973274230957, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 10.1848, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 3.878314256668091, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 10.206, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 4.285410404205322, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 9.9531, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 4.676534175872803, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 9.8563, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 4.5374040603637695, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 9.901, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 4.714870452880859, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 9.942, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 4.6971025466918945, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 9.9516, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 5.068013668060303, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 9.5171, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 5.643128395080566, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 9.5046, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15255759140267686, | |
| "eval_loss": 11.325418472290039, | |
| "eval_runtime": 0.8699, | |
| "eval_samples_per_second": 550.618, | |
| "eval_steps_per_second": 5.748, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 5.951472759246826, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 9.5303, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 5.77167272567749, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 9.5882, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 6.605103492736816, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 9.5135, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 6.943516731262207, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 9.0137, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 7.392699718475342, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 9.0531, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 7.223564147949219, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 9.1135, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 7.3054304122924805, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 9.1281, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 7.477367877960205, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.9097, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 8.256974220275879, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 8.5396, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 7.7928009033203125, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 8.5868, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 8.170751571655273, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 8.6354, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 7.986641883850098, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 8.6714, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 8.704106330871582, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 8.304, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 9.693860054016113, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 8.0954, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 8.443461418151855, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 8.1607, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 10.055970191955566, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 8.1962, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 9.612964630126953, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 8.2455, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 8.767157554626465, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.7366, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 9.725934982299805, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.6859, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 9.836505889892578, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.7574, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15415395808265944, | |
| "eval_loss": 12.814332962036133, | |
| "eval_runtime": 0.8606, | |
| "eval_samples_per_second": 556.563, | |
| "eval_steps_per_second": 5.81, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 9.989017486572266, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.7933, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 10.352441787719727, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.7838, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 9.68741512298584, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.2839, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 9.891133308410645, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.334, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 11.549692153930664, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.3987, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 10.629108428955078, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.4309, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 10.113447189331055, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.315, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 10.212636947631836, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 7.0154, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 10.056076049804688, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 7.0637, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 11.2257080078125, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 7.1062, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 10.78973388671875, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 7.1318, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 8.973546028137207, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.9624, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 11.215229988098145, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.8239, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 10.821270942687988, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.8606, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 10.789911270141602, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.887, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 11.016899108886719, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.9095, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 9.650943756103516, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.7177, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 9.670699119567871, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.6991, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 11.04586410522461, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.7035, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 9.242329597473145, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.719, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15134013681731523, | |
| "eval_loss": 14.174110412597656, | |
| "eval_runtime": 0.8775, | |
| "eval_samples_per_second": 545.883, | |
| "eval_steps_per_second": 5.698, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 6.37711238861084, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.748, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 6.800008773803711, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.5747, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 9.045692443847656, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.5866, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 9.95047664642334, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.6027, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 10.001580238342285, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.6279, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 8.160560607910156, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.5971, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 8.65356159210205, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.5167, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 9.773505210876465, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.5273, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 9.659945487976074, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.5435, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 7.946258068084717, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.562, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 8.297663688659668, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.5008, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 8.944379806518555, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.4634, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 9.064960479736328, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.4673, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 7.672547340393066, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.4689, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 8.471962928771973, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.4705, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 6.679225921630859, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.4459, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 10.19491958618164, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.4203, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.086577415466309, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.4171, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 7.2403364181518555, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.4183, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 6.269439220428467, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.4292, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.15050326632461378, | |
| "eval_loss": 14.587776184082031, | |
| "eval_runtime": 0.8608, | |
| "eval_samples_per_second": 556.467, | |
| "eval_steps_per_second": 5.809, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.631340026855469, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.3794, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 6.36605167388916, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.3895, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 8.747396469116211, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.3771, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 7.057696342468262, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.3772, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 5.619050979614258, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.3749, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 5.275052070617676, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.3415, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 8.303990364074707, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.3393, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 7.621928691864014, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.3525, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 5.604230880737305, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.3384, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 7.690316677093506, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.3327, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 6.64884614944458, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.3133, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 6.544528007507324, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.3259, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 7.8797078132629395, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.3097, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 8.240303039550781, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.3138, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.800259590148926, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.2993, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 9.30471420288086, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.3028, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 6.491007328033447, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.3099, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 5.998480796813965, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.2958, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 7.8639421463012695, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.2853, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.69437063436032e+17, | |
| "train_loss": 8.341223693386434, | |
| "train_runtime": 17177.2378, | |
| "train_samples_per_second": 332.765, | |
| "train_steps_per_second": 3.467 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.69437063436032e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |