| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.2770083102493075, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002770083102493075, | |
| "grad_norm": 15.603601455688477, | |
| "learning_rate": 7.5e-06, | |
| "loss": 1.4608, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00554016620498615, | |
| "grad_norm": 14.760490417480469, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.428, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008310249307479225, | |
| "grad_norm": 19.810142517089844, | |
| "learning_rate": 2.25e-05, | |
| "loss": 1.2831, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0110803324099723, | |
| "grad_norm": 6.9140729904174805, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0688, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013850415512465374, | |
| "grad_norm": 6.4214253425598145, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.2067, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01662049861495845, | |
| "grad_norm": 11.87337589263916, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.4816, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019390581717451522, | |
| "grad_norm": 6.669366359710693, | |
| "learning_rate": 5.25e-05, | |
| "loss": 1.1287, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0221606648199446, | |
| "grad_norm": 8.616178512573242, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2203, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024930747922437674, | |
| "grad_norm": 9.265680313110352, | |
| "learning_rate": 5.999881193903433e-05, | |
| "loss": 1.1897, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.027700831024930747, | |
| "grad_norm": 5.8777852058410645, | |
| "learning_rate": 5.999524785023657e-05, | |
| "loss": 1.0788, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030470914127423823, | |
| "grad_norm": 5.538949012756348, | |
| "learning_rate": 5.998930801589704e-05, | |
| "loss": 1.0736, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0332409972299169, | |
| "grad_norm": 5.686643123626709, | |
| "learning_rate": 5.9980992906474764e-05, | |
| "loss": 1.1573, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.036011080332409975, | |
| "grad_norm": 6.211599349975586, | |
| "learning_rate": 5.9970303180560196e-05, | |
| "loss": 1.1618, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.038781163434903045, | |
| "grad_norm": 5.079687595367432, | |
| "learning_rate": 5.995723968482309e-05, | |
| "loss": 0.9948, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04155124653739612, | |
| "grad_norm": 5.150979042053223, | |
| "learning_rate": 5.994180345394539e-05, | |
| "loss": 1.0328, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0443213296398892, | |
| "grad_norm": 5.891695499420166, | |
| "learning_rate": 5.9923995710539324e-05, | |
| "loss": 1.1854, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04709141274238227, | |
| "grad_norm": 5.33378791809082, | |
| "learning_rate": 5.990381786505055e-05, | |
| "loss": 1.0424, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04986149584487535, | |
| "grad_norm": 4.58151912689209, | |
| "learning_rate": 5.988127151564644e-05, | |
| "loss": 1.0392, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 5.48449182510376, | |
| "learning_rate": 5.9856358448089506e-05, | |
| "loss": 1.147, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.055401662049861494, | |
| "grad_norm": 5.061849117279053, | |
| "learning_rate": 5.9829080635595944e-05, | |
| "loss": 1.1072, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05817174515235457, | |
| "grad_norm": 4.828271389007568, | |
| "learning_rate": 5.979944023867938e-05, | |
| "loss": 1.1132, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.060941828254847646, | |
| "grad_norm": 5.035257816314697, | |
| "learning_rate": 5.976743960497973e-05, | |
| "loss": 1.0272, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06371191135734072, | |
| "grad_norm": 5.55271577835083, | |
| "learning_rate": 5.973308126907723e-05, | |
| "loss": 1.1674, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0664819944598338, | |
| "grad_norm": 4.926880836486816, | |
| "learning_rate": 5.969636795229172e-05, | |
| "loss": 1.0765, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06925207756232687, | |
| "grad_norm": 5.437366962432861, | |
| "learning_rate": 5.965730256246713e-05, | |
| "loss": 1.1344, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07202216066481995, | |
| "grad_norm": 4.4522786140441895, | |
| "learning_rate": 5.9615888193741094e-05, | |
| "loss": 0.9976, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07479224376731301, | |
| "grad_norm": 5.204329967498779, | |
| "learning_rate": 5.9572128126299946e-05, | |
| "loss": 1.1298, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07756232686980609, | |
| "grad_norm": 5.264033317565918, | |
| "learning_rate": 5.9526025826118885e-05, | |
| "loss": 1.0428, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08033240997229917, | |
| "grad_norm": 5.672830581665039, | |
| "learning_rate": 5.947758494468746e-05, | |
| "loss": 1.1636, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08310249307479224, | |
| "grad_norm": 5.798141002655029, | |
| "learning_rate": 5.942680931872036e-05, | |
| "loss": 1.1112, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08587257617728532, | |
| "grad_norm": 4.760410308837891, | |
| "learning_rate": 5.937370296985354e-05, | |
| "loss": 1.0444, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0886426592797784, | |
| "grad_norm": 5.4676313400268555, | |
| "learning_rate": 5.931827010432566e-05, | |
| "loss": 1.0957, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.09141274238227147, | |
| "grad_norm": 4.983479976654053, | |
| "learning_rate": 5.9260515112644995e-05, | |
| "loss": 1.0734, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09418282548476455, | |
| "grad_norm": 5.242843151092529, | |
| "learning_rate": 5.9200442569241606e-05, | |
| "loss": 1.0879, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09695290858725762, | |
| "grad_norm": 5.12119722366333, | |
| "learning_rate": 5.9138057232105084e-05, | |
| "loss": 1.0841, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0997229916897507, | |
| "grad_norm": 5.087031841278076, | |
| "learning_rate": 5.9073364042407705e-05, | |
| "loss": 1.0343, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10249307479224377, | |
| "grad_norm": 5.226377487182617, | |
| "learning_rate": 5.900636812411301e-05, | |
| "loss": 1.0775, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 5.388728141784668, | |
| "learning_rate": 5.893707478357005e-05, | |
| "loss": 1.0439, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10803324099722991, | |
| "grad_norm": 5.6382951736450195, | |
| "learning_rate": 5.886548950909301e-05, | |
| "loss": 1.096, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.11080332409972299, | |
| "grad_norm": 5.046523094177246, | |
| "learning_rate": 5.879161797052658e-05, | |
| "loss": 0.9928, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11357340720221606, | |
| "grad_norm": 5.98545503616333, | |
| "learning_rate": 5.8715466018796865e-05, | |
| "loss": 1.1639, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11634349030470914, | |
| "grad_norm": 5.064982891082764, | |
| "learning_rate": 5.863703968544795e-05, | |
| "loss": 0.9878, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11911357340720222, | |
| "grad_norm": 5.36715841293335, | |
| "learning_rate": 5.8556345182164176e-05, | |
| "loss": 1.056, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.12188365650969529, | |
| "grad_norm": 4.7610368728637695, | |
| "learning_rate": 5.8473388900278206e-05, | |
| "loss": 0.9625, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12465373961218837, | |
| "grad_norm": 5.5188889503479, | |
| "learning_rate": 5.83881774102647e-05, | |
| "loss": 1.0828, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12742382271468145, | |
| "grad_norm": 5.703125476837158, | |
| "learning_rate": 5.8300717461220027e-05, | |
| "loss": 1.0632, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.13019390581717452, | |
| "grad_norm": 5.264132499694824, | |
| "learning_rate": 5.821101598032759e-05, | |
| "loss": 1.0057, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1329639889196676, | |
| "grad_norm": 6.05692720413208, | |
| "learning_rate": 5.811908007230929e-05, | |
| "loss": 1.0565, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13573407202216067, | |
| "grad_norm": 4.773650169372559, | |
| "learning_rate": 5.802491701886268e-05, | |
| "loss": 0.9764, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13850415512465375, | |
| "grad_norm": 5.461302757263184, | |
| "learning_rate": 5.792853427808431e-05, | |
| "loss": 1.0002, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14127423822714683, | |
| "grad_norm": 5.83071231842041, | |
| "learning_rate": 5.7829939483878996e-05, | |
| "loss": 1.0353, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1440443213296399, | |
| "grad_norm": 5.715704917907715, | |
| "learning_rate": 5.772914044535516e-05, | |
| "loss": 0.985, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14681440443213298, | |
| "grad_norm": 5.309004306793213, | |
| "learning_rate": 5.762614514620634e-05, | |
| "loss": 1.0106, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14958448753462603, | |
| "grad_norm": 5.693915843963623, | |
| "learning_rate": 5.752096174407884e-05, | |
| "loss": 1.0049, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1523545706371191, | |
| "grad_norm": 5.644995212554932, | |
| "learning_rate": 5.741359856992561e-05, | |
| "loss": 1.0411, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.15512465373961218, | |
| "grad_norm": 5.765170574188232, | |
| "learning_rate": 5.73040641273464e-05, | |
| "loss": 1.0039, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 5.643161296844482, | |
| "learning_rate": 5.719236709191428e-05, | |
| "loss": 1.0173, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.16066481994459833, | |
| "grad_norm": 5.821686267852783, | |
| "learning_rate": 5.707851631048841e-05, | |
| "loss": 0.9725, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1634349030470914, | |
| "grad_norm": 6.034667015075684, | |
| "learning_rate": 5.6962520800513414e-05, | |
| "loss": 1.0381, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16620498614958448, | |
| "grad_norm": 5.336913585662842, | |
| "learning_rate": 5.6844389749305136e-05, | |
| "loss": 0.9266, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16897506925207756, | |
| "grad_norm": 5.65543270111084, | |
| "learning_rate": 5.672413251332297e-05, | |
| "loss": 0.9768, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.17174515235457063, | |
| "grad_norm": 5.3824615478515625, | |
| "learning_rate": 5.6601758617428766e-05, | |
| "loss": 0.9276, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1745152354570637, | |
| "grad_norm": 5.195915222167969, | |
| "learning_rate": 5.647727775413245e-05, | |
| "loss": 0.9332, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1772853185595568, | |
| "grad_norm": 5.240732669830322, | |
| "learning_rate": 5.6350699782824346e-05, | |
| "loss": 0.9012, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.18005540166204986, | |
| "grad_norm": 5.112541675567627, | |
| "learning_rate": 5.622203472899423e-05, | |
| "loss": 0.9211, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.18282548476454294, | |
| "grad_norm": 6.605805397033691, | |
| "learning_rate": 5.609129278343731e-05, | |
| "loss": 0.9847, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.18559556786703602, | |
| "grad_norm": 5.1862568855285645, | |
| "learning_rate": 5.595848430144705e-05, | |
| "loss": 0.9064, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1883656509695291, | |
| "grad_norm": 6.03863525390625, | |
| "learning_rate": 5.582361980199504e-05, | |
| "loss": 0.9405, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.19113573407202217, | |
| "grad_norm": 6.4853901863098145, | |
| "learning_rate": 5.568670996689773e-05, | |
| "loss": 0.9676, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.19390581717451524, | |
| "grad_norm": 5.415260314941406, | |
| "learning_rate": 5.554776563997056e-05, | |
| "loss": 0.8873, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19667590027700832, | |
| "grad_norm": 5.11320686340332, | |
| "learning_rate": 5.540679782616892e-05, | |
| "loss": 0.8752, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1994459833795014, | |
| "grad_norm": 4.8276214599609375, | |
| "learning_rate": 5.52638176907166e-05, | |
| "loss": 0.8441, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.20221606648199447, | |
| "grad_norm": 5.574111461639404, | |
| "learning_rate": 5.5118836558221475e-05, | |
| "loss": 0.907, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.20498614958448755, | |
| "grad_norm": 5.575868606567383, | |
| "learning_rate": 5.497186591177849e-05, | |
| "loss": 0.8691, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2077562326869806, | |
| "grad_norm": 5.388404846191406, | |
| "learning_rate": 5.4822917392060184e-05, | |
| "loss": 0.8935, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 6.2548747062683105, | |
| "learning_rate": 5.46720027963947e-05, | |
| "loss": 0.8814, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.21329639889196675, | |
| "grad_norm": 5.358329772949219, | |
| "learning_rate": 5.4519134077831395e-05, | |
| "loss": 0.9041, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.21606648199445982, | |
| "grad_norm": 6.0121564865112305, | |
| "learning_rate": 5.4364323344194095e-05, | |
| "loss": 0.8568, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2188365650969529, | |
| "grad_norm": 6.2294020652771, | |
| "learning_rate": 5.420758285712211e-05, | |
| "loss": 0.9205, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.22160664819944598, | |
| "grad_norm": 5.889683723449707, | |
| "learning_rate": 5.404892503109906e-05, | |
| "loss": 0.8264, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.22437673130193905, | |
| "grad_norm": 5.401793956756592, | |
| "learning_rate": 5.388836243246963e-05, | |
| "loss": 0.8574, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.22714681440443213, | |
| "grad_norm": 6.194601058959961, | |
| "learning_rate": 5.372590777844421e-05, | |
| "loss": 0.8742, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2299168975069252, | |
| "grad_norm": 5.504939079284668, | |
| "learning_rate": 5.356157393609167e-05, | |
| "loss": 0.8261, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.23268698060941828, | |
| "grad_norm": 5.818756103515625, | |
| "learning_rate": 5.339537392132025e-05, | |
| "loss": 0.8364, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.23545706371191136, | |
| "grad_norm": 5.8196024894714355, | |
| "learning_rate": 5.322732089784661e-05, | |
| "loss": 0.8577, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.23822714681440443, | |
| "grad_norm": 6.102773189544678, | |
| "learning_rate": 5.305742817615325e-05, | |
| "loss": 0.8708, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2409972299168975, | |
| "grad_norm": 6.225861549377441, | |
| "learning_rate": 5.288570921243423e-05, | |
| "loss": 0.8654, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.24376731301939059, | |
| "grad_norm": 6.0107340812683105, | |
| "learning_rate": 5.2712177607529405e-05, | |
| "loss": 0.8401, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.24653739612188366, | |
| "grad_norm": 6.3431715965271, | |
| "learning_rate": 5.2536847105847185e-05, | |
| "loss": 0.855, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.24930747922437674, | |
| "grad_norm": 5.196923732757568, | |
| "learning_rate": 5.235973159427591e-05, | |
| "loss": 0.756, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2520775623268698, | |
| "grad_norm": 6.157230377197266, | |
| "learning_rate": 5.218084510108397e-05, | |
| "loss": 0.856, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2548476454293629, | |
| "grad_norm": 5.813492774963379, | |
| "learning_rate": 5.200020179480868e-05, | |
| "loss": 0.7829, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.25761772853185594, | |
| "grad_norm": 6.468319416046143, | |
| "learning_rate": 5.181781598313409e-05, | |
| "loss": 0.8327, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.26038781163434904, | |
| "grad_norm": 6.1590423583984375, | |
| "learning_rate": 5.16337021117578e-05, | |
| "loss": 0.8035, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 6.378724575042725, | |
| "learning_rate": 5.14478747632467e-05, | |
| "loss": 0.8134, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2659279778393352, | |
| "grad_norm": 5.47311544418335, | |
| "learning_rate": 5.126034865588208e-05, | |
| "loss": 0.7149, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.26869806094182824, | |
| "grad_norm": 5.790994644165039, | |
| "learning_rate": 5.107113864249381e-05, | |
| "loss": 0.7714, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.27146814404432135, | |
| "grad_norm": 5.880769729614258, | |
| "learning_rate": 5.088025970928399e-05, | |
| "loss": 0.7519, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2742382271468144, | |
| "grad_norm": 5.894231796264648, | |
| "learning_rate": 5.068772697463992e-05, | |
| "loss": 0.8124, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.2770083102493075, | |
| "grad_norm": 6.011329174041748, | |
| "learning_rate": 5.0493555687936704e-05, | |
| "loss": 0.7536, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 361, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2857642925601587e+17, | |
| "train_batch_size": 9, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |