| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 35.0, | |
| "eval_steps": 100, | |
| "global_step": 980, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 9.759769439697266, | |
| "learning_rate": 0.0, | |
| "loss": 2.7414, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 9.035130500793457, | |
| "learning_rate": 1.0204081632653061e-07, | |
| "loss": 2.532, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 8.679366111755371, | |
| "learning_rate": 2.0408163265306121e-07, | |
| "loss": 2.439, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 9.50596809387207, | |
| "learning_rate": 3.0612244897959183e-07, | |
| "loss": 2.5109, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 9.989800453186035, | |
| "learning_rate": 4.0816326530612243e-07, | |
| "loss": 2.5067, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 9.429203033447266, | |
| "learning_rate": 5.102040816326531e-07, | |
| "loss": 2.6609, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 10.171276092529297, | |
| "learning_rate": 6.122448979591837e-07, | |
| "loss": 2.5307, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 9.263684272766113, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 2.6476, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 8.816690444946289, | |
| "learning_rate": 8.163265306122449e-07, | |
| "loss": 2.5897, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.35714285714285715, | |
| "grad_norm": 8.883913040161133, | |
| "learning_rate": 9.183673469387756e-07, | |
| "loss": 2.5036, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.39285714285714285, | |
| "grad_norm": 8.780312538146973, | |
| "learning_rate": 1.0204081632653063e-06, | |
| "loss": 2.3625, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 8.389798164367676, | |
| "learning_rate": 1.122448979591837e-06, | |
| "loss": 2.3566, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4642857142857143, | |
| "grad_norm": 9.69943904876709, | |
| "learning_rate": 1.2244897959183673e-06, | |
| "loss": 2.3067, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.483558654785156, | |
| "learning_rate": 1.3265306122448982e-06, | |
| "loss": 2.4596, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5357142857142857, | |
| "grad_norm": 6.695621967315674, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 2.4559, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 6.5322794914245605, | |
| "learning_rate": 1.5306122448979593e-06, | |
| "loss": 2.2781, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6071428571428571, | |
| "grad_norm": 6.274178504943848, | |
| "learning_rate": 1.6326530612244897e-06, | |
| "loss": 2.3636, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6428571428571429, | |
| "grad_norm": 6.396870136260986, | |
| "learning_rate": 1.7346938775510206e-06, | |
| "loss": 2.4038, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6785714285714286, | |
| "grad_norm": 4.775918960571289, | |
| "learning_rate": 1.8367346938775512e-06, | |
| "loss": 2.2931, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 4.627754211425781, | |
| "learning_rate": 1.938775510204082e-06, | |
| "loss": 2.0346, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.323748588562012, | |
| "learning_rate": 2.0408163265306125e-06, | |
| "loss": 2.2192, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7857142857142857, | |
| "grad_norm": 4.8098955154418945, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 2.1561, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8214285714285714, | |
| "grad_norm": 4.36919641494751, | |
| "learning_rate": 2.244897959183674e-06, | |
| "loss": 2.2667, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 4.16939115524292, | |
| "learning_rate": 2.3469387755102044e-06, | |
| "loss": 2.0838, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 4.2570366859436035, | |
| "learning_rate": 2.4489795918367347e-06, | |
| "loss": 2.102, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.9285714285714286, | |
| "grad_norm": 3.462320566177368, | |
| "learning_rate": 2.5510204081632657e-06, | |
| "loss": 2.0475, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9642857142857143, | |
| "grad_norm": 2.8186793327331543, | |
| "learning_rate": 2.6530612244897964e-06, | |
| "loss": 2.0281, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.044743776321411, | |
| "learning_rate": 2.7551020408163266e-06, | |
| "loss": 2.019, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0357142857142858, | |
| "grad_norm": 2.4265270233154297, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 1.9081, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0714285714285714, | |
| "grad_norm": 2.3639659881591797, | |
| "learning_rate": 2.959183673469388e-06, | |
| "loss": 1.9358, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1071428571428572, | |
| "grad_norm": 2.0860698223114014, | |
| "learning_rate": 3.0612244897959185e-06, | |
| "loss": 1.9145, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 1.9538745880126953, | |
| "learning_rate": 3.1632653061224496e-06, | |
| "loss": 1.7714, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1785714285714286, | |
| "grad_norm": 1.7645140886306763, | |
| "learning_rate": 3.2653061224489794e-06, | |
| "loss": 1.7249, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.2142857142857142, | |
| "grad_norm": 1.7323063611984253, | |
| "learning_rate": 3.3673469387755105e-06, | |
| "loss": 1.7585, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.3425657749176025, | |
| "learning_rate": 3.469387755102041e-06, | |
| "loss": 1.5646, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 1.5468051433563232, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 1.6348, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.3214285714285714, | |
| "grad_norm": 1.569055438041687, | |
| "learning_rate": 3.6734693877551024e-06, | |
| "loss": 1.7431, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.3571428571428572, | |
| "grad_norm": 1.3902649879455566, | |
| "learning_rate": 3.7755102040816327e-06, | |
| "loss": 1.7093, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3928571428571428, | |
| "grad_norm": 1.4763963222503662, | |
| "learning_rate": 3.877551020408164e-06, | |
| "loss": 1.7049, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 1.287192463874817, | |
| "learning_rate": 3.979591836734694e-06, | |
| "loss": 1.6828, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4642857142857144, | |
| "grad_norm": 1.175764799118042, | |
| "learning_rate": 4.081632653061225e-06, | |
| "loss": 1.6261, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.1271580457687378, | |
| "learning_rate": 4.183673469387755e-06, | |
| "loss": 1.6354, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.5357142857142856, | |
| "grad_norm": 1.1049038171768188, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 1.7451, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.9844851493835449, | |
| "learning_rate": 4.3877551020408165e-06, | |
| "loss": 1.5587, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.6071428571428572, | |
| "grad_norm": 0.9872157573699951, | |
| "learning_rate": 4.489795918367348e-06, | |
| "loss": 1.6485, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.6428571428571428, | |
| "grad_norm": 0.9424026012420654, | |
| "learning_rate": 4.591836734693878e-06, | |
| "loss": 1.5298, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.6785714285714286, | |
| "grad_norm": 1.15427827835083, | |
| "learning_rate": 4.693877551020409e-06, | |
| "loss": 1.5205, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 1.1777286529541016, | |
| "learning_rate": 4.795918367346939e-06, | |
| "loss": 1.4686, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.9405274987220764, | |
| "learning_rate": 4.897959183673469e-06, | |
| "loss": 1.5902, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.9954162240028381, | |
| "learning_rate": 5e-06, | |
| "loss": 1.513, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.8214285714285714, | |
| "grad_norm": 1.252631425857544, | |
| "learning_rate": 5.1020408163265315e-06, | |
| "loss": 1.5964, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 1.0796560049057007, | |
| "learning_rate": 5.204081632653062e-06, | |
| "loss": 1.4724, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8928571428571428, | |
| "grad_norm": 1.2125780582427979, | |
| "learning_rate": 5.306122448979593e-06, | |
| "loss": 1.589, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.9285714285714286, | |
| "grad_norm": 0.9952357411384583, | |
| "learning_rate": 5.408163265306123e-06, | |
| "loss": 1.5026, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.9642857142857144, | |
| "grad_norm": 1.0817049741744995, | |
| "learning_rate": 5.510204081632653e-06, | |
| "loss": 1.5709, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.2521437406539917, | |
| "learning_rate": 5.6122448979591834e-06, | |
| "loss": 1.528, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.0357142857142856, | |
| "grad_norm": 1.1188162565231323, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 1.5065, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.0714285714285716, | |
| "grad_norm": 0.8846582174301147, | |
| "learning_rate": 5.816326530612246e-06, | |
| "loss": 1.5931, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.107142857142857, | |
| "grad_norm": 1.2342886924743652, | |
| "learning_rate": 5.918367346938776e-06, | |
| "loss": 1.5268, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 1.337417483329773, | |
| "learning_rate": 6.020408163265307e-06, | |
| "loss": 1.4429, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.1785714285714284, | |
| "grad_norm": 1.1548265218734741, | |
| "learning_rate": 6.122448979591837e-06, | |
| "loss": 1.4426, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.2142857142857144, | |
| "grad_norm": 1.0125023126602173, | |
| "learning_rate": 6.224489795918368e-06, | |
| "loss": 1.4092, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.1323378086090088, | |
| "learning_rate": 6.326530612244899e-06, | |
| "loss": 1.4018, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 1.0492571592330933, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 1.3959, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.3214285714285716, | |
| "grad_norm": 1.0407048463821411, | |
| "learning_rate": 6.530612244897959e-06, | |
| "loss": 1.3596, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.357142857142857, | |
| "grad_norm": 0.8995230197906494, | |
| "learning_rate": 6.63265306122449e-06, | |
| "loss": 1.4646, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.392857142857143, | |
| "grad_norm": 0.8292604088783264, | |
| "learning_rate": 6.734693877551021e-06, | |
| "loss": 1.4945, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 0.9430536031723022, | |
| "learning_rate": 6.836734693877551e-06, | |
| "loss": 1.3554, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.4642857142857144, | |
| "grad_norm": 0.9344185590744019, | |
| "learning_rate": 6.938775510204082e-06, | |
| "loss": 1.3483, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.8723719716072083, | |
| "learning_rate": 7.0408163265306125e-06, | |
| "loss": 1.3678, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.5357142857142856, | |
| "grad_norm": 0.8184205889701843, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 1.4009, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.9736836552619934, | |
| "learning_rate": 7.244897959183675e-06, | |
| "loss": 1.3519, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.607142857142857, | |
| "grad_norm": 0.9170386791229248, | |
| "learning_rate": 7.346938775510205e-06, | |
| "loss": 1.4351, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.642857142857143, | |
| "grad_norm": 0.799105703830719, | |
| "learning_rate": 7.448979591836736e-06, | |
| "loss": 1.4096, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.678571428571429, | |
| "grad_norm": 1.1402575969696045, | |
| "learning_rate": 7.551020408163265e-06, | |
| "loss": 1.4303, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.8628937602043152, | |
| "learning_rate": 7.653061224489796e-06, | |
| "loss": 1.2549, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.0535259246826172, | |
| "learning_rate": 7.755102040816327e-06, | |
| "loss": 1.303, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.7857142857142856, | |
| "grad_norm": 1.0219699144363403, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 1.3732, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.821428571428571, | |
| "grad_norm": 0.8952599167823792, | |
| "learning_rate": 7.959183673469388e-06, | |
| "loss": 1.4765, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.8639037609100342, | |
| "learning_rate": 8.06122448979592e-06, | |
| "loss": 1.3302, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.892857142857143, | |
| "grad_norm": 1.001290202140808, | |
| "learning_rate": 8.16326530612245e-06, | |
| "loss": 1.3691, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.928571428571429, | |
| "grad_norm": 0.9045864939689636, | |
| "learning_rate": 8.26530612244898e-06, | |
| "loss": 1.3872, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.9642857142857144, | |
| "grad_norm": 0.8884599208831787, | |
| "learning_rate": 8.36734693877551e-06, | |
| "loss": 1.3095, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8961983919143677, | |
| "learning_rate": 8.469387755102042e-06, | |
| "loss": 1.3005, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.0357142857142856, | |
| "grad_norm": 0.8967141509056091, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 1.3534, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.0714285714285716, | |
| "grad_norm": 0.7872421741485596, | |
| "learning_rate": 8.673469387755103e-06, | |
| "loss": 1.2637, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.107142857142857, | |
| "grad_norm": 0.7589600682258606, | |
| "learning_rate": 8.775510204081633e-06, | |
| "loss": 1.2964, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 0.8907157778739929, | |
| "learning_rate": 8.877551020408163e-06, | |
| "loss": 1.26, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.1785714285714284, | |
| "grad_norm": 0.8499204516410828, | |
| "learning_rate": 8.979591836734695e-06, | |
| "loss": 1.2085, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.2142857142857144, | |
| "grad_norm": 0.816241979598999, | |
| "learning_rate": 9.081632653061225e-06, | |
| "loss": 1.313, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.9863104820251465, | |
| "learning_rate": 9.183673469387756e-06, | |
| "loss": 1.2019, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 0.8051828145980835, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 1.2688, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.3214285714285716, | |
| "grad_norm": 0.8511350154876709, | |
| "learning_rate": 9.387755102040818e-06, | |
| "loss": 1.2488, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.357142857142857, | |
| "grad_norm": 0.8502587676048279, | |
| "learning_rate": 9.489795918367348e-06, | |
| "loss": 1.1932, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.392857142857143, | |
| "grad_norm": 0.8357100486755371, | |
| "learning_rate": 9.591836734693878e-06, | |
| "loss": 1.2833, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.8697457313537598, | |
| "learning_rate": 9.693877551020408e-06, | |
| "loss": 1.1389, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.4642857142857144, | |
| "grad_norm": 0.8650133609771729, | |
| "learning_rate": 9.795918367346939e-06, | |
| "loss": 1.1914, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.8051957488059998, | |
| "learning_rate": 9.89795918367347e-06, | |
| "loss": 1.1436, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.5357142857142856, | |
| "grad_norm": 0.9405279755592346, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3277, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.7801042795181274, | |
| "learning_rate": 9.999968282268043e-06, | |
| "loss": 1.1546, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "eval_loss": 1.2822846174240112, | |
| "eval_runtime": 0.47, | |
| "eval_samples_per_second": 297.86, | |
| "eval_steps_per_second": 2.128, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.607142857142857, | |
| "grad_norm": 0.861859142780304, | |
| "learning_rate": 9.999873129474573e-06, | |
| "loss": 1.2382, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.642857142857143, | |
| "grad_norm": 0.8621125221252441, | |
| "learning_rate": 9.999714542826806e-06, | |
| "loss": 1.3204, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.678571428571429, | |
| "grad_norm": 0.7925216555595398, | |
| "learning_rate": 9.999492524336743e-06, | |
| "loss": 1.1399, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 0.9186500906944275, | |
| "learning_rate": 9.999207076821155e-06, | |
| "loss": 1.1609, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.9390427470207214, | |
| "learning_rate": 9.99885820390154e-06, | |
| "loss": 1.1709, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.7857142857142856, | |
| "grad_norm": 0.8102933764457703, | |
| "learning_rate": 9.998445910004082e-06, | |
| "loss": 1.2304, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.821428571428571, | |
| "grad_norm": 0.9541983604431152, | |
| "learning_rate": 9.997970200359592e-06, | |
| "loss": 1.2702, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.857142857142857, | |
| "grad_norm": 0.8650714755058289, | |
| "learning_rate": 9.99743108100344e-06, | |
| "loss": 1.2122, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.892857142857143, | |
| "grad_norm": 0.8606563210487366, | |
| "learning_rate": 9.996828558775486e-06, | |
| "loss": 1.2045, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.928571428571429, | |
| "grad_norm": 0.8290749192237854, | |
| "learning_rate": 9.996162641319985e-06, | |
| "loss": 1.1833, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.9642857142857144, | |
| "grad_norm": 0.8623219132423401, | |
| "learning_rate": 9.995433337085492e-06, | |
| "loss": 1.1361, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7909784913063049, | |
| "learning_rate": 9.994640655324758e-06, | |
| "loss": 1.1751, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.035714285714286, | |
| "grad_norm": 0.7679526209831238, | |
| "learning_rate": 9.993784606094612e-06, | |
| "loss": 1.0578, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.071428571428571, | |
| "grad_norm": 0.8180373907089233, | |
| "learning_rate": 9.992865200255829e-06, | |
| "loss": 1.0855, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.107142857142857, | |
| "grad_norm": 0.8258321285247803, | |
| "learning_rate": 9.991882449472994e-06, | |
| "loss": 1.0501, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.142857142857143, | |
| "grad_norm": 0.9095039963722229, | |
| "learning_rate": 9.99083636621436e-06, | |
| "loss": 1.0903, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.178571428571429, | |
| "grad_norm": 0.8757370710372925, | |
| "learning_rate": 9.989726963751683e-06, | |
| "loss": 1.0142, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.214285714285714, | |
| "grad_norm": 0.9213536381721497, | |
| "learning_rate": 9.988554256160052e-06, | |
| "loss": 1.03, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 1.1767222881317139, | |
| "learning_rate": 9.987318258317718e-06, | |
| "loss": 1.0143, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 0.9031118154525757, | |
| "learning_rate": 9.986018985905901e-06, | |
| "loss": 0.9545, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.321428571428571, | |
| "grad_norm": 0.9330540299415588, | |
| "learning_rate": 9.984656455408591e-06, | |
| "loss": 0.969, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.357142857142857, | |
| "grad_norm": 1.1439694166183472, | |
| "learning_rate": 9.983230684112338e-06, | |
| "loss": 1.1886, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.392857142857143, | |
| "grad_norm": 0.9566424489021301, | |
| "learning_rate": 9.981741690106035e-06, | |
| "loss": 0.9773, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.428571428571429, | |
| "grad_norm": 1.067845344543457, | |
| "learning_rate": 9.980189492280688e-06, | |
| "loss": 1.0976, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.464285714285714, | |
| "grad_norm": 1.0497171878814697, | |
| "learning_rate": 9.978574110329174e-06, | |
| "loss": 0.9478, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.9933088421821594, | |
| "learning_rate": 9.976895564745993e-06, | |
| "loss": 0.9368, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 4.535714285714286, | |
| "grad_norm": 0.8733711838722229, | |
| "learning_rate": 9.975153876827008e-06, | |
| "loss": 1.0447, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.9675903916358948, | |
| "learning_rate": 9.973349068669178e-06, | |
| "loss": 1.0044, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 4.607142857142857, | |
| "grad_norm": 1.0770503282546997, | |
| "learning_rate": 9.97148116317027e-06, | |
| "loss": 1.0712, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 4.642857142857143, | |
| "grad_norm": 0.8858775496482849, | |
| "learning_rate": 9.969550184028572e-06, | |
| "loss": 0.9705, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.678571428571429, | |
| "grad_norm": 1.0158929824829102, | |
| "learning_rate": 9.9675561557426e-06, | |
| "loss": 1.0405, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.714285714285714, | |
| "grad_norm": 1.1270577907562256, | |
| "learning_rate": 9.965499103610775e-06, | |
| "loss": 1.0283, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.8913384079933167, | |
| "learning_rate": 9.963379053731104e-06, | |
| "loss": 1.0334, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.785714285714286, | |
| "grad_norm": 1.0675349235534668, | |
| "learning_rate": 9.961196033000862e-06, | |
| "loss": 1.1302, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.821428571428571, | |
| "grad_norm": 0.9325812458992004, | |
| "learning_rate": 9.95895006911623e-06, | |
| "loss": 0.9444, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 1.0121493339538574, | |
| "learning_rate": 9.956641190571967e-06, | |
| "loss": 0.9612, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.892857142857143, | |
| "grad_norm": 0.8960527777671814, | |
| "learning_rate": 9.954269426661023e-06, | |
| "loss": 1.0558, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.928571428571429, | |
| "grad_norm": 0.9022102952003479, | |
| "learning_rate": 9.951834807474191e-06, | |
| "loss": 0.9476, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.964285714285714, | |
| "grad_norm": 0.921174168586731, | |
| "learning_rate": 9.949337363899709e-06, | |
| "loss": 0.994, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.8534979224205017, | |
| "learning_rate": 9.946777127622874e-06, | |
| "loss": 0.8216, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.035714285714286, | |
| "grad_norm": 0.9021621346473694, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.8672, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 5.071428571428571, | |
| "grad_norm": 0.9865545034408569, | |
| "learning_rate": 9.941468407686216e-06, | |
| "loss": 0.824, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 5.107142857142857, | |
| "grad_norm": 0.9824482798576355, | |
| "learning_rate": 9.938719991378614e-06, | |
| "loss": 0.8532, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 1.110496163368225, | |
| "learning_rate": 9.935908917072253e-06, | |
| "loss": 0.9036, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 5.178571428571429, | |
| "grad_norm": 1.095384120941162, | |
| "learning_rate": 9.933035220431489e-06, | |
| "loss": 0.8404, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 5.214285714285714, | |
| "grad_norm": 1.309322714805603, | |
| "learning_rate": 9.930098937915177e-06, | |
| "loss": 0.8397, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 1.5140193700790405, | |
| "learning_rate": 9.927100106776213e-06, | |
| "loss": 0.8126, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 5.285714285714286, | |
| "grad_norm": 1.2346681356430054, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.8103, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 5.321428571428571, | |
| "grad_norm": 1.0885480642318726, | |
| "learning_rate": 9.920914951609189e-06, | |
| "loss": 0.7544, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 1.0442003011703491, | |
| "learning_rate": 9.917728706052765e-06, | |
| "loss": 0.8543, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.392857142857143, | |
| "grad_norm": 1.0954079627990723, | |
| "learning_rate": 9.914480068815964e-06, | |
| "loss": 0.7868, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "grad_norm": 1.0408991575241089, | |
| "learning_rate": 9.91116908111455e-06, | |
| "loss": 0.7768, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 5.464285714285714, | |
| "grad_norm": 1.1352790594100952, | |
| "learning_rate": 9.907795784955327e-06, | |
| "loss": 0.8077, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 1.1276353597640991, | |
| "learning_rate": 9.90436022313562e-06, | |
| "loss": 0.7622, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 5.535714285714286, | |
| "grad_norm": 1.1519376039505005, | |
| "learning_rate": 9.900862439242719e-06, | |
| "loss": 0.8162, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 5.571428571428571, | |
| "grad_norm": 1.187958836555481, | |
| "learning_rate": 9.897302477653334e-06, | |
| "loss": 0.7768, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 5.607142857142857, | |
| "grad_norm": 1.3001772165298462, | |
| "learning_rate": 9.893680383533027e-06, | |
| "loss": 0.8367, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 5.642857142857143, | |
| "grad_norm": 1.199255347251892, | |
| "learning_rate": 9.889996202835642e-06, | |
| "loss": 0.752, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 5.678571428571429, | |
| "grad_norm": 1.1505409479141235, | |
| "learning_rate": 9.88624998230272e-06, | |
| "loss": 0.878, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 1.1740809679031372, | |
| "learning_rate": 9.882441769462911e-06, | |
| "loss": 0.7647, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 1.0937939882278442, | |
| "learning_rate": 9.878571612631364e-06, | |
| "loss": 0.8364, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 5.785714285714286, | |
| "grad_norm": 1.0483719110488892, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.8453, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 5.821428571428571, | |
| "grad_norm": 1.0722523927688599, | |
| "learning_rate": 9.870645664182478e-06, | |
| "loss": 0.76, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 5.857142857142857, | |
| "grad_norm": 1.093754768371582, | |
| "learning_rate": 9.86658997312238e-06, | |
| "loss": 0.7596, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.892857142857143, | |
| "grad_norm": 1.1799498796463013, | |
| "learning_rate": 9.862472539183757e-06, | |
| "loss": 0.8612, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.928571428571429, | |
| "grad_norm": 1.1799345016479492, | |
| "learning_rate": 9.858293414604871e-06, | |
| "loss": 0.8048, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.964285714285714, | |
| "grad_norm": 1.090215802192688, | |
| "learning_rate": 9.854052652406666e-06, | |
| "loss": 0.7531, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.1735352277755737, | |
| "learning_rate": 9.849750306392085e-06, | |
| "loss": 0.6251, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 6.035714285714286, | |
| "grad_norm": 1.2174513339996338, | |
| "learning_rate": 9.84538643114539e-06, | |
| "loss": 0.6708, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 6.071428571428571, | |
| "grad_norm": 1.188399076461792, | |
| "learning_rate": 9.840961082031473e-06, | |
| "loss": 0.6973, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.107142857142857, | |
| "grad_norm": 1.3352035284042358, | |
| "learning_rate": 9.836474315195148e-06, | |
| "loss": 0.6517, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 6.142857142857143, | |
| "grad_norm": 1.464682698249817, | |
| "learning_rate": 9.831926187560441e-06, | |
| "loss": 0.6084, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 6.178571428571429, | |
| "grad_norm": 1.567628264427185, | |
| "learning_rate": 9.827316756829871e-06, | |
| "loss": 0.5943, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 6.214285714285714, | |
| "grad_norm": 1.671661376953125, | |
| "learning_rate": 9.822646081483713e-06, | |
| "loss": 0.6451, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.2910634279251099, | |
| "learning_rate": 9.817914220779258e-06, | |
| "loss": 0.6001, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 1.2736347913742065, | |
| "learning_rate": 9.81312123475006e-06, | |
| "loss": 0.5976, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 6.321428571428571, | |
| "grad_norm": 1.1851528882980347, | |
| "learning_rate": 9.808267184205182e-06, | |
| "loss": 0.5813, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 6.357142857142857, | |
| "grad_norm": 1.202125906944275, | |
| "learning_rate": 9.80335213072841e-06, | |
| "loss": 0.5511, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 6.392857142857143, | |
| "grad_norm": 1.1597651243209839, | |
| "learning_rate": 9.798376136677486e-06, | |
| "loss": 0.5579, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 1.2182499170303345, | |
| "learning_rate": 9.793339265183303e-06, | |
| "loss": 0.527, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 6.464285714285714, | |
| "grad_norm": 1.3637059926986694, | |
| "learning_rate": 9.788241580149123e-06, | |
| "loss": 0.5881, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 1.3525656461715698, | |
| "learning_rate": 9.783083146249749e-06, | |
| "loss": 0.5526, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 6.535714285714286, | |
| "grad_norm": 1.5944671630859375, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.5739, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 6.571428571428571, | |
| "grad_norm": 1.248140573501587, | |
| "learning_rate": 9.77258429440742e-06, | |
| "loss": 0.5514, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 6.607142857142857, | |
| "grad_norm": 1.1553500890731812, | |
| "learning_rate": 9.767244009664376e-06, | |
| "loss": 0.566, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 6.642857142857143, | |
| "grad_norm": 1.4417532682418823, | |
| "learning_rate": 9.761843242454261e-06, | |
| "loss": 0.596, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 6.678571428571429, | |
| "grad_norm": 1.140500783920288, | |
| "learning_rate": 9.75638206129711e-06, | |
| "loss": 0.5865, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 6.714285714285714, | |
| "grad_norm": 1.204921841621399, | |
| "learning_rate": 9.750860535479434e-06, | |
| "loss": 0.6362, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 1.2617542743682861, | |
| "learning_rate": 9.745278735053345e-06, | |
| "loss": 0.5731, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 6.785714285714286, | |
| "grad_norm": 1.2948684692382812, | |
| "learning_rate": 9.73963673083566e-06, | |
| "loss": 0.5541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 6.821428571428571, | |
| "grad_norm": 1.2236193418502808, | |
| "learning_rate": 9.733934594407012e-06, | |
| "loss": 0.5517, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 1.3996909856796265, | |
| "learning_rate": 9.728172398110935e-06, | |
| "loss": 0.5111, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 6.892857142857143, | |
| "grad_norm": 1.3360153436660767, | |
| "learning_rate": 9.722350215052946e-06, | |
| "loss": 0.5726, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 6.928571428571429, | |
| "grad_norm": 1.3151594400405884, | |
| "learning_rate": 9.716468119099626e-06, | |
| "loss": 0.5435, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 6.964285714285714, | |
| "grad_norm": 1.371817708015442, | |
| "learning_rate": 9.710526184877667e-06, | |
| "loss": 0.5764, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.2646558284759521, | |
| "learning_rate": 9.704524487772944e-06, | |
| "loss": 0.4543, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 7.035714285714286, | |
| "grad_norm": 1.474683165550232, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.3993, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 7.071428571428571, | |
| "grad_norm": 1.3176275491714478, | |
| "learning_rate": 9.692342110248802e-06, | |
| "loss": 0.3759, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 7.107142857142857, | |
| "grad_norm": 1.3523168563842773, | |
| "learning_rate": 9.68616158438834e-06, | |
| "loss": 0.35, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 1.6409916877746582, | |
| "learning_rate": 9.679921604761056e-06, | |
| "loss": 0.4067, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "eval_loss": 1.6618373394012451, | |
| "eval_runtime": 0.4663, | |
| "eval_samples_per_second": 300.222, | |
| "eval_steps_per_second": 2.144, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.178571428571429, | |
| "grad_norm": 2.2635273933410645, | |
| "learning_rate": 9.673622250534155e-06, | |
| "loss": 0.3858, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 7.214285714285714, | |
| "grad_norm": 2.0102970600128174, | |
| "learning_rate": 9.66726360162813e-06, | |
| "loss": 0.3662, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "grad_norm": 1.694830298423767, | |
| "learning_rate": 9.660845738715743e-06, | |
| "loss": 0.3926, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 7.285714285714286, | |
| "grad_norm": 1.3782461881637573, | |
| "learning_rate": 9.654368743221022e-06, | |
| "loss": 0.4011, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 7.321428571428571, | |
| "grad_norm": 1.4142459630966187, | |
| "learning_rate": 9.647832697318207e-06, | |
| "loss": 0.3561, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 7.357142857142857, | |
| "grad_norm": 1.3831766843795776, | |
| "learning_rate": 9.641237683930722e-06, | |
| "loss": 0.397, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 7.392857142857143, | |
| "grad_norm": 1.2713834047317505, | |
| "learning_rate": 9.63458378673011e-06, | |
| "loss": 0.3405, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 1.6474286317825317, | |
| "learning_rate": 9.627871090134984e-06, | |
| "loss": 0.327, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 7.464285714285714, | |
| "grad_norm": 1.477697730064392, | |
| "learning_rate": 9.621099679309948e-06, | |
| "loss": 0.3921, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.6822872161865234, | |
| "learning_rate": 9.61426964016452e-06, | |
| "loss": 0.3775, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 7.535714285714286, | |
| "grad_norm": 1.541839838027954, | |
| "learning_rate": 9.60738105935204e-06, | |
| "loss": 0.3586, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 7.571428571428571, | |
| "grad_norm": 1.4712790250778198, | |
| "learning_rate": 9.60043402426857e-06, | |
| "loss": 0.3626, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 7.607142857142857, | |
| "grad_norm": 1.3638650178909302, | |
| "learning_rate": 9.593428623051793e-06, | |
| "loss": 0.3889, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 7.642857142857143, | |
| "grad_norm": 1.3237727880477905, | |
| "learning_rate": 9.58636494457988e-06, | |
| "loss": 0.4192, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 7.678571428571429, | |
| "grad_norm": 1.4845259189605713, | |
| "learning_rate": 9.57924307847038e-06, | |
| "loss": 0.408, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 7.714285714285714, | |
| "grad_norm": 1.4447940587997437, | |
| "learning_rate": 9.572063115079063e-06, | |
| "loss": 0.3425, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "grad_norm": 1.365172266960144, | |
| "learning_rate": 9.564825145498795e-06, | |
| "loss": 0.3447, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 7.785714285714286, | |
| "grad_norm": 1.419386625289917, | |
| "learning_rate": 9.557529261558367e-06, | |
| "loss": 0.3816, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 7.821428571428571, | |
| "grad_norm": 1.5688282251358032, | |
| "learning_rate": 9.550175555821333e-06, | |
| "loss": 0.3568, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 1.4586085081100464, | |
| "learning_rate": 9.542764121584845e-06, | |
| "loss": 0.4127, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 7.892857142857143, | |
| "grad_norm": 1.4946858882904053, | |
| "learning_rate": 9.53529505287845e-06, | |
| "loss": 0.3844, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 7.928571428571429, | |
| "grad_norm": 1.4703893661499023, | |
| "learning_rate": 9.527768444462922e-06, | |
| "loss": 0.3597, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 7.964285714285714, | |
| "grad_norm": 1.5224125385284424, | |
| "learning_rate": 9.520184391829037e-06, | |
| "loss": 0.4019, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.2209713459014893, | |
| "learning_rate": 9.512542991196377e-06, | |
| "loss": 0.2579, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 8.035714285714286, | |
| "grad_norm": 1.3363842964172363, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.2908, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 8.071428571428571, | |
| "grad_norm": 1.3185490369796753, | |
| "learning_rate": 9.497088534449707e-06, | |
| "loss": 0.2495, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 8.107142857142858, | |
| "grad_norm": 1.544686198234558, | |
| "learning_rate": 9.489275674407826e-06, | |
| "loss": 0.2688, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 8.142857142857142, | |
| "grad_norm": 1.9225260019302368, | |
| "learning_rate": 9.481405858508935e-06, | |
| "loss": 0.2262, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 8.178571428571429, | |
| "grad_norm": 1.690416693687439, | |
| "learning_rate": 9.473479186598115e-06, | |
| "loss": 0.2447, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 8.214285714285714, | |
| "grad_norm": 1.738194465637207, | |
| "learning_rate": 9.465495759241793e-06, | |
| "loss": 0.2446, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "grad_norm": 1.5437583923339844, | |
| "learning_rate": 9.457455677726447e-06, | |
| "loss": 0.2381, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "grad_norm": 1.5749611854553223, | |
| "learning_rate": 9.449359044057344e-06, | |
| "loss": 0.2048, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 8.321428571428571, | |
| "grad_norm": 1.3072010278701782, | |
| "learning_rate": 9.441205960957221e-06, | |
| "loss": 0.2285, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 8.357142857142858, | |
| "grad_norm": 1.231365442276001, | |
| "learning_rate": 9.432996531865001e-06, | |
| "loss": 0.2356, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 8.392857142857142, | |
| "grad_norm": 1.3257508277893066, | |
| "learning_rate": 9.424730860934474e-06, | |
| "loss": 0.2461, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 8.428571428571429, | |
| "grad_norm": 1.645837426185608, | |
| "learning_rate": 9.416409053032971e-06, | |
| "loss": 0.2078, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 8.464285714285714, | |
| "grad_norm": 1.4033600091934204, | |
| "learning_rate": 9.408031213740045e-06, | |
| "loss": 0.2085, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 1.4971022605895996, | |
| "learning_rate": 9.399597449346119e-06, | |
| "loss": 0.2101, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 8.535714285714286, | |
| "grad_norm": 1.5445584058761597, | |
| "learning_rate": 9.391107866851143e-06, | |
| "loss": 0.2305, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 1.6261149644851685, | |
| "learning_rate": 9.382562573963238e-06, | |
| "loss": 0.2392, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 8.607142857142858, | |
| "grad_norm": 1.9455870389938354, | |
| "learning_rate": 9.37396167909733e-06, | |
| "loss": 0.2314, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 8.642857142857142, | |
| "grad_norm": 1.6666280031204224, | |
| "learning_rate": 9.365305291373769e-06, | |
| "loss": 0.2831, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 8.678571428571429, | |
| "grad_norm": 1.313746690750122, | |
| "learning_rate": 9.356593520616948e-06, | |
| "loss": 0.2016, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 8.714285714285714, | |
| "grad_norm": 1.4457995891571045, | |
| "learning_rate": 9.347826477353911e-06, | |
| "loss": 0.1895, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 1.6301132440567017, | |
| "learning_rate": 9.33900427281295e-06, | |
| "loss": 0.2404, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 8.785714285714286, | |
| "grad_norm": 1.0887130498886108, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.2037, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 8.821428571428571, | |
| "grad_norm": 1.372310996055603, | |
| "learning_rate": 9.321194828308185e-06, | |
| "loss": 0.2165, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "grad_norm": 1.2410130500793457, | |
| "learning_rate": 9.312207814294454e-06, | |
| "loss": 0.2109, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 8.892857142857142, | |
| "grad_norm": 1.3588364124298096, | |
| "learning_rate": 9.303166090900082e-06, | |
| "loss": 0.212, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 8.928571428571429, | |
| "grad_norm": 1.1505169868469238, | |
| "learning_rate": 9.294069772838253e-06, | |
| "loss": 0.2078, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 8.964285714285714, | |
| "grad_norm": 1.456135869026184, | |
| "learning_rate": 9.284918975514798e-06, | |
| "loss": 0.2207, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.2700921297073364, | |
| "learning_rate": 9.275713815026732e-06, | |
| "loss": 0.1748, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 9.035714285714286, | |
| "grad_norm": 1.0988452434539795, | |
| "learning_rate": 9.266454408160779e-06, | |
| "loss": 0.1457, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 9.071428571428571, | |
| "grad_norm": 1.0384048223495483, | |
| "learning_rate": 9.257140872391895e-06, | |
| "loss": 0.1342, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 9.107142857142858, | |
| "grad_norm": 1.3684717416763306, | |
| "learning_rate": 9.24777332588177e-06, | |
| "loss": 0.1019, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 1.41575026512146, | |
| "learning_rate": 9.238351887477338e-06, | |
| "loss": 0.1322, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 9.178571428571429, | |
| "grad_norm": 2.176236391067505, | |
| "learning_rate": 9.22887667670926e-06, | |
| "loss": 0.1442, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 9.214285714285714, | |
| "grad_norm": 1.456347942352295, | |
| "learning_rate": 9.219347813790416e-06, | |
| "loss": 0.1149, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 9.25, | |
| "grad_norm": 1.751892328262329, | |
| "learning_rate": 9.209765419614375e-06, | |
| "loss": 0.1544, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 1.2597330808639526, | |
| "learning_rate": 9.200129615753858e-06, | |
| "loss": 0.1247, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 9.321428571428571, | |
| "grad_norm": 1.3868889808654785, | |
| "learning_rate": 9.190440524459203e-06, | |
| "loss": 0.1597, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 9.357142857142858, | |
| "grad_norm": 1.349303960800171, | |
| "learning_rate": 9.180698268656814e-06, | |
| "loss": 0.1262, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 9.392857142857142, | |
| "grad_norm": 1.2209405899047852, | |
| "learning_rate": 9.170902971947589e-06, | |
| "loss": 0.1283, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 9.428571428571429, | |
| "grad_norm": 1.1362848281860352, | |
| "learning_rate": 9.16105475860537e-06, | |
| "loss": 0.1142, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 9.464285714285714, | |
| "grad_norm": 1.333030343055725, | |
| "learning_rate": 9.151153753575351e-06, | |
| "loss": 0.1241, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 1.1623438596725464, | |
| "learning_rate": 9.141200082472503e-06, | |
| "loss": 0.1157, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 9.535714285714286, | |
| "grad_norm": 1.3490017652511597, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.1696, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 9.571428571428571, | |
| "grad_norm": 1.5136406421661377, | |
| "learning_rate": 9.121135247847492e-06, | |
| "loss": 0.1364, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 9.607142857142858, | |
| "grad_norm": 1.6852158308029175, | |
| "learning_rate": 9.111024338889748e-06, | |
| "loss": 0.1288, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 9.642857142857142, | |
| "grad_norm": 1.359726071357727, | |
| "learning_rate": 9.10086127298478e-06, | |
| "loss": 0.1219, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 9.678571428571429, | |
| "grad_norm": 1.0993460416793823, | |
| "learning_rate": 9.090646179072352e-06, | |
| "loss": 0.1314, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "grad_norm": 1.2296324968338013, | |
| "learning_rate": 9.080379186752304e-06, | |
| "loss": 0.1295, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "grad_norm": 1.2677600383758545, | |
| "learning_rate": 9.070060426282924e-06, | |
| "loss": 0.1494, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 9.785714285714286, | |
| "grad_norm": 0.98940509557724, | |
| "learning_rate": 9.059690028579285e-06, | |
| "loss": 0.1142, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 9.821428571428571, | |
| "grad_norm": 1.3205757141113281, | |
| "learning_rate": 9.049268125211577e-06, | |
| "loss": 0.1331, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 9.857142857142858, | |
| "grad_norm": 1.6608718633651733, | |
| "learning_rate": 9.038794848403463e-06, | |
| "loss": 0.158, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 9.892857142857142, | |
| "grad_norm": 1.2538622617721558, | |
| "learning_rate": 9.028270331030373e-06, | |
| "loss": 0.1281, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 9.928571428571429, | |
| "grad_norm": 1.2672852277755737, | |
| "learning_rate": 9.017694706617836e-06, | |
| "loss": 0.1477, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 9.964285714285714, | |
| "grad_norm": 1.1793338060379028, | |
| "learning_rate": 9.007068109339783e-06, | |
| "loss": 0.116, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.9627423286437988, | |
| "learning_rate": 8.996390674016839e-06, | |
| "loss": 0.0847, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 10.035714285714286, | |
| "grad_norm": 0.9174171686172485, | |
| "learning_rate": 8.985662536114614e-06, | |
| "loss": 0.0765, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 10.071428571428571, | |
| "grad_norm": 1.147916555404663, | |
| "learning_rate": 8.97488383174199e-06, | |
| "loss": 0.086, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 10.107142857142858, | |
| "grad_norm": 1.289789080619812, | |
| "learning_rate": 8.964054697649389e-06, | |
| "loss": 0.0736, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 10.142857142857142, | |
| "grad_norm": 0.9539849758148193, | |
| "learning_rate": 8.953175271227042e-06, | |
| "loss": 0.0702, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 10.178571428571429, | |
| "grad_norm": 1.0880526304244995, | |
| "learning_rate": 8.94224569050324e-06, | |
| "loss": 0.0662, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 10.214285714285714, | |
| "grad_norm": 1.3260632753372192, | |
| "learning_rate": 8.931266094142588e-06, | |
| "loss": 0.0878, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 10.25, | |
| "grad_norm": 1.0914793014526367, | |
| "learning_rate": 8.920236621444243e-06, | |
| "loss": 0.0716, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 10.285714285714286, | |
| "grad_norm": 1.0627952814102173, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.0769, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 10.321428571428571, | |
| "grad_norm": 0.9562764763832092, | |
| "learning_rate": 8.89802860739326e-06, | |
| "loss": 0.062, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 10.357142857142858, | |
| "grad_norm": 1.1574186086654663, | |
| "learning_rate": 8.88685034779576e-06, | |
| "loss": 0.0857, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 10.392857142857142, | |
| "grad_norm": 1.2380372285842896, | |
| "learning_rate": 8.87562277536726e-06, | |
| "loss": 0.0908, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 10.428571428571429, | |
| "grad_norm": 0.8159970045089722, | |
| "learning_rate": 8.864346032553016e-06, | |
| "loss": 0.068, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 10.464285714285714, | |
| "grad_norm": 1.1167945861816406, | |
| "learning_rate": 8.853020262422111e-06, | |
| "loss": 0.0696, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 1.2967480421066284, | |
| "learning_rate": 8.84164560866564e-06, | |
| "loss": 0.0803, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 10.535714285714286, | |
| "grad_norm": 1.5502207279205322, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.0984, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 10.571428571428571, | |
| "grad_norm": 0.9975540637969971, | |
| "learning_rate": 8.818750228139513e-06, | |
| "loss": 0.0675, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 10.607142857142858, | |
| "grad_norm": 0.9909189939498901, | |
| "learning_rate": 8.807229791845673e-06, | |
| "loss": 0.079, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 10.642857142857142, | |
| "grad_norm": 1.3548370599746704, | |
| "learning_rate": 8.795661052874217e-06, | |
| "loss": 0.0909, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 10.678571428571429, | |
| "grad_norm": 1.3748208284378052, | |
| "learning_rate": 8.78404415799881e-06, | |
| "loss": 0.079, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "grad_norm": 1.6325222253799438, | |
| "learning_rate": 8.772379254604074e-06, | |
| "loss": 0.0978, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "eval_loss": 2.0024502277374268, | |
| "eval_runtime": 0.4759, | |
| "eval_samples_per_second": 294.154, | |
| "eval_steps_per_second": 2.101, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.75, | |
| "grad_norm": 1.1626653671264648, | |
| "learning_rate": 8.76066649068372e-06, | |
| "loss": 0.0926, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 10.785714285714286, | |
| "grad_norm": 1.312601089477539, | |
| "learning_rate": 8.748906014838672e-06, | |
| "loss": 0.0895, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 10.821428571428571, | |
| "grad_norm": 0.9997202754020691, | |
| "learning_rate": 8.737097976275177e-06, | |
| "loss": 0.0765, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 10.857142857142858, | |
| "grad_norm": 0.9879180788993835, | |
| "learning_rate": 8.725242524802919e-06, | |
| "loss": 0.081, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 10.892857142857142, | |
| "grad_norm": 1.0174518823623657, | |
| "learning_rate": 8.713339810833105e-06, | |
| "loss": 0.0714, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 10.928571428571429, | |
| "grad_norm": 1.340936541557312, | |
| "learning_rate": 8.701389985376578e-06, | |
| "loss": 0.0826, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 10.964285714285714, | |
| "grad_norm": 1.0532597303390503, | |
| "learning_rate": 8.689393200041878e-06, | |
| "loss": 0.0914, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.8435774445533752, | |
| "learning_rate": 8.677349607033336e-06, | |
| "loss": 0.0473, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 11.035714285714286, | |
| "grad_norm": 0.7733930349349976, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.0461, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 11.071428571428571, | |
| "grad_norm": 0.7377230525016785, | |
| "learning_rate": 8.653122609779365e-06, | |
| "loss": 0.0506, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 11.107142857142858, | |
| "grad_norm": 0.9703754186630249, | |
| "learning_rate": 8.640939512904097e-06, | |
| "loss": 0.044, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 11.142857142857142, | |
| "grad_norm": 0.9164966940879822, | |
| "learning_rate": 8.62871022309141e-06, | |
| "loss": 0.0442, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 11.178571428571429, | |
| "grad_norm": 0.5867542624473572, | |
| "learning_rate": 8.61643489549544e-06, | |
| "loss": 0.0356, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 11.214285714285714, | |
| "grad_norm": 0.9844388365745544, | |
| "learning_rate": 8.604113685854407e-06, | |
| "loss": 0.0465, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "grad_norm": 0.8281243443489075, | |
| "learning_rate": 8.591746750488639e-06, | |
| "loss": 0.0485, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 11.285714285714286, | |
| "grad_norm": 1.1860142946243286, | |
| "learning_rate": 8.579334246298593e-06, | |
| "loss": 0.0623, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 11.321428571428571, | |
| "grad_norm": 0.7477958798408508, | |
| "learning_rate": 8.566876330762861e-06, | |
| "loss": 0.0485, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 11.357142857142858, | |
| "grad_norm": 1.0370478630065918, | |
| "learning_rate": 8.554373161936176e-06, | |
| "loss": 0.0549, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 11.392857142857142, | |
| "grad_norm": 0.7288575172424316, | |
| "learning_rate": 8.541824898447399e-06, | |
| "loss": 0.0419, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 0.7542941570281982, | |
| "learning_rate": 8.529231699497512e-06, | |
| "loss": 0.0538, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 11.464285714285714, | |
| "grad_norm": 0.7196518778800964, | |
| "learning_rate": 8.516593724857598e-06, | |
| "loss": 0.0476, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.6946846842765808, | |
| "learning_rate": 8.503911134866819e-06, | |
| "loss": 0.048, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 11.535714285714286, | |
| "grad_norm": 0.81142657995224, | |
| "learning_rate": 8.491184090430365e-06, | |
| "loss": 0.052, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 11.571428571428571, | |
| "grad_norm": 0.9150547981262207, | |
| "learning_rate": 8.478412753017433e-06, | |
| "loss": 0.0451, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 11.607142857142858, | |
| "grad_norm": 1.1183067560195923, | |
| "learning_rate": 8.465597284659163e-06, | |
| "loss": 0.0511, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 11.642857142857142, | |
| "grad_norm": 1.272926688194275, | |
| "learning_rate": 8.452737847946597e-06, | |
| "loss": 0.0612, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 11.678571428571429, | |
| "grad_norm": 0.9870100617408752, | |
| "learning_rate": 8.439834606028594e-06, | |
| "loss": 0.0663, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 11.714285714285714, | |
| "grad_norm": 0.9883282780647278, | |
| "learning_rate": 8.426887722609787e-06, | |
| "loss": 0.0538, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 11.75, | |
| "grad_norm": 1.1376686096191406, | |
| "learning_rate": 8.413897361948484e-06, | |
| "loss": 0.059, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 11.785714285714286, | |
| "grad_norm": 1.1064047813415527, | |
| "learning_rate": 8.400863688854598e-06, | |
| "loss": 0.0597, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 11.821428571428571, | |
| "grad_norm": 1.1057697534561157, | |
| "learning_rate": 8.387786868687549e-06, | |
| "loss": 0.0445, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 11.857142857142858, | |
| "grad_norm": 0.7781133651733398, | |
| "learning_rate": 8.374667067354164e-06, | |
| "loss": 0.0616, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 11.892857142857142, | |
| "grad_norm": 0.7614454627037048, | |
| "learning_rate": 8.361504451306585e-06, | |
| "loss": 0.0467, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 11.928571428571429, | |
| "grad_norm": 1.1504416465759277, | |
| "learning_rate": 8.34829918754014e-06, | |
| "loss": 0.0685, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 11.964285714285714, | |
| "grad_norm": 0.943291962146759, | |
| "learning_rate": 8.335051443591236e-06, | |
| "loss": 0.0503, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.6303163766860962, | |
| "learning_rate": 8.321761387535231e-06, | |
| "loss": 0.0429, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 12.035714285714286, | |
| "grad_norm": 0.5599375367164612, | |
| "learning_rate": 8.308429187984298e-06, | |
| "loss": 0.033, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 12.071428571428571, | |
| "grad_norm": 0.5403127074241638, | |
| "learning_rate": 8.295055014085289e-06, | |
| "loss": 0.0379, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 12.107142857142858, | |
| "grad_norm": 0.6652883887290955, | |
| "learning_rate": 8.281639035517591e-06, | |
| "loss": 0.0423, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 12.142857142857142, | |
| "grad_norm": 0.6625171899795532, | |
| "learning_rate": 8.268181422490969e-06, | |
| "loss": 0.0284, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 12.178571428571429, | |
| "grad_norm": 0.6045703291893005, | |
| "learning_rate": 8.254682345743406e-06, | |
| "loss": 0.0302, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 12.214285714285714, | |
| "grad_norm": 0.9115837216377258, | |
| "learning_rate": 8.241141976538944e-06, | |
| "loss": 0.0341, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 12.25, | |
| "grad_norm": 0.650475263595581, | |
| "learning_rate": 8.227560486665498e-06, | |
| "loss": 0.025, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 12.285714285714286, | |
| "grad_norm": 0.8756570816040039, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.0273, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 12.321428571428571, | |
| "grad_norm": 0.5917655229568481, | |
| "learning_rate": 8.200274834669675e-06, | |
| "loss": 0.0299, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 12.357142857142858, | |
| "grad_norm": 1.3080816268920898, | |
| "learning_rate": 8.186571018722894e-06, | |
| "loss": 0.0278, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 12.392857142857142, | |
| "grad_norm": 0.994071364402771, | |
| "learning_rate": 8.172826774453937e-06, | |
| "loss": 0.037, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 12.428571428571429, | |
| "grad_norm": 1.0529603958129883, | |
| "learning_rate": 8.159042276237308e-06, | |
| "loss": 0.0457, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 12.464285714285714, | |
| "grad_norm": 0.8449487686157227, | |
| "learning_rate": 8.145217698958213e-06, | |
| "loss": 0.0334, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.7479754090309143, | |
| "learning_rate": 8.131353218010347e-06, | |
| "loss": 0.0381, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 12.535714285714286, | |
| "grad_norm": 0.4850471317768097, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.0331, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 12.571428571428571, | |
| "grad_norm": 1.2365081310272217, | |
| "learning_rate": 8.10350524921216e-06, | |
| "loss": 0.0379, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 12.607142857142858, | |
| "grad_norm": 0.731280505657196, | |
| "learning_rate": 8.089522114671603e-06, | |
| "loss": 0.0349, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 12.642857142857142, | |
| "grad_norm": 0.6076593995094299, | |
| "learning_rate": 8.075499783077321e-06, | |
| "loss": 0.0354, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 12.678571428571429, | |
| "grad_norm": 0.5255799293518066, | |
| "learning_rate": 8.061438432331935e-06, | |
| "loss": 0.0358, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 12.714285714285714, | |
| "grad_norm": 0.8889122009277344, | |
| "learning_rate": 8.047338240833108e-06, | |
| "loss": 0.0485, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 12.75, | |
| "grad_norm": 0.5412372350692749, | |
| "learning_rate": 8.033199387471278e-06, | |
| "loss": 0.04, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 12.785714285714286, | |
| "grad_norm": 0.8177741765975952, | |
| "learning_rate": 8.019022051627387e-06, | |
| "loss": 0.0285, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 12.821428571428571, | |
| "grad_norm": 0.5282717347145081, | |
| "learning_rate": 8.004806413170613e-06, | |
| "loss": 0.0356, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "grad_norm": 0.7756925225257874, | |
| "learning_rate": 7.99055265245608e-06, | |
| "loss": 0.0436, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 12.892857142857142, | |
| "grad_norm": 0.9730148911476135, | |
| "learning_rate": 7.976260950322572e-06, | |
| "loss": 0.0434, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 12.928571428571429, | |
| "grad_norm": 0.7948063015937805, | |
| "learning_rate": 7.96193148809024e-06, | |
| "loss": 0.0451, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 12.964285714285714, | |
| "grad_norm": 0.5865257978439331, | |
| "learning_rate": 7.9475644475583e-06, | |
| "loss": 0.0291, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.4693900942802429, | |
| "learning_rate": 7.933160011002729e-06, | |
| "loss": 0.0277, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 13.035714285714286, | |
| "grad_norm": 0.3828136920928955, | |
| "learning_rate": 7.918718361173951e-06, | |
| "loss": 0.02, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 13.071428571428571, | |
| "grad_norm": 0.596170961856842, | |
| "learning_rate": 7.904239681294515e-06, | |
| "loss": 0.0185, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 13.107142857142858, | |
| "grad_norm": 0.39570969343185425, | |
| "learning_rate": 7.889724155056776e-06, | |
| "loss": 0.0195, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 13.142857142857142, | |
| "grad_norm": 0.5549917221069336, | |
| "learning_rate": 7.875171966620567e-06, | |
| "loss": 0.0266, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 13.178571428571429, | |
| "grad_norm": 0.42615193128585815, | |
| "learning_rate": 7.860583300610849e-06, | |
| "loss": 0.0193, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 13.214285714285714, | |
| "grad_norm": 0.6593537926673889, | |
| "learning_rate": 7.84595834211538e-06, | |
| "loss": 0.0213, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 13.25, | |
| "grad_norm": 0.49755775928497314, | |
| "learning_rate": 7.83129727668237e-06, | |
| "loss": 0.0235, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 13.285714285714286, | |
| "grad_norm": 0.9553983807563782, | |
| "learning_rate": 7.81660029031811e-06, | |
| "loss": 0.0254, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 13.321428571428571, | |
| "grad_norm": 0.6323786377906799, | |
| "learning_rate": 7.801867569484635e-06, | |
| "loss": 0.0255, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 13.357142857142858, | |
| "grad_norm": 0.7185072302818298, | |
| "learning_rate": 7.78709930109734e-06, | |
| "loss": 0.0277, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 13.392857142857142, | |
| "grad_norm": 0.508331835269928, | |
| "learning_rate": 7.772295672522615e-06, | |
| "loss": 0.0285, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 13.428571428571429, | |
| "grad_norm": 0.5408092141151428, | |
| "learning_rate": 7.75745687157547e-06, | |
| "loss": 0.0236, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 13.464285714285714, | |
| "grad_norm": 0.7089282274246216, | |
| "learning_rate": 7.742583086517151e-06, | |
| "loss": 0.0184, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 0.5609864592552185, | |
| "learning_rate": 7.727674506052744e-06, | |
| "loss": 0.026, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 13.535714285714286, | |
| "grad_norm": 0.5647308826446533, | |
| "learning_rate": 7.712731319328798e-06, | |
| "loss": 0.0251, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 13.571428571428571, | |
| "grad_norm": 0.46377676725387573, | |
| "learning_rate": 7.697753715930906e-06, | |
| "loss": 0.0224, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 13.607142857142858, | |
| "grad_norm": 0.44435882568359375, | |
| "learning_rate": 7.682741885881314e-06, | |
| "loss": 0.0261, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 13.642857142857142, | |
| "grad_norm": 0.9688218832015991, | |
| "learning_rate": 7.667696019636504e-06, | |
| "loss": 0.0418, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 13.678571428571429, | |
| "grad_norm": 0.48680123686790466, | |
| "learning_rate": 7.652616308084774e-06, | |
| "loss": 0.0296, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 13.714285714285714, | |
| "grad_norm": 0.9336219429969788, | |
| "learning_rate": 7.637502942543825e-06, | |
| "loss": 0.0345, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "grad_norm": 0.5719844102859497, | |
| "learning_rate": 7.622356114758328e-06, | |
| "loss": 0.0374, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 13.785714285714286, | |
| "grad_norm": 0.7421361207962036, | |
| "learning_rate": 7.607176016897491e-06, | |
| "loss": 0.0335, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 13.821428571428571, | |
| "grad_norm": 0.4875544011592865, | |
| "learning_rate": 7.591962841552627e-06, | |
| "loss": 0.0312, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 13.857142857142858, | |
| "grad_norm": 0.5115516185760498, | |
| "learning_rate": 7.576716781734699e-06, | |
| "loss": 0.0306, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 13.892857142857142, | |
| "grad_norm": 0.4477997422218323, | |
| "learning_rate": 7.561438030871886e-06, | |
| "loss": 0.0248, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 13.928571428571429, | |
| "grad_norm": 0.35774773359298706, | |
| "learning_rate": 7.546126782807117e-06, | |
| "loss": 0.0243, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 13.964285714285714, | |
| "grad_norm": 0.3818391263484955, | |
| "learning_rate": 7.530783231795615e-06, | |
| "loss": 0.0276, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.2867811620235443, | |
| "learning_rate": 7.515407572502438e-06, | |
| "loss": 0.016, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 14.035714285714286, | |
| "grad_norm": 0.33533164858818054, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0185, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 14.071428571428571, | |
| "grad_norm": 0.27466756105422974, | |
| "learning_rate": 7.484560709765605e-06, | |
| "loss": 0.0171, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 14.107142857142858, | |
| "grad_norm": 0.4225900173187256, | |
| "learning_rate": 7.469089897678958e-06, | |
| "loss": 0.0227, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 14.142857142857142, | |
| "grad_norm": 1.1478960514068604, | |
| "learning_rate": 7.453587760019691e-06, | |
| "loss": 0.0308, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 14.178571428571429, | |
| "grad_norm": 0.3350810408592224, | |
| "learning_rate": 7.438054493464859e-06, | |
| "loss": 0.016, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 14.214285714285714, | |
| "grad_norm": 0.3689461946487427, | |
| "learning_rate": 7.422490295086457e-06, | |
| "loss": 0.0219, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 14.25, | |
| "grad_norm": 0.34574565291404724, | |
| "learning_rate": 7.406895362348916e-06, | |
| "loss": 0.0165, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 0.39184218645095825, | |
| "learning_rate": 7.391269893106592e-06, | |
| "loss": 0.0219, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "eval_loss": 2.3083860874176025, | |
| "eval_runtime": 0.4688, | |
| "eval_samples_per_second": 298.631, | |
| "eval_steps_per_second": 2.133, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.321428571428571, | |
| "grad_norm": 0.5275436043739319, | |
| "learning_rate": 7.375614085601265e-06, | |
| "loss": 0.0188, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 14.357142857142858, | |
| "grad_norm": 0.36986497044563293, | |
| "learning_rate": 7.359928138459615e-06, | |
| "loss": 0.0194, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 14.392857142857142, | |
| "grad_norm": 0.35321933031082153, | |
| "learning_rate": 7.344212250690712e-06, | |
| "loss": 0.0195, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 14.428571428571429, | |
| "grad_norm": 0.5846700072288513, | |
| "learning_rate": 7.328466621683481e-06, | |
| "loss": 0.0199, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 14.464285714285714, | |
| "grad_norm": 0.36776694655418396, | |
| "learning_rate": 7.312691451204178e-06, | |
| "loss": 0.0182, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 14.5, | |
| "grad_norm": 0.7046879529953003, | |
| "learning_rate": 7.296886939393852e-06, | |
| "loss": 0.0241, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 14.535714285714286, | |
| "grad_norm": 0.5038862824440002, | |
| "learning_rate": 7.281053286765816e-06, | |
| "loss": 0.0154, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 14.571428571428571, | |
| "grad_norm": 0.42418766021728516, | |
| "learning_rate": 7.265190694203086e-06, | |
| "loss": 0.0212, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 14.607142857142858, | |
| "grad_norm": 0.5494039058685303, | |
| "learning_rate": 7.249299362955846e-06, | |
| "loss": 0.0254, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 14.642857142857142, | |
| "grad_norm": 0.5212989449501038, | |
| "learning_rate": 7.233379494638891e-06, | |
| "loss": 0.0216, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 14.678571428571429, | |
| "grad_norm": 0.3063950836658478, | |
| "learning_rate": 7.217431291229068e-06, | |
| "loss": 0.0173, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 14.714285714285714, | |
| "grad_norm": 0.26393306255340576, | |
| "learning_rate": 7.201454955062712e-06, | |
| "loss": 0.0169, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 14.75, | |
| "grad_norm": 0.44368988275527954, | |
| "learning_rate": 7.185450688833083e-06, | |
| "loss": 0.0229, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 14.785714285714286, | |
| "grad_norm": 0.4474826753139496, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.0196, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 14.821428571428571, | |
| "grad_norm": 0.39892250299453735, | |
| "learning_rate": 7.153359178726222e-06, | |
| "loss": 0.0239, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 14.857142857142858, | |
| "grad_norm": 0.6791783571243286, | |
| "learning_rate": 7.137272341996958e-06, | |
| "loss": 0.0241, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 14.892857142857142, | |
| "grad_norm": 0.3901786804199219, | |
| "learning_rate": 7.121158389495187e-06, | |
| "loss": 0.0232, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 14.928571428571429, | |
| "grad_norm": 0.3575139045715332, | |
| "learning_rate": 7.10501752566012e-06, | |
| "loss": 0.0225, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 14.964285714285714, | |
| "grad_norm": 0.48408985137939453, | |
| "learning_rate": 7.088849955272396e-06, | |
| "loss": 0.0246, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.30433982610702515, | |
| "learning_rate": 7.072655883451478e-06, | |
| "loss": 0.0183, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 15.035714285714286, | |
| "grad_norm": 0.23058322072029114, | |
| "learning_rate": 7.056435515653059e-06, | |
| "loss": 0.0147, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 15.071428571428571, | |
| "grad_norm": 0.21939268708229065, | |
| "learning_rate": 7.040189057666449e-06, | |
| "loss": 0.0162, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 15.107142857142858, | |
| "grad_norm": 0.30123960971832275, | |
| "learning_rate": 7.023916715611969e-06, | |
| "loss": 0.0152, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 15.142857142857142, | |
| "grad_norm": 0.24728672206401825, | |
| "learning_rate": 7.007618695938334e-06, | |
| "loss": 0.0126, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 15.178571428571429, | |
| "grad_norm": 0.37947019934654236, | |
| "learning_rate": 6.991295205420028e-06, | |
| "loss": 0.0133, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 15.214285714285714, | |
| "grad_norm": 0.2239277958869934, | |
| "learning_rate": 6.974946451154694e-06, | |
| "loss": 0.0134, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 15.25, | |
| "grad_norm": 0.2562693655490875, | |
| "learning_rate": 6.9585726405604915e-06, | |
| "loss": 0.0159, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 15.285714285714286, | |
| "grad_norm": 0.32447507977485657, | |
| "learning_rate": 6.942173981373474e-06, | |
| "loss": 0.0141, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 15.321428571428571, | |
| "grad_norm": 0.30523914098739624, | |
| "learning_rate": 6.925750681644954e-06, | |
| "loss": 0.0145, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 15.357142857142858, | |
| "grad_norm": 0.33100420236587524, | |
| "learning_rate": 6.90930294973886e-06, | |
| "loss": 0.0158, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 15.392857142857142, | |
| "grad_norm": 0.3664904534816742, | |
| "learning_rate": 6.892830994329089e-06, | |
| "loss": 0.0187, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 15.428571428571429, | |
| "grad_norm": 0.4915764331817627, | |
| "learning_rate": 6.876335024396872e-06, | |
| "loss": 0.0141, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 15.464285714285714, | |
| "grad_norm": 0.3460884392261505, | |
| "learning_rate": 6.859815249228106e-06, | |
| "loss": 0.0206, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 15.5, | |
| "grad_norm": 0.5521484017372131, | |
| "learning_rate": 6.8432718784107145e-06, | |
| "loss": 0.0145, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 15.535714285714286, | |
| "grad_norm": 0.6945492625236511, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.0168, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 15.571428571428571, | |
| "grad_norm": 0.4687165915966034, | |
| "learning_rate": 6.81011518967587e-06, | |
| "loss": 0.0184, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 15.607142857142858, | |
| "grad_norm": 0.30982425808906555, | |
| "learning_rate": 6.793502292420402e-06, | |
| "loss": 0.0191, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 15.642857142857142, | |
| "grad_norm": 0.3576700687408447, | |
| "learning_rate": 6.7768666408349445e-06, | |
| "loss": 0.0118, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 15.678571428571429, | |
| "grad_norm": 0.4062334895133972, | |
| "learning_rate": 6.760208445977551e-06, | |
| "loss": 0.0201, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 15.714285714285714, | |
| "grad_norm": 0.44516295194625854, | |
| "learning_rate": 6.743527919192285e-06, | |
| "loss": 0.0183, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 15.75, | |
| "grad_norm": 0.3430921137332916, | |
| "learning_rate": 6.726825272106539e-06, | |
| "loss": 0.0172, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 15.785714285714286, | |
| "grad_norm": 0.34668734669685364, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.0202, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 15.821428571428571, | |
| "grad_norm": 0.34746137261390686, | |
| "learning_rate": 6.693354464943689e-06, | |
| "loss": 0.0175, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 15.857142857142858, | |
| "grad_norm": 0.38835713267326355, | |
| "learning_rate": 6.676586729513823e-06, | |
| "loss": 0.0205, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 15.892857142857142, | |
| "grad_norm": 0.4056047797203064, | |
| "learning_rate": 6.659797723072558e-06, | |
| "loss": 0.0206, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 15.928571428571429, | |
| "grad_norm": 0.3119281530380249, | |
| "learning_rate": 6.642987658623581e-06, | |
| "loss": 0.0207, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 15.964285714285714, | |
| "grad_norm": 0.4128764271736145, | |
| "learning_rate": 6.626156749437736e-06, | |
| "loss": 0.0216, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.4920613765716553, | |
| "learning_rate": 6.609305209050332e-06, | |
| "loss": 0.0178, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 16.035714285714285, | |
| "grad_norm": 0.5353336930274963, | |
| "learning_rate": 6.592433251258423e-06, | |
| "loss": 0.0132, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 16.071428571428573, | |
| "grad_norm": 0.2517736554145813, | |
| "learning_rate": 6.575541090118105e-06, | |
| "loss": 0.0142, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 16.107142857142858, | |
| "grad_norm": 0.23888464272022247, | |
| "learning_rate": 6.558628939941792e-06, | |
| "loss": 0.0139, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 16.142857142857142, | |
| "grad_norm": 0.23904645442962646, | |
| "learning_rate": 6.541697015295503e-06, | |
| "loss": 0.0147, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 16.178571428571427, | |
| "grad_norm": 0.16995219886302948, | |
| "learning_rate": 6.524745530996137e-06, | |
| "loss": 0.0112, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 16.214285714285715, | |
| "grad_norm": 0.301095575094223, | |
| "learning_rate": 6.507774702108748e-06, | |
| "loss": 0.0138, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "grad_norm": 0.19944924116134644, | |
| "learning_rate": 6.490784743943819e-06, | |
| "loss": 0.0111, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 16.285714285714285, | |
| "grad_norm": 0.2589744031429291, | |
| "learning_rate": 6.473775872054522e-06, | |
| "loss": 0.0149, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 16.321428571428573, | |
| "grad_norm": 0.25151684880256653, | |
| "learning_rate": 6.456748302233995e-06, | |
| "loss": 0.0134, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 16.357142857142858, | |
| "grad_norm": 0.23779137432575226, | |
| "learning_rate": 6.439702250512596e-06, | |
| "loss": 0.0153, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 16.392857142857142, | |
| "grad_norm": 0.30735448002815247, | |
| "learning_rate": 6.4226379331551625e-06, | |
| "loss": 0.0133, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 16.428571428571427, | |
| "grad_norm": 0.34077078104019165, | |
| "learning_rate": 6.405555566658276e-06, | |
| "loss": 0.0172, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 16.464285714285715, | |
| "grad_norm": 0.33910149335861206, | |
| "learning_rate": 6.388455367747503e-06, | |
| "loss": 0.013, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 16.5, | |
| "grad_norm": 0.5794575810432434, | |
| "learning_rate": 6.3713375533746525e-06, | |
| "loss": 0.0133, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 16.535714285714285, | |
| "grad_norm": 0.2994963228702545, | |
| "learning_rate": 6.354202340715027e-06, | |
| "loss": 0.0143, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 16.571428571428573, | |
| "grad_norm": 0.2810235023498535, | |
| "learning_rate": 6.337049947164656e-06, | |
| "loss": 0.0153, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 16.607142857142858, | |
| "grad_norm": 0.28916069865226746, | |
| "learning_rate": 6.319880590337549e-06, | |
| "loss": 0.0173, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 16.642857142857142, | |
| "grad_norm": 0.3143618106842041, | |
| "learning_rate": 6.302694488062931e-06, | |
| "loss": 0.0185, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 16.678571428571427, | |
| "grad_norm": 0.24434876441955566, | |
| "learning_rate": 6.2854918583824745e-06, | |
| "loss": 0.014, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 16.714285714285715, | |
| "grad_norm": 0.37958788871765137, | |
| "learning_rate": 6.268272919547537e-06, | |
| "loss": 0.0172, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 16.75, | |
| "grad_norm": 0.21300822496414185, | |
| "learning_rate": 6.251037890016396e-06, | |
| "loss": 0.0128, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 16.785714285714285, | |
| "grad_norm": 0.35435208678245544, | |
| "learning_rate": 6.233786988451468e-06, | |
| "loss": 0.0148, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 16.821428571428573, | |
| "grad_norm": 0.4400753080844879, | |
| "learning_rate": 6.216520433716544e-06, | |
| "loss": 0.0188, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 16.857142857142858, | |
| "grad_norm": 0.30804184079170227, | |
| "learning_rate": 6.199238444874005e-06, | |
| "loss": 0.0172, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 16.892857142857142, | |
| "grad_norm": 0.23466266691684723, | |
| "learning_rate": 6.181941241182044e-06, | |
| "loss": 0.0149, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 16.928571428571427, | |
| "grad_norm": 0.40080544352531433, | |
| "learning_rate": 6.164629042091894e-06, | |
| "loss": 0.0159, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 16.964285714285715, | |
| "grad_norm": 0.29989176988601685, | |
| "learning_rate": 6.1473020672450275e-06, | |
| "loss": 0.0175, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.23383177816867828, | |
| "learning_rate": 6.1299605364703826e-06, | |
| "loss": 0.0137, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 17.035714285714285, | |
| "grad_norm": 0.1993263214826584, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.0126, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 17.071428571428573, | |
| "grad_norm": 0.19340385496616364, | |
| "learning_rate": 6.095234687374085e-06, | |
| "loss": 0.0111, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 17.107142857142858, | |
| "grad_norm": 0.2317463755607605, | |
| "learning_rate": 6.0778508096224985e-06, | |
| "loss": 0.014, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 0.17298246920108795, | |
| "learning_rate": 6.060453257077686e-06, | |
| "loss": 0.011, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 17.178571428571427, | |
| "grad_norm": 0.17967282235622406, | |
| "learning_rate": 6.043042250464005e-06, | |
| "loss": 0.0113, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 17.214285714285715, | |
| "grad_norm": 0.14388182759284973, | |
| "learning_rate": 6.025618010676516e-06, | |
| "loss": 0.0076, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 17.25, | |
| "grad_norm": 0.32818639278411865, | |
| "learning_rate": 6.008180758778167e-06, | |
| "loss": 0.0116, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 17.285714285714285, | |
| "grad_norm": 0.29826852679252625, | |
| "learning_rate": 5.990730715996989e-06, | |
| "loss": 0.0152, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 17.321428571428573, | |
| "grad_norm": 0.2203068882226944, | |
| "learning_rate": 5.973268103723293e-06, | |
| "loss": 0.0116, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 17.357142857142858, | |
| "grad_norm": 0.19426509737968445, | |
| "learning_rate": 5.955793143506863e-06, | |
| "loss": 0.0108, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 17.392857142857142, | |
| "grad_norm": 0.18054582178592682, | |
| "learning_rate": 5.938306057054139e-06, | |
| "loss": 0.0094, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 17.428571428571427, | |
| "grad_norm": 0.23862619698047638, | |
| "learning_rate": 5.920807066225409e-06, | |
| "loss": 0.0135, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 17.464285714285715, | |
| "grad_norm": 0.21144980192184448, | |
| "learning_rate": 5.903296393031996e-06, | |
| "loss": 0.0109, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "grad_norm": 0.32449236512184143, | |
| "learning_rate": 5.885774259633432e-06, | |
| "loss": 0.0151, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 17.535714285714285, | |
| "grad_norm": 0.7206212282180786, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0104, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 17.571428571428573, | |
| "grad_norm": 0.2740480303764343, | |
| "learning_rate": 5.850696501583164e-06, | |
| "loss": 0.0145, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 17.607142857142858, | |
| "grad_norm": 0.24609383940696716, | |
| "learning_rate": 5.8331413219662295e-06, | |
| "loss": 0.0153, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 17.642857142857142, | |
| "grad_norm": 0.296929270029068, | |
| "learning_rate": 5.815575572208042e-06, | |
| "loss": 0.018, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 17.678571428571427, | |
| "grad_norm": 0.7659111022949219, | |
| "learning_rate": 5.797999475166897e-06, | |
| "loss": 0.017, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 17.714285714285715, | |
| "grad_norm": 0.35650327801704407, | |
| "learning_rate": 5.78041325383237e-06, | |
| "loss": 0.0152, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 17.75, | |
| "grad_norm": 0.2899593412876129, | |
| "learning_rate": 5.762817131322482e-06, | |
| "loss": 0.0143, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 17.785714285714285, | |
| "grad_norm": 0.20666421949863434, | |
| "learning_rate": 5.745211330880872e-06, | |
| "loss": 0.0153, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 17.821428571428573, | |
| "grad_norm": 0.2033378630876541, | |
| "learning_rate": 5.7275960758739655e-06, | |
| "loss": 0.014, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 17.857142857142858, | |
| "grad_norm": 0.24258489906787872, | |
| "learning_rate": 5.709971589788136e-06, | |
| "loss": 0.0181, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 17.857142857142858, | |
| "eval_loss": 2.345672607421875, | |
| "eval_runtime": 0.5108, | |
| "eval_samples_per_second": 274.081, | |
| "eval_steps_per_second": 1.958, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 17.892857142857142, | |
| "grad_norm": 0.28151920437812805, | |
| "learning_rate": 5.69233809622687e-06, | |
| "loss": 0.0181, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 17.928571428571427, | |
| "grad_norm": 0.3348303437232971, | |
| "learning_rate": 5.674695818907943e-06, | |
| "loss": 0.0161, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 17.964285714285715, | |
| "grad_norm": 0.7704312801361084, | |
| "learning_rate": 5.65704498166056e-06, | |
| "loss": 0.0176, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.2007305771112442, | |
| "learning_rate": 5.6393858084225305e-06, | |
| "loss": 0.0129, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 18.035714285714285, | |
| "grad_norm": 0.15058767795562744, | |
| "learning_rate": 5.621718523237427e-06, | |
| "loss": 0.0088, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 18.071428571428573, | |
| "grad_norm": 0.16612547636032104, | |
| "learning_rate": 5.604043350251733e-06, | |
| "loss": 0.0099, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 18.107142857142858, | |
| "grad_norm": 0.20791685581207275, | |
| "learning_rate": 5.586360513712011e-06, | |
| "loss": 0.0117, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 18.142857142857142, | |
| "grad_norm": 0.13226307928562164, | |
| "learning_rate": 5.568670237962045e-06, | |
| "loss": 0.009, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 18.178571428571427, | |
| "grad_norm": 0.2213960736989975, | |
| "learning_rate": 5.550972747440007e-06, | |
| "loss": 0.0124, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 18.214285714285715, | |
| "grad_norm": 0.19490720331668854, | |
| "learning_rate": 5.533268266675601e-06, | |
| "loss": 0.0128, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 18.25, | |
| "grad_norm": 0.1481325626373291, | |
| "learning_rate": 5.515557020287219e-06, | |
| "loss": 0.0078, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 18.285714285714285, | |
| "grad_norm": 0.196213498711586, | |
| "learning_rate": 5.497839232979084e-06, | |
| "loss": 0.0119, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 18.321428571428573, | |
| "grad_norm": 0.26103919744491577, | |
| "learning_rate": 5.480115129538409e-06, | |
| "loss": 0.0152, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 18.357142857142858, | |
| "grad_norm": 0.17787081003189087, | |
| "learning_rate": 5.4623849348325396e-06, | |
| "loss": 0.0102, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 18.392857142857142, | |
| "grad_norm": 0.21203866600990295, | |
| "learning_rate": 5.444648873806101e-06, | |
| "loss": 0.0122, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 18.428571428571427, | |
| "grad_norm": 0.1847071498632431, | |
| "learning_rate": 5.426907171478143e-06, | |
| "loss": 0.0099, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 18.464285714285715, | |
| "grad_norm": 0.2914210557937622, | |
| "learning_rate": 5.409160052939292e-06, | |
| "loss": 0.0133, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 18.5, | |
| "grad_norm": 0.30277934670448303, | |
| "learning_rate": 5.391407743348884e-06, | |
| "loss": 0.0165, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 18.535714285714285, | |
| "grad_norm": 0.22359944880008698, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.0121, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 18.571428571428573, | |
| "grad_norm": 0.3102841377258301, | |
| "learning_rate": 5.355888451977204e-06, | |
| "loss": 0.0163, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 18.607142857142858, | |
| "grad_norm": 0.26486146450042725, | |
| "learning_rate": 5.3381219208324755e-06, | |
| "loss": 0.0141, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 18.642857142857142, | |
| "grad_norm": 0.2617660462856293, | |
| "learning_rate": 5.320351099903565e-06, | |
| "loss": 0.0157, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 18.678571428571427, | |
| "grad_norm": 0.21751557290554047, | |
| "learning_rate": 5.302576214650527e-06, | |
| "loss": 0.0136, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 18.714285714285715, | |
| "grad_norm": 0.2128409892320633, | |
| "learning_rate": 5.284797490584979e-06, | |
| "loss": 0.0121, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "grad_norm": 0.21959584951400757, | |
| "learning_rate": 5.267015153267246e-06, | |
| "loss": 0.0125, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 18.785714285714285, | |
| "grad_norm": 0.22850820422172546, | |
| "learning_rate": 5.249229428303486e-06, | |
| "loss": 0.0124, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 18.821428571428573, | |
| "grad_norm": 0.8751614689826965, | |
| "learning_rate": 5.231440541342846e-06, | |
| "loss": 0.0197, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 18.857142857142858, | |
| "grad_norm": 0.18895405530929565, | |
| "learning_rate": 5.213648718074584e-06, | |
| "loss": 0.0122, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 18.892857142857142, | |
| "grad_norm": 0.22877931594848633, | |
| "learning_rate": 5.1958541842252145e-06, | |
| "loss": 0.0136, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 18.928571428571427, | |
| "grad_norm": 0.31640875339508057, | |
| "learning_rate": 5.178057165555636e-06, | |
| "loss": 0.0143, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 18.964285714285715, | |
| "grad_norm": 0.2516610026359558, | |
| "learning_rate": 5.160257887858278e-06, | |
| "loss": 0.0144, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.17745497822761536, | |
| "learning_rate": 5.142456576954225e-06, | |
| "loss": 0.0104, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 19.035714285714285, | |
| "grad_norm": 0.1342853158712387, | |
| "learning_rate": 5.1246534586903655e-06, | |
| "loss": 0.0082, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 19.071428571428573, | |
| "grad_norm": 0.18636493384838104, | |
| "learning_rate": 5.106848758936508e-06, | |
| "loss": 0.0122, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 19.107142857142858, | |
| "grad_norm": 0.14282165467739105, | |
| "learning_rate": 5.089042703582533e-06, | |
| "loss": 0.0101, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 19.142857142857142, | |
| "grad_norm": 0.16291485726833344, | |
| "learning_rate": 5.071235518535516e-06, | |
| "loss": 0.0105, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 19.178571428571427, | |
| "grad_norm": 0.17317119240760803, | |
| "learning_rate": 5.053427429716867e-06, | |
| "loss": 0.0106, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 19.214285714285715, | |
| "grad_norm": 0.14252512156963348, | |
| "learning_rate": 5.0356186630594585e-06, | |
| "loss": 0.0076, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 19.25, | |
| "grad_norm": 0.17922668159008026, | |
| "learning_rate": 5.017809444504768e-06, | |
| "loss": 0.0117, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 19.285714285714285, | |
| "grad_norm": 0.2028946876525879, | |
| "learning_rate": 5e-06, | |
| "loss": 0.012, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 19.321428571428573, | |
| "grad_norm": 0.13679425418376923, | |
| "learning_rate": 4.982190555495236e-06, | |
| "loss": 0.0073, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 19.357142857142858, | |
| "grad_norm": 0.6376967430114746, | |
| "learning_rate": 4.964381336940542e-06, | |
| "loss": 0.0157, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 19.392857142857142, | |
| "grad_norm": 0.19504405558109283, | |
| "learning_rate": 4.946572570283135e-06, | |
| "loss": 0.0098, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 19.428571428571427, | |
| "grad_norm": 0.21073895692825317, | |
| "learning_rate": 4.928764481464485e-06, | |
| "loss": 0.0138, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 19.464285714285715, | |
| "grad_norm": 0.3294600546360016, | |
| "learning_rate": 4.910957296417467e-06, | |
| "loss": 0.0131, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 19.5, | |
| "grad_norm": 0.21520303189754486, | |
| "learning_rate": 4.893151241063493e-06, | |
| "loss": 0.0133, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 19.535714285714285, | |
| "grad_norm": 0.2246185541152954, | |
| "learning_rate": 4.875346541309637e-06, | |
| "loss": 0.0139, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 19.571428571428573, | |
| "grad_norm": 0.15642490983009338, | |
| "learning_rate": 4.857543423045775e-06, | |
| "loss": 0.0105, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 19.607142857142858, | |
| "grad_norm": 0.1819118857383728, | |
| "learning_rate": 4.839742112141725e-06, | |
| "loss": 0.0104, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 19.642857142857142, | |
| "grad_norm": 0.2114618718624115, | |
| "learning_rate": 4.821942834444367e-06, | |
| "loss": 0.0127, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 19.678571428571427, | |
| "grad_norm": 0.2271358221769333, | |
| "learning_rate": 4.804145815774787e-06, | |
| "loss": 0.0106, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 19.714285714285715, | |
| "grad_norm": 0.21014434099197388, | |
| "learning_rate": 4.786351281925417e-06, | |
| "loss": 0.0114, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 19.75, | |
| "grad_norm": 0.21880529820919037, | |
| "learning_rate": 4.768559458657156e-06, | |
| "loss": 0.0127, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 19.785714285714285, | |
| "grad_norm": 0.28759488463401794, | |
| "learning_rate": 4.750770571696514e-06, | |
| "loss": 0.0176, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 19.821428571428573, | |
| "grad_norm": 0.2287236750125885, | |
| "learning_rate": 4.732984846732755e-06, | |
| "loss": 0.0139, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 19.857142857142858, | |
| "grad_norm": 0.16419465839862823, | |
| "learning_rate": 4.7152025094150214e-06, | |
| "loss": 0.0098, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 19.892857142857142, | |
| "grad_norm": 0.18456807732582092, | |
| "learning_rate": 4.697423785349475e-06, | |
| "loss": 0.0112, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 19.928571428571427, | |
| "grad_norm": 0.24935705959796906, | |
| "learning_rate": 4.679648900096436e-06, | |
| "loss": 0.018, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 19.964285714285715, | |
| "grad_norm": 0.208778977394104, | |
| "learning_rate": 4.661878079167527e-06, | |
| "loss": 0.0134, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.16569219529628754, | |
| "learning_rate": 4.644111548022798e-06, | |
| "loss": 0.0077, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 20.035714285714285, | |
| "grad_norm": 0.16942204535007477, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.0101, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 20.071428571428573, | |
| "grad_norm": 0.143929585814476, | |
| "learning_rate": 4.608592256651117e-06, | |
| "loss": 0.0083, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 20.107142857142858, | |
| "grad_norm": 0.18715058267116547, | |
| "learning_rate": 4.5908399470607106e-06, | |
| "loss": 0.0109, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 20.142857142857142, | |
| "grad_norm": 0.14599822461605072, | |
| "learning_rate": 4.573092828521857e-06, | |
| "loss": 0.0101, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 20.178571428571427, | |
| "grad_norm": 0.1562821865081787, | |
| "learning_rate": 4.555351126193901e-06, | |
| "loss": 0.0097, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 20.214285714285715, | |
| "grad_norm": 0.34438711404800415, | |
| "learning_rate": 4.537615065167461e-06, | |
| "loss": 0.011, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "grad_norm": 0.18419024348258972, | |
| "learning_rate": 4.5198848704615915e-06, | |
| "loss": 0.012, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 20.285714285714285, | |
| "grad_norm": 0.1408761739730835, | |
| "learning_rate": 4.502160767020918e-06, | |
| "loss": 0.0086, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 20.321428571428573, | |
| "grad_norm": 0.18469876050949097, | |
| "learning_rate": 4.484442979712783e-06, | |
| "loss": 0.0114, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 20.357142857142858, | |
| "grad_norm": 0.1582385003566742, | |
| "learning_rate": 4.466731733324399e-06, | |
| "loss": 0.0084, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 20.392857142857142, | |
| "grad_norm": 0.16665539145469666, | |
| "learning_rate": 4.449027252559994e-06, | |
| "loss": 0.0095, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 20.428571428571427, | |
| "grad_norm": 0.1675061285495758, | |
| "learning_rate": 4.431329762037958e-06, | |
| "loss": 0.009, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 20.464285714285715, | |
| "grad_norm": 0.22415050864219666, | |
| "learning_rate": 4.413639486287992e-06, | |
| "loss": 0.0147, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 20.5, | |
| "grad_norm": 0.24033626914024353, | |
| "learning_rate": 4.395956649748269e-06, | |
| "loss": 0.0144, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 20.535714285714285, | |
| "grad_norm": 0.18106594681739807, | |
| "learning_rate": 4.3782814767625755e-06, | |
| "loss": 0.0097, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 20.571428571428573, | |
| "grad_norm": 0.18736809492111206, | |
| "learning_rate": 4.3606141915774695e-06, | |
| "loss": 0.0107, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 20.607142857142858, | |
| "grad_norm": 0.1885957419872284, | |
| "learning_rate": 4.342955018339442e-06, | |
| "loss": 0.0125, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 20.642857142857142, | |
| "grad_norm": 0.17403589189052582, | |
| "learning_rate": 4.3253041810920595e-06, | |
| "loss": 0.0097, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 20.678571428571427, | |
| "grad_norm": 0.2727905511856079, | |
| "learning_rate": 4.307661903773129e-06, | |
| "loss": 0.0104, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 20.714285714285715, | |
| "grad_norm": 0.17853574454784393, | |
| "learning_rate": 4.290028410211866e-06, | |
| "loss": 0.0094, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "grad_norm": 0.17175881564617157, | |
| "learning_rate": 4.272403924126035e-06, | |
| "loss": 0.0108, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 20.785714285714285, | |
| "grad_norm": 0.19546657800674438, | |
| "learning_rate": 4.254788669119127e-06, | |
| "loss": 0.0138, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 20.821428571428573, | |
| "grad_norm": 0.2268596589565277, | |
| "learning_rate": 4.237182868677519e-06, | |
| "loss": 0.0121, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 20.857142857142858, | |
| "grad_norm": 0.22344724833965302, | |
| "learning_rate": 4.219586746167632e-06, | |
| "loss": 0.0122, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 20.892857142857142, | |
| "grad_norm": 0.18819448351860046, | |
| "learning_rate": 4.2020005248331056e-06, | |
| "loss": 0.0115, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 20.928571428571427, | |
| "grad_norm": 0.27732348442077637, | |
| "learning_rate": 4.18442442779196e-06, | |
| "loss": 0.0163, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 20.964285714285715, | |
| "grad_norm": 0.18527792394161224, | |
| "learning_rate": 4.166858678033771e-06, | |
| "loss": 0.0113, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.18095916509628296, | |
| "learning_rate": 4.149303498416838e-06, | |
| "loss": 0.0107, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 21.035714285714285, | |
| "grad_norm": 0.1701073944568634, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0105, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 21.071428571428573, | |
| "grad_norm": 0.14660249650478363, | |
| "learning_rate": 4.114225740366569e-06, | |
| "loss": 0.0073, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 21.107142857142858, | |
| "grad_norm": 0.1561962366104126, | |
| "learning_rate": 4.096703606968007e-06, | |
| "loss": 0.0095, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 21.142857142857142, | |
| "grad_norm": 0.15540745854377747, | |
| "learning_rate": 4.079192933774592e-06, | |
| "loss": 0.0096, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 21.178571428571427, | |
| "grad_norm": 0.14585469663143158, | |
| "learning_rate": 4.061693942945863e-06, | |
| "loss": 0.0088, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 21.214285714285715, | |
| "grad_norm": 0.15605901181697845, | |
| "learning_rate": 4.04420685649314e-06, | |
| "loss": 0.0092, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 21.25, | |
| "grad_norm": 0.14791828393936157, | |
| "learning_rate": 4.026731896276708e-06, | |
| "loss": 0.0088, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 21.285714285714285, | |
| "grad_norm": 0.16311927139759064, | |
| "learning_rate": 4.009269284003014e-06, | |
| "loss": 0.0105, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 21.321428571428573, | |
| "grad_norm": 0.1642966866493225, | |
| "learning_rate": 3.991819241221836e-06, | |
| "loss": 0.0096, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 21.357142857142858, | |
| "grad_norm": 0.15627528727054596, | |
| "learning_rate": 3.974381989323484e-06, | |
| "loss": 0.0087, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 21.392857142857142, | |
| "grad_norm": 0.1767517775297165, | |
| "learning_rate": 3.956957749535997e-06, | |
| "loss": 0.0111, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "grad_norm": 0.1780816912651062, | |
| "learning_rate": 3.939546742922318e-06, | |
| "loss": 0.0108, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "eval_loss": 2.483830213546753, | |
| "eval_runtime": 0.4708, | |
| "eval_samples_per_second": 297.392, | |
| "eval_steps_per_second": 2.124, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 21.464285714285715, | |
| "grad_norm": 0.15922360122203827, | |
| "learning_rate": 3.9221491903775014e-06, | |
| "loss": 0.0088, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 21.5, | |
| "grad_norm": 0.14839443564414978, | |
| "learning_rate": 3.904765312625916e-06, | |
| "loss": 0.0093, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 21.535714285714285, | |
| "grad_norm": 0.20598483085632324, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.0122, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 21.571428571428573, | |
| "grad_norm": 0.2388192117214203, | |
| "learning_rate": 3.8700394635296166e-06, | |
| "loss": 0.0101, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 21.607142857142858, | |
| "grad_norm": 0.21544699370861053, | |
| "learning_rate": 3.852697932754974e-06, | |
| "loss": 0.0126, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 21.642857142857142, | |
| "grad_norm": 0.2086128294467926, | |
| "learning_rate": 3.835370957908108e-06, | |
| "loss": 0.0108, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 21.678571428571427, | |
| "grad_norm": 0.1632758378982544, | |
| "learning_rate": 3.818058758817956e-06, | |
| "loss": 0.0091, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 21.714285714285715, | |
| "grad_norm": 0.21566864848136902, | |
| "learning_rate": 3.800761555125997e-06, | |
| "loss": 0.0118, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 21.75, | |
| "grad_norm": 0.1939634531736374, | |
| "learning_rate": 3.783479566283457e-06, | |
| "loss": 0.0107, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 21.785714285714285, | |
| "grad_norm": 0.24293866753578186, | |
| "learning_rate": 3.7662130115485317e-06, | |
| "loss": 0.0147, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 21.821428571428573, | |
| "grad_norm": 0.22918657958507538, | |
| "learning_rate": 3.748962109983605e-06, | |
| "loss": 0.0125, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 21.857142857142858, | |
| "grad_norm": 0.1861945390701294, | |
| "learning_rate": 3.731727080452464e-06, | |
| "loss": 0.011, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 21.892857142857142, | |
| "grad_norm": 0.2174372524023056, | |
| "learning_rate": 3.714508141617527e-06, | |
| "loss": 0.0121, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 21.928571428571427, | |
| "grad_norm": 0.23144198954105377, | |
| "learning_rate": 3.69730551193707e-06, | |
| "loss": 0.0128, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 21.964285714285715, | |
| "grad_norm": 0.20174066722393036, | |
| "learning_rate": 3.6801194096624515e-06, | |
| "loss": 0.0121, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.16544412076473236, | |
| "learning_rate": 3.6629500528353464e-06, | |
| "loss": 0.0098, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 22.035714285714285, | |
| "grad_norm": 0.13743500411510468, | |
| "learning_rate": 3.6457976592849753e-06, | |
| "loss": 0.0082, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 22.071428571428573, | |
| "grad_norm": 0.14166727662086487, | |
| "learning_rate": 3.6286624466253496e-06, | |
| "loss": 0.0079, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 22.107142857142858, | |
| "grad_norm": 0.13137659430503845, | |
| "learning_rate": 3.6115446322525007e-06, | |
| "loss": 0.0074, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 22.142857142857142, | |
| "grad_norm": 0.1719367355108261, | |
| "learning_rate": 3.594444433341725e-06, | |
| "loss": 0.0113, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 22.178571428571427, | |
| "grad_norm": 0.14747175574302673, | |
| "learning_rate": 3.5773620668448384e-06, | |
| "loss": 0.0079, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 22.214285714285715, | |
| "grad_norm": 0.141871839761734, | |
| "learning_rate": 3.560297749487407e-06, | |
| "loss": 0.0077, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 22.25, | |
| "grad_norm": 0.15803030133247375, | |
| "learning_rate": 3.543251697766006e-06, | |
| "loss": 0.0097, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 22.285714285714285, | |
| "grad_norm": 0.12996134161949158, | |
| "learning_rate": 3.526224127945479e-06, | |
| "loss": 0.0074, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 22.321428571428573, | |
| "grad_norm": 0.14572900533676147, | |
| "learning_rate": 3.5092152560561833e-06, | |
| "loss": 0.0087, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 22.357142857142858, | |
| "grad_norm": 0.13245287537574768, | |
| "learning_rate": 3.4922252978912523e-06, | |
| "loss": 0.0069, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 22.392857142857142, | |
| "grad_norm": 0.1103530079126358, | |
| "learning_rate": 3.475254469003865e-06, | |
| "loss": 0.0076, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 22.428571428571427, | |
| "grad_norm": 0.15494345128536224, | |
| "learning_rate": 3.4583029847044996e-06, | |
| "loss": 0.0099, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 22.464285714285715, | |
| "grad_norm": 0.14390411972999573, | |
| "learning_rate": 3.4413710600582096e-06, | |
| "loss": 0.0092, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "grad_norm": 0.18431764841079712, | |
| "learning_rate": 3.424458909881897e-06, | |
| "loss": 0.0113, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 22.535714285714285, | |
| "grad_norm": 0.17182418704032898, | |
| "learning_rate": 3.4075667487415785e-06, | |
| "loss": 0.0102, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 22.571428571428573, | |
| "grad_norm": 0.2069699615240097, | |
| "learning_rate": 3.3906947909496696e-06, | |
| "loss": 0.0126, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 22.607142857142858, | |
| "grad_norm": 0.17551185190677643, | |
| "learning_rate": 3.3738432505622653e-06, | |
| "loss": 0.0099, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 22.642857142857142, | |
| "grad_norm": 0.17832733690738678, | |
| "learning_rate": 3.357012341376421e-06, | |
| "loss": 0.0105, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 22.678571428571427, | |
| "grad_norm": 0.25220856070518494, | |
| "learning_rate": 3.3402022769274422e-06, | |
| "loss": 0.014, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 22.714285714285715, | |
| "grad_norm": 0.1663549244403839, | |
| "learning_rate": 3.3234132704861786e-06, | |
| "loss": 0.0094, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 22.75, | |
| "grad_norm": 0.23084159195423126, | |
| "learning_rate": 3.306645535056312e-06, | |
| "loss": 0.0127, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 22.785714285714285, | |
| "grad_norm": 0.18974913656711578, | |
| "learning_rate": 3.289899283371657e-06, | |
| "loss": 0.0125, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 22.821428571428573, | |
| "grad_norm": 0.17198923230171204, | |
| "learning_rate": 3.273174727893463e-06, | |
| "loss": 0.0098, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 0.2270413190126419, | |
| "learning_rate": 3.2564720808077167e-06, | |
| "loss": 0.0149, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 22.892857142857142, | |
| "grad_norm": 0.20339983701705933, | |
| "learning_rate": 3.2397915540224493e-06, | |
| "loss": 0.0115, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 22.928571428571427, | |
| "grad_norm": 0.2048344910144806, | |
| "learning_rate": 3.2231333591650567e-06, | |
| "loss": 0.0111, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 22.964285714285715, | |
| "grad_norm": 0.19488964974880219, | |
| "learning_rate": 3.2064977075795988e-06, | |
| "loss": 0.0117, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.14864051342010498, | |
| "learning_rate": 3.189884810324133e-06, | |
| "loss": 0.0086, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 23.035714285714285, | |
| "grad_norm": 0.18157891929149628, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.0107, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 23.071428571428573, | |
| "grad_norm": 0.09408193826675415, | |
| "learning_rate": 3.1567281215892868e-06, | |
| "loss": 0.0047, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 23.107142857142858, | |
| "grad_norm": 0.15541407465934753, | |
| "learning_rate": 3.140184750771895e-06, | |
| "loss": 0.009, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 23.142857142857142, | |
| "grad_norm": 0.15836073458194733, | |
| "learning_rate": 3.12366497560313e-06, | |
| "loss": 0.0091, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 23.178571428571427, | |
| "grad_norm": 0.12873704731464386, | |
| "learning_rate": 3.1071690056709125e-06, | |
| "loss": 0.0079, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 23.214285714285715, | |
| "grad_norm": 0.15572570264339447, | |
| "learning_rate": 3.090697050261143e-06, | |
| "loss": 0.0082, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 23.25, | |
| "grad_norm": 0.14964070916175842, | |
| "learning_rate": 3.074249318355046e-06, | |
| "loss": 0.0078, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 23.285714285714285, | |
| "grad_norm": 0.13191375136375427, | |
| "learning_rate": 3.057826018626527e-06, | |
| "loss": 0.0077, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 23.321428571428573, | |
| "grad_norm": 0.13573727011680603, | |
| "learning_rate": 3.0414273594395106e-06, | |
| "loss": 0.0073, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 23.357142857142858, | |
| "grad_norm": 0.17368032038211823, | |
| "learning_rate": 3.0250535488453077e-06, | |
| "loss": 0.0113, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 23.392857142857142, | |
| "grad_norm": 0.14922013878822327, | |
| "learning_rate": 3.008704794579973e-06, | |
| "loss": 0.0079, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 23.428571428571427, | |
| "grad_norm": 0.14999383687973022, | |
| "learning_rate": 2.9923813040616685e-06, | |
| "loss": 0.0088, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 23.464285714285715, | |
| "grad_norm": 0.19925668835639954, | |
| "learning_rate": 2.976083284388031e-06, | |
| "loss": 0.0111, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 23.5, | |
| "grad_norm": 0.19494231045246124, | |
| "learning_rate": 2.959810942333552e-06, | |
| "loss": 0.0117, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 23.535714285714285, | |
| "grad_norm": 0.14841331541538239, | |
| "learning_rate": 2.9435644843469434e-06, | |
| "loss": 0.008, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 23.571428571428573, | |
| "grad_norm": 0.15695911645889282, | |
| "learning_rate": 2.9273441165485227e-06, | |
| "loss": 0.0078, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 23.607142857142858, | |
| "grad_norm": 0.2552220821380615, | |
| "learning_rate": 2.9111500447276053e-06, | |
| "loss": 0.0141, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 23.642857142857142, | |
| "grad_norm": 0.17610006034374237, | |
| "learning_rate": 2.8949824743398804e-06, | |
| "loss": 0.0095, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 23.678571428571427, | |
| "grad_norm": 0.17815729975700378, | |
| "learning_rate": 2.8788416105048124e-06, | |
| "loss": 0.0104, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 23.714285714285715, | |
| "grad_norm": 0.21268025040626526, | |
| "learning_rate": 2.862727658003042e-06, | |
| "loss": 0.0122, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 23.75, | |
| "grad_norm": 0.1655074954032898, | |
| "learning_rate": 2.8466408212737777e-06, | |
| "loss": 0.0091, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 23.785714285714285, | |
| "grad_norm": 0.18781046569347382, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.0115, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 23.821428571428573, | |
| "grad_norm": 0.19021618366241455, | |
| "learning_rate": 2.8145493111669186e-06, | |
| "loss": 0.01, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 23.857142857142858, | |
| "grad_norm": 0.21351297199726105, | |
| "learning_rate": 2.79854504493729e-06, | |
| "loss": 0.0133, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 23.892857142857142, | |
| "grad_norm": 0.17441827058792114, | |
| "learning_rate": 2.782568708770933e-06, | |
| "loss": 0.0094, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 23.928571428571427, | |
| "grad_norm": 0.18894660472869873, | |
| "learning_rate": 2.7666205053611097e-06, | |
| "loss": 0.0126, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 23.964285714285715, | |
| "grad_norm": 0.19022874534130096, | |
| "learning_rate": 2.7507006370441557e-06, | |
| "loss": 0.012, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.17531032860279083, | |
| "learning_rate": 2.734809305796915e-06, | |
| "loss": 0.0101, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 24.035714285714285, | |
| "grad_norm": 0.17306911945343018, | |
| "learning_rate": 2.718946713234185e-06, | |
| "loss": 0.0113, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 24.071428571428573, | |
| "grad_norm": 0.16345195472240448, | |
| "learning_rate": 2.7031130606061486e-06, | |
| "loss": 0.0094, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 24.107142857142858, | |
| "grad_norm": 0.1428937315940857, | |
| "learning_rate": 2.687308548795825e-06, | |
| "loss": 0.0078, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 24.142857142857142, | |
| "grad_norm": 0.1553509682416916, | |
| "learning_rate": 2.67153337831652e-06, | |
| "loss": 0.0085, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 24.178571428571427, | |
| "grad_norm": 0.16722197830677032, | |
| "learning_rate": 2.6557877493092885e-06, | |
| "loss": 0.0098, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 24.214285714285715, | |
| "grad_norm": 0.14916226267814636, | |
| "learning_rate": 2.6400718615403852e-06, | |
| "loss": 0.0075, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 24.25, | |
| "grad_norm": 0.1756383627653122, | |
| "learning_rate": 2.624385914398737e-06, | |
| "loss": 0.0122, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 24.285714285714285, | |
| "grad_norm": 0.12129461765289307, | |
| "learning_rate": 2.608730106893411e-06, | |
| "loss": 0.0067, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 24.321428571428573, | |
| "grad_norm": 0.1598646193742752, | |
| "learning_rate": 2.5931046376510875e-06, | |
| "loss": 0.0077, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 24.357142857142858, | |
| "grad_norm": 0.13373102247714996, | |
| "learning_rate": 2.5775097049135445e-06, | |
| "loss": 0.0075, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 24.392857142857142, | |
| "grad_norm": 0.1555156260728836, | |
| "learning_rate": 2.561945506535144e-06, | |
| "loss": 0.0102, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 24.428571428571427, | |
| "grad_norm": 0.15757839381694794, | |
| "learning_rate": 2.5464122399803126e-06, | |
| "loss": 0.0099, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 24.464285714285715, | |
| "grad_norm": 0.19057762622833252, | |
| "learning_rate": 2.5309101023210426e-06, | |
| "loss": 0.0125, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 24.5, | |
| "grad_norm": 0.11190243065357208, | |
| "learning_rate": 2.5154392902343966e-06, | |
| "loss": 0.0061, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 24.535714285714285, | |
| "grad_norm": 0.1816374808549881, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0104, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 24.571428571428573, | |
| "grad_norm": 0.148861825466156, | |
| "learning_rate": 2.4845924274975625e-06, | |
| "loss": 0.008, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 24.607142857142858, | |
| "grad_norm": 0.14404255151748657, | |
| "learning_rate": 2.4692167682043855e-06, | |
| "loss": 0.0072, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 24.642857142857142, | |
| "grad_norm": 0.20183296501636505, | |
| "learning_rate": 2.4538732171928847e-06, | |
| "loss": 0.0129, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 24.678571428571427, | |
| "grad_norm": 0.16051512956619263, | |
| "learning_rate": 2.4385619691281144e-06, | |
| "loss": 0.0093, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 24.714285714285715, | |
| "grad_norm": 0.18522028625011444, | |
| "learning_rate": 2.4232832182653014e-06, | |
| "loss": 0.01, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 24.75, | |
| "grad_norm": 0.1713511347770691, | |
| "learning_rate": 2.408037158447375e-06, | |
| "loss": 0.0092, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 24.785714285714285, | |
| "grad_norm": 0.18777602910995483, | |
| "learning_rate": 2.39282398310251e-06, | |
| "loss": 0.0099, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 24.821428571428573, | |
| "grad_norm": 0.15512120723724365, | |
| "learning_rate": 2.3776438852416743e-06, | |
| "loss": 0.009, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 24.857142857142858, | |
| "grad_norm": 0.17730368673801422, | |
| "learning_rate": 2.3624970574561773e-06, | |
| "loss": 0.0098, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 24.892857142857142, | |
| "grad_norm": 0.2074640393257141, | |
| "learning_rate": 2.3473836919152267e-06, | |
| "loss": 0.013, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 24.928571428571427, | |
| "grad_norm": 0.17625132203102112, | |
| "learning_rate": 2.332303980363497e-06, | |
| "loss": 0.0098, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 24.964285714285715, | |
| "grad_norm": 0.19898360967636108, | |
| "learning_rate": 2.317258114118686e-06, | |
| "loss": 0.0109, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.19583548605442047, | |
| "learning_rate": 2.3022462840690933e-06, | |
| "loss": 0.0119, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 2.5329432487487793, | |
| "eval_runtime": 0.4711, | |
| "eval_samples_per_second": 297.205, | |
| "eval_steps_per_second": 2.123, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 25.035714285714285, | |
| "grad_norm": 0.1791287660598755, | |
| "learning_rate": 2.2872686806712037e-06, | |
| "loss": 0.0101, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 25.071428571428573, | |
| "grad_norm": 0.15796615183353424, | |
| "learning_rate": 2.272325493947257e-06, | |
| "loss": 0.0078, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 25.107142857142858, | |
| "grad_norm": 0.1644110381603241, | |
| "learning_rate": 2.257416913482853e-06, | |
| "loss": 0.0084, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 25.142857142857142, | |
| "grad_norm": 0.1331474632024765, | |
| "learning_rate": 2.2425431284245302e-06, | |
| "loss": 0.0071, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 25.178571428571427, | |
| "grad_norm": 0.152842178940773, | |
| "learning_rate": 2.2277043274773856e-06, | |
| "loss": 0.0085, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 25.214285714285715, | |
| "grad_norm": 0.15128417313098907, | |
| "learning_rate": 2.2129006989026612e-06, | |
| "loss": 0.0079, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 25.25, | |
| "grad_norm": 0.18646658957004547, | |
| "learning_rate": 2.1981324305153644e-06, | |
| "loss": 0.0105, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 25.285714285714285, | |
| "grad_norm": 0.1679907739162445, | |
| "learning_rate": 2.1833997096818897e-06, | |
| "loss": 0.0099, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 25.321428571428573, | |
| "grad_norm": 0.13102103769779205, | |
| "learning_rate": 2.168702723317632e-06, | |
| "loss": 0.0067, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 25.357142857142858, | |
| "grad_norm": 0.1524365246295929, | |
| "learning_rate": 2.1540416578846207e-06, | |
| "loss": 0.0083, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 25.392857142857142, | |
| "grad_norm": 0.20382919907569885, | |
| "learning_rate": 2.139416699389153e-06, | |
| "loss": 0.0091, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 25.428571428571427, | |
| "grad_norm": 0.14647948741912842, | |
| "learning_rate": 2.1248280333794347e-06, | |
| "loss": 0.0075, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 25.464285714285715, | |
| "grad_norm": 0.15126314759254456, | |
| "learning_rate": 2.1102758449432233e-06, | |
| "loss": 0.0075, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 25.5, | |
| "grad_norm": 0.14693264663219452, | |
| "learning_rate": 2.095760318705487e-06, | |
| "loss": 0.0081, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 25.535714285714285, | |
| "grad_norm": 0.18826764822006226, | |
| "learning_rate": 2.081281638826052e-06, | |
| "loss": 0.01, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 25.571428571428573, | |
| "grad_norm": 0.1897839456796646, | |
| "learning_rate": 2.0668399889972717e-06, | |
| "loss": 0.0136, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 25.607142857142858, | |
| "grad_norm": 0.17151887714862823, | |
| "learning_rate": 2.0524355524417017e-06, | |
| "loss": 0.009, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 25.642857142857142, | |
| "grad_norm": 0.1408870816230774, | |
| "learning_rate": 2.038068511909762e-06, | |
| "loss": 0.0076, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 25.678571428571427, | |
| "grad_norm": 0.2066071778535843, | |
| "learning_rate": 2.0237390496774284e-06, | |
| "loss": 0.0108, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 0.17692652344703674, | |
| "learning_rate": 2.00944734754392e-06, | |
| "loss": 0.009, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 25.75, | |
| "grad_norm": 0.2316911220550537, | |
| "learning_rate": 1.995193586829387e-06, | |
| "loss": 0.0139, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 25.785714285714285, | |
| "grad_norm": 0.1862197071313858, | |
| "learning_rate": 1.980977948372612e-06, | |
| "loss": 0.01, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 25.821428571428573, | |
| "grad_norm": 0.21679328382015228, | |
| "learning_rate": 1.966800612528723e-06, | |
| "loss": 0.0125, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 25.857142857142858, | |
| "grad_norm": 0.19006772339344025, | |
| "learning_rate": 1.952661759166893e-06, | |
| "loss": 0.0115, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 25.892857142857142, | |
| "grad_norm": 0.12936274707317352, | |
| "learning_rate": 1.9385615676680663e-06, | |
| "loss": 0.0064, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 25.928571428571427, | |
| "grad_norm": 0.2046034336090088, | |
| "learning_rate": 1.9245002169226814e-06, | |
| "loss": 0.0117, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 25.964285714285715, | |
| "grad_norm": 0.16442671418190002, | |
| "learning_rate": 1.910477885328399e-06, | |
| "loss": 0.0098, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.15325023233890533, | |
| "learning_rate": 1.8964947507878401e-06, | |
| "loss": 0.0086, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 26.035714285714285, | |
| "grad_norm": 0.17084500193595886, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.0094, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 26.071428571428573, | |
| "grad_norm": 0.1273665577173233, | |
| "learning_rate": 1.8686467819896542e-06, | |
| "loss": 0.007, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 26.107142857142858, | |
| "grad_norm": 0.1403261423110962, | |
| "learning_rate": 1.8547823010417876e-06, | |
| "loss": 0.0071, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 26.142857142857142, | |
| "grad_norm": 0.15290212631225586, | |
| "learning_rate": 1.8409577237626935e-06, | |
| "loss": 0.0092, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 26.178571428571427, | |
| "grad_norm": 0.16614960134029388, | |
| "learning_rate": 1.8271732255460644e-06, | |
| "loss": 0.0095, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 26.214285714285715, | |
| "grad_norm": 0.12265990674495697, | |
| "learning_rate": 1.8134289812771077e-06, | |
| "loss": 0.0068, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 26.25, | |
| "grad_norm": 0.12230860441923141, | |
| "learning_rate": 1.7997251653303249e-06, | |
| "loss": 0.0058, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 26.285714285714285, | |
| "grad_norm": 0.15943826735019684, | |
| "learning_rate": 1.7860619515673034e-06, | |
| "loss": 0.0089, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 26.321428571428573, | |
| "grad_norm": 0.15436235070228577, | |
| "learning_rate": 1.7724395133345025e-06, | |
| "loss": 0.0078, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 26.357142857142858, | |
| "grad_norm": 0.19412128627300262, | |
| "learning_rate": 1.7588580234610592e-06, | |
| "loss": 0.0112, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 26.392857142857142, | |
| "grad_norm": 0.15104462206363678, | |
| "learning_rate": 1.7453176542565958e-06, | |
| "loss": 0.0078, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 26.428571428571427, | |
| "grad_norm": 0.1390601247549057, | |
| "learning_rate": 1.7318185775090336e-06, | |
| "loss": 0.0084, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 26.464285714285715, | |
| "grad_norm": 0.1711408495903015, | |
| "learning_rate": 1.7183609644824096e-06, | |
| "loss": 0.01, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 26.5, | |
| "grad_norm": 0.16743941605091095, | |
| "learning_rate": 1.7049449859147121e-06, | |
| "loss": 0.008, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 26.535714285714285, | |
| "grad_norm": 0.14585740864276886, | |
| "learning_rate": 1.6915708120157042e-06, | |
| "loss": 0.008, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 26.571428571428573, | |
| "grad_norm": 0.17908959090709686, | |
| "learning_rate": 1.67823861246477e-06, | |
| "loss": 0.0108, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 26.607142857142858, | |
| "grad_norm": 0.15890510380268097, | |
| "learning_rate": 1.6649485564087646e-06, | |
| "loss": 0.0088, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 26.642857142857142, | |
| "grad_norm": 0.16608744859695435, | |
| "learning_rate": 1.6517008124598622e-06, | |
| "loss": 0.0102, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 26.678571428571427, | |
| "grad_norm": 0.20837059617042542, | |
| "learning_rate": 1.6384955486934157e-06, | |
| "loss": 0.0123, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 26.714285714285715, | |
| "grad_norm": 0.1581769585609436, | |
| "learning_rate": 1.6253329326458367e-06, | |
| "loss": 0.0086, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 26.75, | |
| "grad_norm": 0.1689825803041458, | |
| "learning_rate": 1.612213131312454e-06, | |
| "loss": 0.0089, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 26.785714285714285, | |
| "grad_norm": 0.18127354979515076, | |
| "learning_rate": 1.5991363111454023e-06, | |
| "loss": 0.009, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 26.821428571428573, | |
| "grad_norm": 0.18165580928325653, | |
| "learning_rate": 1.5861026380515165e-06, | |
| "loss": 0.0098, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 26.857142857142858, | |
| "grad_norm": 0.18343548476696014, | |
| "learning_rate": 1.5731122773902147e-06, | |
| "loss": 0.0102, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 26.892857142857142, | |
| "grad_norm": 0.17770545184612274, | |
| "learning_rate": 1.5601653939714073e-06, | |
| "loss": 0.0098, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 26.928571428571427, | |
| "grad_norm": 0.19391901791095734, | |
| "learning_rate": 1.547262152053406e-06, | |
| "loss": 0.0112, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 26.964285714285715, | |
| "grad_norm": 0.19513018429279327, | |
| "learning_rate": 1.5344027153408375e-06, | |
| "loss": 0.0097, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.15814387798309326, | |
| "learning_rate": 1.5215872469825682e-06, | |
| "loss": 0.0085, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 27.035714285714285, | |
| "grad_norm": 0.17074698209762573, | |
| "learning_rate": 1.5088159095696365e-06, | |
| "loss": 0.0086, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 27.071428571428573, | |
| "grad_norm": 0.16025789082050323, | |
| "learning_rate": 1.4960888651331833e-06, | |
| "loss": 0.0088, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 27.107142857142858, | |
| "grad_norm": 0.1347316950559616, | |
| "learning_rate": 1.4834062751424018e-06, | |
| "loss": 0.008, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 27.142857142857142, | |
| "grad_norm": 0.1638944447040558, | |
| "learning_rate": 1.4707683005024898e-06, | |
| "loss": 0.0093, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 27.178571428571427, | |
| "grad_norm": 0.14700771868228912, | |
| "learning_rate": 1.4581751015526035e-06, | |
| "loss": 0.007, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 27.214285714285715, | |
| "grad_norm": 0.1447134166955948, | |
| "learning_rate": 1.4456268380638262e-06, | |
| "loss": 0.0079, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 27.25, | |
| "grad_norm": 0.1443662941455841, | |
| "learning_rate": 1.4331236692371386e-06, | |
| "loss": 0.0086, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 27.285714285714285, | |
| "grad_norm": 0.17928117513656616, | |
| "learning_rate": 1.4206657537014078e-06, | |
| "loss": 0.0087, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 27.321428571428573, | |
| "grad_norm": 0.13179831206798553, | |
| "learning_rate": 1.4082532495113627e-06, | |
| "loss": 0.0083, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 27.357142857142858, | |
| "grad_norm": 0.12459900975227356, | |
| "learning_rate": 1.3958863141455937e-06, | |
| "loss": 0.0068, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 27.392857142857142, | |
| "grad_norm": 0.17652547359466553, | |
| "learning_rate": 1.38356510450456e-06, | |
| "loss": 0.0098, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 27.428571428571427, | |
| "grad_norm": 0.20520183444023132, | |
| "learning_rate": 1.3712897769085903e-06, | |
| "loss": 0.0126, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 27.464285714285715, | |
| "grad_norm": 0.18517449498176575, | |
| "learning_rate": 1.3590604870959046e-06, | |
| "loss": 0.0093, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "grad_norm": 0.19475194811820984, | |
| "learning_rate": 1.3468773902206378e-06, | |
| "loss": 0.011, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 27.535714285714285, | |
| "grad_norm": 0.1644594371318817, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.0084, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 27.571428571428573, | |
| "grad_norm": 0.16756068170070648, | |
| "learning_rate": 1.322650392966665e-06, | |
| "loss": 0.0082, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 27.607142857142858, | |
| "grad_norm": 0.14502035081386566, | |
| "learning_rate": 1.3106067999581224e-06, | |
| "loss": 0.0084, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 27.642857142857142, | |
| "grad_norm": 0.14674372971057892, | |
| "learning_rate": 1.298610014623423e-06, | |
| "loss": 0.0076, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 27.678571428571427, | |
| "grad_norm": 0.17021062970161438, | |
| "learning_rate": 1.2866601891668945e-06, | |
| "loss": 0.0098, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 27.714285714285715, | |
| "grad_norm": 0.1700037717819214, | |
| "learning_rate": 1.2747574751970826e-06, | |
| "loss": 0.0077, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 27.75, | |
| "grad_norm": 0.19784575700759888, | |
| "learning_rate": 1.2629020237248241e-06, | |
| "loss": 0.0107, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 27.785714285714285, | |
| "grad_norm": 0.15924608707427979, | |
| "learning_rate": 1.2510939851613285e-06, | |
| "loss": 0.0096, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 27.821428571428573, | |
| "grad_norm": 0.1310548335313797, | |
| "learning_rate": 1.239333509316281e-06, | |
| "loss": 0.0067, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 27.857142857142858, | |
| "grad_norm": 0.17363598942756653, | |
| "learning_rate": 1.2276207453959283e-06, | |
| "loss": 0.0088, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 27.892857142857142, | |
| "grad_norm": 0.161713108420372, | |
| "learning_rate": 1.2159558420011907e-06, | |
| "loss": 0.0082, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 27.928571428571427, | |
| "grad_norm": 0.18847158551216125, | |
| "learning_rate": 1.2043389471257833e-06, | |
| "loss": 0.0105, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 27.964285714285715, | |
| "grad_norm": 0.16982363164424896, | |
| "learning_rate": 1.1927702081543279e-06, | |
| "loss": 0.01, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.15337583422660828, | |
| "learning_rate": 1.1812497718604887e-06, | |
| "loss": 0.0082, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 28.035714285714285, | |
| "grad_norm": 0.15224626660346985, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0076, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 28.071428571428573, | |
| "grad_norm": 0.1463271677494049, | |
| "learning_rate": 1.158354391334362e-06, | |
| "loss": 0.0082, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 28.107142857142858, | |
| "grad_norm": 0.1552111953496933, | |
| "learning_rate": 1.1469797375778902e-06, | |
| "loss": 0.0083, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 28.142857142857142, | |
| "grad_norm": 0.14375971257686615, | |
| "learning_rate": 1.1356539674469852e-06, | |
| "loss": 0.0064, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 28.178571428571427, | |
| "grad_norm": 0.14443305134773254, | |
| "learning_rate": 1.1243772246327416e-06, | |
| "loss": 0.009, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 28.214285714285715, | |
| "grad_norm": 0.18736590445041656, | |
| "learning_rate": 1.1131496522042424e-06, | |
| "loss": 0.0109, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 28.25, | |
| "grad_norm": 0.1467347890138626, | |
| "learning_rate": 1.1019713926067394e-06, | |
| "loss": 0.0083, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 28.285714285714285, | |
| "grad_norm": 0.1514614075422287, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.0091, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 28.321428571428573, | |
| "grad_norm": 0.16135913133621216, | |
| "learning_rate": 1.0797633785557582e-06, | |
| "loss": 0.0093, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 28.357142857142858, | |
| "grad_norm": 0.12219899147748947, | |
| "learning_rate": 1.068733905857413e-06, | |
| "loss": 0.0057, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 28.392857142857142, | |
| "grad_norm": 0.1365957111120224, | |
| "learning_rate": 1.0577543094967613e-06, | |
| "loss": 0.0063, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 28.428571428571427, | |
| "grad_norm": 0.14362846314907074, | |
| "learning_rate": 1.0468247287729593e-06, | |
| "loss": 0.0072, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 28.464285714285715, | |
| "grad_norm": 0.20032982528209686, | |
| "learning_rate": 1.0359453023506123e-06, | |
| "loss": 0.0116, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 28.5, | |
| "grad_norm": 0.14230377972126007, | |
| "learning_rate": 1.0251161682580125e-06, | |
| "loss": 0.0074, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 28.535714285714285, | |
| "grad_norm": 0.17326034605503082, | |
| "learning_rate": 1.0143374638853892e-06, | |
| "loss": 0.0099, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 0.1817915439605713, | |
| "learning_rate": 1.0036093259831624e-06, | |
| "loss": 0.0106, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "eval_loss": 2.567107677459717, | |
| "eval_runtime": 0.4721, | |
| "eval_samples_per_second": 296.568, | |
| "eval_steps_per_second": 2.118, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 28.607142857142858, | |
| "grad_norm": 0.16237416863441467, | |
| "learning_rate": 9.929318906602176e-07, | |
| "loss": 0.0088, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 28.642857142857142, | |
| "grad_norm": 0.15039542317390442, | |
| "learning_rate": 9.823052933821643e-07, | |
| "loss": 0.0067, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 28.678571428571427, | |
| "grad_norm": 0.1549944132566452, | |
| "learning_rate": 9.717296689696283e-07, | |
| "loss": 0.0081, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 28.714285714285715, | |
| "grad_norm": 0.161531463265419, | |
| "learning_rate": 9.612051515965388e-07, | |
| "loss": 0.0069, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 28.75, | |
| "grad_norm": 0.1960175782442093, | |
| "learning_rate": 9.507318747884243e-07, | |
| "loss": 0.0107, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 28.785714285714285, | |
| "grad_norm": 0.23234865069389343, | |
| "learning_rate": 9.403099714207175e-07, | |
| "loss": 0.0137, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 28.821428571428573, | |
| "grad_norm": 0.16583284735679626, | |
| "learning_rate": 9.299395737170758e-07, | |
| "loss": 0.0096, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 28.857142857142858, | |
| "grad_norm": 0.14131559431552887, | |
| "learning_rate": 9.196208132476963e-07, | |
| "loss": 0.0068, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 28.892857142857142, | |
| "grad_norm": 0.18351143598556519, | |
| "learning_rate": 9.093538209276487e-07, | |
| "loss": 0.0074, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 28.928571428571427, | |
| "grad_norm": 0.18050339818000793, | |
| "learning_rate": 8.991387270152202e-07, | |
| "loss": 0.0105, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 28.964285714285715, | |
| "grad_norm": 0.16577477753162384, | |
| "learning_rate": 8.88975661110254e-07, | |
| "loss": 0.0079, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.1450648307800293, | |
| "learning_rate": 8.78864752152509e-07, | |
| "loss": 0.007, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 29.035714285714285, | |
| "grad_norm": 0.17009346187114716, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.0091, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 29.071428571428573, | |
| "grad_norm": 0.17218393087387085, | |
| "learning_rate": 8.587999175274986e-07, | |
| "loss": 0.0087, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 29.107142857142858, | |
| "grad_norm": 0.17585642635822296, | |
| "learning_rate": 8.488462464246495e-07, | |
| "loss": 0.0089, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 29.142857142857142, | |
| "grad_norm": 0.14578095078468323, | |
| "learning_rate": 8.389452413946314e-07, | |
| "loss": 0.0077, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 29.178571428571427, | |
| "grad_norm": 0.16501793265342712, | |
| "learning_rate": 8.290970280524124e-07, | |
| "loss": 0.0088, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 29.214285714285715, | |
| "grad_norm": 0.1998208463191986, | |
| "learning_rate": 8.193017313431872e-07, | |
| "loss": 0.0106, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 29.25, | |
| "grad_norm": 0.13896793127059937, | |
| "learning_rate": 8.095594755407971e-07, | |
| "loss": 0.0068, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 29.285714285714285, | |
| "grad_norm": 0.1557644158601761, | |
| "learning_rate": 7.99870384246143e-07, | |
| "loss": 0.0085, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 29.321428571428573, | |
| "grad_norm": 0.1574714481830597, | |
| "learning_rate": 7.902345803856265e-07, | |
| "loss": 0.0075, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 29.357142857142858, | |
| "grad_norm": 0.15094228088855743, | |
| "learning_rate": 7.806521862095834e-07, | |
| "loss": 0.0074, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 29.392857142857142, | |
| "grad_norm": 0.13187389075756073, | |
| "learning_rate": 7.711233232907401e-07, | |
| "loss": 0.0068, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 29.428571428571427, | |
| "grad_norm": 0.1584806889295578, | |
| "learning_rate": 7.616481125226632e-07, | |
| "loss": 0.0087, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 29.464285714285715, | |
| "grad_norm": 0.17923088371753693, | |
| "learning_rate": 7.522266741182305e-07, | |
| "loss": 0.0106, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 29.5, | |
| "grad_norm": 0.15848712623119354, | |
| "learning_rate": 7.42859127608106e-07, | |
| "loss": 0.0094, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 29.535714285714285, | |
| "grad_norm": 0.21657797694206238, | |
| "learning_rate": 7.33545591839222e-07, | |
| "loss": 0.0114, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 29.571428571428573, | |
| "grad_norm": 0.14506831765174866, | |
| "learning_rate": 7.242861849732696e-07, | |
| "loss": 0.0069, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 29.607142857142858, | |
| "grad_norm": 0.14687834680080414, | |
| "learning_rate": 7.150810244852036e-07, | |
| "loss": 0.0075, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 29.642857142857142, | |
| "grad_norm": 0.1327245533466339, | |
| "learning_rate": 7.059302271617485e-07, | |
| "loss": 0.0067, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 29.678571428571427, | |
| "grad_norm": 0.17178060114383698, | |
| "learning_rate": 6.968339090999188e-07, | |
| "loss": 0.0094, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 29.714285714285715, | |
| "grad_norm": 0.13178479671478271, | |
| "learning_rate": 6.877921857055476e-07, | |
| "loss": 0.0068, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 29.75, | |
| "grad_norm": 0.17513161897659302, | |
| "learning_rate": 6.78805171691817e-07, | |
| "loss": 0.009, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 29.785714285714285, | |
| "grad_norm": 0.16417403519153595, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.0088, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 29.821428571428573, | |
| "grad_norm": 0.1668260246515274, | |
| "learning_rate": 6.609957271870505e-07, | |
| "loss": 0.0087, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 29.857142857142858, | |
| "grad_norm": 0.13266846537590027, | |
| "learning_rate": 6.521735226460901e-07, | |
| "loss": 0.0065, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 29.892857142857142, | |
| "grad_norm": 0.18691495060920715, | |
| "learning_rate": 6.43406479383053e-07, | |
| "loss": 0.0107, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 29.928571428571427, | |
| "grad_norm": 0.1541401743888855, | |
| "learning_rate": 6.346947086262323e-07, | |
| "loss": 0.007, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 29.964285714285715, | |
| "grad_norm": 0.19448424875736237, | |
| "learning_rate": 6.260383209026704e-07, | |
| "loss": 0.0103, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.17279107868671417, | |
| "learning_rate": 6.174374260367611e-07, | |
| "loss": 0.0093, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 30.035714285714285, | |
| "grad_norm": 0.1862161010503769, | |
| "learning_rate": 6.088921331488568e-07, | |
| "loss": 0.0097, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 30.071428571428573, | |
| "grad_norm": 0.1417967826128006, | |
| "learning_rate": 6.004025506538813e-07, | |
| "loss": 0.0065, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 30.107142857142858, | |
| "grad_norm": 0.15217995643615723, | |
| "learning_rate": 5.919687862599549e-07, | |
| "loss": 0.008, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 30.142857142857142, | |
| "grad_norm": 0.1934983730316162, | |
| "learning_rate": 5.835909469670292e-07, | |
| "loss": 0.0091, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 30.178571428571427, | |
| "grad_norm": 0.11829309910535812, | |
| "learning_rate": 5.752691390655279e-07, | |
| "loss": 0.005, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 30.214285714285715, | |
| "grad_norm": 0.156152606010437, | |
| "learning_rate": 5.670034681349995e-07, | |
| "loss": 0.008, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 30.25, | |
| "grad_norm": 0.12935622036457062, | |
| "learning_rate": 5.587940390427804e-07, | |
| "loss": 0.0068, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 30.285714285714285, | |
| "grad_norm": 0.17266200482845306, | |
| "learning_rate": 5.506409559426573e-07, | |
| "loss": 0.0101, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 30.321428571428573, | |
| "grad_norm": 0.1491808146238327, | |
| "learning_rate": 5.425443222735527e-07, | |
| "loss": 0.0078, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 30.357142857142858, | |
| "grad_norm": 0.18063697218894958, | |
| "learning_rate": 5.345042407582079e-07, | |
| "loss": 0.009, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 30.392857142857142, | |
| "grad_norm": 0.13006682693958282, | |
| "learning_rate": 5.265208134018851e-07, | |
| "loss": 0.0063, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 30.428571428571427, | |
| "grad_norm": 0.16981807351112366, | |
| "learning_rate": 5.185941414910673e-07, | |
| "loss": 0.0095, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 30.464285714285715, | |
| "grad_norm": 0.16083835065364838, | |
| "learning_rate": 5.107243255921746e-07, | |
| "loss": 0.008, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 30.5, | |
| "grad_norm": 0.16899225115776062, | |
| "learning_rate": 5.029114655502937e-07, | |
| "loss": 0.0081, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 30.535714285714285, | |
| "grad_norm": 0.16869334876537323, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.0092, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 30.571428571428573, | |
| "grad_norm": 0.19343996047973633, | |
| "learning_rate": 4.874570088036252e-07, | |
| "loss": 0.01, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 30.607142857142858, | |
| "grad_norm": 0.1612548828125, | |
| "learning_rate": 4.798156081709638e-07, | |
| "loss": 0.0085, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 30.642857142857142, | |
| "grad_norm": 0.18179935216903687, | |
| "learning_rate": 4.722315555370793e-07, | |
| "loss": 0.0095, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 30.678571428571427, | |
| "grad_norm": 0.16212452948093414, | |
| "learning_rate": 4.647049471215498e-07, | |
| "loss": 0.0079, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 30.714285714285715, | |
| "grad_norm": 0.15953488647937775, | |
| "learning_rate": 4.5723587841515707e-07, | |
| "loss": 0.0082, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 30.75, | |
| "grad_norm": 0.17954635620117188, | |
| "learning_rate": 4.4982444417866753e-07, | |
| "loss": 0.0088, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 30.785714285714285, | |
| "grad_norm": 0.15342117846012115, | |
| "learning_rate": 4.4247073844163434e-07, | |
| "loss": 0.0096, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 30.821428571428573, | |
| "grad_norm": 0.1617392897605896, | |
| "learning_rate": 4.351748545012058e-07, | |
| "loss": 0.0086, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 30.857142857142858, | |
| "grad_norm": 0.17065538465976715, | |
| "learning_rate": 4.279368849209381e-07, | |
| "loss": 0.0096, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 30.892857142857142, | |
| "grad_norm": 0.15124979615211487, | |
| "learning_rate": 4.2075692152962145e-07, | |
| "loss": 0.0085, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 30.928571428571427, | |
| "grad_norm": 0.17575320601463318, | |
| "learning_rate": 4.136350554201196e-07, | |
| "loss": 0.0087, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 30.964285714285715, | |
| "grad_norm": 0.15727750957012177, | |
| "learning_rate": 4.0657137694820826e-07, | |
| "loss": 0.0081, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.17699168622493744, | |
| "learning_rate": 3.9956597573142966e-07, | |
| "loss": 0.0089, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 31.035714285714285, | |
| "grad_norm": 0.1416083723306656, | |
| "learning_rate": 3.9261894064796136e-07, | |
| "loss": 0.0079, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 31.071428571428573, | |
| "grad_norm": 0.14622581005096436, | |
| "learning_rate": 3.8573035983548167e-07, | |
| "loss": 0.0073, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 31.107142857142858, | |
| "grad_norm": 0.14799736440181732, | |
| "learning_rate": 3.789003206900538e-07, | |
| "loss": 0.0072, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 31.142857142857142, | |
| "grad_norm": 0.15523618459701538, | |
| "learning_rate": 3.7212890986501773e-07, | |
| "loss": 0.0082, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 31.178571428571427, | |
| "grad_norm": 0.17029112577438354, | |
| "learning_rate": 3.6541621326989183e-07, | |
| "loss": 0.0107, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 31.214285714285715, | |
| "grad_norm": 0.15336446464061737, | |
| "learning_rate": 3.5876231606927936e-07, | |
| "loss": 0.0071, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "grad_norm": 0.17205260694026947, | |
| "learning_rate": 3.5216730268179346e-07, | |
| "loss": 0.0085, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 31.285714285714285, | |
| "grad_norm": 0.13131991028785706, | |
| "learning_rate": 3.4563125677897936e-07, | |
| "loss": 0.0075, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 31.321428571428573, | |
| "grad_norm": 0.1458451747894287, | |
| "learning_rate": 3.3915426128425744e-07, | |
| "loss": 0.0077, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 31.357142857142858, | |
| "grad_norm": 0.14860883355140686, | |
| "learning_rate": 3.327363983718723e-07, | |
| "loss": 0.007, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 31.392857142857142, | |
| "grad_norm": 0.18567205965518951, | |
| "learning_rate": 3.263777494658449e-07, | |
| "loss": 0.0106, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 0.16162905097007751, | |
| "learning_rate": 3.200783952389447e-07, | |
| "loss": 0.0091, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 31.464285714285715, | |
| "grad_norm": 0.16891200840473175, | |
| "learning_rate": 3.138384156116614e-07, | |
| "loss": 0.0083, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 31.5, | |
| "grad_norm": 0.1782492995262146, | |
| "learning_rate": 3.076578897511978e-07, | |
| "loss": 0.0083, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 31.535714285714285, | |
| "grad_norm": 0.1816483587026596, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0093, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 31.571428571428573, | |
| "grad_norm": 0.1914282590150833, | |
| "learning_rate": 2.954755122270564e-07, | |
| "loss": 0.0107, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 31.607142857142858, | |
| "grad_norm": 0.16533005237579346, | |
| "learning_rate": 2.894738151223331e-07, | |
| "loss": 0.0094, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 31.642857142857142, | |
| "grad_norm": 0.12326805293560028, | |
| "learning_rate": 2.835318809003751e-07, | |
| "loss": 0.0054, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 31.678571428571427, | |
| "grad_norm": 0.15294621884822845, | |
| "learning_rate": 2.776497849470544e-07, | |
| "loss": 0.0071, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 31.714285714285715, | |
| "grad_norm": 0.15630222856998444, | |
| "learning_rate": 2.71827601889067e-07, | |
| "loss": 0.0069, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 31.75, | |
| "grad_norm": 0.15990734100341797, | |
| "learning_rate": 2.6606540559298956e-07, | |
| "loss": 0.0079, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 31.785714285714285, | |
| "grad_norm": 0.15448947250843048, | |
| "learning_rate": 2.6036326916434153e-07, | |
| "loss": 0.0081, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 31.821428571428573, | |
| "grad_norm": 0.14279963076114655, | |
| "learning_rate": 2.547212649466568e-07, | |
| "loss": 0.0067, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 31.857142857142858, | |
| "grad_norm": 0.16762612760066986, | |
| "learning_rate": 2.491394645205669e-07, | |
| "loss": 0.0084, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 31.892857142857142, | |
| "grad_norm": 0.1591629981994629, | |
| "learning_rate": 2.436179387028903e-07, | |
| "loss": 0.0079, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 31.928571428571427, | |
| "grad_norm": 0.16853336989879608, | |
| "learning_rate": 2.3815675754573885e-07, | |
| "loss": 0.0088, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 31.964285714285715, | |
| "grad_norm": 0.19901008903980255, | |
| "learning_rate": 2.3275599033562414e-07, | |
| "loss": 0.0102, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.12494003027677536, | |
| "learning_rate": 2.274157055925802e-07, | |
| "loss": 0.007, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 32.035714285714285, | |
| "grad_norm": 0.16306279599666595, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.0093, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 32.07142857142857, | |
| "grad_norm": 0.1226641833782196, | |
| "learning_rate": 2.1691685375025362e-07, | |
| "loss": 0.007, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 32.107142857142854, | |
| "grad_norm": 0.16900447010993958, | |
| "learning_rate": 2.117584198508771e-07, | |
| "loss": 0.0079, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 32.142857142857146, | |
| "grad_norm": 0.16100150346755981, | |
| "learning_rate": 2.0666073481669714e-07, | |
| "loss": 0.0086, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 32.142857142857146, | |
| "eval_loss": 2.5882554054260254, | |
| "eval_runtime": 0.4733, | |
| "eval_samples_per_second": 295.815, | |
| "eval_steps_per_second": 2.113, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 32.17857142857143, | |
| "grad_norm": 0.16585837304592133, | |
| "learning_rate": 2.016238633225165e-07, | |
| "loss": 0.008, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 32.214285714285715, | |
| "grad_norm": 0.15317648649215698, | |
| "learning_rate": 1.9664786927159064e-07, | |
| "loss": 0.008, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 32.25, | |
| "grad_norm": 0.18521113693714142, | |
| "learning_rate": 1.9173281579481896e-07, | |
| "loss": 0.01, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 32.285714285714285, | |
| "grad_norm": 0.1476742923259735, | |
| "learning_rate": 1.8687876524993987e-07, | |
| "loss": 0.0065, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 32.32142857142857, | |
| "grad_norm": 0.15810611844062805, | |
| "learning_rate": 1.820857792207431e-07, | |
| "loss": 0.0072, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 32.357142857142854, | |
| "grad_norm": 0.20258426666259766, | |
| "learning_rate": 1.7735391851628814e-07, | |
| "loss": 0.0105, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 32.392857142857146, | |
| "grad_norm": 0.16195912659168243, | |
| "learning_rate": 1.7268324317012974e-07, | |
| "loss": 0.0076, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 32.42857142857143, | |
| "grad_norm": 0.15066063404083252, | |
| "learning_rate": 1.680738124395598e-07, | |
| "loss": 0.0069, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 32.464285714285715, | |
| "grad_norm": 0.13371895253658295, | |
| "learning_rate": 1.6352568480485277e-07, | |
| "loss": 0.0063, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "grad_norm": 0.1613711714744568, | |
| "learning_rate": 1.5903891796852756e-07, | |
| "loss": 0.0098, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 32.535714285714285, | |
| "grad_norm": 0.16096334159374237, | |
| "learning_rate": 1.5461356885461077e-07, | |
| "loss": 0.0068, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 32.57142857142857, | |
| "grad_norm": 0.14560391008853912, | |
| "learning_rate": 1.5024969360791564e-07, | |
| "loss": 0.0064, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 32.607142857142854, | |
| "grad_norm": 0.16068080067634583, | |
| "learning_rate": 1.4594734759333484e-07, | |
| "loss": 0.0075, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 32.642857142857146, | |
| "grad_norm": 0.1491153985261917, | |
| "learning_rate": 1.4170658539512993e-07, | |
| "loss": 0.008, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 32.67857142857143, | |
| "grad_norm": 0.16911394894123077, | |
| "learning_rate": 1.375274608162447e-07, | |
| "loss": 0.0085, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 32.714285714285715, | |
| "grad_norm": 0.14449740946292877, | |
| "learning_rate": 1.3341002687762062e-07, | |
| "loss": 0.0074, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 32.75, | |
| "grad_norm": 0.1513504683971405, | |
| "learning_rate": 1.2935433581752365e-07, | |
| "loss": 0.0082, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 32.785714285714285, | |
| "grad_norm": 0.1615770161151886, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.008, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 32.82142857142857, | |
| "grad_norm": 0.18504442274570465, | |
| "learning_rate": 1.2142838736863562e-07, | |
| "loss": 0.0097, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "grad_norm": 0.17463642358779907, | |
| "learning_rate": 1.175582305370887e-07, | |
| "loss": 0.0094, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 32.892857142857146, | |
| "grad_norm": 0.16344715654850006, | |
| "learning_rate": 1.1375001769728e-07, | |
| "loss": 0.0092, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 32.92857142857143, | |
| "grad_norm": 0.11732794344425201, | |
| "learning_rate": 1.1000379716435916e-07, | |
| "loss": 0.0069, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 32.964285714285715, | |
| "grad_norm": 0.1688617467880249, | |
| "learning_rate": 1.0631961646697387e-07, | |
| "loss": 0.0101, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.15955369174480438, | |
| "learning_rate": 1.0269752234666642e-07, | |
| "loss": 0.009, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 33.035714285714285, | |
| "grad_norm": 0.14984415471553802, | |
| "learning_rate": 9.913756075728088e-08, | |
| "loss": 0.0078, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 33.07142857142857, | |
| "grad_norm": 0.20997703075408936, | |
| "learning_rate": 9.563977686438019e-08, | |
| "loss": 0.0104, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 33.107142857142854, | |
| "grad_norm": 0.17377646267414093, | |
| "learning_rate": 9.22042150446728e-08, | |
| "loss": 0.0093, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 33.142857142857146, | |
| "grad_norm": 0.15954937040805817, | |
| "learning_rate": 8.883091888545136e-08, | |
| "loss": 0.0077, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 33.17857142857143, | |
| "grad_norm": 0.18202899396419525, | |
| "learning_rate": 8.551993118403656e-08, | |
| "loss": 0.0105, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 33.214285714285715, | |
| "grad_norm": 0.16078250110149384, | |
| "learning_rate": 8.227129394723643e-08, | |
| "loss": 0.0082, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 33.25, | |
| "grad_norm": 0.15054886043071747, | |
| "learning_rate": 7.908504839081343e-08, | |
| "loss": 0.0071, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 33.285714285714285, | |
| "grad_norm": 0.1575610637664795, | |
| "learning_rate": 7.59612349389599e-08, | |
| "loss": 0.007, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 33.32142857142857, | |
| "grad_norm": 0.13964974880218506, | |
| "learning_rate": 7.289989322378732e-08, | |
| "loss": 0.0062, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 33.357142857142854, | |
| "grad_norm": 0.17797306180000305, | |
| "learning_rate": 6.990106208482227e-08, | |
| "loss": 0.0077, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 33.392857142857146, | |
| "grad_norm": 0.1711435765028, | |
| "learning_rate": 6.696477956851356e-08, | |
| "loss": 0.0095, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 33.42857142857143, | |
| "grad_norm": 0.19469492137432098, | |
| "learning_rate": 6.409108292774912e-08, | |
| "loss": 0.0109, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 33.464285714285715, | |
| "grad_norm": 0.16585536301136017, | |
| "learning_rate": 6.12800086213866e-08, | |
| "loss": 0.0083, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 33.5, | |
| "grad_norm": 0.1359519064426422, | |
| "learning_rate": 5.853159231378469e-08, | |
| "loss": 0.0078, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 33.535714285714285, | |
| "grad_norm": 0.1809779703617096, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.0104, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 33.57142857142857, | |
| "grad_norm": 0.13768798112869263, | |
| "learning_rate": 5.322287237712664e-08, | |
| "loss": 0.0055, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 33.607142857142854, | |
| "grad_norm": 0.16895705461502075, | |
| "learning_rate": 5.0662636100292094e-08, | |
| "loss": 0.0087, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 33.642857142857146, | |
| "grad_norm": 0.12327758222818375, | |
| "learning_rate": 4.8165192525809754e-08, | |
| "loss": 0.0063, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 33.67857142857143, | |
| "grad_norm": 0.15800786018371582, | |
| "learning_rate": 4.573057333897679e-08, | |
| "loss": 0.0072, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 33.714285714285715, | |
| "grad_norm": 0.1666928380727768, | |
| "learning_rate": 4.335880942803405e-08, | |
| "loss": 0.0088, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 33.75, | |
| "grad_norm": 0.16826018691062927, | |
| "learning_rate": 4.104993088376974e-08, | |
| "loss": 0.0081, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 33.785714285714285, | |
| "grad_norm": 0.12341219186782837, | |
| "learning_rate": 3.8803966999139686e-08, | |
| "loss": 0.0059, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 33.82142857142857, | |
| "grad_norm": 0.14304673671722412, | |
| "learning_rate": 3.662094626889656e-08, | |
| "loss": 0.0067, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 33.857142857142854, | |
| "grad_norm": 0.15671806037425995, | |
| "learning_rate": 3.450089638922738e-08, | |
| "loss": 0.008, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 33.892857142857146, | |
| "grad_norm": 0.16353370249271393, | |
| "learning_rate": 3.2443844257400434e-08, | |
| "loss": 0.008, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 33.92857142857143, | |
| "grad_norm": 0.19006143510341644, | |
| "learning_rate": 3.044981597142837e-08, | |
| "loss": 0.0096, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 33.964285714285715, | |
| "grad_norm": 0.16587476432323456, | |
| "learning_rate": 2.8518836829732332e-08, | |
| "loss": 0.0081, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.16104213893413544, | |
| "learning_rate": 2.6650931330823305e-08, | |
| "loss": 0.0083, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 34.035714285714285, | |
| "grad_norm": 0.1580321341753006, | |
| "learning_rate": 2.4846123172992953e-08, | |
| "loss": 0.0086, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 34.07142857142857, | |
| "grad_norm": 0.1464652419090271, | |
| "learning_rate": 2.3104435254008852e-08, | |
| "loss": 0.0073, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 34.107142857142854, | |
| "grad_norm": 0.12108955532312393, | |
| "learning_rate": 2.1425889670827483e-08, | |
| "loss": 0.0053, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 34.142857142857146, | |
| "grad_norm": 0.1350407600402832, | |
| "learning_rate": 1.981050771931281e-08, | |
| "loss": 0.007, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 34.17857142857143, | |
| "grad_norm": 0.16024035215377808, | |
| "learning_rate": 1.8258309893965375e-08, | |
| "loss": 0.0087, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 34.214285714285715, | |
| "grad_norm": 0.1773962676525116, | |
| "learning_rate": 1.6769315887662508e-08, | |
| "loss": 0.0082, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 34.25, | |
| "grad_norm": 0.17347432672977448, | |
| "learning_rate": 1.5343544591409632e-08, | |
| "loss": 0.0088, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 0.17487084865570068, | |
| "learning_rate": 1.3981014094099354e-08, | |
| "loss": 0.0097, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 34.32142857142857, | |
| "grad_norm": 0.18157510459423065, | |
| "learning_rate": 1.2681741682282755e-08, | |
| "loss": 0.01, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 34.357142857142854, | |
| "grad_norm": 0.177931547164917, | |
| "learning_rate": 1.1445743839949008e-08, | |
| "loss": 0.0096, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 34.392857142857146, | |
| "grad_norm": 0.17463018000125885, | |
| "learning_rate": 1.0273036248318325e-08, | |
| "loss": 0.0098, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 34.42857142857143, | |
| "grad_norm": 0.17054671049118042, | |
| "learning_rate": 9.163633785639892e-09, | |
| "loss": 0.0084, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 34.464285714285715, | |
| "grad_norm": 0.17844708263874054, | |
| "learning_rate": 8.117550527005913e-09, | |
| "loss": 0.0089, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 34.5, | |
| "grad_norm": 0.14375606179237366, | |
| "learning_rate": 7.13479974417175e-09, | |
| "loss": 0.0073, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 34.535714285714285, | |
| "grad_norm": 0.1640142798423767, | |
| "learning_rate": 6.215393905388278e-09, | |
| "loss": 0.0083, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 34.57142857142857, | |
| "grad_norm": 0.13746048510074615, | |
| "learning_rate": 5.359344675242018e-09, | |
| "loss": 0.007, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 34.607142857142854, | |
| "grad_norm": 0.18680772185325623, | |
| "learning_rate": 4.56666291450858e-09, | |
| "loss": 0.0104, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 34.642857142857146, | |
| "grad_norm": 0.17095445096492767, | |
| "learning_rate": 3.837358680016112e-09, | |
| "loss": 0.0087, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 34.67857142857143, | |
| "grad_norm": 0.13912023603916168, | |
| "learning_rate": 3.1714412245148486e-09, | |
| "loss": 0.0065, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 34.714285714285715, | |
| "grad_norm": 0.16117946803569794, | |
| "learning_rate": 2.568918996560532e-09, | |
| "loss": 0.0088, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 34.75, | |
| "grad_norm": 0.15318843722343445, | |
| "learning_rate": 2.029799640409502e-09, | |
| "loss": 0.009, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 34.785714285714285, | |
| "grad_norm": 0.15008534491062164, | |
| "learning_rate": 1.5540899959187727e-09, | |
| "loss": 0.0072, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 34.82142857142857, | |
| "grad_norm": 0.13134929537773132, | |
| "learning_rate": 1.1417960984605459e-09, | |
| "loss": 0.0056, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 34.857142857142854, | |
| "grad_norm": 0.14197276532649994, | |
| "learning_rate": 7.92923178845606e-10, | |
| "loss": 0.0073, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 34.892857142857146, | |
| "grad_norm": 0.15681394934654236, | |
| "learning_rate": 5.07475663257262e-10, | |
| "loss": 0.0084, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 34.92857142857143, | |
| "grad_norm": 0.13825960457324982, | |
| "learning_rate": 2.854571731947253e-10, | |
| "loss": 0.0059, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 34.964285714285715, | |
| "grad_norm": 0.14714203774929047, | |
| "learning_rate": 1.2687052542759148e-10, | |
| "loss": 0.008, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 0.14494210481643677, | |
| "learning_rate": 3.171773195809191e-11, | |
| "loss": 0.0075, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "step": 980, | |
| "total_flos": 1.12061156850611e+18, | |
| "train_loss": 0.2908882966189056, | |
| "train_runtime": 4060.664, | |
| "train_samples_per_second": 119.463, | |
| "train_steps_per_second": 0.241 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 980, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 35, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.12061156850611e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |