| { | |
| "best_metric": 2.505129098892212, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 1.9002375296912115, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009501187648456057, | |
| "grad_norm": 19.4837589263916, | |
| "learning_rate": 1e-05, | |
| "loss": 9.9265, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009501187648456057, | |
| "eval_loss": 2.819766044616699, | |
| "eval_runtime": 2.1979, | |
| "eval_samples_per_second": 80.988, | |
| "eval_steps_per_second": 20.474, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019002375296912115, | |
| "grad_norm": 22.4416561126709, | |
| "learning_rate": 2e-05, | |
| "loss": 10.3355, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.028503562945368172, | |
| "grad_norm": 25.827529907226562, | |
| "learning_rate": 3e-05, | |
| "loss": 10.0698, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03800475059382423, | |
| "grad_norm": 26.252370834350586, | |
| "learning_rate": 4e-05, | |
| "loss": 10.0949, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.047505938242280284, | |
| "grad_norm": 33.103885650634766, | |
| "learning_rate": 5e-05, | |
| "loss": 10.576, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.057007125890736345, | |
| "grad_norm": 29.274099349975586, | |
| "learning_rate": 6e-05, | |
| "loss": 10.198, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0665083135391924, | |
| "grad_norm": 33.24345397949219, | |
| "learning_rate": 7e-05, | |
| "loss": 10.1067, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07600950118764846, | |
| "grad_norm": 34.14030838012695, | |
| "learning_rate": 8e-05, | |
| "loss": 9.9379, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0855106888361045, | |
| "grad_norm": 34.2035026550293, | |
| "learning_rate": 9e-05, | |
| "loss": 10.0649, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09501187648456057, | |
| "grad_norm": 40.71099090576172, | |
| "learning_rate": 0.0001, | |
| "loss": 10.1645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10451306413301663, | |
| "grad_norm": 39.27701187133789, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 10.0236, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11401425178147269, | |
| "grad_norm": 47.291786193847656, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 10.1491, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12351543942992874, | |
| "grad_norm": 45.70542907714844, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.9557, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1330166270783848, | |
| "grad_norm": 50.600215911865234, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.0959, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14251781472684086, | |
| "grad_norm": 65.09989166259766, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 10.775, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15201900237529692, | |
| "grad_norm": 70.45303344726562, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 10.8423, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16152019002375298, | |
| "grad_norm": 65.23307037353516, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 11.3941, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.171021377672209, | |
| "grad_norm": 71.38343048095703, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 10.6975, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18052256532066507, | |
| "grad_norm": 65.44233703613281, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 10.5171, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19002375296912113, | |
| "grad_norm": 70.14350128173828, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 11.3004, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1995249406175772, | |
| "grad_norm": 89.68062591552734, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 11.6624, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.20902612826603326, | |
| "grad_norm": 74.7454833984375, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 11.6524, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.21852731591448932, | |
| "grad_norm": 74.62859344482422, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 11.2738, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22802850356294538, | |
| "grad_norm": 83.69624328613281, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 11.8601, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2375296912114014, | |
| "grad_norm": 82.94718170166016, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 11.9405, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24703087885985747, | |
| "grad_norm": 165.92579650878906, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 13.8785, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25653206650831356, | |
| "grad_norm": 32.7075080871582, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 10.7774, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2660332541567696, | |
| "grad_norm": 46.25209045410156, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 10.9704, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2755344418052256, | |
| "grad_norm": 45.33772277832031, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 10.5613, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2850356294536817, | |
| "grad_norm": 37.04106521606445, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 10.2978, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29453681710213775, | |
| "grad_norm": 44.480587005615234, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 10.2833, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30403800475059384, | |
| "grad_norm": 43.32684326171875, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 10.4677, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31353919239904987, | |
| "grad_norm": 45.136192321777344, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 9.8608, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.32304038004750596, | |
| "grad_norm": 51.23587417602539, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 9.8421, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.332541567695962, | |
| "grad_norm": 54.82752227783203, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 9.8299, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.342042755344418, | |
| "grad_norm": 49.9445686340332, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 10.1103, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3515439429928741, | |
| "grad_norm": 55.708702087402344, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 9.434, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36104513064133015, | |
| "grad_norm": 58.81883239746094, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 10.0032, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.37054631828978624, | |
| "grad_norm": 54.223934173583984, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 9.1359, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38004750593824227, | |
| "grad_norm": 118.3580322265625, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.052, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38954869358669836, | |
| "grad_norm": 67.40377807617188, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.8413, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3990498812351544, | |
| "grad_norm": 81.58238220214844, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 11.0454, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4085510688836104, | |
| "grad_norm": 123.13668823242188, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 10.3621, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4180522565320665, | |
| "grad_norm": 86.48877716064453, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 10.5499, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.42755344418052255, | |
| "grad_norm": 93.03791809082031, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 10.4301, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43705463182897863, | |
| "grad_norm": 137.51673889160156, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 11.3399, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.44655581947743467, | |
| "grad_norm": 93.7792739868164, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 11.0524, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.45605700712589076, | |
| "grad_norm": 108.02003479003906, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 10.8861, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4655581947743468, | |
| "grad_norm": 120.84715270996094, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 11.4739, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4750593824228028, | |
| "grad_norm": 126.59524536132812, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 11.6551, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4750593824228028, | |
| "eval_loss": 2.5943713188171387, | |
| "eval_runtime": 2.1978, | |
| "eval_samples_per_second": 80.988, | |
| "eval_steps_per_second": 20.475, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4845605700712589, | |
| "grad_norm": 107.98383331298828, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 11.9595, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.49406175771971494, | |
| "grad_norm": 133.92140197753906, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 12.7066, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.503562945368171, | |
| "grad_norm": 51.18816375732422, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 10.8455, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5130641330166271, | |
| "grad_norm": 64.95140838623047, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 10.932, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5225653206650831, | |
| "grad_norm": 75.63224029541016, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 10.6041, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5320665083135392, | |
| "grad_norm": 75.99925231933594, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 10.7956, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5415676959619953, | |
| "grad_norm": 57.267391204833984, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 10.6428, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5510688836104513, | |
| "grad_norm": 86.51301574707031, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 10.1122, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5605700712589073, | |
| "grad_norm": 76.13331604003906, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.4613, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5700712589073634, | |
| "grad_norm": 91.9851303100586, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 9.1572, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5795724465558195, | |
| "grad_norm": 75.63456726074219, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 9.9452, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5890736342042755, | |
| "grad_norm": 93.57852172851562, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 10.0992, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5985748218527316, | |
| "grad_norm": 171.41571044921875, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 9.409, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6080760095011877, | |
| "grad_norm": 92.99851989746094, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 9.5722, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6175771971496437, | |
| "grad_norm": 147.53897094726562, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 9.6792, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6270783847980997, | |
| "grad_norm": 84.4323959350586, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 9.5559, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6365795724465558, | |
| "grad_norm": 115.26848602294922, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.7097, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6460807600950119, | |
| "grad_norm": 136.4612274169922, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 10.2567, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6555819477434679, | |
| "grad_norm": 95.94424438476562, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 9.9999, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.665083135391924, | |
| "grad_norm": 89.12234497070312, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 10.6295, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6745843230403801, | |
| "grad_norm": 117.38805389404297, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 10.7291, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.684085510688836, | |
| "grad_norm": 121.39945220947266, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 11.4512, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6935866983372921, | |
| "grad_norm": 130.9851837158203, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.6927, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7030878859857482, | |
| "grad_norm": 119.76856994628906, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 10.5335, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7125890736342043, | |
| "grad_norm": 145.76394653320312, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 10.8515, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7220902612826603, | |
| "grad_norm": 148.53482055664062, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 12.1, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7315914489311164, | |
| "grad_norm": 135.47276306152344, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 12.3097, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7410926365795725, | |
| "grad_norm": 137.0022735595703, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 12.0694, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7505938242280285, | |
| "grad_norm": 61.83055114746094, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 10.3654, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7600950118764845, | |
| "grad_norm": 71.3388671875, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 10.3184, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7695961995249406, | |
| "grad_norm": 79.79215240478516, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 10.4616, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7790973871733967, | |
| "grad_norm": 86.5008544921875, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 10.4754, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7885985748218527, | |
| "grad_norm": 84.40261840820312, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 10.3465, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7980997624703088, | |
| "grad_norm": 80.28064727783203, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 10.656, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8076009501187649, | |
| "grad_norm": 75.19422912597656, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 9.9042, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8171021377672208, | |
| "grad_norm": 100.35276794433594, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 9.8205, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8266033254156769, | |
| "grad_norm": 89.78428649902344, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 9.4354, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.836104513064133, | |
| "grad_norm": 143.8874053955078, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 9.5833, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8456057007125891, | |
| "grad_norm": 111.96574401855469, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 9.9287, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8551068883610451, | |
| "grad_norm": 76.66552734375, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 9.8425, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8646080760095012, | |
| "grad_norm": 72.75141906738281, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 9.1628, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8741092636579573, | |
| "grad_norm": 97.06959533691406, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 9.7337, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8836104513064132, | |
| "grad_norm": 98.35889434814453, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 9.8502, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8931116389548693, | |
| "grad_norm": 103.33940124511719, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 10.127, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9026128266033254, | |
| "grad_norm": 81.66262817382812, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 10.549, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9121140142517815, | |
| "grad_norm": 96.3223648071289, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 11.2002, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9216152019002375, | |
| "grad_norm": 94.24899291992188, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 10.5699, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9311163895486936, | |
| "grad_norm": 141.52479553222656, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 11.5338, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9406175771971497, | |
| "grad_norm": 124.4282455444336, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 10.8015, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "grad_norm": 119.40655517578125, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 10.5854, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "eval_loss": 2.5514986515045166, | |
| "eval_runtime": 2.2035, | |
| "eval_samples_per_second": 80.781, | |
| "eval_steps_per_second": 20.422, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9596199524940617, | |
| "grad_norm": 119.67656707763672, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 11.156, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9691211401425178, | |
| "grad_norm": 96.49195098876953, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 11.5515, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9786223277909739, | |
| "grad_norm": 113.34687042236328, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 11.5258, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9881235154394299, | |
| "grad_norm": 153.2520751953125, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 11.5513, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.997624703087886, | |
| "grad_norm": 95.41917419433594, | |
| "learning_rate": 5e-05, | |
| "loss": 9.7464, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.007125890736342, | |
| "grad_norm": 132.74461364746094, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 11.098, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0166270783847982, | |
| "grad_norm": 57.19962692260742, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 10.9078, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0261282660332542, | |
| "grad_norm": 83.69065856933594, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 10.2019, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.03562945368171, | |
| "grad_norm": 71.2281265258789, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 10.0873, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0451306413301662, | |
| "grad_norm": 81.04923248291016, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 10.5363, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0546318289786223, | |
| "grad_norm": 81.03862762451172, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 10.4283, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0641330166270784, | |
| "grad_norm": 75.69035339355469, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 9.5565, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0736342042755345, | |
| "grad_norm": 85.27928924560547, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 9.4412, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0831353919239906, | |
| "grad_norm": 79.6570816040039, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 9.5959, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0926365795724466, | |
| "grad_norm": 75.83023834228516, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 9.5345, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1021377672209025, | |
| "grad_norm": 97.34774017333984, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 9.6553, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1116389548693586, | |
| "grad_norm": 80.85506439208984, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 9.6712, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1211401425178147, | |
| "grad_norm": 72.7548599243164, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 9.0978, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1306413301662708, | |
| "grad_norm": 154.8693389892578, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 9.2004, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1401425178147269, | |
| "grad_norm": 107.32441711425781, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 9.7621, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.149643705463183, | |
| "grad_norm": 132.4986572265625, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 10.3251, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.159144893111639, | |
| "grad_norm": 118.84049987792969, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 10.1513, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.168646080760095, | |
| "grad_norm": 143.0657958984375, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 10.1979, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.178147268408551, | |
| "grad_norm": 97.22129821777344, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 10.2347, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.187648456057007, | |
| "grad_norm": 94.11260986328125, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 10.5843, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1971496437054632, | |
| "grad_norm": 135.90623474121094, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 10.5021, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2066508313539193, | |
| "grad_norm": 101.69560241699219, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 10.8865, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2161520190023754, | |
| "grad_norm": 110.10270690917969, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 10.4683, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2256532066508314, | |
| "grad_norm": 87.06689453125, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 11.0287, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2351543942992875, | |
| "grad_norm": 96.41451263427734, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 11.7243, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2446555819477434, | |
| "grad_norm": 257.0499267578125, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 11.351, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2541567695961995, | |
| "grad_norm": 65.98271179199219, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 10.5907, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2636579572446556, | |
| "grad_norm": 53.33887481689453, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 10.3695, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2731591448931117, | |
| "grad_norm": 55.912235260009766, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 10.1185, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2826603325415677, | |
| "grad_norm": 77.50757598876953, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 10.3438, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2921615201900236, | |
| "grad_norm": 56.65907287597656, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 9.9854, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3016627078384797, | |
| "grad_norm": 55.6379508972168, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 9.9772, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3111638954869358, | |
| "grad_norm": 97.31382751464844, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 9.5288, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3206650831353919, | |
| "grad_norm": 66.84870147705078, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 9.0591, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.330166270783848, | |
| "grad_norm": 64.66853332519531, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 9.5168, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.339667458432304, | |
| "grad_norm": 66.04855346679688, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 9.3083, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3491686460807601, | |
| "grad_norm": 77.19223022460938, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 9.5059, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3586698337292162, | |
| "grad_norm": 92.79340362548828, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 9.479, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3681710213776723, | |
| "grad_norm": 64.79487609863281, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 9.0669, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3776722090261282, | |
| "grad_norm": 73.25862884521484, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 9.504, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3871733966745843, | |
| "grad_norm": 74.61793518066406, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 9.475, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.3966745843230404, | |
| "grad_norm": 134.77496337890625, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 10.5703, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4061757719714965, | |
| "grad_norm": 116.8616943359375, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 9.4287, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4156769596199525, | |
| "grad_norm": 83.34971618652344, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 9.9763, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4251781472684084, | |
| "grad_norm": 99.0321044921875, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 10.2022, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4251781472684084, | |
| "eval_loss": 2.5159730911254883, | |
| "eval_runtime": 2.2, | |
| "eval_samples_per_second": 80.909, | |
| "eval_steps_per_second": 20.454, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4346793349168645, | |
| "grad_norm": 84.7126235961914, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 10.5227, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4441805225653206, | |
| "grad_norm": 122.6131362915039, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 11.4731, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4536817102137767, | |
| "grad_norm": 156.39503479003906, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 9.7612, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4631828978622328, | |
| "grad_norm": 159.140869140625, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 10.4874, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4726840855106889, | |
| "grad_norm": 112.50142669677734, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 10.6872, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.482185273159145, | |
| "grad_norm": 169.0659637451172, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 10.9691, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.491686460807601, | |
| "grad_norm": 188.6661834716797, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 11.2024, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5011876484560571, | |
| "grad_norm": 89.59786224365234, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 11.2119, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5106888361045132, | |
| "grad_norm": 59.78520965576172, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 10.6282, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.520190023752969, | |
| "grad_norm": 83.67992401123047, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 10.4133, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5296912114014252, | |
| "grad_norm": 83.5473403930664, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 10.1326, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5391923990498813, | |
| "grad_norm": 78.61830139160156, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 9.7638, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5486935866983373, | |
| "grad_norm": 76.97246551513672, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 10.4008, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5581947743467932, | |
| "grad_norm": 86.40713500976562, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 9.4252, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5676959619952493, | |
| "grad_norm": 75.06586456298828, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 9.1285, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5771971496437054, | |
| "grad_norm": 75.3528060913086, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 8.99, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.5866983372921615, | |
| "grad_norm": 94.358642578125, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 9.7196, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.5961995249406176, | |
| "grad_norm": 78.89218139648438, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 9.2189, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6057007125890737, | |
| "grad_norm": 84.5966796875, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 9.1692, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6152019002375297, | |
| "grad_norm": 77.07421112060547, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 8.9361, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6247030878859858, | |
| "grad_norm": 96.34380340576172, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 9.1843, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.634204275534442, | |
| "grad_norm": 67.07677459716797, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 9.53, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.643705463182898, | |
| "grad_norm": 141.90057373046875, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 10.2738, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6532066508313539, | |
| "grad_norm": 104.226318359375, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 10.0341, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.66270783847981, | |
| "grad_norm": 101.72418212890625, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 10.2237, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.672209026128266, | |
| "grad_norm": 134.41738891601562, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 10.7662, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6817102137767221, | |
| "grad_norm": 126.75146484375, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 10.7483, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.691211401425178, | |
| "grad_norm": 95.49859619140625, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.4185, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.700712589073634, | |
| "grad_norm": 112.31136322021484, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 10.8657, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7102137767220902, | |
| "grad_norm": 143.15719604492188, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 11.0102, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7197149643705463, | |
| "grad_norm": 140.19906616210938, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 10.485, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7292161520190024, | |
| "grad_norm": 165.3531951904297, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 11.084, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7387173396674585, | |
| "grad_norm": 246.70103454589844, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 11.2965, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7482185273159145, | |
| "grad_norm": 190.2755126953125, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 10.836, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7577197149643706, | |
| "grad_norm": 54.2744255065918, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.4541, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7672209026128267, | |
| "grad_norm": 60.90446853637695, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 10.2176, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7767220902612828, | |
| "grad_norm": 64.88226318359375, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 10.1289, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.7862232779097387, | |
| "grad_norm": 64.57106018066406, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 10.4699, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.7957244655581948, | |
| "grad_norm": 71.23513793945312, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 10.5886, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8052256532066508, | |
| "grad_norm": 62.82306671142578, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 10.136, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.814726840855107, | |
| "grad_norm": 72.74017333984375, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 9.192, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8242280285035628, | |
| "grad_norm": 62.476192474365234, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 9.0637, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.833729216152019, | |
| "grad_norm": 79.42888641357422, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 9.5623, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.843230403800475, | |
| "grad_norm": 80.78561401367188, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 9.71, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.852731591448931, | |
| "grad_norm": 87.67532348632812, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 9.0287, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8622327790973872, | |
| "grad_norm": 84.16883087158203, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 9.629, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8717339667458432, | |
| "grad_norm": 100.6878890991211, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 9.2654, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8812351543942993, | |
| "grad_norm": 86.63809204101562, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 9.8003, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.8907363420427554, | |
| "grad_norm": 149.69566345214844, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 9.5615, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9002375296912115, | |
| "grad_norm": 103.23074340820312, | |
| "learning_rate": 0.0, | |
| "loss": 9.9939, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9002375296912115, | |
| "eval_loss": 2.505129098892212, | |
| "eval_runtime": 2.202, | |
| "eval_samples_per_second": 80.835, | |
| "eval_steps_per_second": 20.436, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5377451077074944e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |