| { | |
| "best_global_step": 10000, | |
| "best_metric": 9.169220924377441, | |
| "best_model_checkpoint": "distilbert_rand_20_v2/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": 2.839494466781616, | |
| "learning_rate": 5e-06, | |
| "loss": 12.295, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 2.4242842197418213, | |
| "learning_rate": 1e-05, | |
| "loss": 10.5891, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.7410306930541992, | |
| "learning_rate": 1.5e-05, | |
| "loss": 10.0631, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.7641091346740723, | |
| "learning_rate": 2e-05, | |
| "loss": 9.9119, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 2.0337624549865723, | |
| "learning_rate": 2.5e-05, | |
| "loss": 9.7769, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 2.2860662937164307, | |
| "learning_rate": 3e-05, | |
| "loss": 9.6625, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 2.1490907669067383, | |
| "learning_rate": 3.5e-05, | |
| "loss": 9.5659, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.8452446460723877, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4905, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 2.248866558074951, | |
| "learning_rate": 4.5e-05, | |
| "loss": 9.4311, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 2.205521583557129, | |
| "learning_rate": 5e-05, | |
| "loss": 9.3799, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 2.088688850402832, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 9.3392, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.5509796142578125, | |
| "learning_rate": 6e-05, | |
| "loss": 9.304, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 2.811840772628784, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 9.2708, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 2.690173864364624, | |
| "learning_rate": 7e-05, | |
| "loss": 9.2417, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 2.387943983078003, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 9.2027, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 2.3690969944000244, | |
| "learning_rate": 8e-05, | |
| "loss": 9.1765, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 3.117220163345337, | |
| "learning_rate": 8.5e-05, | |
| "loss": 9.162, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 3.0220932960510254, | |
| "learning_rate": 9e-05, | |
| "loss": 9.1394, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 3.0078608989715576, | |
| "learning_rate": 9.5e-05, | |
| "loss": 9.1265, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 3.7584919929504395, | |
| "learning_rate": 0.0001, | |
| "loss": 9.0191, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15056067225020173, | |
| "eval_loss": 9.169220924377441, | |
| "eval_runtime": 0.8946, | |
| "eval_samples_per_second": 535.46, | |
| "eval_steps_per_second": 5.589, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.652250289916992, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 9.0207, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.9136319160461426, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 9.02, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 3.528905153274536, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 9.0215, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 3.356295347213745, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 8.9843, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 3.516859531402588, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 8.864, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 3.8037853240966797, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 8.8851, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.843942880630493, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 8.8948, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.761806011199951, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 8.8942, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 3.2125625610351562, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 8.8272, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.9537875652313232, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 8.7602, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.773268461227417, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 8.776, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.131239891052246, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 8.7919, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 4.001720428466797, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 8.7982, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.204725503921509, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 8.6755, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 3.509864568710327, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 8.632, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 3.831369638442993, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 8.6673, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 3.5109853744506836, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 8.6815, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.758779764175415, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 8.6855, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 3.7631304264068604, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 8.4605, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.984078884124756, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 8.4686, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1512668705631471, | |
| "eval_loss": 9.461175918579102, | |
| "eval_runtime": 0.7985, | |
| "eval_samples_per_second": 599.851, | |
| "eval_steps_per_second": 6.261, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 4.056648254394531, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 8.4752, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 4.289050579071045, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 8.5025, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 4.93964147567749, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 8.462, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 5.5340776443481445, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 8.1657, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 5.351807594299316, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 8.1843, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 5.357039451599121, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 8.2259, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 5.8689751625061035, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 8.2237, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 5.938015460968018, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.0956, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 6.712480545043945, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 7.874, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 5.908409595489502, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 7.8939, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 5.897430419921875, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 7.9246, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 6.488427639007568, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 7.953, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 7.22756290435791, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.7254, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 7.378805637359619, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.6003, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 6.353366374969482, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.6439, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 8.001075744628906, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.655, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 6.13035774230957, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.6915, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 7.017744064331055, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.386, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 8.221517562866211, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.3464, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 7.083924293518066, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.4003, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.1536782606262417, | |
| "eval_loss": 10.904170036315918, | |
| "eval_runtime": 0.802, | |
| "eval_samples_per_second": 597.253, | |
| "eval_steps_per_second": 6.234, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 8.425161361694336, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.4267, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 8.940080642700195, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.4089, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 8.634459495544434, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.0836, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 7.144838333129883, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.1341, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 8.895588874816895, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.1816, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 7.708846092224121, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.1807, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 8.748336791992188, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.0967, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 9.473174095153809, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.8908, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 8.400636672973633, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.9241, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 8.310261726379395, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.9588, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 9.908066749572754, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.967, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 9.162622451782227, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.8265, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 9.188130378723145, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.7211, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 8.647799491882324, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.7462, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 8.854975700378418, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.774, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 9.705062866210938, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.7905, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 8.551140785217285, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.6067, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 8.363558769226074, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.5872, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 10.148078918457031, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.5978, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 11.427934646606445, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.6138, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15131210048222496, | |
| "eval_loss": 12.232389450073242, | |
| "eval_runtime": 0.7967, | |
| "eval_samples_per_second": 601.267, | |
| "eval_steps_per_second": 6.276, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 8.80695629119873, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.6346, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 6.59151554107666, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.4606, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 9.461859703063965, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.4748, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 8.955385208129883, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.4959, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 9.80893611907959, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.5126, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 7.227355003356934, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.4837, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 4.860254764556885, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.3988, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 7.772268295288086, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.406, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 9.385651588439941, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.4196, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 9.672471046447754, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.4373, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 8.28320598602295, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.3845, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 7.17496395111084, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.3499, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 9.062373161315918, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.3459, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 7.60693883895874, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.35, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 7.330896377563477, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.3586, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 8.745220184326172, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.3213, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 6.972262859344482, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.3084, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 6.277470588684082, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.3001, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 5.862802982330322, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.3026, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 7.365855693817139, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.312, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1510920458687302, | |
| "eval_loss": 12.813960075378418, | |
| "eval_runtime": 0.8028, | |
| "eval_samples_per_second": 596.652, | |
| "eval_steps_per_second": 6.228, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.677397727966309, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.2693, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 9.135784149169922, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.2795, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 7.724148750305176, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.2656, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 5.842483997344971, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.2682, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 7.466794013977051, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.2664, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 4.366174697875977, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.2354, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 7.936322212219238, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.2374, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 7.256034851074219, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.2416, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 6.694550514221191, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.2335, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 5.7823486328125, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.225, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 8.693979263305664, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.2114, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 6.171530723571777, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.2232, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 5.320813179016113, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.204, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 7.7308526039123535, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.2156, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 5.540814399719238, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.2013, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 9.152900695800781, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.2049, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 6.086564064025879, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.2016, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 7.7911200523376465, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.1998, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 4.825611591339111, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.1863, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 7.68356804521728e+17, | |
| "train_loss": 7.6570130803823275, | |
| "train_runtime": 16738.0825, | |
| "train_samples_per_second": 341.495, | |
| "train_steps_per_second": 3.558 | |
| } | |
| ], | |
| "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": 7.68356804521728e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |