| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 110, | |
| "global_step": 438, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00228310502283105, | |
| "grad_norm": 49763.3828125, | |
| "learning_rate": 1e-05, | |
| "loss": 1.458, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00228310502283105, | |
| "eval_loss": 1.3469293117523193, | |
| "eval_runtime": 5.899, | |
| "eval_samples_per_second": 16.952, | |
| "eval_steps_per_second": 8.476, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0045662100456621, | |
| "grad_norm": 47645.4765625, | |
| "learning_rate": 2e-05, | |
| "loss": 1.663, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00684931506849315, | |
| "grad_norm": 53930.67578125, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2375, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0091324200913242, | |
| "grad_norm": 79495.359375, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4099, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01141552511415525, | |
| "grad_norm": 51649.53125, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3848, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0136986301369863, | |
| "grad_norm": 62501.76171875, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3113, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01598173515981735, | |
| "grad_norm": 56830.796875, | |
| "learning_rate": 7e-05, | |
| "loss": 1.394, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0182648401826484, | |
| "grad_norm": 69029.3671875, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1999, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02054794520547945, | |
| "grad_norm": 77872.7421875, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4121, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0228310502283105, | |
| "grad_norm": 67018.5234375, | |
| "learning_rate": 0.0001, | |
| "loss": 1.5292, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02511415525114155, | |
| "grad_norm": 52883.80078125, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 1.1904, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0273972602739726, | |
| "grad_norm": 89700.34375, | |
| "learning_rate": 0.00012, | |
| "loss": 1.6428, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02968036529680365, | |
| "grad_norm": 70959.6875, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 1.8412, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0319634703196347, | |
| "grad_norm": 123958.625, | |
| "learning_rate": 0.00014, | |
| "loss": 1.3549, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03424657534246575, | |
| "grad_norm": 55201.86328125, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.717, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0365296803652968, | |
| "grad_norm": 79495.921875, | |
| "learning_rate": 0.00016, | |
| "loss": 2.5554, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03881278538812785, | |
| "grad_norm": 94823.359375, | |
| "learning_rate": 0.00017, | |
| "loss": 1.3731, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0410958904109589, | |
| "grad_norm": 78739.7890625, | |
| "learning_rate": 0.00018, | |
| "loss": 0.9813, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.04337899543378995, | |
| "grad_norm": 70018.5234375, | |
| "learning_rate": 0.00019, | |
| "loss": 1.3944, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.045662100456621, | |
| "grad_norm": 93789.5625, | |
| "learning_rate": 0.0002, | |
| "loss": 1.3604, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04794520547945205, | |
| "grad_norm": 69996.328125, | |
| "learning_rate": 0.0001999971756719333, | |
| "loss": 0.7975, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0502283105022831, | |
| "grad_norm": 62440.68359375, | |
| "learning_rate": 0.00019998870284726968, | |
| "loss": 1.3632, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05251141552511415, | |
| "grad_norm": 57081.87890625, | |
| "learning_rate": 0.00019997458200460993, | |
| "loss": 1.181, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0547945205479452, | |
| "grad_norm": 68026.21875, | |
| "learning_rate": 0.00019995481394159188, | |
| "loss": 1.2883, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05707762557077625, | |
| "grad_norm": 68773.109375, | |
| "learning_rate": 0.0001999293997748454, | |
| "loss": 1.209, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0593607305936073, | |
| "grad_norm": 54953.078125, | |
| "learning_rate": 0.00019989834093992945, | |
| "loss": 1.2207, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06164383561643835, | |
| "grad_norm": 68915.734375, | |
| "learning_rate": 0.00019986163919125075, | |
| "loss": 1.1395, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0639269406392694, | |
| "grad_norm": 77723.328125, | |
| "learning_rate": 0.00019981929660196492, | |
| "loss": 1.4178, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06621004566210045, | |
| "grad_norm": 57558.64453125, | |
| "learning_rate": 0.0001997713155638592, | |
| "loss": 1.3991, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0684931506849315, | |
| "grad_norm": 57487.20703125, | |
| "learning_rate": 0.00019971769878721743, | |
| "loss": 1.2361, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07077625570776255, | |
| "grad_norm": 56958.12109375, | |
| "learning_rate": 0.000199658449300667, | |
| "loss": 1.0886, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0730593607305936, | |
| "grad_norm": 69826.890625, | |
| "learning_rate": 0.00019959357045100764, | |
| "loss": 1.2462, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07534246575342465, | |
| "grad_norm": 83606.59375, | |
| "learning_rate": 0.00019952306590302247, | |
| "loss": 1.5512, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0776255707762557, | |
| "grad_norm": 96047.1171875, | |
| "learning_rate": 0.00019944693963927092, | |
| "loss": 1.2642, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07990867579908675, | |
| "grad_norm": 75340.171875, | |
| "learning_rate": 0.00019936519595986394, | |
| "loss": 1.3945, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0821917808219178, | |
| "grad_norm": 53603.67578125, | |
| "learning_rate": 0.00019927783948222084, | |
| "loss": 0.8141, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08447488584474885, | |
| "grad_norm": 60950.3359375, | |
| "learning_rate": 0.00019918487514080865, | |
| "loss": 1.1856, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0867579908675799, | |
| "grad_norm": 57694.49609375, | |
| "learning_rate": 0.00019908630818686338, | |
| "loss": 0.5208, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08904109589041095, | |
| "grad_norm": 83846.890625, | |
| "learning_rate": 0.0001989821441880933, | |
| "loss": 1.2401, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.091324200913242, | |
| "grad_norm": 60294.91796875, | |
| "learning_rate": 0.00019887238902836448, | |
| "loss": 1.3362, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09360730593607305, | |
| "grad_norm": 55908.76953125, | |
| "learning_rate": 0.00019875704890736853, | |
| "loss": 1.0295, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0958904109589041, | |
| "grad_norm": 79842.359375, | |
| "learning_rate": 0.00019863613034027224, | |
| "loss": 1.3764, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09817351598173515, | |
| "grad_norm": 53915.5703125, | |
| "learning_rate": 0.0001985096401573497, | |
| "loss": 1.1399, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1004566210045662, | |
| "grad_norm": 56798.44921875, | |
| "learning_rate": 0.00019837758550359636, | |
| "loss": 0.9945, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.10273972602739725, | |
| "grad_norm": 72487.9921875, | |
| "learning_rate": 0.0001982399738383255, | |
| "loss": 1.4328, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1050228310502283, | |
| "grad_norm": 56317.5546875, | |
| "learning_rate": 0.00019809681293474693, | |
| "loss": 1.2217, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10730593607305935, | |
| "grad_norm": 108634.3203125, | |
| "learning_rate": 0.0001979481108795278, | |
| "loss": 1.5159, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1095890410958904, | |
| "grad_norm": 77436.7578125, | |
| "learning_rate": 0.00019779387607233586, | |
| "loss": 1.051, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.11187214611872145, | |
| "grad_norm": 61779.69921875, | |
| "learning_rate": 0.00019763411722536502, | |
| "loss": 1.2487, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1141552511415525, | |
| "grad_norm": 74456.3359375, | |
| "learning_rate": 0.00019746884336284317, | |
| "loss": 1.5362, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11643835616438356, | |
| "grad_norm": 65273.90625, | |
| "learning_rate": 0.00019729806382052248, | |
| "loss": 1.125, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1187214611872146, | |
| "grad_norm": 74326.5390625, | |
| "learning_rate": 0.00019712178824515212, | |
| "loss": 1.4872, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.12100456621004566, | |
| "grad_norm": 66311.375, | |
| "learning_rate": 0.00019694002659393305, | |
| "loss": 1.3129, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.1232876712328767, | |
| "grad_norm": 93956.140625, | |
| "learning_rate": 0.00019675278913395606, | |
| "loss": 1.1963, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12557077625570776, | |
| "grad_norm": 54108.7421875, | |
| "learning_rate": 0.0001965600864416213, | |
| "loss": 1.3549, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1278538812785388, | |
| "grad_norm": 82672.5234375, | |
| "learning_rate": 0.00019636192940204134, | |
| "loss": 1.3415, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.13013698630136986, | |
| "grad_norm": 79705.0234375, | |
| "learning_rate": 0.00019615832920842586, | |
| "loss": 1.0711, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.1324200913242009, | |
| "grad_norm": 66569.171875, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 1.5752, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13470319634703196, | |
| "grad_norm": 54818.6953125, | |
| "learning_rate": 0.0001957348456686032, | |
| "loss": 1.2086, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.136986301369863, | |
| "grad_norm": 84023.5625, | |
| "learning_rate": 0.00019551498624352496, | |
| "loss": 1.2713, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13926940639269406, | |
| "grad_norm": 58488.77734375, | |
| "learning_rate": 0.00019528973150531787, | |
| "loss": 1.1957, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1415525114155251, | |
| "grad_norm": 59256.1328125, | |
| "learning_rate": 0.00019505909417784754, | |
| "loss": 1.1863, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.14383561643835616, | |
| "grad_norm": 58009.8359375, | |
| "learning_rate": 0.00019482308728902356, | |
| "loss": 1.0046, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1461187214611872, | |
| "grad_norm": 58970.30859375, | |
| "learning_rate": 0.00019458172417006347, | |
| "loss": 1.3599, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.14840182648401826, | |
| "grad_norm": 75271.78125, | |
| "learning_rate": 0.00019433501845473995, | |
| "loss": 1.175, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.1506849315068493, | |
| "grad_norm": 60125.44140625, | |
| "learning_rate": 0.00019408298407861042, | |
| "loss": 1.2096, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.15296803652968036, | |
| "grad_norm": 62565.88671875, | |
| "learning_rate": 0.00019382563527823026, | |
| "loss": 1.0284, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1552511415525114, | |
| "grad_norm": 64562.3359375, | |
| "learning_rate": 0.00019356298659034817, | |
| "loss": 1.1955, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15753424657534246, | |
| "grad_norm": 61627.109375, | |
| "learning_rate": 0.00019329505285108542, | |
| "loss": 1.1498, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.1598173515981735, | |
| "grad_norm": 65598.3515625, | |
| "learning_rate": 0.00019302184919509755, | |
| "loss": 1.046, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16210045662100456, | |
| "grad_norm": 56694.41015625, | |
| "learning_rate": 0.00019274339105471971, | |
| "loss": 0.6779, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1643835616438356, | |
| "grad_norm": 82800.1015625, | |
| "learning_rate": 0.00019245969415909465, | |
| "loss": 1.2381, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 44540.94140625, | |
| "learning_rate": 0.00019217077453328449, | |
| "loss": 1.0871, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.1689497716894977, | |
| "grad_norm": 62427.3125, | |
| "learning_rate": 0.0001918766484973654, | |
| "loss": 1.4182, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.17123287671232876, | |
| "grad_norm": 62395.83203125, | |
| "learning_rate": 0.00019157733266550575, | |
| "loss": 1.1391, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1735159817351598, | |
| "grad_norm": 86479.0546875, | |
| "learning_rate": 0.0001912728439450276, | |
| "loss": 1.4769, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.17579908675799086, | |
| "grad_norm": 52077.796875, | |
| "learning_rate": 0.00019096319953545185, | |
| "loss": 0.9951, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1780821917808219, | |
| "grad_norm": 65181.8671875, | |
| "learning_rate": 0.0001906484169275263, | |
| "loss": 1.1435, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.18036529680365296, | |
| "grad_norm": 61927.88671875, | |
| "learning_rate": 0.00019032851390223812, | |
| "loss": 1.3638, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.182648401826484, | |
| "grad_norm": 65557.1796875, | |
| "learning_rate": 0.00019000350852980909, | |
| "loss": 1.0751, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18493150684931506, | |
| "grad_norm": 67026.625, | |
| "learning_rate": 0.00018967341916867518, | |
| "loss": 1.4392, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.1872146118721461, | |
| "grad_norm": 65608.7578125, | |
| "learning_rate": 0.00018933826446444933, | |
| "loss": 1.4204, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18949771689497716, | |
| "grad_norm": 62439.32421875, | |
| "learning_rate": 0.0001889980633488683, | |
| "loss": 1.6256, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1917808219178082, | |
| "grad_norm": 58315.05078125, | |
| "learning_rate": 0.00018865283503872324, | |
| "loss": 1.3327, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.19406392694063926, | |
| "grad_norm": 63276.6953125, | |
| "learning_rate": 0.00018830259903477426, | |
| "loss": 1.3005, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1963470319634703, | |
| "grad_norm": 60068.3125, | |
| "learning_rate": 0.0001879473751206489, | |
| "loss": 1.3073, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.19863013698630136, | |
| "grad_norm": 60654.34375, | |
| "learning_rate": 0.0001875871833617246, | |
| "loss": 1.1668, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2009132420091324, | |
| "grad_norm": 49492.6875, | |
| "learning_rate": 0.0001872220441039952, | |
| "loss": 0.8938, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.20319634703196346, | |
| "grad_norm": 82807.6953125, | |
| "learning_rate": 0.0001868519779729218, | |
| "loss": 1.014, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2054794520547945, | |
| "grad_norm": 50824.89453125, | |
| "learning_rate": 0.0001864770058722676, | |
| "loss": 1.0941, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20776255707762556, | |
| "grad_norm": 53105.984375, | |
| "learning_rate": 0.00018609714898291718, | |
| "loss": 0.7681, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2100456621004566, | |
| "grad_norm": 57766.73046875, | |
| "learning_rate": 0.00018571242876167996, | |
| "loss": 1.0531, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.21232876712328766, | |
| "grad_norm": 66334.5625, | |
| "learning_rate": 0.0001853228669400784, | |
| "loss": 1.2699, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2146118721461187, | |
| "grad_norm": 54520.6015625, | |
| "learning_rate": 0.00018492848552312014, | |
| "loss": 1.4723, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.21689497716894976, | |
| "grad_norm": 75962.671875, | |
| "learning_rate": 0.00018452930678805536, | |
| "loss": 1.379, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2191780821917808, | |
| "grad_norm": 57191.44921875, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 1.3189, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.22146118721461186, | |
| "grad_norm": 63262.0625, | |
| "learning_rate": 0.00018371664782625287, | |
| "loss": 0.9332, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.2237442922374429, | |
| "grad_norm": 55938.12890625, | |
| "learning_rate": 0.00018330321350382544, | |
| "loss": 1.3675, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.22602739726027396, | |
| "grad_norm": 48929.921875, | |
| "learning_rate": 0.00018288507366931905, | |
| "loss": 1.0751, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.228310502283105, | |
| "grad_norm": 60707.87890625, | |
| "learning_rate": 0.00018246225194201517, | |
| "loss": 1.2708, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23059360730593606, | |
| "grad_norm": 59498.01171875, | |
| "learning_rate": 0.00018203477220565912, | |
| "loss": 1.3661, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.2328767123287671, | |
| "grad_norm": 52212.64453125, | |
| "learning_rate": 0.00018160265860711134, | |
| "loss": 0.9827, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.23515981735159816, | |
| "grad_norm": 47361.78515625, | |
| "learning_rate": 0.00018116593555498307, | |
| "loss": 1.0684, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2374429223744292, | |
| "grad_norm": 63993.5078125, | |
| "learning_rate": 0.0001807246277182578, | |
| "loss": 1.2669, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23972602739726026, | |
| "grad_norm": 55152.2578125, | |
| "learning_rate": 0.0001802787600248977, | |
| "loss": 0.9088, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2420091324200913, | |
| "grad_norm": 49352.40625, | |
| "learning_rate": 0.0001798283576604356, | |
| "loss": 1.0416, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.24429223744292236, | |
| "grad_norm": 60283.234375, | |
| "learning_rate": 0.0001793734460665523, | |
| "loss": 1.3194, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2465753424657534, | |
| "grad_norm": 46655.90234375, | |
| "learning_rate": 0.00017891405093963938, | |
| "loss": 0.8581, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24885844748858446, | |
| "grad_norm": 65966.4609375, | |
| "learning_rate": 0.0001784501982293479, | |
| "loss": 0.9858, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2511415525114155, | |
| "grad_norm": 72021.5703125, | |
| "learning_rate": 0.00017798191413712243, | |
| "loss": 1.0411, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2511415525114155, | |
| "eval_loss": 1.1379607915878296, | |
| "eval_runtime": 6.0407, | |
| "eval_samples_per_second": 16.554, | |
| "eval_steps_per_second": 8.277, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2534246575342466, | |
| "grad_norm": 60144.625, | |
| "learning_rate": 0.0001775092251147211, | |
| "loss": 1.1175, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2557077625570776, | |
| "grad_norm": 55304.96484375, | |
| "learning_rate": 0.0001770321578627213, | |
| "loss": 0.7432, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2579908675799087, | |
| "grad_norm": 50044.91015625, | |
| "learning_rate": 0.00017655073932901168, | |
| "loss": 0.8335, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2602739726027397, | |
| "grad_norm": 63828.10546875, | |
| "learning_rate": 0.0001760649967072697, | |
| "loss": 1.1305, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2625570776255708, | |
| "grad_norm": 56665.87890625, | |
| "learning_rate": 0.00017557495743542585, | |
| "loss": 1.1719, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2648401826484018, | |
| "grad_norm": 63755.87109375, | |
| "learning_rate": 0.00017508064919411344, | |
| "loss": 1.0547, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.2671232876712329, | |
| "grad_norm": 55144.74609375, | |
| "learning_rate": 0.00017458209990510527, | |
| "loss": 0.8684, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2694063926940639, | |
| "grad_norm": 68380.3125, | |
| "learning_rate": 0.00017407933772973637, | |
| "loss": 1.1672, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.271689497716895, | |
| "grad_norm": 67942.5, | |
| "learning_rate": 0.00017357239106731317, | |
| "loss": 1.3715, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.273972602739726, | |
| "grad_norm": 50505.95703125, | |
| "learning_rate": 0.00017306128855350942, | |
| "loss": 0.9512, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2762557077625571, | |
| "grad_norm": 56973.859375, | |
| "learning_rate": 0.0001725460590587486, | |
| "loss": 0.9059, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2785388127853881, | |
| "grad_norm": 47352.71484375, | |
| "learning_rate": 0.00017202673168657318, | |
| "loss": 0.9492, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2808219178082192, | |
| "grad_norm": 60938.4921875, | |
| "learning_rate": 0.0001715033357720006, | |
| "loss": 1.1549, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2831050228310502, | |
| "grad_norm": 50557.2265625, | |
| "learning_rate": 0.00017097590087986633, | |
| "loss": 1.1143, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2853881278538813, | |
| "grad_norm": 61211.25390625, | |
| "learning_rate": 0.00017044445680315372, | |
| "loss": 1.1829, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2876712328767123, | |
| "grad_norm": 58093.75, | |
| "learning_rate": 0.00016990903356131124, | |
| "loss": 1.1329, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2899543378995434, | |
| "grad_norm": 62758.921875, | |
| "learning_rate": 0.00016936966139855663, | |
| "loss": 0.8082, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2922374429223744, | |
| "grad_norm": 55444.03515625, | |
| "learning_rate": 0.00016882637078216868, | |
| "loss": 1.0763, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.2945205479452055, | |
| "grad_norm": 45902.125, | |
| "learning_rate": 0.0001682791924007661, | |
| "loss": 0.5558, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2968036529680365, | |
| "grad_norm": 67659.0546875, | |
| "learning_rate": 0.00016772815716257412, | |
| "loss": 0.8364, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2990867579908676, | |
| "grad_norm": 58708.359375, | |
| "learning_rate": 0.0001671732961936785, | |
| "loss": 1.1031, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3013698630136986, | |
| "grad_norm": 44826.8671875, | |
| "learning_rate": 0.00016661464083626734, | |
| "loss": 0.6981, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3036529680365297, | |
| "grad_norm": 49845.3125, | |
| "learning_rate": 0.00016605222264686086, | |
| "loss": 0.932, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3059360730593607, | |
| "grad_norm": 68260.5703125, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 1.1092, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3082191780821918, | |
| "grad_norm": 51835.33203125, | |
| "learning_rate": 0.00016491622505909482, | |
| "loss": 1.175, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3105022831050228, | |
| "grad_norm": 47168.08203125, | |
| "learning_rate": 0.00016434270982933273, | |
| "loss": 1.1088, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3127853881278539, | |
| "grad_norm": 65501.48828125, | |
| "learning_rate": 0.0001637655601011454, | |
| "loss": 1.0026, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3150684931506849, | |
| "grad_norm": 58316.41015625, | |
| "learning_rate": 0.00016318480847573642, | |
| "loss": 1.073, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.317351598173516, | |
| "grad_norm": 47404.8125, | |
| "learning_rate": 0.00016260048775776804, | |
| "loss": 0.9182, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.319634703196347, | |
| "grad_norm": 67868.96875, | |
| "learning_rate": 0.00016201263095350833, | |
| "loss": 1.2406, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3219178082191781, | |
| "grad_norm": 62392.62109375, | |
| "learning_rate": 0.0001614212712689668, | |
| "loss": 0.6007, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.3242009132420091, | |
| "grad_norm": 44980.81640625, | |
| "learning_rate": 0.00016082644210801844, | |
| "loss": 0.9601, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3264840182648402, | |
| "grad_norm": 53141.99609375, | |
| "learning_rate": 0.00016022817707051724, | |
| "loss": 0.7017, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3287671232876712, | |
| "grad_norm": 54526.64453125, | |
| "learning_rate": 0.00015962650995039783, | |
| "loss": 1.1726, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3310502283105023, | |
| "grad_norm": 60199.36328125, | |
| "learning_rate": 0.00015902147473376694, | |
| "loss": 1.0703, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 62003.74609375, | |
| "learning_rate": 0.00015841310559698343, | |
| "loss": 1.384, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3356164383561644, | |
| "grad_norm": 74924.28125, | |
| "learning_rate": 0.0001578014369047279, | |
| "loss": 1.4119, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3378995433789954, | |
| "grad_norm": 53506.546875, | |
| "learning_rate": 0.00015718650320806142, | |
| "loss": 1.1047, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3401826484018265, | |
| "grad_norm": 95636.2578125, | |
| "learning_rate": 0.00015656833924247398, | |
| "loss": 1.2457, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3424657534246575, | |
| "grad_norm": 84311.3671875, | |
| "learning_rate": 0.00015594697992592232, | |
| "loss": 1.8571, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3447488584474886, | |
| "grad_norm": 49623.90625, | |
| "learning_rate": 0.00015532246035685756, | |
| "loss": 0.8779, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3470319634703196, | |
| "grad_norm": 75495.09375, | |
| "learning_rate": 0.00015469481581224272, | |
| "loss": 1.1597, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3493150684931507, | |
| "grad_norm": 61961.66796875, | |
| "learning_rate": 0.00015406408174555976, | |
| "loss": 1.1428, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.3515981735159817, | |
| "grad_norm": 63387.83203125, | |
| "learning_rate": 0.0001534302937848073, | |
| "loss": 1.2391, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.3538812785388128, | |
| "grad_norm": 59753.67578125, | |
| "learning_rate": 0.00015279348773048786, | |
| "loss": 1.6021, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3561643835616438, | |
| "grad_norm": 61863.0078125, | |
| "learning_rate": 0.00015215369955358566, | |
| "loss": 1.241, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.3584474885844749, | |
| "grad_norm": 57081.796875, | |
| "learning_rate": 0.0001515109653935348, | |
| "loss": 1.1971, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3607305936073059, | |
| "grad_norm": 71862.75, | |
| "learning_rate": 0.00015086532155617784, | |
| "loss": 1.1196, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.363013698630137, | |
| "grad_norm": 53156.65234375, | |
| "learning_rate": 0.00015021680451171498, | |
| "loss": 1.1128, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.365296803652968, | |
| "grad_norm": 65439.45703125, | |
| "learning_rate": 0.00014956545089264407, | |
| "loss": 1.1221, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3675799086757991, | |
| "grad_norm": 61590.7734375, | |
| "learning_rate": 0.0001489112974916912, | |
| "loss": 1.1785, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.3698630136986301, | |
| "grad_norm": 44939.07421875, | |
| "learning_rate": 0.00014825438125973264, | |
| "loss": 1.1794, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3721461187214612, | |
| "grad_norm": 52118.67578125, | |
| "learning_rate": 0.00014759473930370736, | |
| "loss": 0.9673, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.3744292237442922, | |
| "grad_norm": 56795.2734375, | |
| "learning_rate": 0.0001469324088845212, | |
| "loss": 1.0183, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.3767123287671233, | |
| "grad_norm": 54740.015625, | |
| "learning_rate": 0.00014626742741494206, | |
| "loss": 1.2797, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3789954337899543, | |
| "grad_norm": 54072.75, | |
| "learning_rate": 0.00014559983245748638, | |
| "loss": 1.1121, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3812785388127854, | |
| "grad_norm": 63319.68359375, | |
| "learning_rate": 0.00014492966172229777, | |
| "loss": 1.1224, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3835616438356164, | |
| "grad_norm": 70814.8828125, | |
| "learning_rate": 0.00014425695306501658, | |
| "loss": 1.5157, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3858447488584475, | |
| "grad_norm": 57420.37890625, | |
| "learning_rate": 0.00014358174448464154, | |
| "loss": 1.213, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3881278538812785, | |
| "grad_norm": 53243.4375, | |
| "learning_rate": 0.00014290407412138366, | |
| "loss": 1.3539, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3904109589041096, | |
| "grad_norm": 55964.484375, | |
| "learning_rate": 0.00014222398025451135, | |
| "loss": 1.0541, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3926940639269406, | |
| "grad_norm": 60372.61328125, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.9876, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3949771689497717, | |
| "grad_norm": 78811.0546875, | |
| "learning_rate": 0.0001408566758093048, | |
| "loss": 1.1479, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3972602739726027, | |
| "grad_norm": 68956.203125, | |
| "learning_rate": 0.00014016954246529696, | |
| "loss": 1.3174, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3995433789954338, | |
| "grad_norm": 54806.1328125, | |
| "learning_rate": 0.00013948014008196487, | |
| "loss": 1.1019, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4018264840182648, | |
| "grad_norm": 62154.96875, | |
| "learning_rate": 0.0001387885076012785, | |
| "loss": 1.2393, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4041095890410959, | |
| "grad_norm": 53926.55078125, | |
| "learning_rate": 0.00013809468409117846, | |
| "loss": 1.1682, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4063926940639269, | |
| "grad_norm": 55390.63671875, | |
| "learning_rate": 0.00013739870874336898, | |
| "loss": 1.0232, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.408675799086758, | |
| "grad_norm": 60079.51953125, | |
| "learning_rate": 0.00013670062087110422, | |
| "loss": 1.2492, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.410958904109589, | |
| "grad_norm": 68658.15625, | |
| "learning_rate": 0.00013600045990696762, | |
| "loss": 1.432, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4132420091324201, | |
| "grad_norm": 79412.75, | |
| "learning_rate": 0.0001352982654006444, | |
| "loss": 1.3679, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4155251141552511, | |
| "grad_norm": 76066.828125, | |
| "learning_rate": 0.00013459407701668763, | |
| "loss": 1.3032, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4178082191780822, | |
| "grad_norm": 74717.1328125, | |
| "learning_rate": 0.00013388793453227767, | |
| "loss": 1.1265, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4200913242009132, | |
| "grad_norm": 83329.0859375, | |
| "learning_rate": 0.0001331798778349752, | |
| "loss": 1.5738, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4223744292237443, | |
| "grad_norm": 56585.3203125, | |
| "learning_rate": 0.00013246994692046836, | |
| "loss": 1.1222, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4246575342465753, | |
| "grad_norm": 66546.9140625, | |
| "learning_rate": 0.00013175818189031327, | |
| "loss": 1.1622, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4269406392694064, | |
| "grad_norm": 89381.2734375, | |
| "learning_rate": 0.00013104462294966896, | |
| "loss": 1.0249, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4292237442922374, | |
| "grad_norm": 61278.7734375, | |
| "learning_rate": 0.00013032931040502627, | |
| "loss": 0.9255, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4315068493150685, | |
| "grad_norm": 73469.1484375, | |
| "learning_rate": 0.00012961228466193116, | |
| "loss": 1.2164, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4337899543378995, | |
| "grad_norm": 56180.0546875, | |
| "learning_rate": 0.00012889358622270223, | |
| "loss": 0.8783, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4360730593607306, | |
| "grad_norm": 55383.29296875, | |
| "learning_rate": 0.00012817325568414297, | |
| "loss": 1.0513, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4383561643835616, | |
| "grad_norm": 67318.65625, | |
| "learning_rate": 0.00012745133373524853, | |
| "loss": 1.2075, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4406392694063927, | |
| "grad_norm": 41490.49609375, | |
| "learning_rate": 0.0001267278611549073, | |
| "loss": 0.4758, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4429223744292237, | |
| "grad_norm": 56853.50390625, | |
| "learning_rate": 0.00012600287880959763, | |
| "loss": 1.2679, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.4452054794520548, | |
| "grad_norm": 34868.66796875, | |
| "learning_rate": 0.0001252764276510792, | |
| "loss": 0.5454, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4474885844748858, | |
| "grad_norm": 53731.953125, | |
| "learning_rate": 0.00012454854871407994, | |
| "loss": 1.2682, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.4497716894977169, | |
| "grad_norm": 57530.828125, | |
| "learning_rate": 0.00012381928311397806, | |
| "loss": 1.1872, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.4520547945205479, | |
| "grad_norm": 59639.96875, | |
| "learning_rate": 0.0001230886720444796, | |
| "loss": 1.2913, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.454337899543379, | |
| "grad_norm": 76171.09375, | |
| "learning_rate": 0.00012235675677529158, | |
| "loss": 1.314, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.45662100456621, | |
| "grad_norm": 73176.8984375, | |
| "learning_rate": 0.00012162357864979072, | |
| "loss": 0.9921, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4589041095890411, | |
| "grad_norm": 44535.90625, | |
| "learning_rate": 0.00012088917908268821, | |
| "loss": 0.9701, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4611872146118721, | |
| "grad_norm": 60352.7734375, | |
| "learning_rate": 0.00012015359955769021, | |
| "loss": 1.0526, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4634703196347032, | |
| "grad_norm": 60158.63671875, | |
| "learning_rate": 0.00011941688162515467, | |
| "loss": 0.9069, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.4657534246575342, | |
| "grad_norm": 56008.59375, | |
| "learning_rate": 0.00011867906689974428, | |
| "loss": 1.1581, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.4680365296803653, | |
| "grad_norm": 60848.1328125, | |
| "learning_rate": 0.00011794019705807584, | |
| "loss": 1.4732, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4703196347031963, | |
| "grad_norm": 73643.15625, | |
| "learning_rate": 0.00011720031383636585, | |
| "loss": 1.4934, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4726027397260274, | |
| "grad_norm": 60684.5625, | |
| "learning_rate": 0.00011645945902807341, | |
| "loss": 0.9241, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.4748858447488584, | |
| "grad_norm": 56336.984375, | |
| "learning_rate": 0.00011571767448153901, | |
| "loss": 0.7535, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.4771689497716895, | |
| "grad_norm": 57490.16015625, | |
| "learning_rate": 0.00011497500209762102, | |
| "loss": 1.0042, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.4794520547945205, | |
| "grad_norm": 58042.80859375, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 1.0596, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4817351598173516, | |
| "grad_norm": 72169.4375, | |
| "learning_rate": 0.00011348716166945195, | |
| "loss": 1.259, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.4840182648401826, | |
| "grad_norm": 60079.5859375, | |
| "learning_rate": 0.0001127420776681905, | |
| "loss": 1.189, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.4863013698630137, | |
| "grad_norm": 67140.265625, | |
| "learning_rate": 0.00011199627391077732, | |
| "loss": 1.4022, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4885844748858447, | |
| "grad_norm": 52046.58984375, | |
| "learning_rate": 0.00011124979252510208, | |
| "loss": 1.3088, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.4908675799086758, | |
| "grad_norm": 61540.34375, | |
| "learning_rate": 0.0001105026756773314, | |
| "loss": 1.1, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4931506849315068, | |
| "grad_norm": 54976.57421875, | |
| "learning_rate": 0.00010975496556952682, | |
| "loss": 0.9573, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4954337899543379, | |
| "grad_norm": 47444.56640625, | |
| "learning_rate": 0.00010900670443726135, | |
| "loss": 0.7793, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4977168949771689, | |
| "grad_norm": 114338.7734375, | |
| "learning_rate": 0.00010825793454723325, | |
| "loss": 1.4479, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 55353.40625, | |
| "learning_rate": 0.00010750869819487883, | |
| "loss": 1.1219, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.502283105022831, | |
| "grad_norm": 56450.48046875, | |
| "learning_rate": 0.00010675903770198333, | |
| "loss": 1.143, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.502283105022831, | |
| "eval_loss": 1.120017647743225, | |
| "eval_runtime": 6.072, | |
| "eval_samples_per_second": 16.469, | |
| "eval_steps_per_second": 8.235, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5045662100456622, | |
| "grad_norm": 60988.2421875, | |
| "learning_rate": 0.00010600899541429004, | |
| "loss": 1.1982, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.5068493150684932, | |
| "grad_norm": 66918.2578125, | |
| "learning_rate": 0.00010525861369910877, | |
| "loss": 1.2181, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.5091324200913242, | |
| "grad_norm": 79952.6640625, | |
| "learning_rate": 0.00010450793494292224, | |
| "loss": 1.1986, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.5114155251141552, | |
| "grad_norm": 66536.671875, | |
| "learning_rate": 0.00010375700154899208, | |
| "loss": 1.5032, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.5136986301369864, | |
| "grad_norm": 71625.8984375, | |
| "learning_rate": 0.00010300585593496348, | |
| "loss": 1.3998, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5159817351598174, | |
| "grad_norm": 66669.5703125, | |
| "learning_rate": 0.00010225454053046921, | |
| "loss": 0.8312, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5182648401826484, | |
| "grad_norm": 39306.3359375, | |
| "learning_rate": 0.00010150309777473306, | |
| "loss": 0.6047, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5205479452054794, | |
| "grad_norm": 53523.4609375, | |
| "learning_rate": 0.0001007515701141722, | |
| "loss": 1.1629, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5228310502283106, | |
| "grad_norm": 74079.578125, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0456, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5251141552511416, | |
| "grad_norm": 62511.48046875, | |
| "learning_rate": 9.924842988582782e-05, | |
| "loss": 1.0753, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5273972602739726, | |
| "grad_norm": 53424.6328125, | |
| "learning_rate": 9.849690222526698e-05, | |
| "loss": 1.0496, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5296803652968036, | |
| "grad_norm": 76170.8671875, | |
| "learning_rate": 9.77454594695308e-05, | |
| "loss": 1.291, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5319634703196348, | |
| "grad_norm": 54918.74609375, | |
| "learning_rate": 9.699414406503654e-05, | |
| "loss": 1.0644, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5342465753424658, | |
| "grad_norm": 90057.484375, | |
| "learning_rate": 9.624299845100795e-05, | |
| "loss": 1.6448, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5365296803652968, | |
| "grad_norm": 50608.3125, | |
| "learning_rate": 9.549206505707777e-05, | |
| "loss": 1.0954, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5388127853881278, | |
| "grad_norm": 75716.2109375, | |
| "learning_rate": 9.474138630089124e-05, | |
| "loss": 1.2342, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.541095890410959, | |
| "grad_norm": 43616.71875, | |
| "learning_rate": 9.399100458570997e-05, | |
| "loss": 0.8373, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.54337899543379, | |
| "grad_norm": 51082.984375, | |
| "learning_rate": 9.324096229801674e-05, | |
| "loss": 1.0579, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.545662100456621, | |
| "grad_norm": 64326.8359375, | |
| "learning_rate": 9.249130180512118e-05, | |
| "loss": 1.1863, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.547945205479452, | |
| "grad_norm": 63323.40234375, | |
| "learning_rate": 9.174206545276677e-05, | |
| "loss": 1.1812, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5502283105022832, | |
| "grad_norm": 57428.61328125, | |
| "learning_rate": 9.099329556273866e-05, | |
| "loss": 1.0751, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5525114155251142, | |
| "grad_norm": 78367.109375, | |
| "learning_rate": 9.024503443047319e-05, | |
| "loss": 1.0871, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5547945205479452, | |
| "grad_norm": 56003.375, | |
| "learning_rate": 8.949732432266866e-05, | |
| "loss": 1.2125, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5570776255707762, | |
| "grad_norm": 50056.81640625, | |
| "learning_rate": 8.875020747489794e-05, | |
| "loss": 1.1738, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5593607305936074, | |
| "grad_norm": 54685.7265625, | |
| "learning_rate": 8.800372608922271e-05, | |
| "loss": 1.2343, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5616438356164384, | |
| "grad_norm": 75873.0, | |
| "learning_rate": 8.72579223318095e-05, | |
| "loss": 1.4155, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5639269406392694, | |
| "grad_norm": 73813.1171875, | |
| "learning_rate": 8.651283833054809e-05, | |
| "loss": 1.3641, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5662100456621004, | |
| "grad_norm": 52799.9140625, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 1.0217, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.5684931506849316, | |
| "grad_norm": 53224.50390625, | |
| "learning_rate": 8.5024997902379e-05, | |
| "loss": 1.1589, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.5707762557077626, | |
| "grad_norm": 101657.953125, | |
| "learning_rate": 8.428232551846101e-05, | |
| "loss": 1.5682, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5730593607305936, | |
| "grad_norm": 61659.9453125, | |
| "learning_rate": 8.35405409719266e-05, | |
| "loss": 1.0378, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5753424657534246, | |
| "grad_norm": 56497.0859375, | |
| "learning_rate": 8.279968616363418e-05, | |
| "loss": 1.2885, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5776255707762558, | |
| "grad_norm": 64090.8359375, | |
| "learning_rate": 8.205980294192421e-05, | |
| "loss": 1.3739, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5799086757990868, | |
| "grad_norm": 56467.49609375, | |
| "learning_rate": 8.132093310025571e-05, | |
| "loss": 1.0083, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5821917808219178, | |
| "grad_norm": 66332.6328125, | |
| "learning_rate": 8.058311837484535e-05, | |
| "loss": 1.2237, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5844748858447488, | |
| "grad_norm": 68149.140625, | |
| "learning_rate": 7.984640044230983e-05, | |
| "loss": 0.8762, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.58675799086758, | |
| "grad_norm": 56869.44140625, | |
| "learning_rate": 7.911082091731181e-05, | |
| "loss": 1.4265, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.589041095890411, | |
| "grad_norm": 66520.6640625, | |
| "learning_rate": 7.837642135020929e-05, | |
| "loss": 1.3615, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.591324200913242, | |
| "grad_norm": 50649.890625, | |
| "learning_rate": 7.764324322470841e-05, | |
| "loss": 1.1493, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.593607305936073, | |
| "grad_norm": 55484.89453125, | |
| "learning_rate": 7.691132795552043e-05, | |
| "loss": 1.125, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5958904109589042, | |
| "grad_norm": 69880.4765625, | |
| "learning_rate": 7.618071688602199e-05, | |
| "loss": 1.0974, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5981735159817352, | |
| "grad_norm": 69534.0234375, | |
| "learning_rate": 7.54514512859201e-05, | |
| "loss": 1.4658, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.6004566210045662, | |
| "grad_norm": 63003.56640625, | |
| "learning_rate": 7.472357234892082e-05, | |
| "loss": 1.3326, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.6027397260273972, | |
| "grad_norm": 48729.6796875, | |
| "learning_rate": 7.399712119040238e-05, | |
| "loss": 0.9631, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.6050228310502284, | |
| "grad_norm": 53327.13671875, | |
| "learning_rate": 7.327213884509272e-05, | |
| "loss": 0.8053, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6073059360730594, | |
| "grad_norm": 56525.19921875, | |
| "learning_rate": 7.254866626475152e-05, | |
| "loss": 0.9795, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.6095890410958904, | |
| "grad_norm": 54734.92578125, | |
| "learning_rate": 7.182674431585704e-05, | |
| "loss": 1.0737, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.6118721461187214, | |
| "grad_norm": 62297.49609375, | |
| "learning_rate": 7.110641377729778e-05, | |
| "loss": 1.3582, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.6141552511415526, | |
| "grad_norm": 69470.6484375, | |
| "learning_rate": 7.038771533806884e-05, | |
| "loss": 1.261, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.6164383561643836, | |
| "grad_norm": 75963.8046875, | |
| "learning_rate": 6.967068959497376e-05, | |
| "loss": 1.2154, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6187214611872146, | |
| "grad_norm": 66193.421875, | |
| "learning_rate": 6.895537705033108e-05, | |
| "loss": 1.2711, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6210045662100456, | |
| "grad_norm": 53985.34765625, | |
| "learning_rate": 6.824181810968675e-05, | |
| "loss": 0.9666, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6232876712328768, | |
| "grad_norm": 54472.45703125, | |
| "learning_rate": 6.753005307953167e-05, | |
| "loss": 0.7533, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.6255707762557078, | |
| "grad_norm": 56472.04296875, | |
| "learning_rate": 6.682012216502484e-05, | |
| "loss": 1.1015, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6278538812785388, | |
| "grad_norm": 59576.98046875, | |
| "learning_rate": 6.611206546772237e-05, | |
| "loss": 1.243, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6301369863013698, | |
| "grad_norm": 46372.140625, | |
| "learning_rate": 6.54059229833124e-05, | |
| "loss": 0.919, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.632420091324201, | |
| "grad_norm": 67258.171875, | |
| "learning_rate": 6.47017345993556e-05, | |
| "loss": 1.2923, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.634703196347032, | |
| "grad_norm": 97251.375, | |
| "learning_rate": 6.39995400930324e-05, | |
| "loss": 1.3966, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.636986301369863, | |
| "grad_norm": 55997.97265625, | |
| "learning_rate": 6.329937912889582e-05, | |
| "loss": 0.7084, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.639269406392694, | |
| "grad_norm": 52094.359375, | |
| "learning_rate": 6.260129125663106e-05, | |
| "loss": 0.9639, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6415525114155252, | |
| "grad_norm": 59095.65625, | |
| "learning_rate": 6.190531590882159e-05, | |
| "loss": 1.2343, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6438356164383562, | |
| "grad_norm": 50110.4375, | |
| "learning_rate": 6.121149239872151e-05, | |
| "loss": 1.1458, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6461187214611872, | |
| "grad_norm": 61831.3359375, | |
| "learning_rate": 6.051985991803517e-05, | |
| "loss": 1.1047, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6484018264840182, | |
| "grad_norm": 68382.28125, | |
| "learning_rate": 5.983045753470308e-05, | |
| "loss": 1.0395, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6506849315068494, | |
| "grad_norm": 52874.1796875, | |
| "learning_rate": 5.9143324190695196e-05, | |
| "loss": 1.1445, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6529680365296804, | |
| "grad_norm": 53148.92578125, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 1.1164, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6552511415525114, | |
| "grad_norm": 49849.4765625, | |
| "learning_rate": 5.777601974548866e-05, | |
| "loss": 0.8629, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6575342465753424, | |
| "grad_norm": 55761.171875, | |
| "learning_rate": 5.709592587861637e-05, | |
| "loss": 1.0324, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6598173515981736, | |
| "grad_norm": 47947.84765625, | |
| "learning_rate": 5.6418255515358486e-05, | |
| "loss": 0.9341, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6621004566210046, | |
| "grad_norm": 69258.109375, | |
| "learning_rate": 5.574304693498346e-05, | |
| "loss": 1.1231, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6643835616438356, | |
| "grad_norm": 48883.62109375, | |
| "learning_rate": 5.507033827770225e-05, | |
| "loss": 1.0446, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 48850.08203125, | |
| "learning_rate": 5.4400167542513636e-05, | |
| "loss": 1.0385, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6689497716894978, | |
| "grad_norm": 58763.58984375, | |
| "learning_rate": 5.3732572585057974e-05, | |
| "loss": 1.3623, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.6712328767123288, | |
| "grad_norm": 53041.3046875, | |
| "learning_rate": 5.306759111547881e-05, | |
| "loss": 1.1103, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6735159817351598, | |
| "grad_norm": 54536.75390625, | |
| "learning_rate": 5.240526069629265e-05, | |
| "loss": 1.359, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6757990867579908, | |
| "grad_norm": 55320.109375, | |
| "learning_rate": 5.174561874026741e-05, | |
| "loss": 1.1521, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.678082191780822, | |
| "grad_norm": 56988.97265625, | |
| "learning_rate": 5.108870250830882e-05, | |
| "loss": 1.2104, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.680365296803653, | |
| "grad_norm": 65441.95703125, | |
| "learning_rate": 5.0434549107355944e-05, | |
| "loss": 1.0381, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.682648401826484, | |
| "grad_norm": 54248.94921875, | |
| "learning_rate": 4.978319548828504e-05, | |
| "loss": 1.0877, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.684931506849315, | |
| "grad_norm": 62817.375, | |
| "learning_rate": 4.9134678443822166e-05, | |
| "loss": 1.1492, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6872146118721462, | |
| "grad_norm": 55446.65234375, | |
| "learning_rate": 4.8489034606465225e-05, | |
| "loss": 1.078, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6894977168949772, | |
| "grad_norm": 61516.6484375, | |
| "learning_rate": 4.784630044641435e-05, | |
| "loss": 1.0592, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6917808219178082, | |
| "grad_norm": 53431.09375, | |
| "learning_rate": 4.7206512269512124e-05, | |
| "loss": 1.2311, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6940639269406392, | |
| "grad_norm": 68350.4609375, | |
| "learning_rate": 4.65697062151927e-05, | |
| "loss": 0.9918, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6963470319634704, | |
| "grad_norm": 53722.33984375, | |
| "learning_rate": 4.593591825444028e-05, | |
| "loss": 1.0563, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6986301369863014, | |
| "grad_norm": 73647.765625, | |
| "learning_rate": 4.530518418775733e-05, | |
| "loss": 1.1067, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.7009132420091324, | |
| "grad_norm": 60410.765625, | |
| "learning_rate": 4.4677539643142454e-05, | |
| "loss": 1.2064, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.7031963470319634, | |
| "grad_norm": 74342.1328125, | |
| "learning_rate": 4.40530200740777e-05, | |
| "loss": 1.2019, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.7054794520547946, | |
| "grad_norm": 58122.41796875, | |
| "learning_rate": 4.343166075752605e-05, | |
| "loss": 1.0697, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.7077625570776256, | |
| "grad_norm": 68661.671875, | |
| "learning_rate": 4.281349679193861e-05, | |
| "loss": 1.444, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7100456621004566, | |
| "grad_norm": 60156.05859375, | |
| "learning_rate": 4.2198563095272116e-05, | |
| "loss": 1.0489, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.7123287671232876, | |
| "grad_norm": 72793.8359375, | |
| "learning_rate": 4.158689440301657e-05, | |
| "loss": 1.225, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.7146118721461188, | |
| "grad_norm": 61000.60546875, | |
| "learning_rate": 4.097852526623307e-05, | |
| "loss": 1.1954, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.7168949771689498, | |
| "grad_norm": 60486.58203125, | |
| "learning_rate": 4.0373490049602204e-05, | |
| "loss": 1.0551, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.7191780821917808, | |
| "grad_norm": 91112.53125, | |
| "learning_rate": 3.977182292948283e-05, | |
| "loss": 1.2928, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7214611872146118, | |
| "grad_norm": 60165.15234375, | |
| "learning_rate": 3.9173557891981573e-05, | |
| "loss": 0.9197, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.723744292237443, | |
| "grad_norm": 63943.1875, | |
| "learning_rate": 3.857872873103322e-05, | |
| "loss": 1.0148, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.726027397260274, | |
| "grad_norm": 66577.5234375, | |
| "learning_rate": 3.7987369046491684e-05, | |
| "loss": 1.1527, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.728310502283105, | |
| "grad_norm": 46133.828125, | |
| "learning_rate": 3.7399512242231995e-05, | |
| "loss": 0.6874, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.730593607305936, | |
| "grad_norm": 54615.4296875, | |
| "learning_rate": 3.6815191524263624e-05, | |
| "loss": 1.2463, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7328767123287672, | |
| "grad_norm": 60014.6015625, | |
| "learning_rate": 3.623443989885462e-05, | |
| "loss": 1.3196, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7351598173515982, | |
| "grad_norm": 53860.69140625, | |
| "learning_rate": 3.565729017066729e-05, | |
| "loss": 1.0468, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7374429223744292, | |
| "grad_norm": 55772.921875, | |
| "learning_rate": 3.508377494090521e-05, | |
| "loss": 1.0851, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7397260273972602, | |
| "grad_norm": 53785.3203125, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 1.2084, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7420091324200914, | |
| "grad_norm": 57143.98046875, | |
| "learning_rate": 3.394777735313919e-05, | |
| "loss": 1.031, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7442922374429224, | |
| "grad_norm": 60805.40625, | |
| "learning_rate": 3.338535916373266e-05, | |
| "loss": 1.2843, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7465753424657534, | |
| "grad_norm": 56165.6875, | |
| "learning_rate": 3.2826703806321525e-05, | |
| "loss": 0.9815, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7488584474885844, | |
| "grad_norm": 65290.7421875, | |
| "learning_rate": 3.227184283742591e-05, | |
| "loss": 1.2205, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7511415525114156, | |
| "grad_norm": 59236.3828125, | |
| "learning_rate": 3.17208075992339e-05, | |
| "loss": 1.0087, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "grad_norm": 68966.4921875, | |
| "learning_rate": 3.117362921783134e-05, | |
| "loss": 1.2744, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7534246575342466, | |
| "eval_loss": 1.1098405122756958, | |
| "eval_runtime": 5.9855, | |
| "eval_samples_per_second": 16.707, | |
| "eval_steps_per_second": 8.354, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7557077625570776, | |
| "grad_norm": 57247.7265625, | |
| "learning_rate": 3.063033860144339e-05, | |
| "loss": 1.3947, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.7579908675799086, | |
| "grad_norm": 60022.890625, | |
| "learning_rate": 3.0090966438688772e-05, | |
| "loss": 1.4142, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7602739726027398, | |
| "grad_norm": 48251.8359375, | |
| "learning_rate": 2.9555543196846292e-05, | |
| "loss": 0.9917, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.7625570776255708, | |
| "grad_norm": 50926.33203125, | |
| "learning_rate": 2.9024099120133673e-05, | |
| "loss": 1.0497, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.7648401826484018, | |
| "grad_norm": 51535.24609375, | |
| "learning_rate": 2.8496664227999415e-05, | |
| "loss": 1.0956, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7671232876712328, | |
| "grad_norm": 47902.59375, | |
| "learning_rate": 2.7973268313426837e-05, | |
| "loss": 1.182, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.769406392694064, | |
| "grad_norm": 57668.80859375, | |
| "learning_rate": 2.745394094125141e-05, | |
| "loss": 1.1184, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.771689497716895, | |
| "grad_norm": 65797.921875, | |
| "learning_rate": 2.6938711446490606e-05, | |
| "loss": 1.524, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.773972602739726, | |
| "grad_norm": 44397.484375, | |
| "learning_rate": 2.6427608932686843e-05, | |
| "loss": 1.0515, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.776255707762557, | |
| "grad_norm": 88163.90625, | |
| "learning_rate": 2.5920662270263653e-05, | |
| "loss": 1.291, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7785388127853882, | |
| "grad_norm": 91966.3046875, | |
| "learning_rate": 2.5417900094894744e-05, | |
| "loss": 1.5279, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7808219178082192, | |
| "grad_norm": 54689.46875, | |
| "learning_rate": 2.4919350805886577e-05, | |
| "loss": 1.0487, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.7831050228310502, | |
| "grad_norm": 60095.0703125, | |
| "learning_rate": 2.4425042564574184e-05, | |
| "loss": 1.3099, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7853881278538812, | |
| "grad_norm": 54892.97265625, | |
| "learning_rate": 2.3935003292730296e-05, | |
| "loss": 1.1576, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7876712328767124, | |
| "grad_norm": 99210.53125, | |
| "learning_rate": 2.344926067098836e-05, | |
| "loss": 1.846, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7899543378995434, | |
| "grad_norm": 65337.87890625, | |
| "learning_rate": 2.2967842137278706e-05, | |
| "loss": 1.1789, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7922374429223744, | |
| "grad_norm": 71690.0625, | |
| "learning_rate": 2.2490774885278908e-05, | |
| "loss": 1.3072, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.7945205479452054, | |
| "grad_norm": 58516.3828125, | |
| "learning_rate": 2.201808586287757e-05, | |
| "loss": 1.0443, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7968036529680366, | |
| "grad_norm": 72086.8828125, | |
| "learning_rate": 2.15498017706521e-05, | |
| "loss": 1.1584, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.7990867579908676, | |
| "grad_norm": 58905.08984375, | |
| "learning_rate": 2.1085949060360654e-05, | |
| "loss": 1.2844, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8013698630136986, | |
| "grad_norm": 56423.5703125, | |
| "learning_rate": 2.0626553933447734e-05, | |
| "loss": 1.1755, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.8036529680365296, | |
| "grad_norm": 55340.45703125, | |
| "learning_rate": 2.01716423395644e-05, | |
| "loss": 1.08, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.8059360730593608, | |
| "grad_norm": 56130.5859375, | |
| "learning_rate": 1.9721239975102313e-05, | |
| "loss": 1.1129, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.8082191780821918, | |
| "grad_norm": 58951.8046875, | |
| "learning_rate": 1.9275372281742242e-05, | |
| "loss": 0.9749, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.8105022831050228, | |
| "grad_norm": 58070.171875, | |
| "learning_rate": 1.8834064445016953e-05, | |
| "loss": 1.3564, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8127853881278538, | |
| "grad_norm": 56061.4921875, | |
| "learning_rate": 1.839734139288868e-05, | |
| "loss": 1.2425, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.815068493150685, | |
| "grad_norm": 56343.59375, | |
| "learning_rate": 1.7965227794340877e-05, | |
| "loss": 0.9662, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.817351598173516, | |
| "grad_norm": 61289.234375, | |
| "learning_rate": 1.753774805798486e-05, | |
| "loss": 1.1741, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.819634703196347, | |
| "grad_norm": 98183.703125, | |
| "learning_rate": 1.7114926330680957e-05, | |
| "loss": 0.9826, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.821917808219178, | |
| "grad_norm": 55233.2890625, | |
| "learning_rate": 1.6696786496174578e-05, | |
| "loss": 1.0732, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8242009132420092, | |
| "grad_norm": 45176.41015625, | |
| "learning_rate": 1.6283352173747145e-05, | |
| "loss": 1.0048, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.8264840182648402, | |
| "grad_norm": 66357.2109375, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 1.1619, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8287671232876712, | |
| "grad_norm": 57376.734375, | |
| "learning_rate": 1.5470693211944643e-05, | |
| "loss": 1.1175, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8310502283105022, | |
| "grad_norm": 54212.546875, | |
| "learning_rate": 1.5071514476879878e-05, | |
| "loss": 0.9726, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 65802.8515625, | |
| "learning_rate": 1.4677133059921632e-05, | |
| "loss": 1.3092, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8356164383561644, | |
| "grad_norm": 49888.99609375, | |
| "learning_rate": 1.4287571238320053e-05, | |
| "loss": 0.5516, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8378995433789954, | |
| "grad_norm": 66406.1875, | |
| "learning_rate": 1.3902851017082864e-05, | |
| "loss": 0.9952, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8401826484018264, | |
| "grad_norm": 57395.77734375, | |
| "learning_rate": 1.3522994127732414e-05, | |
| "loss": 1.1279, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8424657534246576, | |
| "grad_norm": 79530.40625, | |
| "learning_rate": 1.3148022027078222e-05, | |
| "loss": 1.8326, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8447488584474886, | |
| "grad_norm": 63298.96875, | |
| "learning_rate": 1.2777955896004812e-05, | |
| "loss": 1.0041, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8470319634703196, | |
| "grad_norm": 74275.5234375, | |
| "learning_rate": 1.2412816638275404e-05, | |
| "loss": 1.3752, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8493150684931506, | |
| "grad_norm": 57181.0390625, | |
| "learning_rate": 1.2052624879351104e-05, | |
| "loss": 1.3001, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8515981735159818, | |
| "grad_norm": 56421.484375, | |
| "learning_rate": 1.1697400965225747e-05, | |
| "loss": 1.1496, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.8538812785388128, | |
| "grad_norm": 55708.265625, | |
| "learning_rate": 1.134716496127679e-05, | |
| "loss": 1.0953, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8561643835616438, | |
| "grad_norm": 64520.75, | |
| "learning_rate": 1.1001936651131717e-05, | |
| "loss": 0.9648, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8584474885844748, | |
| "grad_norm": 67208.2578125, | |
| "learning_rate": 1.0661735535550666e-05, | |
| "loss": 1.465, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.860730593607306, | |
| "grad_norm": 69055.765625, | |
| "learning_rate": 1.0326580831324817e-05, | |
| "loss": 1.1437, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.863013698630137, | |
| "grad_norm": 56862.21875, | |
| "learning_rate": 9.996491470190917e-06, | |
| "loss": 1.1829, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.865296803652968, | |
| "grad_norm": 43474.109375, | |
| "learning_rate": 9.671486097761917e-06, | |
| "loss": 0.7937, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.867579908675799, | |
| "grad_norm": 58035.9140625, | |
| "learning_rate": 9.351583072473713e-06, | |
| "loss": 1.2742, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8698630136986302, | |
| "grad_norm": 59717.421875, | |
| "learning_rate": 9.036800464548157e-06, | |
| "loss": 1.245, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.8721461187214612, | |
| "grad_norm": 100361.84375, | |
| "learning_rate": 8.727156054972374e-06, | |
| "loss": 0.9628, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.8744292237442922, | |
| "grad_norm": 46780.90234375, | |
| "learning_rate": 8.422667334494249e-06, | |
| "loss": 0.7594, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8767123287671232, | |
| "grad_norm": 58426.484375, | |
| "learning_rate": 8.123351502634625e-06, | |
| "loss": 1.3035, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.8789954337899544, | |
| "grad_norm": 55850.3046875, | |
| "learning_rate": 7.82922546671555e-06, | |
| "loss": 1.3021, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8812785388127854, | |
| "grad_norm": 65162.09375, | |
| "learning_rate": 7.54030584090537e-06, | |
| "loss": 1.214, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8835616438356164, | |
| "grad_norm": 50473.97265625, | |
| "learning_rate": 7.256608945280319e-06, | |
| "loss": 1.1281, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.8858447488584474, | |
| "grad_norm": 50409.5390625, | |
| "learning_rate": 6.97815080490245e-06, | |
| "loss": 1.2092, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.8881278538812786, | |
| "grad_norm": 56921.63671875, | |
| "learning_rate": 6.704947148914609e-06, | |
| "loss": 1.0073, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.8904109589041096, | |
| "grad_norm": 72200.1875, | |
| "learning_rate": 6.437013409651849e-06, | |
| "loss": 1.0158, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8926940639269406, | |
| "grad_norm": 48175.20703125, | |
| "learning_rate": 6.174364721769743e-06, | |
| "loss": 0.939, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.8949771689497716, | |
| "grad_norm": 53166.9296875, | |
| "learning_rate": 5.917015921389568e-06, | |
| "loss": 1.0468, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8972602739726028, | |
| "grad_norm": 58858.93359375, | |
| "learning_rate": 5.664981545260073e-06, | |
| "loss": 1.0791, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8995433789954338, | |
| "grad_norm": 49101.5078125, | |
| "learning_rate": 5.418275829936537e-06, | |
| "loss": 1.1875, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.9018264840182648, | |
| "grad_norm": 65572.046875, | |
| "learning_rate": 5.176912710976467e-06, | |
| "loss": 1.326, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9041095890410958, | |
| "grad_norm": 61459.63671875, | |
| "learning_rate": 4.940905822152453e-06, | |
| "loss": 1.0855, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.906392694063927, | |
| "grad_norm": 64900.1796875, | |
| "learning_rate": 4.710268494682146e-06, | |
| "loss": 1.3393, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.908675799086758, | |
| "grad_norm": 52331.64453125, | |
| "learning_rate": 4.485013756475076e-06, | |
| "loss": 1.0434, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.910958904109589, | |
| "grad_norm": 71229.609375, | |
| "learning_rate": 4.2651543313968145e-06, | |
| "loss": 1.2118, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.91324200913242, | |
| "grad_norm": 63443.84375, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 1.1458, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9155251141552512, | |
| "grad_norm": 64409.73828125, | |
| "learning_rate": 3.841670791574137e-06, | |
| "loss": 1.1705, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.9178082191780822, | |
| "grad_norm": 81092.15625, | |
| "learning_rate": 3.638070597958665e-06, | |
| "loss": 1.2991, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.9200913242009132, | |
| "grad_norm": 41609.2109375, | |
| "learning_rate": 3.4399135583787043e-06, | |
| "loss": 0.4622, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.9223744292237442, | |
| "grad_norm": 58560.71875, | |
| "learning_rate": 3.2472108660439706e-06, | |
| "loss": 0.8759, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.9246575342465754, | |
| "grad_norm": 65780.9609375, | |
| "learning_rate": 3.059973406066963e-06, | |
| "loss": 1.3583, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9269406392694064, | |
| "grad_norm": 53281.921875, | |
| "learning_rate": 2.878211754847926e-06, | |
| "loss": 0.971, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.9292237442922374, | |
| "grad_norm": 62492.41015625, | |
| "learning_rate": 2.7019361794775156e-06, | |
| "loss": 1.1152, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.9315068493150684, | |
| "grad_norm": 57761.875, | |
| "learning_rate": 2.5311566371568507e-06, | |
| "loss": 1.1735, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9337899543378996, | |
| "grad_norm": 53913.1640625, | |
| "learning_rate": 2.365882774634998e-06, | |
| "loss": 0.9759, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9360730593607306, | |
| "grad_norm": 57200.4375, | |
| "learning_rate": 2.206123927664161e-06, | |
| "loss": 1.262, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9383561643835616, | |
| "grad_norm": 64508.68359375, | |
| "learning_rate": 2.0518891204722168e-06, | |
| "loss": 1.1443, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.9406392694063926, | |
| "grad_norm": 68100.71875, | |
| "learning_rate": 1.903187065253076e-06, | |
| "loss": 1.156, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.9429223744292238, | |
| "grad_norm": 46483.26171875, | |
| "learning_rate": 1.7600261616745106e-06, | |
| "loss": 0.9084, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.9452054794520548, | |
| "grad_norm": 75558.0078125, | |
| "learning_rate": 1.6224144964036681e-06, | |
| "loss": 1.3323, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.9474885844748858, | |
| "grad_norm": 61868.96484375, | |
| "learning_rate": 1.4903598426503241e-06, | |
| "loss": 1.6048, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9497716894977168, | |
| "grad_norm": 48915.76953125, | |
| "learning_rate": 1.3638696597277679e-06, | |
| "loss": 0.9012, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.952054794520548, | |
| "grad_norm": 75366.640625, | |
| "learning_rate": 1.2429510926314836e-06, | |
| "loss": 1.132, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.954337899543379, | |
| "grad_norm": 60648.18359375, | |
| "learning_rate": 1.1276109716355287e-06, | |
| "loss": 1.0748, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.95662100456621, | |
| "grad_norm": 55365.9765625, | |
| "learning_rate": 1.0178558119067315e-06, | |
| "loss": 0.827, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.958904109589041, | |
| "grad_norm": 58780.57421875, | |
| "learning_rate": 9.136918131366412e-07, | |
| "loss": 1.2993, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9611872146118722, | |
| "grad_norm": 47850.671875, | |
| "learning_rate": 8.151248591913518e-07, | |
| "loss": 1.0673, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9634703196347032, | |
| "grad_norm": 50755.59765625, | |
| "learning_rate": 7.221605177791691e-07, | |
| "loss": 1.0819, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.9657534246575342, | |
| "grad_norm": 68484.7265625, | |
| "learning_rate": 6.348040401360833e-07, | |
| "loss": 1.3769, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.9680365296803652, | |
| "grad_norm": 59223.92578125, | |
| "learning_rate": 5.530603607290851e-07, | |
| "loss": 1.3159, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.9703196347031964, | |
| "grad_norm": 63650.15234375, | |
| "learning_rate": 4.76934096977566e-07, | |
| "loss": 1.3134, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9726027397260274, | |
| "grad_norm": 62978.29296875, | |
| "learning_rate": 4.0642954899238197e-07, | |
| "loss": 0.9593, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.9748858447488584, | |
| "grad_norm": 63336.55078125, | |
| "learning_rate": 3.415506993330153e-07, | |
| "loss": 1.1585, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.9771689497716894, | |
| "grad_norm": 58021.15625, | |
| "learning_rate": 2.8230121278257637e-07, | |
| "loss": 1.2236, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.9794520547945206, | |
| "grad_norm": 56498.8515625, | |
| "learning_rate": 2.2868443614082469e-07, | |
| "loss": 1.0606, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.9817351598173516, | |
| "grad_norm": 48720.328125, | |
| "learning_rate": 1.8070339803509807e-07, | |
| "loss": 1.123, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9840182648401826, | |
| "grad_norm": 60693.24609375, | |
| "learning_rate": 1.3836080874926049e-07, | |
| "loss": 0.5269, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.9863013698630136, | |
| "grad_norm": 123484.75, | |
| "learning_rate": 1.0165906007056914e-07, | |
| "loss": 1.2241, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.9885844748858448, | |
| "grad_norm": 52265.00390625, | |
| "learning_rate": 7.060022515460451e-08, | |
| "loss": 0.8173, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.9908675799086758, | |
| "grad_norm": 54763.09375, | |
| "learning_rate": 4.518605840815315e-08, | |
| "loss": 1.2235, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.9931506849315068, | |
| "grad_norm": 67981.8515625, | |
| "learning_rate": 2.5417995390086824e-08, | |
| "loss": 1.3369, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9954337899543378, | |
| "grad_norm": 53286.59375, | |
| "learning_rate": 1.129715273033849e-08, | |
| "loss": 0.9584, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.997716894977169, | |
| "grad_norm": 54882.32421875, | |
| "learning_rate": 2.824328066730608e-09, | |
| "loss": 1.0711, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 54870.58984375, | |
| "learning_rate": 0.0, | |
| "loss": 1.1062, | |
| "step": 438 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 438, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 7972410614906880.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |