| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.978930307941653, | |
| "eval_steps": 500, | |
| "global_step": 153, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.019448946515397084, | |
| "grad_norm": 6.861098289489746, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.2627, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03889789303079417, | |
| "grad_norm": 6.737518310546875, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.2374, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05834683954619125, | |
| "grad_norm": 6.920078754425049, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 1.2864, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07779578606158834, | |
| "grad_norm": 6.555400371551514, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.2097, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09724473257698542, | |
| "grad_norm": 6.175917625427246, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.1939, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1166936790923825, | |
| "grad_norm": 4.607879638671875, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.1667, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.13614262560777957, | |
| "grad_norm": 2.9558024406433105, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 1.1707, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15559157212317667, | |
| "grad_norm": 2.6493217945098877, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0802, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.17504051863857376, | |
| "grad_norm": 4.4779953956604, | |
| "learning_rate": 5.625e-06, | |
| "loss": 1.1318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.19448946515397084, | |
| "grad_norm": 4.794122695922852, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.162, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.21393841166936792, | |
| "grad_norm": 4.527707099914551, | |
| "learning_rate": 6.875e-06, | |
| "loss": 1.1654, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.233387358184765, | |
| "grad_norm": 3.3236396312713623, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.0345, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.25283630470016205, | |
| "grad_norm": 3.3806722164154053, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 1.0519, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.27228525121555913, | |
| "grad_norm": 2.476033926010132, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.0189, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2917341977309562, | |
| "grad_norm": 2.0179576873779297, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.9914, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.31118314424635335, | |
| "grad_norm": 1.9065077304840088, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0208, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.33063209076175043, | |
| "grad_norm": 1.8468468189239502, | |
| "learning_rate": 9.998685442495921e-06, | |
| "loss": 0.9893, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3500810372771475, | |
| "grad_norm": 1.5280859470367432, | |
| "learning_rate": 9.994742461208251e-06, | |
| "loss": 0.9482, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3695299837925446, | |
| "grad_norm": 1.369767665863037, | |
| "learning_rate": 9.988173129447251e-06, | |
| "loss": 0.9942, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3889789303079417, | |
| "grad_norm": 1.207467794418335, | |
| "learning_rate": 9.978980901518663e-06, | |
| "loss": 0.9676, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.40842787682333875, | |
| "grad_norm": 1.4069736003875732, | |
| "learning_rate": 9.96717061090737e-06, | |
| "loss": 0.9322, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.42787682333873583, | |
| "grad_norm": 1.1611406803131104, | |
| "learning_rate": 9.95274846773583e-06, | |
| "loss": 0.9786, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.4473257698541329, | |
| "grad_norm": 1.1889172792434692, | |
| "learning_rate": 9.935722055498655e-06, | |
| "loss": 0.9256, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.46677471636953, | |
| "grad_norm": 1.1732343435287476, | |
| "learning_rate": 9.916100327075038e-06, | |
| "loss": 0.8946, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4862236628849271, | |
| "grad_norm": 1.3214962482452393, | |
| "learning_rate": 9.893893600021112e-06, | |
| "loss": 0.9087, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5056726094003241, | |
| "grad_norm": 1.1180990934371948, | |
| "learning_rate": 9.869113551144754e-06, | |
| "loss": 0.9407, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5251215559157212, | |
| "grad_norm": 0.9630658626556396, | |
| "learning_rate": 9.841773210365646e-06, | |
| "loss": 0.9329, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5445705024311183, | |
| "grad_norm": 0.9198795557022095, | |
| "learning_rate": 9.811886953863841e-06, | |
| "loss": 0.9378, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5640194489465153, | |
| "grad_norm": 0.9277356266975403, | |
| "learning_rate": 9.779470496520442e-06, | |
| "loss": 0.9195, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5834683954619124, | |
| "grad_norm": 0.80616694688797, | |
| "learning_rate": 9.744540883654348e-06, | |
| "loss": 0.932, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6029173419773096, | |
| "grad_norm": 0.8931137323379517, | |
| "learning_rate": 9.707116482059447e-06, | |
| "loss": 0.9024, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6223662884927067, | |
| "grad_norm": 0.8059653043746948, | |
| "learning_rate": 9.667216970346916e-06, | |
| "loss": 0.9376, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.6418152350081038, | |
| "grad_norm": 0.7000671625137329, | |
| "learning_rate": 9.624863328597767e-06, | |
| "loss": 0.8836, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6612641815235009, | |
| "grad_norm": 0.788222074508667, | |
| "learning_rate": 9.580077827331038e-06, | |
| "loss": 0.8801, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6807131280388979, | |
| "grad_norm": 0.960985004901886, | |
| "learning_rate": 9.532884015793432e-06, | |
| "loss": 0.8777, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.700162074554295, | |
| "grad_norm": 0.9860551357269287, | |
| "learning_rate": 9.48330670957659e-06, | |
| "loss": 0.905, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7196110210696921, | |
| "grad_norm": 0.8046407699584961, | |
| "learning_rate": 9.431371977568483e-06, | |
| "loss": 0.9257, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7390599675850892, | |
| "grad_norm": 0.9749986529350281, | |
| "learning_rate": 9.377107128245782e-06, | |
| "loss": 0.9026, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7585089141004863, | |
| "grad_norm": 1.0727578401565552, | |
| "learning_rate": 9.32054069531444e-06, | |
| "loss": 0.9418, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7779578606158833, | |
| "grad_norm": 0.8145267963409424, | |
| "learning_rate": 9.261702422706014e-06, | |
| "loss": 0.906, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7974068071312804, | |
| "grad_norm": 0.8878953456878662, | |
| "learning_rate": 9.200623248937619e-06, | |
| "loss": 0.911, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8168557536466775, | |
| "grad_norm": 0.8581916689872742, | |
| "learning_rate": 9.13733529084374e-06, | |
| "loss": 0.8873, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8363047001620746, | |
| "grad_norm": 0.6927335858345032, | |
| "learning_rate": 9.071871826688472e-06, | |
| "loss": 0.8726, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8557536466774717, | |
| "grad_norm": 0.6785300970077515, | |
| "learning_rate": 9.004267278667032e-06, | |
| "loss": 0.9282, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8752025931928687, | |
| "grad_norm": 0.735817015171051, | |
| "learning_rate": 8.934557194805787e-06, | |
| "loss": 0.9498, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8946515397082658, | |
| "grad_norm": 0.6314980387687683, | |
| "learning_rate": 8.862778230270276e-06, | |
| "loss": 0.8851, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.9141004862236629, | |
| "grad_norm": 0.6786671280860901, | |
| "learning_rate": 8.788968128091084e-06, | |
| "loss": 0.8834, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.93354943273906, | |
| "grad_norm": 0.7222251892089844, | |
| "learning_rate": 8.71316569931769e-06, | |
| "loss": 0.8915, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.9529983792544571, | |
| "grad_norm": 0.7341165542602539, | |
| "learning_rate": 8.635410802610724e-06, | |
| "loss": 0.8392, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9724473257698542, | |
| "grad_norm": 0.7214636206626892, | |
| "learning_rate": 8.555744323283364e-06, | |
| "loss": 0.8845, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9918962722852512, | |
| "grad_norm": 0.848508358001709, | |
| "learning_rate": 8.474208151802898e-06, | |
| "loss": 0.8581, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.012965964343598, | |
| "grad_norm": 1.2223843336105347, | |
| "learning_rate": 8.390845161763756e-06, | |
| "loss": 1.4462, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.032414910858995, | |
| "grad_norm": 0.6608096361160278, | |
| "learning_rate": 8.305699187343586e-06, | |
| "loss": 0.9298, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.0518638573743921, | |
| "grad_norm": 0.7156009674072266, | |
| "learning_rate": 8.218815000254233e-06, | |
| "loss": 0.7472, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0713128038897892, | |
| "grad_norm": 0.7926570177078247, | |
| "learning_rate": 8.130238286199747e-06, | |
| "loss": 0.8235, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.0907617504051863, | |
| "grad_norm": 0.8245845437049866, | |
| "learning_rate": 8.04001562085379e-06, | |
| "loss": 0.9211, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.1102106969205834, | |
| "grad_norm": 0.6690172553062439, | |
| "learning_rate": 7.948194445369065e-06, | |
| "loss": 0.9178, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.1296596434359805, | |
| "grad_norm": 0.8226907849311829, | |
| "learning_rate": 7.85482304143168e-06, | |
| "loss": 0.7913, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.1491085899513775, | |
| "grad_norm": 0.9551813006401062, | |
| "learning_rate": 7.759950505873523e-06, | |
| "loss": 0.8701, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.1685575364667746, | |
| "grad_norm": 0.6174091696739197, | |
| "learning_rate": 7.66362672485601e-06, | |
| "loss": 0.8091, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1880064829821717, | |
| "grad_norm": 0.8523845076560974, | |
| "learning_rate": 7.565902347638806e-06, | |
| "loss": 0.8805, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.2074554294975688, | |
| "grad_norm": 0.6692298054695129, | |
| "learning_rate": 7.466828759947271e-06, | |
| "loss": 0.8076, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.2269043760129659, | |
| "grad_norm": 0.6524062156677246, | |
| "learning_rate": 7.366458056952668e-06, | |
| "loss": 0.7676, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.246353322528363, | |
| "grad_norm": 0.7085561752319336, | |
| "learning_rate": 7.264843015879321e-06, | |
| "loss": 0.9008, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.26580226904376, | |
| "grad_norm": 0.6387475728988647, | |
| "learning_rate": 7.162037068253141e-06, | |
| "loss": 0.7623, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2852512155591573, | |
| "grad_norm": 0.6962757110595703, | |
| "learning_rate": 7.058094271806091e-06, | |
| "loss": 0.8989, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.3047001620745542, | |
| "grad_norm": 0.6996948719024658, | |
| "learning_rate": 6.953069282051397e-06, | |
| "loss": 0.8975, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3241491085899515, | |
| "grad_norm": 0.6775169968605042, | |
| "learning_rate": 6.84701732354442e-06, | |
| "loss": 0.744, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.3435980551053484, | |
| "grad_norm": 0.5788021087646484, | |
| "learning_rate": 6.7399941608443096e-06, | |
| "loss": 0.8693, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.3630470016207457, | |
| "grad_norm": 0.797266960144043, | |
| "learning_rate": 6.632056069191723e-06, | |
| "loss": 0.8185, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3824959481361425, | |
| "grad_norm": 0.6203992366790771, | |
| "learning_rate": 6.523259804918001e-06, | |
| "loss": 0.8219, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.4019448946515398, | |
| "grad_norm": 0.5873000621795654, | |
| "learning_rate": 6.413662575601391e-06, | |
| "loss": 0.8675, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.4213938411669367, | |
| "grad_norm": 0.5779026746749878, | |
| "learning_rate": 6.303322009985984e-06, | |
| "loss": 0.8004, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.440842787682334, | |
| "grad_norm": 0.7285134792327881, | |
| "learning_rate": 6.1922961276791925e-06, | |
| "loss": 0.8649, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.4602917341977308, | |
| "grad_norm": 0.5539451837539673, | |
| "learning_rate": 6.08064330864371e-06, | |
| "loss": 0.7551, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4797406807131281, | |
| "grad_norm": 0.530315101146698, | |
| "learning_rate": 5.968422262499983e-06, | |
| "loss": 0.798, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.499189627228525, | |
| "grad_norm": 0.6449223756790161, | |
| "learning_rate": 5.85569199765534e-06, | |
| "loss": 0.8498, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.5186385737439223, | |
| "grad_norm": 0.5452909469604492, | |
| "learning_rate": 5.7425117902760195e-06, | |
| "loss": 0.7277, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.5380875202593192, | |
| "grad_norm": 0.6189656853675842, | |
| "learning_rate": 5.628941153118388e-06, | |
| "loss": 0.8525, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.5575364667747165, | |
| "grad_norm": 0.5398715734481812, | |
| "learning_rate": 5.515039804235772e-06, | |
| "loss": 0.7927, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.5769854132901133, | |
| "grad_norm": 0.5609137415885925, | |
| "learning_rate": 5.400867635577335e-06, | |
| "loss": 0.8495, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.5964343598055106, | |
| "grad_norm": 0.6827127933502197, | |
| "learning_rate": 5.2864846814955e-06, | |
| "loss": 0.7973, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.6158833063209075, | |
| "grad_norm": 0.6233929395675659, | |
| "learning_rate": 5.17195108717852e-06, | |
| "loss": 0.8542, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.6353322528363048, | |
| "grad_norm": 0.46808454394340515, | |
| "learning_rate": 5.057327077024745e-06, | |
| "loss": 0.7751, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.6547811993517016, | |
| "grad_norm": 0.5673578381538391, | |
| "learning_rate": 4.942672922975255e-06, | |
| "loss": 0.7718, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.674230145867099, | |
| "grad_norm": 0.6883957386016846, | |
| "learning_rate": 4.82804891282148e-06, | |
| "loss": 0.7642, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.6936790923824958, | |
| "grad_norm": 0.6758309006690979, | |
| "learning_rate": 4.713515318504501e-06, | |
| "loss": 0.9642, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.7131280388978931, | |
| "grad_norm": 0.4832862913608551, | |
| "learning_rate": 4.599132364422666e-06, | |
| "loss": 0.7228, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.73257698541329, | |
| "grad_norm": 0.7116863131523132, | |
| "learning_rate": 4.4849601957642295e-06, | |
| "loss": 0.8748, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.7520259319286873, | |
| "grad_norm": 0.5377712845802307, | |
| "learning_rate": 4.371058846881614e-06, | |
| "loss": 0.8228, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.7714748784440841, | |
| "grad_norm": 0.49350258708000183, | |
| "learning_rate": 4.257488209723981e-06, | |
| "loss": 0.7398, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.7909238249594814, | |
| "grad_norm": 0.5975490212440491, | |
| "learning_rate": 4.1443080023446605e-06, | |
| "loss": 0.8464, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.8103727714748783, | |
| "grad_norm": 0.5417091846466064, | |
| "learning_rate": 4.0315777375000185e-06, | |
| "loss": 0.8294, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.8298217179902756, | |
| "grad_norm": 0.5055018663406372, | |
| "learning_rate": 3.9193566913562915e-06, | |
| "loss": 0.7465, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.8492706645056725, | |
| "grad_norm": 0.5458024144172668, | |
| "learning_rate": 3.807703872320809e-06, | |
| "loss": 0.8468, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.8687196110210698, | |
| "grad_norm": 0.5673462748527527, | |
| "learning_rate": 3.6966779900140193e-06, | |
| "loss": 0.83, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.8881685575364666, | |
| "grad_norm": 0.4504033625125885, | |
| "learning_rate": 3.586337424398609e-06, | |
| "loss": 0.8286, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.907617504051864, | |
| "grad_norm": 0.5393473505973816, | |
| "learning_rate": 3.4767401950820003e-06, | |
| "loss": 0.7724, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.9270664505672608, | |
| "grad_norm": 0.6027775406837463, | |
| "learning_rate": 3.3679439308082777e-06, | |
| "loss": 0.8476, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.946515397082658, | |
| "grad_norm": 0.5444150567054749, | |
| "learning_rate": 3.260005839155691e-06, | |
| "loss": 0.7916, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.965964343598055, | |
| "grad_norm": 0.5643745064735413, | |
| "learning_rate": 3.152982676455581e-06, | |
| "loss": 0.7754, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.9854132901134522, | |
| "grad_norm": 0.6249105334281921, | |
| "learning_rate": 3.046930717948604e-06, | |
| "loss": 0.8918, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.006482982171799, | |
| "grad_norm": 0.8598921895027161, | |
| "learning_rate": 2.9419057281939106e-06, | |
| "loss": 1.2974, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.025931928687196, | |
| "grad_norm": 0.5371376276016235, | |
| "learning_rate": 2.8379629317468604e-06, | |
| "loss": 0.7535, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.0453808752025933, | |
| "grad_norm": 0.5264458656311035, | |
| "learning_rate": 2.7351569841206792e-06, | |
| "loss": 0.7966, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.06482982171799, | |
| "grad_norm": 0.537636399269104, | |
| "learning_rate": 2.633541943047334e-06, | |
| "loss": 0.8117, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.0842787682333874, | |
| "grad_norm": 0.49134597182273865, | |
| "learning_rate": 2.53317124005273e-06, | |
| "loss": 0.8327, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.1037277147487843, | |
| "grad_norm": 0.5109895467758179, | |
| "learning_rate": 2.4340976523611957e-06, | |
| "loss": 0.7766, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.1231766612641816, | |
| "grad_norm": 0.5269736051559448, | |
| "learning_rate": 2.3363732751439926e-06, | |
| "loss": 0.7372, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.1426256077795784, | |
| "grad_norm": 0.5074533820152283, | |
| "learning_rate": 2.240049494126479e-06, | |
| "loss": 0.7629, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.1620745542949757, | |
| "grad_norm": 0.5118080973625183, | |
| "learning_rate": 2.1451769585683196e-06, | |
| "loss": 0.6913, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.1815235008103726, | |
| "grad_norm": 0.5797163248062134, | |
| "learning_rate": 2.0518055546309362e-06, | |
| "loss": 0.8875, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.20097244732577, | |
| "grad_norm": 0.5009273886680603, | |
| "learning_rate": 1.9599843791462123e-06, | |
| "loss": 0.8202, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.2204213938411668, | |
| "grad_norm": 0.4513159394264221, | |
| "learning_rate": 1.8697617138002545e-06, | |
| "loss": 0.6574, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.239870340356564, | |
| "grad_norm": 0.483067125082016, | |
| "learning_rate": 1.7811849997457681e-06, | |
| "loss": 0.766, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.259319286871961, | |
| "grad_norm": 0.5190160274505615, | |
| "learning_rate": 1.6943008126564164e-06, | |
| "loss": 0.86, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.2787682333873582, | |
| "grad_norm": 0.45720207691192627, | |
| "learning_rate": 1.609154838236246e-06, | |
| "loss": 0.7715, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.298217179902755, | |
| "grad_norm": 0.46229684352874756, | |
| "learning_rate": 1.5257918481971028e-06, | |
| "loss": 0.7503, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.3176661264181524, | |
| "grad_norm": 0.4342985153198242, | |
| "learning_rate": 1.4442556767166371e-06, | |
| "loss": 0.7868, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.3371150729335493, | |
| "grad_norm": 0.4891180098056793, | |
| "learning_rate": 1.3645891973892772e-06, | |
| "loss": 0.7287, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.3565640194489466, | |
| "grad_norm": 0.451238214969635, | |
| "learning_rate": 1.2868343006823113e-06, | |
| "loss": 0.7135, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.3760129659643434, | |
| "grad_norm": 0.5156052112579346, | |
| "learning_rate": 1.211031871908916e-06, | |
| "loss": 0.8523, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.3954619124797407, | |
| "grad_norm": 0.4339147210121155, | |
| "learning_rate": 1.137221769729725e-06, | |
| "loss": 0.7157, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.4149108589951376, | |
| "grad_norm": 0.4493626356124878, | |
| "learning_rate": 1.065442805194214e-06, | |
| "loss": 0.8451, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.434359805510535, | |
| "grad_norm": 0.4212321937084198, | |
| "learning_rate": 9.957327213329687e-07, | |
| "loss": 0.7459, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.4538087520259317, | |
| "grad_norm": 0.49532240629196167, | |
| "learning_rate": 9.281281733115288e-07, | |
| "loss": 0.8074, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.473257698541329, | |
| "grad_norm": 0.4537409543991089, | |
| "learning_rate": 8.626647091562612e-07, | |
| "loss": 0.7581, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.492706645056726, | |
| "grad_norm": 0.4133566617965698, | |
| "learning_rate": 7.993767510623834e-07, | |
| "loss": 0.761, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.512155591572123, | |
| "grad_norm": 0.47344374656677246, | |
| "learning_rate": 7.382975772939866e-07, | |
| "loss": 0.783, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.53160453808752, | |
| "grad_norm": 0.4699486494064331, | |
| "learning_rate": 6.794593046855613e-07, | |
| "loss": 0.7739, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.5510534846029174, | |
| "grad_norm": 0.4478330910205841, | |
| "learning_rate": 6.228928717542205e-07, | |
| "loss": 0.7713, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.5705024311183147, | |
| "grad_norm": 0.43865975737571716, | |
| "learning_rate": 5.686280224315189e-07, | |
| "loss": 0.7643, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.5899513776337115, | |
| "grad_norm": 0.45029687881469727, | |
| "learning_rate": 5.166932904234101e-07, | |
| "loss": 0.8228, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.6094003241491084, | |
| "grad_norm": 0.43190592527389526, | |
| "learning_rate": 4.671159842065698e-07, | |
| "loss": 0.8042, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.6288492706645057, | |
| "grad_norm": 0.4283352494239807, | |
| "learning_rate": 4.199221726689634e-07, | |
| "loss": 0.6469, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.648298217179903, | |
| "grad_norm": 0.444757342338562, | |
| "learning_rate": 3.751366714022342e-07, | |
| "loss": 0.84, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.6677471636953, | |
| "grad_norm": 0.4214431941509247, | |
| "learning_rate": 3.3278302965308593e-07, | |
| "loss": 0.8341, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.6871961102106967, | |
| "grad_norm": 0.44082775712013245, | |
| "learning_rate": 2.928835179405548e-07, | |
| "loss": 0.7816, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.706645056726094, | |
| "grad_norm": 0.4211124777793884, | |
| "learning_rate": 2.5545911634565266e-07, | |
| "loss": 0.7624, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.7260940032414913, | |
| "grad_norm": 0.4169267416000366, | |
| "learning_rate": 2.205295034795596e-07, | |
| "loss": 0.7452, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.745542949756888, | |
| "grad_norm": 0.4236930012702942, | |
| "learning_rate": 1.881130461361591e-07, | |
| "loss": 0.7443, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.764991896272285, | |
| "grad_norm": 0.4934384822845459, | |
| "learning_rate": 1.5822678963435479e-07, | |
| "loss": 0.8023, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.7844408427876823, | |
| "grad_norm": 0.43219056725502014, | |
| "learning_rate": 1.3088644885524637e-07, | |
| "loss": 0.7411, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.8038897893030796, | |
| "grad_norm": 0.4079039990901947, | |
| "learning_rate": 1.0610639997888917e-07, | |
| "loss": 0.742, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.8233387358184765, | |
| "grad_norm": 0.39644140005111694, | |
| "learning_rate": 8.38996729249636e-08, | |
| "loss": 0.7157, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.8427876823338734, | |
| "grad_norm": 0.4225790500640869, | |
| "learning_rate": 6.427794450134529e-08, | |
| "loss": 0.8331, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.8622366288492707, | |
| "grad_norm": 0.4538624882698059, | |
| "learning_rate": 4.72515322641709e-08, | |
| "loss": 0.9018, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.881685575364668, | |
| "grad_norm": 0.4045908451080322, | |
| "learning_rate": 3.282938909263122e-08, | |
| "loss": 0.6865, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.901134521880065, | |
| "grad_norm": 0.4469752609729767, | |
| "learning_rate": 2.101909848133743e-08, | |
| "loss": 0.7197, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.9205834683954617, | |
| "grad_norm": 0.520662248134613, | |
| "learning_rate": 1.1826870552749669e-08, | |
| "loss": 0.8062, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.940032414910859, | |
| "grad_norm": 0.42809349298477173, | |
| "learning_rate": 5.257538791749173e-09, | |
| "loss": 0.799, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.9594813614262563, | |
| "grad_norm": 0.4747503101825714, | |
| "learning_rate": 1.3145575040801605e-09, | |
| "loss": 0.8033, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.978930307941653, | |
| "grad_norm": 0.41074883937835693, | |
| "learning_rate": 0.0, | |
| "loss": 0.6416, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.978930307941653, | |
| "step": 153, | |
| "total_flos": 132192675135488.0, | |
| "train_loss": 0.8684253560172187, | |
| "train_runtime": 7585.4265, | |
| "train_samples_per_second": 1.952, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 153, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 132192675135488.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |