| { | |
| "best_metric": 7.744482040405273, | |
| "best_model_checkpoint": "tiny_bert_rand_5_v1/checkpoint-20000", | |
| "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": 2.296325922012329, | |
| "learning_rate": 5e-06, | |
| "loss": 11.3941, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.128565788269043, | |
| "learning_rate": 1e-05, | |
| "loss": 9.8095, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.2296267747879028, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.8105, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.2943933010101318, | |
| "learning_rate": 2e-05, | |
| "loss": 8.6094, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 0.8330835700035095, | |
| "learning_rate": 2.5e-05, | |
| "loss": 8.4914, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.40715491771698, | |
| "learning_rate": 3e-05, | |
| "loss": 8.3893, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.4087949991226196, | |
| "learning_rate": 3.5e-05, | |
| "loss": 8.3034, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.6161038875579834, | |
| "learning_rate": 4e-05, | |
| "loss": 8.2222, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.071186900138855, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.1479, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.3544769287109375, | |
| "learning_rate": 5e-05, | |
| "loss": 8.0885, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.1537610292434692, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 8.0462, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 1.2293767929077148, | |
| "learning_rate": 6e-05, | |
| "loss": 8.0097, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.351851463317871, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 7.9774, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 1.8882099390029907, | |
| "learning_rate": 7e-05, | |
| "loss": 7.9489, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.214233636856079, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 7.9149, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 1.7951445579528809, | |
| "learning_rate": 8e-05, | |
| "loss": 7.8879, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 1.2143298387527466, | |
| "learning_rate": 8.5e-05, | |
| "loss": 7.8731, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 1.1728650331497192, | |
| "learning_rate": 9e-05, | |
| "loss": 7.8471, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.3379387855529785, | |
| "learning_rate": 9.5e-05, | |
| "loss": 7.8277, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.231257200241089, | |
| "learning_rate": 0.0001, | |
| "loss": 7.7645, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15067197195247503, | |
| "eval_loss": 7.783882141113281, | |
| "eval_runtime": 0.7102, | |
| "eval_samples_per_second": 674.483, | |
| "eval_steps_per_second": 7.041, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 1.648000717163086, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 7.7614, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.4700474739074707, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 7.7482, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 1.8770407438278198, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 7.7463, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 1.7996339797973633, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 7.7213, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 1.8939964771270752, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 7.6707, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.41560959815979, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 7.6691, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 1.3531227111816406, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 7.67, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 1.5285168886184692, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 7.6661, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.4376230239868164, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 7.6371, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 1.776288628578186, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 7.611, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 1.978027105331421, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 7.6098, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 1.6938847303390503, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 7.6109, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.155803680419922, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 7.6092, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.119781970977783, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 7.5721, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 1.7056002616882324, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 7.5552, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.276197671890259, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 7.5622, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 1.658673882484436, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 7.565, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.461730480194092, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 7.5638, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 1.8282390832901, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 7.5167, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.56581449508667, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 7.5214, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15255759140267686, | |
| "eval_loss": 7.744482040405273, | |
| "eval_runtime": 0.6877, | |
| "eval_samples_per_second": 696.504, | |
| "eval_steps_per_second": 7.27, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.0388848781585693, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 7.5131, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 1.852391242980957, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 7.5251, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 2.0081534385681152, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 7.5127, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 2.1764028072357178, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 7.4671, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 2.2152099609375, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 7.4709, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 2.148036003112793, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 7.477, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 1.8313390016555786, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 7.4735, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 1.8666034936904907, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 7.4446, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 1.9521493911743164, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.4135, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.6559360027313232, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.4122, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 2.373878002166748, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 7.4207, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 1.8715249300003052, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 7.427, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 2.4790990352630615, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.3754, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 2.7676239013671875, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.3462, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 2.7188451290130615, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.3556, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 3.002106189727783, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.3596, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 2.4832465648651123, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.3633, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 3.367208957672119, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.275, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 2.0318615436553955, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.2689, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 2.4273407459259033, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.2816, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15353834960964827, | |
| "eval_loss": 7.907983303070068, | |
| "eval_runtime": 0.7121, | |
| "eval_samples_per_second": 672.625, | |
| "eval_steps_per_second": 7.021, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 2.8654825687408447, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.2815, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 3.3296194076538086, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.2818, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 3.609745979309082, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.1648, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 4.198052406311035, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.1789, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 4.00711727142334, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.1951, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 3.822047233581543, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.2034, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 3.3816990852355957, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.1622, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 3.332118034362793, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 7.0627, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 4.029077053070068, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 7.0832, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 3.8176214694976807, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 7.0973, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 4.951391696929932, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 7.1071, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 4.661400318145752, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 7.0199, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 4.034855365753174, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.9707, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 4.592999458312988, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.9801, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 4.279331684112549, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 7.0003, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 4.479630470275879, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 7.0133, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 4.38205623626709, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.8985, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 5.646921157836914, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.8804, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 5.613239288330078, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.893, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 5.378262996673584, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.9052, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15134013681731523, | |
| "eval_loss": 8.425745964050293, | |
| "eval_runtime": 0.7013, | |
| "eval_samples_per_second": 683.007, | |
| "eval_steps_per_second": 7.13, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 4.779518127441406, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.9154, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 6.122113227844238, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.7719, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 6.412014484405518, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.7918, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 5.896696090698242, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.809, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 5.116279602050781, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.8147, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 5.979940414428711, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.7859, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 5.590941905975342, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.6989, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 5.550159454345703, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.7156, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 7.668396472930908, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.7305, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 5.5945916175842285, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.7375, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 6.6677165031433105, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.6784, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 6.023616313934326, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.6276, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 6.2595696449279785, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.6418, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 6.898407459259033, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.6483, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 7.0118913650512695, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.6534, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 7.157012939453125, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.5934, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 7.569313049316406, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.5774, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 5.967202663421631, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.5749, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 7.105442047119141, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.5829, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 6.400492191314697, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.5942, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1511761572321754, | |
| "eval_loss": 8.98286247253418, | |
| "eval_runtime": 0.6926, | |
| "eval_samples_per_second": 691.566, | |
| "eval_steps_per_second": 7.219, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 6.482155799865723, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.5127, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 6.460001468658447, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.5204, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 8.494600296020508, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.5306, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 6.867691516876221, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.5288, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 6.988873481750488, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.5146, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 5.474238872528076, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.4608, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 6.552234172821045, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.4694, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 6.802420139312744, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.4798, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 6.713523864746094, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.4744, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 8.395071029663086, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.4556, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 7.446748733520508, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.4272, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 5.941086292266846, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.4356, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 7.137095928192139, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.43, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 7.807168960571289, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.4359, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.458153247833252, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.4146, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 6.837516784667969, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.4083, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 7.141629695892334, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.4043, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 7.046724319458008, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.4063, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 7.116812705993652, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.4027, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.049796905526016e+17, | |
| "train_loss": 7.277992412269265, | |
| "train_runtime": 11581.9358, | |
| "train_samples_per_second": 493.525, | |
| "train_steps_per_second": 5.142 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.049796905526016e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |