| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 11.899159663865547, | |
| "eval_steps": 500, | |
| "global_step": 708, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01680672268907563, | |
| "grad_norm": 4.385648622708784, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 4.108, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03361344537815126, | |
| "grad_norm": 4.427354400402251, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 4.1105, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05042016806722689, | |
| "grad_norm": 3.8339824048548925, | |
| "learning_rate": 2.7272727272727272e-06, | |
| "loss": 3.8298, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06722689075630252, | |
| "grad_norm": 4.9404739040660814, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 4.4266, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08403361344537816, | |
| "grad_norm": 3.723380522292492, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 3.6814, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10084033613445378, | |
| "grad_norm": 10.136378288136632, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 6.7591, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 3.750584042652294, | |
| "learning_rate": 6.363636363636364e-06, | |
| "loss": 3.6869, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13445378151260504, | |
| "grad_norm": 10.256473970042155, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 6.8465, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15126050420168066, | |
| "grad_norm": 5.299551412905417, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 4.627, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16806722689075632, | |
| "grad_norm": 4.073848965470436, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 3.7922, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.18487394957983194, | |
| "grad_norm": 6.270475941131987, | |
| "learning_rate": 1e-05, | |
| "loss": 5.0956, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.20168067226890757, | |
| "grad_norm": 5.6458603155827225, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 4.6846, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2184873949579832, | |
| "grad_norm": 4.136016188838338, | |
| "learning_rate": 1.181818181818182e-05, | |
| "loss": 3.9917, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 3.6135748045353333, | |
| "learning_rate": 1.2727272727272728e-05, | |
| "loss": 3.6259, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.25210084033613445, | |
| "grad_norm": 4.964675562915649, | |
| "learning_rate": 1.3636363636363637e-05, | |
| "loss": 4.2151, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2689075630252101, | |
| "grad_norm": 6.524726574282932, | |
| "learning_rate": 1.4545454545454546e-05, | |
| "loss": 5.0345, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 5.639830925355285, | |
| "learning_rate": 1.5454545454545454e-05, | |
| "loss": 4.4856, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3025210084033613, | |
| "grad_norm": 3.8533215935243077, | |
| "learning_rate": 1.6363636363636366e-05, | |
| "loss": 3.7071, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.31932773109243695, | |
| "grad_norm": 5.161749059636034, | |
| "learning_rate": 1.7272727272727274e-05, | |
| "loss": 4.2947, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.33613445378151263, | |
| "grad_norm": 3.6136568344159885, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 3.4683, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 4.435498267084917, | |
| "learning_rate": 1.9090909090909094e-05, | |
| "loss": 3.859, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3697478991596639, | |
| "grad_norm": 3.919091925431486, | |
| "learning_rate": 2e-05, | |
| "loss": 3.5794, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3865546218487395, | |
| "grad_norm": 3.488028179094918, | |
| "learning_rate": 1.9999895137366746e-05, | |
| "loss": 3.2719, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.40336134453781514, | |
| "grad_norm": 10.328288130878647, | |
| "learning_rate": 1.9999580551666215e-05, | |
| "loss": 5.7591, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.42016806722689076, | |
| "grad_norm": 3.774723233998757, | |
| "learning_rate": 1.9999056249496065e-05, | |
| "loss": 3.3805, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4369747899159664, | |
| "grad_norm": 6.265735848819648, | |
| "learning_rate": 1.9998322241852238e-05, | |
| "loss": 4.2421, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.453781512605042, | |
| "grad_norm": 5.405371137822227, | |
| "learning_rate": 1.999737854412873e-05, | |
| "loss": 3.8983, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 3.7669212684787263, | |
| "learning_rate": 1.9996225176117264e-05, | |
| "loss": 3.1467, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.48739495798319327, | |
| "grad_norm": 5.5431024884923845, | |
| "learning_rate": 1.999486216200688e-05, | |
| "loss": 3.5476, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5042016806722689, | |
| "grad_norm": 5.394079800803299, | |
| "learning_rate": 1.9993289530383433e-05, | |
| "loss": 3.4722, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5210084033613446, | |
| "grad_norm": 5.971573494363461, | |
| "learning_rate": 1.999150731422898e-05, | |
| "loss": 3.6263, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5378151260504201, | |
| "grad_norm": 7.378751228659765, | |
| "learning_rate": 1.9989515550921088e-05, | |
| "loss": 3.5546, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5546218487394958, | |
| "grad_norm": 5.09952164148768, | |
| "learning_rate": 1.998731428223208e-05, | |
| "loss": 3.0747, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 4.765579668290796, | |
| "learning_rate": 1.9984903554328116e-05, | |
| "loss": 2.7589, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 7.224492836222785, | |
| "learning_rate": 1.9982283417768247e-05, | |
| "loss": 2.871, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6050420168067226, | |
| "grad_norm": 12.931777704179952, | |
| "learning_rate": 1.9979453927503366e-05, | |
| "loss": 3.215, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6218487394957983, | |
| "grad_norm": 3.6837055989248717, | |
| "learning_rate": 1.9976415142875022e-05, | |
| "loss": 2.4582, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6386554621848739, | |
| "grad_norm": 3.163696044731996, | |
| "learning_rate": 1.9973167127614218e-05, | |
| "loss": 2.4035, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6554621848739496, | |
| "grad_norm": 2.6130379292490127, | |
| "learning_rate": 1.9969709949840034e-05, | |
| "loss": 2.27, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6722689075630253, | |
| "grad_norm": 3.0774669175573903, | |
| "learning_rate": 1.9966043682058223e-05, | |
| "loss": 2.2352, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6890756302521008, | |
| "grad_norm": 3.3703328756445514, | |
| "learning_rate": 1.9962168401159685e-05, | |
| "loss": 2.3572, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 6.674678012817149, | |
| "learning_rate": 1.995808418841885e-05, | |
| "loss": 2.5806, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7226890756302521, | |
| "grad_norm": 2.7692115403847932, | |
| "learning_rate": 1.9953791129491985e-05, | |
| "loss": 2.2257, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7394957983193278, | |
| "grad_norm": 3.4385080781188013, | |
| "learning_rate": 1.9949289314415373e-05, | |
| "loss": 2.3251, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7563025210084033, | |
| "grad_norm": 3.513307097744062, | |
| "learning_rate": 1.994457883760346e-05, | |
| "loss": 2.2311, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.773109243697479, | |
| "grad_norm": 3.485460514370931, | |
| "learning_rate": 1.993965979784684e-05, | |
| "loss": 2.0451, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7899159663865546, | |
| "grad_norm": 2.6480026940047967, | |
| "learning_rate": 1.9934532298310206e-05, | |
| "loss": 2.1224, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8067226890756303, | |
| "grad_norm": 3.191875488563921, | |
| "learning_rate": 1.9929196446530184e-05, | |
| "loss": 2.1002, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 3.3443355134212833, | |
| "learning_rate": 1.992365235441306e-05, | |
| "loss": 2.0233, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8403361344537815, | |
| "grad_norm": 2.5999640350239948, | |
| "learning_rate": 1.991790013823246e-05, | |
| "loss": 1.966, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 6.089294042169637, | |
| "learning_rate": 1.991193991862689e-05, | |
| "loss": 1.8622, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8739495798319328, | |
| "grad_norm": 3.2204266121939362, | |
| "learning_rate": 1.9905771820597214e-05, | |
| "loss": 1.8818, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8907563025210085, | |
| "grad_norm": 3.3930839641806942, | |
| "learning_rate": 1.989939597350403e-05, | |
| "loss": 1.7539, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.907563025210084, | |
| "grad_norm": 3.449905117774067, | |
| "learning_rate": 1.9892812511064962e-05, | |
| "loss": 1.7651, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9243697478991597, | |
| "grad_norm": 3.3847593536709097, | |
| "learning_rate": 1.9886021571351854e-05, | |
| "loss": 1.6882, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 3.845735022445724, | |
| "learning_rate": 1.9879023296787866e-05, | |
| "loss": 1.719, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.957983193277311, | |
| "grad_norm": 2.8219721101939386, | |
| "learning_rate": 1.9871817834144506e-05, | |
| "loss": 1.7362, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9747899159663865, | |
| "grad_norm": 2.499202155153014, | |
| "learning_rate": 1.9864405334538518e-05, | |
| "loss": 1.7014, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9915966386554622, | |
| "grad_norm": 2.511155724416345, | |
| "learning_rate": 1.9856785953428757e-05, | |
| "loss": 1.7195, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9915966386554622, | |
| "eval_loss": 1.2069061994552612, | |
| "eval_runtime": 115.8201, | |
| "eval_samples_per_second": 0.812, | |
| "eval_steps_per_second": 0.812, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0084033613445378, | |
| "grad_norm": 5.357847266371679, | |
| "learning_rate": 1.9848959850612895e-05, | |
| "loss": 1.1994, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0252100840336134, | |
| "grad_norm": 2.5172607304803383, | |
| "learning_rate": 1.9840927190224083e-05, | |
| "loss": 1.5096, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.0420168067226891, | |
| "grad_norm": 2.383993749480795, | |
| "learning_rate": 1.9832688140727502e-05, | |
| "loss": 1.5336, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0588235294117647, | |
| "grad_norm": 5.571125580296848, | |
| "learning_rate": 1.982424287491684e-05, | |
| "loss": 1.0043, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.0756302521008403, | |
| "grad_norm": 2.2388809453197083, | |
| "learning_rate": 1.9815591569910654e-05, | |
| "loss": 1.2905, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.092436974789916, | |
| "grad_norm": 2.4116257513181356, | |
| "learning_rate": 1.9806734407148674e-05, | |
| "loss": 1.5209, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1092436974789917, | |
| "grad_norm": 2.268214660771143, | |
| "learning_rate": 1.9797671572387985e-05, | |
| "loss": 1.3414, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1260504201680672, | |
| "grad_norm": 2.1242648315918036, | |
| "learning_rate": 1.978840325569912e-05, | |
| "loss": 1.3624, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 6.45315294142602, | |
| "learning_rate": 1.977892965146211e-05, | |
| "loss": 0.7577, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.1596638655462184, | |
| "grad_norm": 2.0834553976731596, | |
| "learning_rate": 1.976925095836236e-05, | |
| "loss": 1.242, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 2.1599404414408934, | |
| "learning_rate": 1.975936737938653e-05, | |
| "loss": 1.2737, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1932773109243697, | |
| "grad_norm": 3.6995588982952263, | |
| "learning_rate": 1.9749279121818235e-05, | |
| "loss": 0.7993, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2100840336134453, | |
| "grad_norm": 2.352377611171032, | |
| "learning_rate": 1.9738986397233736e-05, | |
| "loss": 0.7334, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.226890756302521, | |
| "grad_norm": 2.152202331744084, | |
| "learning_rate": 1.9728489421497465e-05, | |
| "loss": 1.3078, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2436974789915967, | |
| "grad_norm": 2.5203922797690015, | |
| "learning_rate": 1.9717788414757523e-05, | |
| "loss": 1.1814, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2605042016806722, | |
| "grad_norm": 3.4705153272314475, | |
| "learning_rate": 1.9706883601441066e-05, | |
| "loss": 1.3192, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.2773109243697478, | |
| "grad_norm": 1.8630027919414862, | |
| "learning_rate": 1.969577521024958e-05, | |
| "loss": 0.7795, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.2941176470588236, | |
| "grad_norm": 2.9345669701977912, | |
| "learning_rate": 1.9684463474154095e-05, | |
| "loss": 1.2665, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.3109243697478992, | |
| "grad_norm": 1.4751109531953661, | |
| "learning_rate": 1.9672948630390296e-05, | |
| "loss": 0.6425, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3277310924369747, | |
| "grad_norm": 3.300164935057302, | |
| "learning_rate": 1.9661230920453553e-05, | |
| "loss": 1.2188, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.3445378151260505, | |
| "grad_norm": 2.835411015713917, | |
| "learning_rate": 1.964931059009385e-05, | |
| "loss": 1.263, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.361344537815126, | |
| "grad_norm": 2.235484344805954, | |
| "learning_rate": 1.9637187889310632e-05, | |
| "loss": 1.0342, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.3781512605042017, | |
| "grad_norm": 1.8454261871804942, | |
| "learning_rate": 1.9624863072347565e-05, | |
| "loss": 1.0816, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.3949579831932772, | |
| "grad_norm": 1.8700087156945158, | |
| "learning_rate": 1.96123363976872e-05, | |
| "loss": 1.0611, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 1.776825090533238, | |
| "learning_rate": 1.9599608128045554e-05, | |
| "loss": 1.0875, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 3.2507229249266074, | |
| "learning_rate": 1.9586678530366607e-05, | |
| "loss": 0.6209, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.4453781512605042, | |
| "grad_norm": 2.0636370694690442, | |
| "learning_rate": 1.9573547875816685e-05, | |
| "loss": 1.0625, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.46218487394958, | |
| "grad_norm": 1.941254802441877, | |
| "learning_rate": 1.9560216439778795e-05, | |
| "loss": 1.086, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4789915966386555, | |
| "grad_norm": 1.9461861491812582, | |
| "learning_rate": 1.954668450184683e-05, | |
| "loss": 0.9987, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.495798319327731, | |
| "grad_norm": 2.064972172013034, | |
| "learning_rate": 1.9532952345819723e-05, | |
| "loss": 1.0663, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5126050420168067, | |
| "grad_norm": 1.778380925304329, | |
| "learning_rate": 1.951902025969548e-05, | |
| "loss": 0.9426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5294117647058822, | |
| "grad_norm": 2.361605116359202, | |
| "learning_rate": 1.950488853566515e-05, | |
| "loss": 0.9878, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.5462184873949578, | |
| "grad_norm": 1.7448015788038604, | |
| "learning_rate": 1.949055747010669e-05, | |
| "loss": 1.054, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.5630252100840336, | |
| "grad_norm": 2.411683085008472, | |
| "learning_rate": 1.9476027363578754e-05, | |
| "loss": 0.9234, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.5798319327731094, | |
| "grad_norm": 1.8173611017503086, | |
| "learning_rate": 1.946129852081439e-05, | |
| "loss": 0.9983, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.596638655462185, | |
| "grad_norm": 1.9295587283934226, | |
| "learning_rate": 1.9446371250714645e-05, | |
| "loss": 0.9747, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6134453781512605, | |
| "grad_norm": 1.772861667848063, | |
| "learning_rate": 1.943124586634209e-05, | |
| "loss": 0.8761, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.6302521008403361, | |
| "grad_norm": 4.814779501362561, | |
| "learning_rate": 1.941592268491425e-05, | |
| "loss": 0.6477, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 1.836972617211195, | |
| "learning_rate": 1.9400402027796955e-05, | |
| "loss": 0.924, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.6638655462184873, | |
| "grad_norm": 2.3766647197122595, | |
| "learning_rate": 1.9384684220497605e-05, | |
| "loss": 0.7468, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.680672268907563, | |
| "grad_norm": 1.6832736777019341, | |
| "learning_rate": 1.936876959265833e-05, | |
| "loss": 0.8439, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6974789915966386, | |
| "grad_norm": 1.819587589935648, | |
| "learning_rate": 1.9352658478049085e-05, | |
| "loss": 0.7831, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 6.083715872130075, | |
| "learning_rate": 1.9336351214560648e-05, | |
| "loss": 0.5828, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.73109243697479, | |
| "grad_norm": 2.0455350806117827, | |
| "learning_rate": 1.9319848144197543e-05, | |
| "loss": 0.8423, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.7478991596638656, | |
| "grad_norm": 2.575532108553041, | |
| "learning_rate": 1.9303149613070852e-05, | |
| "loss": 0.8617, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 1.698878909041017, | |
| "learning_rate": 1.928625597139096e-05, | |
| "loss": 0.7528, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.7815126050420167, | |
| "grad_norm": 4.480689810270073, | |
| "learning_rate": 1.926916757346022e-05, | |
| "loss": 0.4962, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.7983193277310925, | |
| "grad_norm": 2.43491273709643, | |
| "learning_rate": 1.9251884777665513e-05, | |
| "loss": 0.6916, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.815126050420168, | |
| "grad_norm": 2.9374762528937945, | |
| "learning_rate": 1.9234407946470735e-05, | |
| "loss": 0.7218, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.8319327731092439, | |
| "grad_norm": 3.110996772971968, | |
| "learning_rate": 1.9216737446409192e-05, | |
| "loss": 0.3829, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.8487394957983194, | |
| "grad_norm": 1.963122392871237, | |
| "learning_rate": 1.919887364807592e-05, | |
| "loss": 0.7205, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.865546218487395, | |
| "grad_norm": 3.066257574584323, | |
| "learning_rate": 1.9180816926119903e-05, | |
| "loss": 0.5874, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 2.2609617726833946, | |
| "learning_rate": 1.9162567659236227e-05, | |
| "loss": 0.5984, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.8991596638655461, | |
| "grad_norm": 2.6110497038118727, | |
| "learning_rate": 1.9144126230158127e-05, | |
| "loss": 0.6068, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.9159663865546217, | |
| "grad_norm": 2.2601235737134195, | |
| "learning_rate": 1.9125493025648963e-05, | |
| "loss": 0.6273, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.9327731092436975, | |
| "grad_norm": 2.271778447040359, | |
| "learning_rate": 1.9106668436494113e-05, | |
| "loss": 0.6891, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.949579831932773, | |
| "grad_norm": 2.0592194165628275, | |
| "learning_rate": 1.908765285749278e-05, | |
| "loss": 0.6283, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.9663865546218489, | |
| "grad_norm": 2.096696308042481, | |
| "learning_rate": 1.9068446687449698e-05, | |
| "loss": 0.5497, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.9831932773109244, | |
| "grad_norm": 2.4126456796916744, | |
| "learning_rate": 1.9049050329166778e-05, | |
| "loss": 0.5112, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.4662803136656954, | |
| "learning_rate": 1.9029464189434663e-05, | |
| "loss": 0.5049, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.4023877680301666, | |
| "eval_runtime": 94.9843, | |
| "eval_samples_per_second": 0.99, | |
| "eval_steps_per_second": 0.99, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.0168067226890756, | |
| "grad_norm": 2.736890471961608, | |
| "learning_rate": 1.900968867902419e-05, | |
| "loss": 0.4793, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.033613445378151, | |
| "grad_norm": 2.030226042772956, | |
| "learning_rate": 1.8989724212677784e-05, | |
| "loss": 0.4857, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.0504201680672267, | |
| "grad_norm": 2.308935605416641, | |
| "learning_rate": 1.8969571209100738e-05, | |
| "loss": 0.3551, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0672268907563027, | |
| "grad_norm": 1.9661040290247784, | |
| "learning_rate": 1.8949230090952463e-05, | |
| "loss": 0.4291, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.0840336134453783, | |
| "grad_norm": 2.0802991306768854, | |
| "learning_rate": 1.8928701284837597e-05, | |
| "loss": 0.4988, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.100840336134454, | |
| "grad_norm": 2.0446185078523804, | |
| "learning_rate": 1.890798522129708e-05, | |
| "loss": 0.3199, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 2.403214003507413, | |
| "learning_rate": 1.8887082334799098e-05, | |
| "loss": 0.4125, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.134453781512605, | |
| "grad_norm": 1.7254339304792963, | |
| "learning_rate": 1.8865993063730003e-05, | |
| "loss": 0.3017, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.1512605042016806, | |
| "grad_norm": 2.3578101043917643, | |
| "learning_rate": 1.884471785038509e-05, | |
| "loss": 0.228, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.168067226890756, | |
| "grad_norm": 1.6260308846139646, | |
| "learning_rate": 1.882325714095934e-05, | |
| "loss": 0.1768, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.184873949579832, | |
| "grad_norm": 1.9572713847794219, | |
| "learning_rate": 1.8801611385538047e-05, | |
| "loss": 0.3387, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2016806722689077, | |
| "grad_norm": 2.0610001563380127, | |
| "learning_rate": 1.8779781038087406e-05, | |
| "loss": 0.3104, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.2184873949579833, | |
| "grad_norm": 2.770005426395026, | |
| "learning_rate": 1.875776655644495e-05, | |
| "loss": 0.3574, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.235294117647059, | |
| "grad_norm": 3.4273616297199796, | |
| "learning_rate": 1.8735568402309987e-05, | |
| "loss": 0.3576, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.2521008403361344, | |
| "grad_norm": 2.0797522448134242, | |
| "learning_rate": 1.8713187041233896e-05, | |
| "loss": 0.289, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.26890756302521, | |
| "grad_norm": 5.8224416033882065, | |
| "learning_rate": 1.8690622942610367e-05, | |
| "loss": 0.231, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 1.805558912475057, | |
| "learning_rate": 1.8667876579665556e-05, | |
| "loss": 0.2549, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.302521008403361, | |
| "grad_norm": 2.1386013924513407, | |
| "learning_rate": 1.8644948429448174e-05, | |
| "loss": 0.3056, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.3193277310924367, | |
| "grad_norm": 2.891103001226989, | |
| "learning_rate": 1.862183897281946e-05, | |
| "loss": 0.2917, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.3361344537815127, | |
| "grad_norm": 1.6366849032923414, | |
| "learning_rate": 1.8598548694443102e-05, | |
| "loss": 0.185, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 3.3075808645565825, | |
| "learning_rate": 1.8575078082775096e-05, | |
| "loss": 0.272, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.369747899159664, | |
| "grad_norm": 2.271589816301807, | |
| "learning_rate": 1.8551427630053464e-05, | |
| "loss": 0.245, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.3865546218487395, | |
| "grad_norm": 1.84180801145786, | |
| "learning_rate": 1.8527597832287954e-05, | |
| "loss": 0.2091, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.403361344537815, | |
| "grad_norm": 3.766252312196367, | |
| "learning_rate": 1.8503589189249637e-05, | |
| "loss": 0.2895, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.4201680672268906, | |
| "grad_norm": 1.873377028811922, | |
| "learning_rate": 1.847940220446042e-05, | |
| "loss": 0.2368, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.4369747899159666, | |
| "grad_norm": 4.545298766002122, | |
| "learning_rate": 1.845503738518249e-05, | |
| "loss": 0.1958, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.453781512605042, | |
| "grad_norm": 3.1348145302552903, | |
| "learning_rate": 1.843049524240766e-05, | |
| "loss": 0.2098, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.4705882352941178, | |
| "grad_norm": 1.6186560450788932, | |
| "learning_rate": 1.8405776290846672e-05, | |
| "loss": 0.1754, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.4873949579831933, | |
| "grad_norm": 3.382962591579758, | |
| "learning_rate": 1.8380881048918406e-05, | |
| "loss": 0.2035, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.504201680672269, | |
| "grad_norm": 2.623712266453748, | |
| "learning_rate": 1.8355810038738986e-05, | |
| "loss": 0.2092, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.5210084033613445, | |
| "grad_norm": 2.2054319523334676, | |
| "learning_rate": 1.8330563786110837e-05, | |
| "loss": 0.1898, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.53781512605042, | |
| "grad_norm": 4.2861344256493785, | |
| "learning_rate": 1.830514282051166e-05, | |
| "loss": 0.1886, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.5546218487394956, | |
| "grad_norm": 2.651873913211827, | |
| "learning_rate": 1.8279547675083343e-05, | |
| "loss": 0.1824, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 1.8547906501685159, | |
| "learning_rate": 1.8253778886620754e-05, | |
| "loss": 0.1705, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 3.9635058525876277, | |
| "learning_rate": 1.822783699556049e-05, | |
| "loss": 0.1838, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.6050420168067228, | |
| "grad_norm": 2.3948916184720606, | |
| "learning_rate": 1.820172254596956e-05, | |
| "loss": 0.1778, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.6218487394957983, | |
| "grad_norm": 2.3071047956425264, | |
| "learning_rate": 1.817543608553395e-05, | |
| "loss": 0.1708, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.638655462184874, | |
| "grad_norm": 2.108074459212226, | |
| "learning_rate": 1.814897816554715e-05, | |
| "loss": 0.1749, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.6554621848739495, | |
| "grad_norm": 1.0692114014911545, | |
| "learning_rate": 1.8122349340898596e-05, | |
| "loss": 0.1373, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.6722689075630255, | |
| "grad_norm": 2.0009322090894592, | |
| "learning_rate": 1.809555017006202e-05, | |
| "loss": 0.1709, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.689075630252101, | |
| "grad_norm": 2.1909981915140375, | |
| "learning_rate": 1.8068581215083752e-05, | |
| "loss": 0.1484, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.7058823529411766, | |
| "grad_norm": 2.103149621290993, | |
| "learning_rate": 1.804144304157091e-05, | |
| "loss": 0.1482, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.722689075630252, | |
| "grad_norm": 1.5470172068234862, | |
| "learning_rate": 1.8014136218679566e-05, | |
| "loss": 0.1489, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.7394957983193278, | |
| "grad_norm": 1.8588243552375616, | |
| "learning_rate": 1.7986661319102795e-05, | |
| "loss": 0.1526, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.7563025210084033, | |
| "grad_norm": 1.1673797887420725, | |
| "learning_rate": 1.7959018919058654e-05, | |
| "loss": 0.1361, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.773109243697479, | |
| "grad_norm": 0.9871009564387679, | |
| "learning_rate": 1.7931209598278117e-05, | |
| "loss": 0.1168, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.7899159663865545, | |
| "grad_norm": 1.375240831135664, | |
| "learning_rate": 1.7903233939992904e-05, | |
| "loss": 0.1271, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.80672268907563, | |
| "grad_norm": 1.1795213655735046, | |
| "learning_rate": 1.787509253092326e-05, | |
| "loss": 0.1092, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 3.071368428139141, | |
| "learning_rate": 1.784678596126563e-05, | |
| "loss": 0.286, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.8403361344537816, | |
| "grad_norm": 1.0579777160189892, | |
| "learning_rate": 1.78183148246803e-05, | |
| "loss": 0.1197, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 1.6336798015628837, | |
| "learning_rate": 1.7789679718278944e-05, | |
| "loss": 0.1282, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.8739495798319328, | |
| "grad_norm": 1.8315662561658703, | |
| "learning_rate": 1.7760881242612096e-05, | |
| "loss": 0.1429, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.8907563025210083, | |
| "grad_norm": 2.9860725552192404, | |
| "learning_rate": 1.773192000165655e-05, | |
| "loss": 0.2833, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.907563025210084, | |
| "grad_norm": 1.04161581184864, | |
| "learning_rate": 1.7702796602802705e-05, | |
| "loss": 0.1042, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.92436974789916, | |
| "grad_norm": 1.2933586996454134, | |
| "learning_rate": 1.7673511656841822e-05, | |
| "loss": 0.1382, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 1.8373790721021168, | |
| "learning_rate": 1.7644065777953206e-05, | |
| "loss": 0.2055, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.957983193277311, | |
| "grad_norm": 1.695542614817302, | |
| "learning_rate": 1.7614459583691346e-05, | |
| "loss": 0.144, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.9747899159663866, | |
| "grad_norm": 1.2079374495539366, | |
| "learning_rate": 1.758469369497293e-05, | |
| "loss": 0.1286, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.991596638655462, | |
| "grad_norm": 7.433145063552697, | |
| "learning_rate": 1.7554768736063858e-05, | |
| "loss": 0.2421, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.991596638655462, | |
| "eval_loss": 0.18344487249851227, | |
| "eval_runtime": 97.2976, | |
| "eval_samples_per_second": 0.966, | |
| "eval_steps_per_second": 0.966, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.008403361344538, | |
| "grad_norm": 1.936304205658518, | |
| "learning_rate": 1.7524685334566126e-05, | |
| "loss": 0.0924, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.0252100840336134, | |
| "grad_norm": 3.65130263061498, | |
| "learning_rate": 1.7494444121404673e-05, | |
| "loss": 0.168, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.042016806722689, | |
| "grad_norm": 6.416735934730296, | |
| "learning_rate": 1.746404573081415e-05, | |
| "loss": 0.1856, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 2.2813578076866623, | |
| "learning_rate": 1.7433490800325614e-05, | |
| "loss": 0.1393, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.0756302521008405, | |
| "grad_norm": 2.595639669498612, | |
| "learning_rate": 1.7402779970753156e-05, | |
| "loss": 0.1428, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.092436974789916, | |
| "grad_norm": 1.767980230153596, | |
| "learning_rate": 1.7371913886180473e-05, | |
| "loss": 0.1331, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.1092436974789917, | |
| "grad_norm": 2.5213252449194252, | |
| "learning_rate": 1.7340893193947342e-05, | |
| "loss": 0.1252, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.1260504201680672, | |
| "grad_norm": 2.1170186194258167, | |
| "learning_rate": 1.7309718544636057e-05, | |
| "loss": 0.1108, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 2.0268322211129592, | |
| "learning_rate": 1.7278390592057785e-05, | |
| "loss": 0.1561, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.1596638655462184, | |
| "grad_norm": 1.7261621681645014, | |
| "learning_rate": 1.7246909993238844e-05, | |
| "loss": 0.1246, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.176470588235294, | |
| "grad_norm": 1.3950437881445872, | |
| "learning_rate": 1.7215277408406932e-05, | |
| "loss": 0.0998, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.19327731092437, | |
| "grad_norm": 1.793336566476121, | |
| "learning_rate": 1.7183493500977277e-05, | |
| "loss": 0.128, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.2100840336134455, | |
| "grad_norm": 1.35402748785192, | |
| "learning_rate": 1.7151558937538725e-05, | |
| "loss": 0.0979, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.226890756302521, | |
| "grad_norm": 1.9875540888658763, | |
| "learning_rate": 1.7119474387839764e-05, | |
| "loss": 0.1477, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.2436974789915967, | |
| "grad_norm": 3.4010639220257755, | |
| "learning_rate": 1.708724052477446e-05, | |
| "loss": 0.2346, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.2605042016806722, | |
| "grad_norm": 1.1526809505800015, | |
| "learning_rate": 1.7054858024368365e-05, | |
| "loss": 0.1189, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.277310924369748, | |
| "grad_norm": 1.8401498885713787, | |
| "learning_rate": 1.7022327565764336e-05, | |
| "loss": 0.1321, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 1.823800189990081, | |
| "learning_rate": 1.6989649831208286e-05, | |
| "loss": 0.1276, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.310924369747899, | |
| "grad_norm": 2.177879884931455, | |
| "learning_rate": 1.6956825506034866e-05, | |
| "loss": 0.1367, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.327731092436975, | |
| "grad_norm": 4.377960471738762, | |
| "learning_rate": 1.6923855278653114e-05, | |
| "loss": 0.177, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.3445378151260505, | |
| "grad_norm": 1.530590314148621, | |
| "learning_rate": 1.6890739840532004e-05, | |
| "loss": 0.1344, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.361344537815126, | |
| "grad_norm": 1.3469335679962628, | |
| "learning_rate": 1.6857479886185942e-05, | |
| "loss": 0.1052, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.3781512605042017, | |
| "grad_norm": 1.6139041164207482, | |
| "learning_rate": 1.682407611316021e-05, | |
| "loss": 0.1475, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.3949579831932772, | |
| "grad_norm": 1.6495915167747617, | |
| "learning_rate": 1.6790529222016328e-05, | |
| "loss": 0.0972, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.411764705882353, | |
| "grad_norm": 1.8398639057645052, | |
| "learning_rate": 1.6756839916317358e-05, | |
| "loss": 0.1084, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 2.0107317404674236, | |
| "learning_rate": 1.672300890261317e-05, | |
| "loss": 0.1102, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.4453781512605044, | |
| "grad_norm": 1.9565606900612817, | |
| "learning_rate": 1.6689036890425596e-05, | |
| "loss": 0.1139, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.46218487394958, | |
| "grad_norm": 1.639534786311468, | |
| "learning_rate": 1.665492459223357e-05, | |
| "loss": 0.0918, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 3.4789915966386555, | |
| "grad_norm": 5.45699346508005, | |
| "learning_rate": 1.6620672723458167e-05, | |
| "loss": 0.1591, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 3.495798319327731, | |
| "grad_norm": 2.0885082670769965, | |
| "learning_rate": 1.658628200244763e-05, | |
| "loss": 0.1118, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 3.5126050420168067, | |
| "grad_norm": 1.972708418564533, | |
| "learning_rate": 1.6551753150462258e-05, | |
| "loss": 0.1533, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 1.3414901525438894, | |
| "learning_rate": 1.6517086891659335e-05, | |
| "loss": 0.0973, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.546218487394958, | |
| "grad_norm": 2.260816117595264, | |
| "learning_rate": 1.6482283953077887e-05, | |
| "loss": 0.1141, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 3.5630252100840334, | |
| "grad_norm": 1.3728281964965443, | |
| "learning_rate": 1.644734506462347e-05, | |
| "loss": 0.1069, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.5798319327731094, | |
| "grad_norm": 1.647487452421848, | |
| "learning_rate": 1.641227095905286e-05, | |
| "loss": 0.1317, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 3.596638655462185, | |
| "grad_norm": 1.555286447904776, | |
| "learning_rate": 1.637706237195867e-05, | |
| "loss": 0.0997, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 3.6134453781512605, | |
| "grad_norm": 1.4295747968285732, | |
| "learning_rate": 1.6341720041753924e-05, | |
| "loss": 0.1116, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.630252100840336, | |
| "grad_norm": 1.4982199659580737, | |
| "learning_rate": 1.6306244709656597e-05, | |
| "loss": 0.1014, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 3.6470588235294117, | |
| "grad_norm": 3.3567838619912513, | |
| "learning_rate": 1.6270637119674023e-05, | |
| "loss": 0.1758, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 3.6638655462184873, | |
| "grad_norm": 1.549085403651707, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 0.0874, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 3.6806722689075633, | |
| "grad_norm": 2.1336886853349295, | |
| "learning_rate": 1.6199028155935793e-05, | |
| "loss": 0.1467, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 3.697478991596639, | |
| "grad_norm": 2.306782696878204, | |
| "learning_rate": 1.6163028284001034e-05, | |
| "loss": 0.1186, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 1.520623845541843, | |
| "learning_rate": 1.612689915779134e-05, | |
| "loss": 0.0923, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 3.73109243697479, | |
| "grad_norm": 1.147756639869472, | |
| "learning_rate": 1.6090641535025773e-05, | |
| "loss": 0.0854, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 3.7478991596638656, | |
| "grad_norm": 1.2212881766323158, | |
| "learning_rate": 1.605425617611829e-05, | |
| "loss": 0.0868, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 1.374752115412969, | |
| "learning_rate": 1.6017743844161802e-05, | |
| "loss": 0.0857, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 3.7815126050420167, | |
| "grad_norm": 1.0655866618063339, | |
| "learning_rate": 1.598110530491216e-05, | |
| "loss": 0.0915, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.7983193277310923, | |
| "grad_norm": 2.573005569504097, | |
| "learning_rate": 1.5944341326772112e-05, | |
| "loss": 0.0879, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 3.815126050420168, | |
| "grad_norm": 7.560760061208962, | |
| "learning_rate": 1.5907452680775164e-05, | |
| "loss": 0.2092, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 3.831932773109244, | |
| "grad_norm": 1.9398630844655902, | |
| "learning_rate": 1.587044014056943e-05, | |
| "loss": 0.1255, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 3.8487394957983194, | |
| "grad_norm": 1.6555395105411446, | |
| "learning_rate": 1.583330448240139e-05, | |
| "loss": 0.104, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 3.865546218487395, | |
| "grad_norm": 2.7598515858574775, | |
| "learning_rate": 1.5796046485099633e-05, | |
| "loss": 0.1696, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.8823529411764706, | |
| "grad_norm": 2.3775725159616377, | |
| "learning_rate": 1.57586669300585e-05, | |
| "loss": 0.1144, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 3.899159663865546, | |
| "grad_norm": 7.982198513719052, | |
| "learning_rate": 1.5721166601221697e-05, | |
| "loss": 0.1838, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 3.9159663865546217, | |
| "grad_norm": 2.9062227838430847, | |
| "learning_rate": 1.5683546285065878e-05, | |
| "loss": 0.1136, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 3.9327731092436977, | |
| "grad_norm": 2.4135228480944035, | |
| "learning_rate": 1.5645806770584122e-05, | |
| "loss": 0.0948, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.9495798319327733, | |
| "grad_norm": 3.0055227991300333, | |
| "learning_rate": 1.5607948849269404e-05, | |
| "loss": 0.1268, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.966386554621849, | |
| "grad_norm": 5.10158046985291, | |
| "learning_rate": 1.5569973315097985e-05, | |
| "loss": 0.152, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.9831932773109244, | |
| "grad_norm": 2.436291138010729, | |
| "learning_rate": 1.5531880964512773e-05, | |
| "loss": 0.1158, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 4.281661543875596, | |
| "learning_rate": 1.54936725964066e-05, | |
| "loss": 0.1045, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.12479228526353836, | |
| "eval_runtime": 94.0271, | |
| "eval_samples_per_second": 1.0, | |
| "eval_steps_per_second": 1.0, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 4.016806722689076, | |
| "grad_norm": 1.912176029492126, | |
| "learning_rate": 1.5455349012105488e-05, | |
| "loss": 0.135, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 4.033613445378151, | |
| "grad_norm": 2.215120261832349, | |
| "learning_rate": 1.5416911015351827e-05, | |
| "loss": 0.127, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.050420168067227, | |
| "grad_norm": 1.52744144609313, | |
| "learning_rate": 1.5378359412287537e-05, | |
| "loss": 0.1167, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 4.067226890756302, | |
| "grad_norm": 1.9530294241000907, | |
| "learning_rate": 1.5339695011437128e-05, | |
| "loss": 0.1071, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 4.084033613445378, | |
| "grad_norm": 2.215735360838886, | |
| "learning_rate": 1.530091862369078e-05, | |
| "loss": 0.1498, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 4.100840336134453, | |
| "grad_norm": 2.2899623389238246, | |
| "learning_rate": 1.526203106228733e-05, | |
| "loss": 0.1452, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 4.117647058823529, | |
| "grad_norm": 2.366322124903928, | |
| "learning_rate": 1.5223033142797183e-05, | |
| "loss": 0.0997, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.1344537815126055, | |
| "grad_norm": 2.180155804240841, | |
| "learning_rate": 1.5183925683105254e-05, | |
| "loss": 0.1103, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 4.151260504201681, | |
| "grad_norm": 1.9915581272469869, | |
| "learning_rate": 1.5144709503393773e-05, | |
| "loss": 0.1178, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 4.168067226890757, | |
| "grad_norm": 2.0239687647120905, | |
| "learning_rate": 1.5105385426125123e-05, | |
| "loss": 0.0996, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 4.184873949579832, | |
| "grad_norm": 1.86641722627037, | |
| "learning_rate": 1.5065954276024561e-05, | |
| "loss": 0.1104, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 4.201680672268908, | |
| "grad_norm": 1.6792110703483962, | |
| "learning_rate": 1.5026416880062932e-05, | |
| "loss": 0.1237, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.218487394957983, | |
| "grad_norm": 1.2608521044993408, | |
| "learning_rate": 1.4986774067439327e-05, | |
| "loss": 0.1011, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 2.5144065390038617, | |
| "learning_rate": 1.4947026669563687e-05, | |
| "loss": 0.1515, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.2521008403361344, | |
| "grad_norm": 1.5804219840273723, | |
| "learning_rate": 1.4907175520039381e-05, | |
| "loss": 0.1163, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 4.26890756302521, | |
| "grad_norm": 2.2244097182086273, | |
| "learning_rate": 1.4867221454645696e-05, | |
| "loss": 0.1188, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 1.64104407678126, | |
| "learning_rate": 1.482716531132034e-05, | |
| "loss": 0.0965, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.302521008403361, | |
| "grad_norm": 0.8551932780454022, | |
| "learning_rate": 1.4787007930141841e-05, | |
| "loss": 0.0655, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 4.319327731092437, | |
| "grad_norm": 1.752648892273715, | |
| "learning_rate": 1.4746750153311951e-05, | |
| "loss": 0.0946, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 4.336134453781512, | |
| "grad_norm": 5.829984337336428, | |
| "learning_rate": 1.4706392825137962e-05, | |
| "loss": 0.1826, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 4.352941176470588, | |
| "grad_norm": 2.302171396641537, | |
| "learning_rate": 1.4665936792015021e-05, | |
| "loss": 0.0909, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 4.369747899159664, | |
| "grad_norm": 1.2086828157662006, | |
| "learning_rate": 1.4625382902408356e-05, | |
| "loss": 0.0972, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.38655462184874, | |
| "grad_norm": 1.3490901275333183, | |
| "learning_rate": 1.4584732006835495e-05, | |
| "loss": 0.0823, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 4.4033613445378155, | |
| "grad_norm": 3.3951376429875393, | |
| "learning_rate": 1.4543984957848438e-05, | |
| "loss": 0.1492, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 4.420168067226891, | |
| "grad_norm": 2.4333139248557165, | |
| "learning_rate": 1.4503142610015751e-05, | |
| "loss": 0.0974, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 4.436974789915967, | |
| "grad_norm": 1.818247800901397, | |
| "learning_rate": 1.4462205819904658e-05, | |
| "loss": 0.0813, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 4.453781512605042, | |
| "grad_norm": 2.761939597967942, | |
| "learning_rate": 1.4421175446063086e-05, | |
| "loss": 0.1356, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 2.3235891826621895, | |
| "learning_rate": 1.4380052349001647e-05, | |
| "loss": 0.1393, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 4.487394957983193, | |
| "grad_norm": 1.4778394471726972, | |
| "learning_rate": 1.4338837391175582e-05, | |
| "loss": 0.1124, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 4.504201680672269, | |
| "grad_norm": 1.6212161850248412, | |
| "learning_rate": 1.42975314369667e-05, | |
| "loss": 0.0834, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 4.5210084033613445, | |
| "grad_norm": 1.513176697844455, | |
| "learning_rate": 1.4256135352665217e-05, | |
| "loss": 0.1053, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 4.53781512605042, | |
| "grad_norm": 1.5842489277719052, | |
| "learning_rate": 1.4214650006451622e-05, | |
| "loss": 0.0793, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.554621848739496, | |
| "grad_norm": 2.882811290847085, | |
| "learning_rate": 1.4173076268378443e-05, | |
| "loss": 0.0944, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 2.540707077443368, | |
| "learning_rate": 1.4131415010352007e-05, | |
| "loss": 0.0929, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 4.588235294117647, | |
| "grad_norm": 2.224835930797884, | |
| "learning_rate": 1.408966710611416e-05, | |
| "loss": 0.0965, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 4.605042016806722, | |
| "grad_norm": 1.603233390085239, | |
| "learning_rate": 1.4047833431223938e-05, | |
| "loss": 0.1082, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 4.621848739495798, | |
| "grad_norm": 1.7397244673026468, | |
| "learning_rate": 1.4005914863039203e-05, | |
| "loss": 0.0765, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.6386554621848735, | |
| "grad_norm": 1.9121905502244794, | |
| "learning_rate": 1.3963912280698238e-05, | |
| "loss": 0.0852, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 4.65546218487395, | |
| "grad_norm": 1.8233199956684913, | |
| "learning_rate": 1.3921826565101325e-05, | |
| "loss": 0.104, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 4.6722689075630255, | |
| "grad_norm": 0.7406392209229766, | |
| "learning_rate": 1.3879658598892254e-05, | |
| "loss": 0.0597, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 4.689075630252101, | |
| "grad_norm": 7.419162565626874, | |
| "learning_rate": 1.3837409266439818e-05, | |
| "loss": 0.209, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 1.1738246179937548, | |
| "learning_rate": 1.3795079453819276e-05, | |
| "loss": 0.0996, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.722689075630252, | |
| "grad_norm": 0.7316310386069824, | |
| "learning_rate": 1.3752670048793744e-05, | |
| "loss": 0.0573, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 4.739495798319328, | |
| "grad_norm": 1.0245543216679647, | |
| "learning_rate": 1.37101819407956e-05, | |
| "loss": 0.0798, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 4.756302521008403, | |
| "grad_norm": 1.2604913385475165, | |
| "learning_rate": 1.366761602090782e-05, | |
| "loss": 0.082, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 4.773109243697479, | |
| "grad_norm": 1.6057488873958579, | |
| "learning_rate": 1.3624973181845302e-05, | |
| "loss": 0.0765, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 4.7899159663865545, | |
| "grad_norm": 2.0379166749602895, | |
| "learning_rate": 1.3582254317936117e-05, | |
| "loss": 0.1001, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 4.80672268907563, | |
| "grad_norm": 1.7603012476865616, | |
| "learning_rate": 1.3539460325102779e-05, | |
| "loss": 0.1126, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 4.823529411764706, | |
| "grad_norm": 1.4253082853875376, | |
| "learning_rate": 1.349659210084344e-05, | |
| "loss": 0.0978, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 4.840336134453781, | |
| "grad_norm": 1.1284525438619117, | |
| "learning_rate": 1.3453650544213078e-05, | |
| "loss": 0.0695, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 0.9802953497383712, | |
| "learning_rate": 1.3410636555804634e-05, | |
| "loss": 0.0751, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 4.873949579831933, | |
| "grad_norm": 1.3612858881668406, | |
| "learning_rate": 1.3367551037730129e-05, | |
| "loss": 0.0966, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.890756302521009, | |
| "grad_norm": 1.0157890782271555, | |
| "learning_rate": 1.3324394893601734e-05, | |
| "loss": 0.059, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 4.907563025210084, | |
| "grad_norm": 1.4054910887816698, | |
| "learning_rate": 1.3281169028512838e-05, | |
| "loss": 0.0709, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 4.92436974789916, | |
| "grad_norm": 0.8662398448794504, | |
| "learning_rate": 1.3237874349019041e-05, | |
| "loss": 0.0756, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 1.3483246839258962, | |
| "learning_rate": 1.319451176311917e-05, | |
| "loss": 0.0769, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 4.957983193277311, | |
| "grad_norm": 2.4153558320453614, | |
| "learning_rate": 1.315108218023621e-05, | |
| "loss": 0.1131, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 4.974789915966387, | |
| "grad_norm": 2.555834294764848, | |
| "learning_rate": 1.3107586511198243e-05, | |
| "loss": 0.141, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 4.991596638655462, | |
| "grad_norm": 4.147522093497237, | |
| "learning_rate": 1.306402566821935e-05, | |
| "loss": 0.1456, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 4.991596638655462, | |
| "eval_loss": 0.1449918895959854, | |
| "eval_runtime": 97.1487, | |
| "eval_samples_per_second": 0.968, | |
| "eval_steps_per_second": 0.968, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 5.008403361344538, | |
| "grad_norm": 1.6155708244562734, | |
| "learning_rate": 1.302040056488047e-05, | |
| "loss": 0.0717, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 5.025210084033613, | |
| "grad_norm": 1.144238074710071, | |
| "learning_rate": 1.297671211611025e-05, | |
| "loss": 0.0719, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 5.042016806722689, | |
| "grad_norm": 2.0576085887744795, | |
| "learning_rate": 1.2932961238165837e-05, | |
| "loss": 0.1152, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.0588235294117645, | |
| "grad_norm": 1.0604218669017003, | |
| "learning_rate": 1.2889148848613695e-05, | |
| "loss": 0.0712, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 5.07563025210084, | |
| "grad_norm": 1.1699347626659127, | |
| "learning_rate": 1.2845275866310325e-05, | |
| "loss": 0.061, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 5.092436974789916, | |
| "grad_norm": 1.1371190586967697, | |
| "learning_rate": 1.2801343211383021e-05, | |
| "loss": 0.0747, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 5.109243697478991, | |
| "grad_norm": 1.8857383359198918, | |
| "learning_rate": 1.2757351805210557e-05, | |
| "loss": 0.1447, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 5.126050420168067, | |
| "grad_norm": 1.0626942422062413, | |
| "learning_rate": 1.2713302570403872e-05, | |
| "loss": 0.0657, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 1.3367342967159013, | |
| "learning_rate": 1.2669196430786715e-05, | |
| "loss": 0.086, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 5.159663865546219, | |
| "grad_norm": 4.036414956554653, | |
| "learning_rate": 1.2625034311376276e-05, | |
| "loss": 0.1317, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 2.092631890687451, | |
| "learning_rate": 1.258081713836378e-05, | |
| "loss": 0.0958, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 5.19327731092437, | |
| "grad_norm": 2.470224875937704, | |
| "learning_rate": 1.2536545839095074e-05, | |
| "loss": 0.0764, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 5.2100840336134455, | |
| "grad_norm": 2.6817499179267483, | |
| "learning_rate": 1.2492221342051153e-05, | |
| "loss": 0.1124, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.226890756302521, | |
| "grad_norm": 2.000873702339644, | |
| "learning_rate": 1.2447844576828719e-05, | |
| "loss": 0.1067, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 5.243697478991597, | |
| "grad_norm": 2.039025163817311, | |
| "learning_rate": 1.2403416474120657e-05, | |
| "loss": 0.0703, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 5.260504201680672, | |
| "grad_norm": 1.2573161613968584, | |
| "learning_rate": 1.2358937965696538e-05, | |
| "loss": 0.0694, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 5.277310924369748, | |
| "grad_norm": 1.6533379980735152, | |
| "learning_rate": 1.2314409984383066e-05, | |
| "loss": 0.0828, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 5.294117647058823, | |
| "grad_norm": 2.0132269342271467, | |
| "learning_rate": 1.2269833464044514e-05, | |
| "loss": 0.064, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 5.310924369747899, | |
| "grad_norm": 1.6870312479105358, | |
| "learning_rate": 1.2225209339563144e-05, | |
| "loss": 0.1093, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 5.3277310924369745, | |
| "grad_norm": 1.2264520173137188, | |
| "learning_rate": 1.2180538546819595e-05, | |
| "loss": 0.0844, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 5.34453781512605, | |
| "grad_norm": 1.6926310927686044, | |
| "learning_rate": 1.2135822022673263e-05, | |
| "loss": 0.0663, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 5.361344537815126, | |
| "grad_norm": 4.790942950250594, | |
| "learning_rate": 1.2091060704942636e-05, | |
| "loss": 0.1585, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 5.378151260504202, | |
| "grad_norm": 1.0161253142944433, | |
| "learning_rate": 1.204625553238565e-05, | |
| "loss": 0.0723, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.394957983193278, | |
| "grad_norm": 1.1558751249048937, | |
| "learning_rate": 1.200140744467997e-05, | |
| "loss": 0.084, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 0.9419572615825091, | |
| "learning_rate": 1.195651738240332e-05, | |
| "loss": 0.0571, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "grad_norm": 1.622044146409352, | |
| "learning_rate": 1.1911586287013726e-05, | |
| "loss": 0.0698, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 5.445378151260504, | |
| "grad_norm": 1.7252519658503152, | |
| "learning_rate": 1.1866615100829777e-05, | |
| "loss": 0.0727, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 5.46218487394958, | |
| "grad_norm": 1.2769460474277, | |
| "learning_rate": 1.1821604767010883e-05, | |
| "loss": 0.0728, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 5.4789915966386555, | |
| "grad_norm": 1.6489519703945064, | |
| "learning_rate": 1.1776556229537461e-05, | |
| "loss": 0.0712, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 5.495798319327731, | |
| "grad_norm": 1.9144641595351362, | |
| "learning_rate": 1.1731470433191173e-05, | |
| "loss": 0.0932, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 5.512605042016807, | |
| "grad_norm": 1.0334398447359328, | |
| "learning_rate": 1.1686348323535078e-05, | |
| "loss": 0.0391, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 5.529411764705882, | |
| "grad_norm": 2.4736197087343195, | |
| "learning_rate": 1.1641190846893824e-05, | |
| "loss": 0.0955, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 5.546218487394958, | |
| "grad_norm": 1.3189455008208653, | |
| "learning_rate": 1.1595998950333794e-05, | |
| "loss": 0.0565, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.563025210084033, | |
| "grad_norm": 1.7006720766337666, | |
| "learning_rate": 1.1550773581643245e-05, | |
| "loss": 0.0786, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 5.579831932773109, | |
| "grad_norm": 4.52735397655905, | |
| "learning_rate": 1.1505515689312424e-05, | |
| "loss": 0.1432, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 5.5966386554621845, | |
| "grad_norm": 3.0704885963032953, | |
| "learning_rate": 1.1460226222513682e-05, | |
| "loss": 0.1099, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 5.61344537815126, | |
| "grad_norm": 4.068391506810968, | |
| "learning_rate": 1.1414906131081575e-05, | |
| "loss": 0.1559, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 5.630252100840336, | |
| "grad_norm": 1.6178193211411984, | |
| "learning_rate": 1.1369556365492924e-05, | |
| "loss": 0.0881, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 1.472096896311054, | |
| "learning_rate": 1.1324177876846897e-05, | |
| "loss": 0.0668, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 5.663865546218488, | |
| "grad_norm": 0.9377593241374005, | |
| "learning_rate": 1.1278771616845061e-05, | |
| "loss": 0.0528, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 5.680672268907563, | |
| "grad_norm": 1.60191420846021, | |
| "learning_rate": 1.1233338537771408e-05, | |
| "loss": 0.0839, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 5.697478991596639, | |
| "grad_norm": 2.4242623809016757, | |
| "learning_rate": 1.1187879592472402e-05, | |
| "loss": 0.0765, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 3.3339459249935075, | |
| "learning_rate": 1.1142395734336986e-05, | |
| "loss": 0.1304, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.73109243697479, | |
| "grad_norm": 2.7416841232755824, | |
| "learning_rate": 1.1096887917276585e-05, | |
| "loss": 0.0776, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 5.7478991596638656, | |
| "grad_norm": 1.8254292684914686, | |
| "learning_rate": 1.1051357095705102e-05, | |
| "loss": 0.0679, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 5.764705882352941, | |
| "grad_norm": 1.807300374973592, | |
| "learning_rate": 1.1005804224518912e-05, | |
| "loss": 0.0769, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 5.781512605042017, | |
| "grad_norm": 1.8476121472163518, | |
| "learning_rate": 1.0960230259076819e-05, | |
| "loss": 0.0777, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 5.798319327731092, | |
| "grad_norm": 2.2451749634606317, | |
| "learning_rate": 1.0914636155180025e-05, | |
| "loss": 0.0843, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 5.815126050420168, | |
| "grad_norm": 2.016326952399232, | |
| "learning_rate": 1.0869022869052091e-05, | |
| "loss": 0.0852, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 5.831932773109243, | |
| "grad_norm": 1.502892225736369, | |
| "learning_rate": 1.0823391357318876e-05, | |
| "loss": 0.0637, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 5.848739495798319, | |
| "grad_norm": 2.2571728633731376, | |
| "learning_rate": 1.0777742576988474e-05, | |
| "loss": 0.0667, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 5.865546218487395, | |
| "grad_norm": 2.3431094429758264, | |
| "learning_rate": 1.0732077485431152e-05, | |
| "loss": 0.0984, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 1.5381903109004802, | |
| "learning_rate": 1.0686397040359253e-05, | |
| "loss": 0.0712, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.899159663865547, | |
| "grad_norm": 1.36348896527395, | |
| "learning_rate": 1.064070219980713e-05, | |
| "loss": 0.0736, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 5.915966386554622, | |
| "grad_norm": 2.8850647864954384, | |
| "learning_rate": 1.059499392211105e-05, | |
| "loss": 0.1959, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 5.932773109243698, | |
| "grad_norm": 2.2743426048039814, | |
| "learning_rate": 1.0549273165889079e-05, | |
| "loss": 0.0661, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 5.949579831932773, | |
| "grad_norm": 2.1602273939968106, | |
| "learning_rate": 1.0503540890020997e-05, | |
| "loss": 0.0978, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 5.966386554621849, | |
| "grad_norm": 1.6282000235425853, | |
| "learning_rate": 1.0457798053628181e-05, | |
| "loss": 0.0712, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 5.983193277310924, | |
| "grad_norm": 2.2108120483379072, | |
| "learning_rate": 1.0412045616053486e-05, | |
| "loss": 0.0959, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 6.96305244714079, | |
| "learning_rate": 1.0366284536841124e-05, | |
| "loss": 0.1132, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.1090986505150795, | |
| "eval_runtime": 95.9195, | |
| "eval_samples_per_second": 0.98, | |
| "eval_steps_per_second": 0.98, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 6.016806722689076, | |
| "grad_norm": 1.4364004042145684, | |
| "learning_rate": 1.0320515775716556e-05, | |
| "loss": 0.0628, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 6.033613445378151, | |
| "grad_norm": 1.9184329896408099, | |
| "learning_rate": 1.0274740292566335e-05, | |
| "loss": 0.085, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 6.050420168067227, | |
| "grad_norm": 2.722228602616158, | |
| "learning_rate": 1.0228959047418005e-05, | |
| "loss": 0.0764, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.067226890756302, | |
| "grad_norm": 1.9487119713360619, | |
| "learning_rate": 1.0183173000419954e-05, | |
| "loss": 0.0735, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 6.084033613445378, | |
| "grad_norm": 2.4618304983550883, | |
| "learning_rate": 1.0137383111821267e-05, | |
| "loss": 0.0742, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 6.100840336134453, | |
| "grad_norm": 3.82982664079687, | |
| "learning_rate": 1.009159034195161e-05, | |
| "loss": 0.1549, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 2.5393832652005686, | |
| "learning_rate": 1.0045795651201062e-05, | |
| "loss": 0.072, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 6.1344537815126055, | |
| "grad_norm": 2.1988376418173305, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0786, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 6.151260504201681, | |
| "grad_norm": 3.668071708848647, | |
| "learning_rate": 9.954204348798938e-06, | |
| "loss": 0.0928, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 6.168067226890757, | |
| "grad_norm": 2.378790242128915, | |
| "learning_rate": 9.908409658048395e-06, | |
| "loss": 0.0717, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 6.184873949579832, | |
| "grad_norm": 1.8836633579149347, | |
| "learning_rate": 9.862616888178733e-06, | |
| "loss": 0.0519, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 6.201680672268908, | |
| "grad_norm": 2.181673653848983, | |
| "learning_rate": 9.816826999580049e-06, | |
| "loss": 0.0748, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 6.218487394957983, | |
| "grad_norm": 2.544264789630926, | |
| "learning_rate": 9.771040952581998e-06, | |
| "loss": 0.1029, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.235294117647059, | |
| "grad_norm": 2.673802179214641, | |
| "learning_rate": 9.72525970743367e-06, | |
| "loss": 0.0857, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 6.2521008403361344, | |
| "grad_norm": 2.601200264033827, | |
| "learning_rate": 9.67948422428345e-06, | |
| "loss": 0.0886, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 6.26890756302521, | |
| "grad_norm": 1.9627943528177962, | |
| "learning_rate": 9.633715463158881e-06, | |
| "loss": 0.1016, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 1.0216877914513534, | |
| "learning_rate": 9.587954383946518e-06, | |
| "loss": 0.0672, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 6.302521008403361, | |
| "grad_norm": 1.8373707532976224, | |
| "learning_rate": 9.542201946371819e-06, | |
| "loss": 0.0754, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 6.319327731092437, | |
| "grad_norm": 0.9199770052216976, | |
| "learning_rate": 9.496459109979004e-06, | |
| "loss": 0.0655, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 6.336134453781512, | |
| "grad_norm": 1.9486911830327862, | |
| "learning_rate": 9.450726834110923e-06, | |
| "loss": 0.0682, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 1.4805691280913602, | |
| "learning_rate": 9.405006077888954e-06, | |
| "loss": 0.0487, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 6.369747899159664, | |
| "grad_norm": 1.4042709384224348, | |
| "learning_rate": 9.359297800192873e-06, | |
| "loss": 0.0689, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 6.38655462184874, | |
| "grad_norm": 3.6158792324022775, | |
| "learning_rate": 9.313602959640754e-06, | |
| "loss": 0.1121, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.4033613445378155, | |
| "grad_norm": 1.5349914188342018, | |
| "learning_rate": 9.267922514568853e-06, | |
| "loss": 0.0548, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 6.420168067226891, | |
| "grad_norm": 5.196823890519191, | |
| "learning_rate": 9.22225742301153e-06, | |
| "loss": 0.1468, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 6.436974789915967, | |
| "grad_norm": 2.4164444219519816, | |
| "learning_rate": 9.176608642681127e-06, | |
| "loss": 0.0689, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 6.453781512605042, | |
| "grad_norm": 1.0767734429156512, | |
| "learning_rate": 9.13097713094791e-06, | |
| "loss": 0.0537, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 6.470588235294118, | |
| "grad_norm": 1.1364625205117547, | |
| "learning_rate": 9.085363844819979e-06, | |
| "loss": 0.0505, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 6.487394957983193, | |
| "grad_norm": 1.2535232324818428, | |
| "learning_rate": 9.039769740923183e-06, | |
| "loss": 0.0518, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 6.504201680672269, | |
| "grad_norm": 3.057282990063465, | |
| "learning_rate": 8.99419577548109e-06, | |
| "loss": 0.1124, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 6.5210084033613445, | |
| "grad_norm": 1.9870672030743288, | |
| "learning_rate": 8.948642904294901e-06, | |
| "loss": 0.0534, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 6.53781512605042, | |
| "grad_norm": 4.506177944111496, | |
| "learning_rate": 8.90311208272342e-06, | |
| "loss": 0.1007, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 6.554621848739496, | |
| "grad_norm": 4.687056129955279, | |
| "learning_rate": 8.857604265663016e-06, | |
| "loss": 0.1015, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.571428571428571, | |
| "grad_norm": 2.793771357084546, | |
| "learning_rate": 8.812120407527603e-06, | |
| "loss": 0.0791, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 1.5631834583542967, | |
| "learning_rate": 8.766661462228593e-06, | |
| "loss": 0.0583, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 6.605042016806722, | |
| "grad_norm": 2.503384804703961, | |
| "learning_rate": 8.721228383154939e-06, | |
| "loss": 0.0819, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 6.621848739495798, | |
| "grad_norm": 1.298763753508359, | |
| "learning_rate": 8.675822123153104e-06, | |
| "loss": 0.0452, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 6.6386554621848735, | |
| "grad_norm": 3.468345102178359, | |
| "learning_rate": 8.630443634507077e-06, | |
| "loss": 0.0893, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 6.65546218487395, | |
| "grad_norm": 2.526122457983386, | |
| "learning_rate": 8.585093868918426e-06, | |
| "loss": 0.1011, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 6.6722689075630255, | |
| "grad_norm": 2.075764833700991, | |
| "learning_rate": 8.539773777486321e-06, | |
| "loss": 0.0752, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 6.689075630252101, | |
| "grad_norm": 1.3259731049756247, | |
| "learning_rate": 8.494484310687581e-06, | |
| "loss": 0.0536, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 6.705882352941177, | |
| "grad_norm": 1.5262707963177293, | |
| "learning_rate": 8.44922641835676e-06, | |
| "loss": 0.0649, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 6.722689075630252, | |
| "grad_norm": 1.3923407644979977, | |
| "learning_rate": 8.404001049666211e-06, | |
| "loss": 0.0781, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.739495798319328, | |
| "grad_norm": 1.474705597824637, | |
| "learning_rate": 8.35880915310618e-06, | |
| "loss": 0.0614, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 6.756302521008403, | |
| "grad_norm": 1.5133797172950414, | |
| "learning_rate": 8.313651676464924e-06, | |
| "loss": 0.0588, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 6.773109243697479, | |
| "grad_norm": 1.5937067083804244, | |
| "learning_rate": 8.26852956680883e-06, | |
| "loss": 0.0658, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 6.7899159663865545, | |
| "grad_norm": 3.2146725324389447, | |
| "learning_rate": 8.223443770462539e-06, | |
| "loss": 0.0967, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 6.80672268907563, | |
| "grad_norm": 1.6054189964924985, | |
| "learning_rate": 8.17839523298912e-06, | |
| "loss": 0.0676, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 1.3282376180631894, | |
| "learning_rate": 8.133384899170224e-06, | |
| "loss": 0.0373, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 6.840336134453781, | |
| "grad_norm": 1.1953714752813585, | |
| "learning_rate": 8.08841371298628e-06, | |
| "loss": 0.0591, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 2.9661126814818393, | |
| "learning_rate": 8.043482617596681e-06, | |
| "loss": 0.141, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 6.873949579831933, | |
| "grad_norm": 1.3266256289537959, | |
| "learning_rate": 7.99859255532003e-06, | |
| "loss": 0.053, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 6.890756302521009, | |
| "grad_norm": 2.1914466357984574, | |
| "learning_rate": 7.953744467614356e-06, | |
| "loss": 0.0778, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.907563025210084, | |
| "grad_norm": 4.698036043159151, | |
| "learning_rate": 7.908939295057362e-06, | |
| "loss": 0.1673, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 6.92436974789916, | |
| "grad_norm": 2.0220996669989555, | |
| "learning_rate": 7.864177977326739e-06, | |
| "loss": 0.0706, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "grad_norm": 0.863174779051157, | |
| "learning_rate": 7.819461453180403e-06, | |
| "loss": 0.0536, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 6.957983193277311, | |
| "grad_norm": 1.162974910961646, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 0.053, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 6.974789915966387, | |
| "grad_norm": 1.5401943180276223, | |
| "learning_rate": 7.730166535955489e-06, | |
| "loss": 0.0562, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 6.991596638655462, | |
| "grad_norm": 1.0451220578083906, | |
| "learning_rate": 7.685590015616939e-06, | |
| "loss": 0.061, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 6.991596638655462, | |
| "eval_loss": 0.11021654307842255, | |
| "eval_runtime": 96.6092, | |
| "eval_samples_per_second": 0.973, | |
| "eval_steps_per_second": 0.973, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 7.008403361344538, | |
| "grad_norm": 1.9748689801680077, | |
| "learning_rate": 7.641062034303464e-06, | |
| "loss": 0.0744, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 7.025210084033613, | |
| "grad_norm": 1.407098109780849, | |
| "learning_rate": 7.596583525879344e-06, | |
| "loss": 0.0575, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 7.042016806722689, | |
| "grad_norm": 3.002224281961417, | |
| "learning_rate": 7.5521554231712845e-06, | |
| "loss": 0.0761, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 1.3836323412125109, | |
| "learning_rate": 7.507778657948847e-06, | |
| "loss": 0.0508, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 7.07563025210084, | |
| "grad_norm": 1.5627121625333806, | |
| "learning_rate": 7.463454160904928e-06, | |
| "loss": 0.0619, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 7.092436974789916, | |
| "grad_norm": 1.7887521098557067, | |
| "learning_rate": 7.419182861636218e-06, | |
| "loss": 0.0566, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 7.109243697478991, | |
| "grad_norm": 1.6937210033771022, | |
| "learning_rate": 7.374965688623726e-06, | |
| "loss": 0.0601, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 7.126050420168067, | |
| "grad_norm": 2.35276516070813, | |
| "learning_rate": 7.3308035692132896e-06, | |
| "loss": 0.0778, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 1.7638699741053976, | |
| "learning_rate": 7.286697429596135e-06, | |
| "loss": 0.0622, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.159663865546219, | |
| "grad_norm": 2.20424078040461, | |
| "learning_rate": 7.242648194789447e-06, | |
| "loss": 0.0597, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 7.176470588235294, | |
| "grad_norm": 2.50814340199316, | |
| "learning_rate": 7.19865678861698e-06, | |
| "loss": 0.0686, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 7.19327731092437, | |
| "grad_norm": 1.448580373256141, | |
| "learning_rate": 7.154724133689677e-06, | |
| "loss": 0.0312, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 7.2100840336134455, | |
| "grad_norm": 2.1552454847239075, | |
| "learning_rate": 7.110851151386306e-06, | |
| "loss": 0.0547, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 7.226890756302521, | |
| "grad_norm": 2.9974503334538602, | |
| "learning_rate": 7.067038761834164e-06, | |
| "loss": 0.0881, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.243697478991597, | |
| "grad_norm": 0.9735826198652626, | |
| "learning_rate": 7.023287883889753e-06, | |
| "loss": 0.039, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 7.260504201680672, | |
| "grad_norm": 2.5579376905692293, | |
| "learning_rate": 6.979599435119531e-06, | |
| "loss": 0.0919, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 7.277310924369748, | |
| "grad_norm": 2.4454794736196, | |
| "learning_rate": 6.935974331780653e-06, | |
| "loss": 0.0544, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 1.4023653427232834, | |
| "learning_rate": 6.892413488801762e-06, | |
| "loss": 0.0626, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 7.310924369747899, | |
| "grad_norm": 1.36640843494829, | |
| "learning_rate": 6.848917819763794e-06, | |
| "loss": 0.061, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 7.3277310924369745, | |
| "grad_norm": 1.6526923758851404, | |
| "learning_rate": 6.805488236880831e-06, | |
| "loss": 0.0628, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 7.34453781512605, | |
| "grad_norm": 1.5372290912683437, | |
| "learning_rate": 6.76212565098096e-06, | |
| "loss": 0.06, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 7.361344537815126, | |
| "grad_norm": 1.1019297711397662, | |
| "learning_rate": 6.718830971487165e-06, | |
| "loss": 0.0534, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 7.378151260504202, | |
| "grad_norm": 1.474792358247268, | |
| "learning_rate": 6.675605106398269e-06, | |
| "loss": 0.0498, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 7.394957983193278, | |
| "grad_norm": 1.5934869270544132, | |
| "learning_rate": 6.632448962269873e-06, | |
| "loss": 0.0579, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.411764705882353, | |
| "grad_norm": 1.5423848913860474, | |
| "learning_rate": 6.589363444195367e-06, | |
| "loss": 0.0836, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 0.6052937489588177, | |
| "learning_rate": 6.546349455786926e-06, | |
| "loss": 0.0235, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 7.445378151260504, | |
| "grad_norm": 0.8930183656715149, | |
| "learning_rate": 6.503407899156565e-06, | |
| "loss": 0.0536, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 7.46218487394958, | |
| "grad_norm": 1.9378732272470842, | |
| "learning_rate": 6.460539674897226e-06, | |
| "loss": 0.0713, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 7.4789915966386555, | |
| "grad_norm": 0.9653273021580584, | |
| "learning_rate": 6.417745682063884e-06, | |
| "loss": 0.0572, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 7.495798319327731, | |
| "grad_norm": 5.150704023745161, | |
| "learning_rate": 6.3750268181547e-06, | |
| "loss": 0.1011, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 7.512605042016807, | |
| "grad_norm": 2.663053864822168, | |
| "learning_rate": 6.3323839790921785e-06, | |
| "loss": 0.1071, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 1.8175181083869496, | |
| "learning_rate": 6.289818059204404e-06, | |
| "loss": 0.0409, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 7.546218487394958, | |
| "grad_norm": 1.2608121221023865, | |
| "learning_rate": 6.24732995120626e-06, | |
| "loss": 0.0447, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 7.563025210084033, | |
| "grad_norm": 1.558905828362155, | |
| "learning_rate": 6.204920546180728e-06, | |
| "loss": 0.0629, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.579831932773109, | |
| "grad_norm": 3.0836236832919774, | |
| "learning_rate": 6.162590733560183e-06, | |
| "loss": 0.0714, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 7.5966386554621845, | |
| "grad_norm": 1.9457063009832318, | |
| "learning_rate": 6.120341401107751e-06, | |
| "loss": 0.0591, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 7.61344537815126, | |
| "grad_norm": 2.8239440684779, | |
| "learning_rate": 6.078173434898679e-06, | |
| "loss": 0.0598, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 7.630252100840336, | |
| "grad_norm": 1.6266933180384318, | |
| "learning_rate": 6.036087719301763e-06, | |
| "loss": 0.0622, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 7.647058823529412, | |
| "grad_norm": 1.5079333156466859, | |
| "learning_rate": 5.994085136960801e-06, | |
| "loss": 0.0662, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 7.663865546218488, | |
| "grad_norm": 2.513503195610099, | |
| "learning_rate": 5.952166568776062e-06, | |
| "loss": 0.0741, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 7.680672268907563, | |
| "grad_norm": 2.942516455656699, | |
| "learning_rate": 5.910332893885842e-06, | |
| "loss": 0.0768, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 7.697478991596639, | |
| "grad_norm": 0.9029614933690709, | |
| "learning_rate": 5.868584989647994e-06, | |
| "loss": 0.0465, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 7.714285714285714, | |
| "grad_norm": 1.188640664205059, | |
| "learning_rate": 5.826923731621562e-06, | |
| "loss": 0.0507, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 7.73109243697479, | |
| "grad_norm": 1.2597887336311209, | |
| "learning_rate": 5.785349993548382e-06, | |
| "loss": 0.0442, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.7478991596638656, | |
| "grad_norm": 5.82455886129036, | |
| "learning_rate": 5.743864647334789e-06, | |
| "loss": 0.154, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 1.2355207857604258, | |
| "learning_rate": 5.702468563033307e-06, | |
| "loss": 0.0472, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 7.781512605042017, | |
| "grad_norm": 1.673091621193694, | |
| "learning_rate": 5.66116260882442e-06, | |
| "loss": 0.0714, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 7.798319327731092, | |
| "grad_norm": 1.2037358486316267, | |
| "learning_rate": 5.6199476509983546e-06, | |
| "loss": 0.0576, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 7.815126050420168, | |
| "grad_norm": 1.4923009661165723, | |
| "learning_rate": 5.5788245539369144e-06, | |
| "loss": 0.0546, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 7.831932773109243, | |
| "grad_norm": 7.004074578858643, | |
| "learning_rate": 5.537794180095341e-06, | |
| "loss": 0.0931, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 7.848739495798319, | |
| "grad_norm": 1.9350012591991381, | |
| "learning_rate": 5.496857389984251e-06, | |
| "loss": 0.0501, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 7.865546218487395, | |
| "grad_norm": 2.0402004080487575, | |
| "learning_rate": 5.456015042151563e-06, | |
| "loss": 0.0548, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 7.882352941176471, | |
| "grad_norm": 1.5295516734084738, | |
| "learning_rate": 5.415267993164504e-06, | |
| "loss": 0.0529, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 7.899159663865547, | |
| "grad_norm": 1.8934498235843238, | |
| "learning_rate": 5.37461709759165e-06, | |
| "loss": 0.0541, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 7.915966386554622, | |
| "grad_norm": 2.3477637506052935, | |
| "learning_rate": 5.334063207984983e-06, | |
| "loss": 0.091, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 7.932773109243698, | |
| "grad_norm": 2.068442722717763, | |
| "learning_rate": 5.2936071748620386e-06, | |
| "loss": 0.0659, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 7.949579831932773, | |
| "grad_norm": 1.5583233740963265, | |
| "learning_rate": 5.253249846688053e-06, | |
| "loss": 0.0582, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 7.966386554621849, | |
| "grad_norm": 1.3210668874752445, | |
| "learning_rate": 5.21299206985816e-06, | |
| "loss": 0.0457, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 7.983193277310924, | |
| "grad_norm": 1.8720045795583338, | |
| "learning_rate": 5.172834688679665e-06, | |
| "loss": 0.0485, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.717240919628548, | |
| "learning_rate": 5.132778545354305e-06, | |
| "loss": 0.0528, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.09206734597682953, | |
| "eval_runtime": 95.9403, | |
| "eval_samples_per_second": 0.98, | |
| "eval_steps_per_second": 0.98, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 8.016806722689076, | |
| "grad_norm": 1.5169536755225592, | |
| "learning_rate": 5.092824479960625e-06, | |
| "loss": 0.0474, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 8.033613445378151, | |
| "grad_norm": 1.7357799184112437, | |
| "learning_rate": 5.0529733304363145e-06, | |
| "loss": 0.0431, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 8.050420168067227, | |
| "grad_norm": 1.3870139040501723, | |
| "learning_rate": 5.013225932560679e-06, | |
| "loss": 0.0423, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 8.067226890756302, | |
| "grad_norm": 3.302734792491234, | |
| "learning_rate": 4.973583119937072e-06, | |
| "loss": 0.0655, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 8.084033613445378, | |
| "grad_norm": 1.8227280505702954, | |
| "learning_rate": 4.934045723975441e-06, | |
| "loss": 0.0731, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 8.100840336134453, | |
| "grad_norm": 1.3256962048722176, | |
| "learning_rate": 4.894614573874877e-06, | |
| "loss": 0.0432, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 8.117647058823529, | |
| "grad_norm": 1.427387472053757, | |
| "learning_rate": 4.85529049660623e-06, | |
| "loss": 0.0549, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 8.134453781512605, | |
| "grad_norm": 1.722178796219511, | |
| "learning_rate": 4.81607431689475e-06, | |
| "loss": 0.057, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 8.15126050420168, | |
| "grad_norm": 1.4946268224043258, | |
| "learning_rate": 4.776966857202816e-06, | |
| "loss": 0.0501, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 8.168067226890756, | |
| "grad_norm": 1.4410674537931873, | |
| "learning_rate": 4.737968937712674e-06, | |
| "loss": 0.0472, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 8.184873949579831, | |
| "grad_norm": 4.079218048537049, | |
| "learning_rate": 4.699081376309218e-06, | |
| "loss": 0.0669, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 8.201680672268907, | |
| "grad_norm": 2.445073900654683, | |
| "learning_rate": 4.660304988562877e-06, | |
| "loss": 0.0642, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 8.218487394957982, | |
| "grad_norm": 6.718089966341145, | |
| "learning_rate": 4.621640587712468e-06, | |
| "loss": 0.1271, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 2.8570358473448256, | |
| "learning_rate": 4.583088984648172e-06, | |
| "loss": 0.0629, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 8.252100840336134, | |
| "grad_norm": 1.9274312241325455, | |
| "learning_rate": 4.544650987894514e-06, | |
| "loss": 0.0367, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 8.268907563025211, | |
| "grad_norm": 1.7166820491309733, | |
| "learning_rate": 4.5063274035934016e-06, | |
| "loss": 0.0451, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "grad_norm": 1.3196462720335975, | |
| "learning_rate": 4.468119035487231e-06, | |
| "loss": 0.0451, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 8.302521008403362, | |
| "grad_norm": 1.5329980369426424, | |
| "learning_rate": 4.430026684902017e-06, | |
| "loss": 0.0382, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 8.319327731092438, | |
| "grad_norm": 1.7844728493839146, | |
| "learning_rate": 4.392051150730602e-06, | |
| "loss": 0.0346, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 8.336134453781513, | |
| "grad_norm": 1.233877897725148, | |
| "learning_rate": 4.354193229415882e-06, | |
| "loss": 0.0347, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 8.352941176470589, | |
| "grad_norm": 1.0911759571068154, | |
| "learning_rate": 4.3164537149341246e-06, | |
| "loss": 0.0534, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 8.369747899159664, | |
| "grad_norm": 2.1095342838945093, | |
| "learning_rate": 4.278833398778306e-06, | |
| "loss": 0.0531, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 8.38655462184874, | |
| "grad_norm": 2.3616676446482647, | |
| "learning_rate": 4.241333069941503e-06, | |
| "loss": 0.0569, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 8.403361344537815, | |
| "grad_norm": 1.3977212089055078, | |
| "learning_rate": 4.203953514900366e-06, | |
| "loss": 0.0353, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.420168067226891, | |
| "grad_norm": 2.6496004578015575, | |
| "learning_rate": 4.166695517598611e-06, | |
| "loss": 0.1042, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 8.436974789915967, | |
| "grad_norm": 2.06089201000074, | |
| "learning_rate": 4.129559859430573e-06, | |
| "loss": 0.0968, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 8.453781512605042, | |
| "grad_norm": 1.826445704568566, | |
| "learning_rate": 4.092547319224837e-06, | |
| "loss": 0.0516, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 3.8412502857797195, | |
| "learning_rate": 4.055658673227891e-06, | |
| "loss": 0.0884, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 8.487394957983193, | |
| "grad_norm": 0.9584236905314262, | |
| "learning_rate": 4.01889469508784e-06, | |
| "loss": 0.0249, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 8.504201680672269, | |
| "grad_norm": 3.291735033394008, | |
| "learning_rate": 3.982256155838199e-06, | |
| "loss": 0.0854, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 8.521008403361344, | |
| "grad_norm": 4.532957288732598, | |
| "learning_rate": 3.945743823881713e-06, | |
| "loss": 0.0629, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 8.53781512605042, | |
| "grad_norm": 4.273353415832438, | |
| "learning_rate": 3.909358464974228e-06, | |
| "loss": 0.0624, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 8.554621848739496, | |
| "grad_norm": 1.437659985078597, | |
| "learning_rate": 3.873100842208661e-06, | |
| "loss": 0.044, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 2.6929955709108584, | |
| "learning_rate": 3.836971715998968e-06, | |
| "loss": 0.0598, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 8.588235294117647, | |
| "grad_norm": 1.9024448417705355, | |
| "learning_rate": 3.8009718440642128e-06, | |
| "loss": 0.0603, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 8.605042016806722, | |
| "grad_norm": 0.7656090871294385, | |
| "learning_rate": 3.7651019814126656e-06, | |
| "loss": 0.0271, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 8.621848739495798, | |
| "grad_norm": 1.498792756768189, | |
| "learning_rate": 3.729362880325983e-06, | |
| "loss": 0.0406, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 8.638655462184873, | |
| "grad_norm": 1.914650360335185, | |
| "learning_rate": 3.693755290343409e-06, | |
| "loss": 0.0538, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 8.655462184873949, | |
| "grad_norm": 1.097214252517178, | |
| "learning_rate": 3.658279958246075e-06, | |
| "loss": 0.0381, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 8.672268907563025, | |
| "grad_norm": 2.861008119753485, | |
| "learning_rate": 3.622937628041334e-06, | |
| "loss": 0.0808, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 8.6890756302521, | |
| "grad_norm": 1.9341336610200206, | |
| "learning_rate": 3.587729040947141e-06, | |
| "loss": 0.0343, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 2.42481473153521, | |
| "learning_rate": 3.5526549353765294e-06, | |
| "loss": 0.1077, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 8.722689075630251, | |
| "grad_norm": 2.4065694655289973, | |
| "learning_rate": 3.5177160469221184e-06, | |
| "loss": 0.0641, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 8.739495798319329, | |
| "grad_norm": 1.8534338905199441, | |
| "learning_rate": 3.4829131083406684e-06, | |
| "loss": 0.0529, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.756302521008404, | |
| "grad_norm": 2.072202396225646, | |
| "learning_rate": 3.448246849537741e-06, | |
| "loss": 0.0565, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 8.77310924369748, | |
| "grad_norm": 1.5479608735352894, | |
| "learning_rate": 3.413717997552376e-06, | |
| "loss": 0.0411, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 8.789915966386555, | |
| "grad_norm": 2.6939787148898016, | |
| "learning_rate": 3.379327276541834e-06, | |
| "loss": 0.0722, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 8.806722689075631, | |
| "grad_norm": 1.5450816630775992, | |
| "learning_rate": 3.3450754077664337e-06, | |
| "loss": 0.0544, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 1.5918689109996174, | |
| "learning_rate": 3.310963109574408e-06, | |
| "loss": 0.0407, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 8.840336134453782, | |
| "grad_norm": 1.5593134309915293, | |
| "learning_rate": 3.2769910973868314e-06, | |
| "loss": 0.0418, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "grad_norm": 1.9817294913123196, | |
| "learning_rate": 3.243160083682645e-06, | |
| "loss": 0.0529, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 8.873949579831933, | |
| "grad_norm": 1.695431393346344, | |
| "learning_rate": 3.209470777983675e-06, | |
| "loss": 0.0488, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 8.890756302521009, | |
| "grad_norm": 1.8840574831179149, | |
| "learning_rate": 3.1759238868397925e-06, | |
| "loss": 0.0529, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 8.907563025210084, | |
| "grad_norm": 5.952373137687765, | |
| "learning_rate": 3.1425201138140592e-06, | |
| "loss": 0.0808, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 8.92436974789916, | |
| "grad_norm": 1.290713251381066, | |
| "learning_rate": 3.1092601594679993e-06, | |
| "loss": 0.0439, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.9948068099144867, | |
| "learning_rate": 3.0761447213468888e-06, | |
| "loss": 0.0405, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 8.957983193277311, | |
| "grad_norm": 1.9536076330602832, | |
| "learning_rate": 3.0431744939651365e-06, | |
| "loss": 0.0538, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 8.974789915966387, | |
| "grad_norm": 1.6523207102197872, | |
| "learning_rate": 3.0103501687917192e-06, | |
| "loss": 0.0544, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 8.991596638655462, | |
| "grad_norm": 0.9746183785405975, | |
| "learning_rate": 2.9776724342356654e-06, | |
| "loss": 0.0531, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 8.991596638655462, | |
| "eval_loss": 0.09362534433603287, | |
| "eval_runtime": 84.973, | |
| "eval_samples_per_second": 1.106, | |
| "eval_steps_per_second": 1.106, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 9.008403361344538, | |
| "grad_norm": 3.1103066899739282, | |
| "learning_rate": 2.945141975631637e-06, | |
| "loss": 0.0553, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 9.025210084033613, | |
| "grad_norm": 2.3282836310480435, | |
| "learning_rate": 2.912759475225546e-06, | |
| "loss": 0.0532, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 9.042016806722689, | |
| "grad_norm": 1.1197327667128945, | |
| "learning_rate": 2.88052561216024e-06, | |
| "loss": 0.0308, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 9.058823529411764, | |
| "grad_norm": 1.4098107198655643, | |
| "learning_rate": 2.8484410624612744e-06, | |
| "loss": 0.0364, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 9.07563025210084, | |
| "grad_norm": 4.438658787202018, | |
| "learning_rate": 2.8165064990227255e-06, | |
| "loss": 0.0561, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 9.092436974789916, | |
| "grad_norm": 1.9493994527119616, | |
| "learning_rate": 2.7847225915930697e-06, | |
| "loss": 0.0601, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 9.109243697478991, | |
| "grad_norm": 1.8527667273461481, | |
| "learning_rate": 2.7530900067611577e-06, | |
| "loss": 0.0531, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 9.126050420168067, | |
| "grad_norm": 1.701462207843824, | |
| "learning_rate": 2.7216094079422185e-06, | |
| "loss": 0.0472, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 1.8914078563116579, | |
| "learning_rate": 2.6902814553639443e-06, | |
| "loss": 0.0551, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 9.159663865546218, | |
| "grad_norm": 2.591885452858992, | |
| "learning_rate": 2.6591068060526626e-06, | |
| "loss": 0.0395, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 3.074064642620058, | |
| "learning_rate": 2.62808611381953e-06, | |
| "loss": 0.0619, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 9.193277310924369, | |
| "grad_norm": 1.6148671787143798, | |
| "learning_rate": 2.597220029246846e-06, | |
| "loss": 0.0369, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 9.210084033613445, | |
| "grad_norm": 5.145830985825016, | |
| "learning_rate": 2.5665091996743898e-06, | |
| "loss": 0.0912, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 9.22689075630252, | |
| "grad_norm": 1.1197523582174669, | |
| "learning_rate": 2.5359542691858542e-06, | |
| "loss": 0.0411, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 9.243697478991596, | |
| "grad_norm": 1.5754449262763028, | |
| "learning_rate": 2.5055558785953304e-06, | |
| "loss": 0.0502, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 9.260504201680673, | |
| "grad_norm": 2.843721070657444, | |
| "learning_rate": 2.4753146654338765e-06, | |
| "loss": 0.0835, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 9.277310924369749, | |
| "grad_norm": 2.2196607101882817, | |
| "learning_rate": 2.4452312639361462e-06, | |
| "loss": 0.0465, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 9.294117647058824, | |
| "grad_norm": 2.166755095388008, | |
| "learning_rate": 2.415306305027072e-06, | |
| "loss": 0.0421, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 9.3109243697479, | |
| "grad_norm": 3.8956621071776945, | |
| "learning_rate": 2.3855404163086558e-06, | |
| "loss": 0.0456, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 9.327731092436975, | |
| "grad_norm": 1.198392921249873, | |
| "learning_rate": 2.355934222046794e-06, | |
| "loss": 0.0368, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 9.344537815126051, | |
| "grad_norm": 2.081788407790618, | |
| "learning_rate": 2.32648834315818e-06, | |
| "loss": 0.0742, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 9.361344537815127, | |
| "grad_norm": 1.8547943707777044, | |
| "learning_rate": 2.2972033971972953e-06, | |
| "loss": 0.0404, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 9.378151260504202, | |
| "grad_norm": 2.06348184199221, | |
| "learning_rate": 2.2680799983434532e-06, | |
| "loss": 0.0362, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 9.394957983193278, | |
| "grad_norm": 1.6015036795043573, | |
| "learning_rate": 2.239118757387907e-06, | |
| "loss": 0.0461, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 2.500319456131113, | |
| "learning_rate": 2.2103202817210555e-06, | |
| "loss": 0.0436, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.428571428571429, | |
| "grad_norm": 2.5704589340698036, | |
| "learning_rate": 2.1816851753197023e-06, | |
| "loss": 0.0797, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 9.445378151260504, | |
| "grad_norm": 1.1623903030627807, | |
| "learning_rate": 2.1532140387343736e-06, | |
| "loss": 0.0434, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 9.46218487394958, | |
| "grad_norm": 1.7348909728003972, | |
| "learning_rate": 2.1249074690767434e-06, | |
| "loss": 0.0513, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 9.478991596638656, | |
| "grad_norm": 1.443247675887106, | |
| "learning_rate": 2.096766060007096e-06, | |
| "loss": 0.0436, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 9.495798319327731, | |
| "grad_norm": 1.719154980437653, | |
| "learning_rate": 2.068790401721886e-06, | |
| "loss": 0.0473, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 9.512605042016807, | |
| "grad_norm": 1.3479219484263167, | |
| "learning_rate": 2.040981080941349e-06, | |
| "loss": 0.0403, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 9.529411764705882, | |
| "grad_norm": 1.7448556207352623, | |
| "learning_rate": 2.013338680897209e-06, | |
| "loss": 0.0374, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 9.546218487394958, | |
| "grad_norm": 1.778610985288892, | |
| "learning_rate": 1.9858637813204352e-06, | |
| "loss": 0.0481, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 9.563025210084033, | |
| "grad_norm": 1.023079090275057, | |
| "learning_rate": 1.958556958429092e-06, | |
| "loss": 0.043, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 9.579831932773109, | |
| "grad_norm": 2.2504655920316927, | |
| "learning_rate": 1.9314187849162523e-06, | |
| "loss": 0.104, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 9.596638655462185, | |
| "grad_norm": 4.623708198206026, | |
| "learning_rate": 1.904449829937981e-06, | |
| "loss": 0.0878, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 9.61344537815126, | |
| "grad_norm": 1.1241851785983472, | |
| "learning_rate": 1.8776506591014054e-06, | |
| "loss": 0.0358, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 9.630252100840336, | |
| "grad_norm": 1.9153117494048388, | |
| "learning_rate": 1.851021834452853e-06, | |
| "loss": 0.0523, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 4.390748395794137, | |
| "learning_rate": 1.8245639144660532e-06, | |
| "loss": 0.0744, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 9.663865546218487, | |
| "grad_norm": 2.7441495241393783, | |
| "learning_rate": 1.7982774540304404e-06, | |
| "loss": 0.0492, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 9.680672268907562, | |
| "grad_norm": 1.3900528175419058, | |
| "learning_rate": 1.772163004439511e-06, | |
| "loss": 0.0479, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 9.697478991596638, | |
| "grad_norm": 2.77436372781934, | |
| "learning_rate": 1.7462211133792484e-06, | |
| "loss": 0.0621, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "grad_norm": 1.560534629765185, | |
| "learning_rate": 1.720452324916656e-06, | |
| "loss": 0.0357, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 9.731092436974789, | |
| "grad_norm": 2.4373786388347294, | |
| "learning_rate": 1.6948571794883406e-06, | |
| "loss": 0.0497, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 9.747899159663866, | |
| "grad_norm": 1.3388696128320787, | |
| "learning_rate": 1.6694362138891674e-06, | |
| "loss": 0.0409, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.764705882352942, | |
| "grad_norm": 2.602632435071536, | |
| "learning_rate": 1.6441899612610178e-06, | |
| "loss": 0.0557, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 9.781512605042018, | |
| "grad_norm": 2.241392316929053, | |
| "learning_rate": 1.6191189510815942e-06, | |
| "loss": 0.0852, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 9.798319327731093, | |
| "grad_norm": 1.2305602460104232, | |
| "learning_rate": 1.5942237091533297e-06, | |
| "loss": 0.0424, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 9.815126050420169, | |
| "grad_norm": 1.2682290196019605, | |
| "learning_rate": 1.5695047575923462e-06, | |
| "loss": 0.0497, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 9.831932773109244, | |
| "grad_norm": 3.1108653972627924, | |
| "learning_rate": 1.5449626148175144e-06, | |
| "loss": 0.0943, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 9.84873949579832, | |
| "grad_norm": 1.29875959685456, | |
| "learning_rate": 1.5205977955395812e-06, | |
| "loss": 0.0239, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 9.865546218487395, | |
| "grad_norm": 1.255438999538065, | |
| "learning_rate": 1.4964108107503638e-06, | |
| "loss": 0.0432, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 1.6849212493860068, | |
| "learning_rate": 1.4724021677120491e-06, | |
| "loss": 0.0498, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 9.899159663865547, | |
| "grad_norm": 1.543073795782045, | |
| "learning_rate": 1.4485723699465392e-06, | |
| "loss": 0.0436, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 9.915966386554622, | |
| "grad_norm": 1.6326111794747336, | |
| "learning_rate": 1.4249219172249051e-06, | |
| "loss": 0.0448, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 9.932773109243698, | |
| "grad_norm": 1.6795029810703082, | |
| "learning_rate": 1.4014513055568978e-06, | |
| "loss": 0.0483, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 9.949579831932773, | |
| "grad_norm": 5.509418938412195, | |
| "learning_rate": 1.3781610271805436e-06, | |
| "loss": 0.1002, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 9.966386554621849, | |
| "grad_norm": 1.9231485974986042, | |
| "learning_rate": 1.3550515705518263e-06, | |
| "loss": 0.0406, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 9.983193277310924, | |
| "grad_norm": 2.6221792896880918, | |
| "learning_rate": 1.3321234203344435e-06, | |
| "loss": 0.0369, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.378390284106787, | |
| "learning_rate": 1.3093770573896369e-06, | |
| "loss": 0.0475, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.08747600764036179, | |
| "eval_runtime": 83.5907, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 10.016806722689076, | |
| "grad_norm": 1.340634522580302, | |
| "learning_rate": 1.286812958766106e-06, | |
| "loss": 0.0302, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 10.033613445378151, | |
| "grad_norm": 1.836375096978306, | |
| "learning_rate": 1.2644315976900145e-06, | |
| "loss": 0.0546, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 10.050420168067227, | |
| "grad_norm": 4.930603999286428, | |
| "learning_rate": 1.242233443555051e-06, | |
| "loss": 0.0534, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 10.067226890756302, | |
| "grad_norm": 2.204329623971633, | |
| "learning_rate": 1.220218961912597e-06, | |
| "loss": 0.0588, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 10.084033613445378, | |
| "grad_norm": 3.9225927136224987, | |
| "learning_rate": 1.1983886144619527e-06, | |
| "loss": 0.0427, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.100840336134453, | |
| "grad_norm": 1.5324167733429486, | |
| "learning_rate": 1.1767428590406648e-06, | |
| "loss": 0.0269, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 1.7250937116057505, | |
| "learning_rate": 1.1552821496149136e-06, | |
| "loss": 0.0535, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 10.134453781512605, | |
| "grad_norm": 0.7987608814501375, | |
| "learning_rate": 1.134006936269999e-06, | |
| "loss": 0.0173, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 10.15126050420168, | |
| "grad_norm": 2.1053285256762733, | |
| "learning_rate": 1.1129176652009043e-06, | |
| "loss": 0.0406, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 10.168067226890756, | |
| "grad_norm": 2.1300541811924094, | |
| "learning_rate": 1.0920147787029233e-06, | |
| "loss": 0.0461, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 10.184873949579831, | |
| "grad_norm": 1.5136074758546754, | |
| "learning_rate": 1.0712987151624056e-06, | |
| "loss": 0.0433, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 10.201680672268907, | |
| "grad_norm": 1.6819131803547542, | |
| "learning_rate": 1.05076990904754e-06, | |
| "loss": 0.0529, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 10.218487394957982, | |
| "grad_norm": 3.0644142026231727, | |
| "learning_rate": 1.0304287908992626e-06, | |
| "loss": 0.0532, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 10.235294117647058, | |
| "grad_norm": 1.59650378921116, | |
| "learning_rate": 1.010275787322219e-06, | |
| "loss": 0.0538, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 10.252100840336134, | |
| "grad_norm": 1.5363947337082766, | |
| "learning_rate": 9.903113209758098e-07, | |
| "loss": 0.0428, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 10.268907563025211, | |
| "grad_norm": 1.5003741433198152, | |
| "learning_rate": 9.705358105653373e-07, | |
| "loss": 0.0322, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 10.285714285714286, | |
| "grad_norm": 2.0150139046727453, | |
| "learning_rate": 9.509496708332233e-07, | |
| "loss": 0.0395, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 10.302521008403362, | |
| "grad_norm": 2.2695792176582885, | |
| "learning_rate": 9.315533125503051e-07, | |
| "loss": 0.0755, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 10.319327731092438, | |
| "grad_norm": 2.7525380883317134, | |
| "learning_rate": 9.123471425072205e-07, | |
| "loss": 0.0816, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 10.336134453781513, | |
| "grad_norm": 1.9575555600743102, | |
| "learning_rate": 8.933315635058881e-07, | |
| "loss": 0.0444, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 1.3340430154540783, | |
| "learning_rate": 8.745069743510393e-07, | |
| "loss": 0.0289, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 10.369747899159664, | |
| "grad_norm": 2.350943233326136, | |
| "learning_rate": 8.558737698418762e-07, | |
| "loss": 0.0557, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 10.38655462184874, | |
| "grad_norm": 1.3949514445639306, | |
| "learning_rate": 8.374323407637741e-07, | |
| "loss": 0.0327, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 10.403361344537815, | |
| "grad_norm": 1.3745535851759583, | |
| "learning_rate": 8.191830738800977e-07, | |
| "loss": 0.0412, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 10.420168067226891, | |
| "grad_norm": 2.797313621901212, | |
| "learning_rate": 8.01126351924082e-07, | |
| "loss": 0.0415, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 10.436974789915967, | |
| "grad_norm": 2.0008993409572753, | |
| "learning_rate": 7.83262553590809e-07, | |
| "loss": 0.0473, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 10.453781512605042, | |
| "grad_norm": 2.566675704823962, | |
| "learning_rate": 7.655920535292682e-07, | |
| "loss": 0.0342, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 10.470588235294118, | |
| "grad_norm": 1.3899500927307256, | |
| "learning_rate": 7.48115222334489e-07, | |
| "loss": 0.0508, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 10.487394957983193, | |
| "grad_norm": 2.160264462063064, | |
| "learning_rate": 7.308324265397837e-07, | |
| "loss": 0.0535, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 10.504201680672269, | |
| "grad_norm": 1.6549288751982645, | |
| "learning_rate": 7.137440286090436e-07, | |
| "loss": 0.0527, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 10.521008403361344, | |
| "grad_norm": 1.7616896227790626, | |
| "learning_rate": 6.968503869291521e-07, | |
| "loss": 0.0463, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 10.53781512605042, | |
| "grad_norm": 1.882435848120671, | |
| "learning_rate": 6.8015185580246e-07, | |
| "loss": 0.0791, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 10.554621848739496, | |
| "grad_norm": 2.1866189366482485, | |
| "learning_rate": 6.636487854393536e-07, | |
| "loss": 0.0595, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 10.571428571428571, | |
| "grad_norm": 0.8846616976894125, | |
| "learning_rate": 6.473415219509182e-07, | |
| "loss": 0.0297, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 2.3038569022667614, | |
| "learning_rate": 6.31230407341672e-07, | |
| "loss": 0.0679, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 10.605042016806722, | |
| "grad_norm": 1.0106801357197905, | |
| "learning_rate": 6.153157795023956e-07, | |
| "loss": 0.0271, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 10.621848739495798, | |
| "grad_norm": 1.9606151371435605, | |
| "learning_rate": 5.995979722030443e-07, | |
| "loss": 0.0566, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 10.638655462184873, | |
| "grad_norm": 8.812038870074772, | |
| "learning_rate": 5.840773150857526e-07, | |
| "loss": 0.0936, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 10.655462184873949, | |
| "grad_norm": 2.37646496100217, | |
| "learning_rate": 5.687541336579127e-07, | |
| "loss": 0.0689, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 10.672268907563025, | |
| "grad_norm": 1.722317758769946, | |
| "learning_rate": 5.536287492853575e-07, | |
| "loss": 0.0409, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 10.6890756302521, | |
| "grad_norm": 1.7758828341496589, | |
| "learning_rate": 5.387014791856127e-07, | |
| "loss": 0.0501, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 10.705882352941176, | |
| "grad_norm": 0.9972447472445873, | |
| "learning_rate": 5.239726364212494e-07, | |
| "loss": 0.0354, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 10.722689075630251, | |
| "grad_norm": 1.8890386119269527, | |
| "learning_rate": 5.094425298933136e-07, | |
| "loss": 0.0375, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 10.739495798319329, | |
| "grad_norm": 3.0981629606543426, | |
| "learning_rate": 4.951114643348531e-07, | |
| "loss": 0.0618, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 10.756302521008404, | |
| "grad_norm": 3.3069946004768576, | |
| "learning_rate": 4.809797403045224e-07, | |
| "loss": 0.0738, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.77310924369748, | |
| "grad_norm": 1.0677359119397434, | |
| "learning_rate": 4.670476541802782e-07, | |
| "loss": 0.0349, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 10.789915966386555, | |
| "grad_norm": 1.8374696207253192, | |
| "learning_rate": 4.533154981531718e-07, | |
| "loss": 0.0431, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 10.806722689075631, | |
| "grad_norm": 3.948719353562884, | |
| "learning_rate": 4.397835602212064e-07, | |
| "loss": 0.0533, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 1.572765088896342, | |
| "learning_rate": 4.264521241833153e-07, | |
| "loss": 0.0425, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 10.840336134453782, | |
| "grad_norm": 1.2552552948039755, | |
| "learning_rate": 4.133214696333943e-07, | |
| "loss": 0.0326, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 10.857142857142858, | |
| "grad_norm": 1.4869473389842671, | |
| "learning_rate": 4.003918719544464e-07, | |
| "loss": 0.0334, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 10.873949579831933, | |
| "grad_norm": 1.4103343408301399, | |
| "learning_rate": 3.876636023128022e-07, | |
| "loss": 0.035, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 10.890756302521009, | |
| "grad_norm": 1.491957032559717, | |
| "learning_rate": 3.7513692765243637e-07, | |
| "loss": 0.0446, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 10.907563025210084, | |
| "grad_norm": 5.277585616799005, | |
| "learning_rate": 3.628121106893701e-07, | |
| "loss": 0.063, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 10.92436974789916, | |
| "grad_norm": 2.3243782807441837, | |
| "learning_rate": 3.50689409906152e-07, | |
| "loss": 0.0408, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "grad_norm": 5.5511386963840605, | |
| "learning_rate": 3.3876907954644933e-07, | |
| "loss": 0.0991, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 10.957983193277311, | |
| "grad_norm": 0.9258571015003478, | |
| "learning_rate": 3.2705136960970554e-07, | |
| "loss": 0.0324, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 10.974789915966387, | |
| "grad_norm": 3.7891311679004582, | |
| "learning_rate": 3.1553652584590864e-07, | |
| "loss": 0.0649, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 10.991596638655462, | |
| "grad_norm": 1.3430987417489015, | |
| "learning_rate": 3.0422478975042245e-07, | |
| "loss": 0.0329, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 10.991596638655462, | |
| "eval_loss": 0.08618413656949997, | |
| "eval_runtime": 84.2423, | |
| "eval_samples_per_second": 1.116, | |
| "eval_steps_per_second": 1.116, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 11.008403361344538, | |
| "grad_norm": 10.112136406292526, | |
| "learning_rate": 2.931163985589369e-07, | |
| "loss": 0.0792, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 11.025210084033613, | |
| "grad_norm": 1.2237182031841543, | |
| "learning_rate": 2.8221158524248003e-07, | |
| "loss": 0.0375, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 11.042016806722689, | |
| "grad_norm": 1.8203551353192673, | |
| "learning_rate": 2.7151057850253957e-07, | |
| "loss": 0.0529, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 1.6032849699615903, | |
| "learning_rate": 2.6101360276626795e-07, | |
| "loss": 0.0446, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 11.07563025210084, | |
| "grad_norm": 1.545240215165262, | |
| "learning_rate": 2.507208781817638e-07, | |
| "loss": 0.0356, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 11.092436974789916, | |
| "grad_norm": 1.5501298244234751, | |
| "learning_rate": 2.406326206134724e-07, | |
| "loss": 0.0413, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 11.109243697478991, | |
| "grad_norm": 3.6911299360545278, | |
| "learning_rate": 2.3074904163764012e-07, | |
| "loss": 0.1112, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 11.126050420168067, | |
| "grad_norm": 1.668296969639716, | |
| "learning_rate": 2.210703485378929e-07, | |
| "loss": 0.0343, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 11.142857142857142, | |
| "grad_norm": 1.464185072769606, | |
| "learning_rate": 2.115967443008804e-07, | |
| "loss": 0.042, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 11.159663865546218, | |
| "grad_norm": 3.21826813941365, | |
| "learning_rate": 2.0232842761201854e-07, | |
| "loss": 0.0393, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 11.176470588235293, | |
| "grad_norm": 1.5268309356925824, | |
| "learning_rate": 1.9326559285132495e-07, | |
| "loss": 0.0396, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 11.193277310924369, | |
| "grad_norm": 2.7559911444343625, | |
| "learning_rate": 1.844084300893456e-07, | |
| "loss": 0.0378, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 11.210084033613445, | |
| "grad_norm": 2.064362647787835, | |
| "learning_rate": 1.7575712508316244e-07, | |
| "loss": 0.0692, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 11.22689075630252, | |
| "grad_norm": 2.5064352487208676, | |
| "learning_rate": 1.673118592724987e-07, | |
| "loss": 0.0625, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 11.243697478991596, | |
| "grad_norm": 2.7390411371250774, | |
| "learning_rate": 1.5907280977591866e-07, | |
| "loss": 0.0737, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 11.260504201680673, | |
| "grad_norm": 0.9309858940918065, | |
| "learning_rate": 1.5104014938710498e-07, | |
| "loss": 0.0266, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 11.277310924369749, | |
| "grad_norm": 1.851032261256866, | |
| "learning_rate": 1.4321404657124393e-07, | |
| "loss": 0.0493, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 1.0812191705502274, | |
| "learning_rate": 1.3559466546148369e-07, | |
| "loss": 0.0434, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 11.3109243697479, | |
| "grad_norm": 1.8327569259876093, | |
| "learning_rate": 1.2818216585549824e-07, | |
| "loss": 0.0449, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 11.327731092436975, | |
| "grad_norm": 1.160640896943556, | |
| "learning_rate": 1.209767032121345e-07, | |
| "loss": 0.0328, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 11.344537815126051, | |
| "grad_norm": 2.0693866133338017, | |
| "learning_rate": 1.1397842864814712e-07, | |
| "loss": 0.0599, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 11.361344537815127, | |
| "grad_norm": 1.8624412914202513, | |
| "learning_rate": 1.0718748893503883e-07, | |
| "loss": 0.0458, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 11.378151260504202, | |
| "grad_norm": 2.0035143994192377, | |
| "learning_rate": 1.0060402649597178e-07, | |
| "loss": 0.0472, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 11.394957983193278, | |
| "grad_norm": 2.6901112576784447, | |
| "learning_rate": 9.422817940278773e-08, | |
| "loss": 0.0458, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 11.411764705882353, | |
| "grad_norm": 1.4284628255192529, | |
| "learning_rate": 8.806008137311028e-08, | |
| "loss": 0.0327, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 2.222615104628747, | |
| "learning_rate": 8.209986176753947e-08, | |
| "loss": 0.041, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 11.445378151260504, | |
| "grad_norm": 3.9740102271718634, | |
| "learning_rate": 7.634764558693941e-08, | |
| "loss": 0.0729, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 11.46218487394958, | |
| "grad_norm": 1.4918545836760815, | |
| "learning_rate": 7.080355346981815e-08, | |
| "loss": 0.0566, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 11.478991596638656, | |
| "grad_norm": 2.521127424464659, | |
| "learning_rate": 6.546770168979421e-08, | |
| "loss": 0.0437, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 11.495798319327731, | |
| "grad_norm": 1.2741993338648574, | |
| "learning_rate": 6.034020215316184e-08, | |
| "loss": 0.0349, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 11.512605042016807, | |
| "grad_norm": 1.296314142560303, | |
| "learning_rate": 5.5421162396542824e-08, | |
| "loss": 0.0356, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 1.8595898721766801, | |
| "learning_rate": 5.071068558462733e-08, | |
| "loss": 0.0559, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 11.546218487394958, | |
| "grad_norm": 8.704714098773897, | |
| "learning_rate": 4.6208870508017703e-08, | |
| "loss": 0.101, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 11.563025210084033, | |
| "grad_norm": 2.032442231084466, | |
| "learning_rate": 4.191581158115021e-08, | |
| "loss": 0.0478, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 11.579831932773109, | |
| "grad_norm": 9.418069463610777, | |
| "learning_rate": 3.783159884031773e-08, | |
| "loss": 0.1048, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 11.596638655462185, | |
| "grad_norm": 1.0069915518246801, | |
| "learning_rate": 3.3956317941779004e-08, | |
| "loss": 0.032, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 11.61344537815126, | |
| "grad_norm": 1.3869570014003472, | |
| "learning_rate": 3.029005015996789e-08, | |
| "loss": 0.0276, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 11.630252100840336, | |
| "grad_norm": 2.0275074366047128, | |
| "learning_rate": 2.6832872385783583e-08, | |
| "loss": 0.041, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 11.647058823529411, | |
| "grad_norm": 3.7065994847690407, | |
| "learning_rate": 2.3584857124977488e-08, | |
| "loss": 0.058, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 11.663865546218487, | |
| "grad_norm": 1.1492957912068942, | |
| "learning_rate": 2.054607249663665e-08, | |
| "loss": 0.04, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 11.680672268907562, | |
| "grad_norm": 7.151682674273455, | |
| "learning_rate": 1.7716582231752656e-08, | |
| "loss": 0.077, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 11.697478991596638, | |
| "grad_norm": 1.5415699598561934, | |
| "learning_rate": 1.509644567188717e-08, | |
| "loss": 0.041, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 11.714285714285714, | |
| "grad_norm": 1.7646669493531948, | |
| "learning_rate": 1.2685717767921823e-08, | |
| "loss": 0.0518, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 11.731092436974789, | |
| "grad_norm": 4.9082532974240465, | |
| "learning_rate": 1.048444907891244e-08, | |
| "loss": 0.1108, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 11.747899159663866, | |
| "grad_norm": 3.994721765198761, | |
| "learning_rate": 8.492685771025466e-09, | |
| "loss": 0.0613, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 2.688716545259077, | |
| "learning_rate": 6.710469616569848e-09, | |
| "loss": 0.0534, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.781512605042018, | |
| "grad_norm": 2.097132472666691, | |
| "learning_rate": 5.137837993121064e-09, | |
| "loss": 0.0534, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 11.798319327731093, | |
| "grad_norm": 1.036594664039175, | |
| "learning_rate": 3.774823882738421e-09, | |
| "loss": 0.0338, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 11.815126050420169, | |
| "grad_norm": 2.2273501309394907, | |
| "learning_rate": 2.6214558712722714e-09, | |
| "loss": 0.0337, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 11.831932773109244, | |
| "grad_norm": 1.4175794307129967, | |
| "learning_rate": 1.677758147762276e-09, | |
| "loss": 0.0304, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 11.84873949579832, | |
| "grad_norm": 2.118091027580824, | |
| "learning_rate": 9.43750503935581e-10, | |
| "loss": 0.0483, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 11.865546218487395, | |
| "grad_norm": 7.711002671886926, | |
| "learning_rate": 4.1944833378604334e-10, | |
| "loss": 0.1102, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 11.882352941176471, | |
| "grad_norm": 1.7414362067953533, | |
| "learning_rate": 1.0486263325559798e-10, | |
| "loss": 0.0357, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 11.899159663865547, | |
| "grad_norm": 1.2682884045214264, | |
| "learning_rate": 0.0, | |
| "loss": 0.0266, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 11.899159663865547, | |
| "eval_loss": 0.08401793986558914, | |
| "eval_runtime": 85.0205, | |
| "eval_samples_per_second": 1.106, | |
| "eval_steps_per_second": 1.106, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 11.899159663865547, | |
| "step": 708, | |
| "total_flos": 356037728403456.0, | |
| "train_loss": 0.4260152625183096, | |
| "train_runtime": 5257.2185, | |
| "train_samples_per_second": 1.08, | |
| "train_steps_per_second": 0.135 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 708, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 200.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 356037728403456.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |