| { | |
| "best_metric": 2.4192488193511963, | |
| "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": 19.052827835083008, | |
| "learning_rate": 1e-05, | |
| "loss": 9.9265, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009501187648456057, | |
| "eval_loss": 2.819766044616699, | |
| "eval_runtime": 2.2024, | |
| "eval_samples_per_second": 80.823, | |
| "eval_steps_per_second": 20.433, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019002375296912115, | |
| "grad_norm": 22.11795425415039, | |
| "learning_rate": 2e-05, | |
| "loss": 10.3355, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.028503562945368172, | |
| "grad_norm": 23.46179962158203, | |
| "learning_rate": 3e-05, | |
| "loss": 10.0643, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03800475059382423, | |
| "grad_norm": 25.683902740478516, | |
| "learning_rate": 4e-05, | |
| "loss": 10.0876, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.047505938242280284, | |
| "grad_norm": 28.117198944091797, | |
| "learning_rate": 5e-05, | |
| "loss": 10.5635, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.057007125890736345, | |
| "grad_norm": 29.595441818237305, | |
| "learning_rate": 6e-05, | |
| "loss": 10.2137, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0665083135391924, | |
| "grad_norm": 31.624820709228516, | |
| "learning_rate": 7e-05, | |
| "loss": 10.1075, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07600950118764846, | |
| "grad_norm": 35.10360336303711, | |
| "learning_rate": 8e-05, | |
| "loss": 9.9741, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0855106888361045, | |
| "grad_norm": 30.275375366210938, | |
| "learning_rate": 9e-05, | |
| "loss": 10.0326, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09501187648456057, | |
| "grad_norm": 42.54201889038086, | |
| "learning_rate": 0.0001, | |
| "loss": 10.1282, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10451306413301663, | |
| "grad_norm": 38.00617599487305, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 10.0324, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11401425178147269, | |
| "grad_norm": 43.97023391723633, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 10.1721, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12351543942992874, | |
| "grad_norm": 43.446868896484375, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.9704, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1330166270783848, | |
| "grad_norm": 41.41438293457031, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.1132, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14251781472684086, | |
| "grad_norm": 54.14292907714844, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 10.735, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.15201900237529692, | |
| "grad_norm": 59.9361572265625, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 10.8552, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16152019002375298, | |
| "grad_norm": 65.35136413574219, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 11.5271, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.171021377672209, | |
| "grad_norm": 60.65117645263672, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 10.6617, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.18052256532066507, | |
| "grad_norm": 66.33427429199219, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 10.5953, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.19002375296912113, | |
| "grad_norm": 78.8612289428711, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 11.4295, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1995249406175772, | |
| "grad_norm": 81.67987823486328, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 11.7918, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.20902612826603326, | |
| "grad_norm": 82.05351257324219, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 11.693, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.21852731591448932, | |
| "grad_norm": 72.68646240234375, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 11.4251, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22802850356294538, | |
| "grad_norm": 116.53019714355469, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 12.0219, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2375296912114014, | |
| "grad_norm": 91.26791381835938, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 12.1224, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24703087885985747, | |
| "grad_norm": 142.7079315185547, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 13.8356, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25653206650831356, | |
| "grad_norm": 31.430696487426758, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 10.7579, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2660332541567696, | |
| "grad_norm": 38.121150970458984, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 11.0067, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2755344418052256, | |
| "grad_norm": 39.036582946777344, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 10.5208, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2850356294536817, | |
| "grad_norm": 36.783042907714844, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 10.3499, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29453681710213775, | |
| "grad_norm": 38.77050018310547, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 10.3018, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.30403800475059384, | |
| "grad_norm": 43.17377471923828, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 10.5024, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.31353919239904987, | |
| "grad_norm": 50.610130310058594, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 9.9037, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.32304038004750596, | |
| "grad_norm": 50.932167053222656, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 10.0025, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.332541567695962, | |
| "grad_norm": 63.05067443847656, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 9.9396, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.342042755344418, | |
| "grad_norm": 60.02093505859375, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 10.2482, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3515439429928741, | |
| "grad_norm": 61.30398941040039, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 9.506, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.36104513064133015, | |
| "grad_norm": 58.13288497924805, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 10.1374, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.37054631828978624, | |
| "grad_norm": 51.784263610839844, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 9.2826, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.38004750593824227, | |
| "grad_norm": 71.26021575927734, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.074, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.38954869358669836, | |
| "grad_norm": 63.67902755737305, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.9213, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3990498812351544, | |
| "grad_norm": 82.36457061767578, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 11.0074, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4085510688836104, | |
| "grad_norm": 71.11475372314453, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 10.3478, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4180522565320665, | |
| "grad_norm": 57.912925720214844, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 10.5365, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.42755344418052255, | |
| "grad_norm": 72.04336547851562, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 10.5838, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.43705463182897863, | |
| "grad_norm": 82.88634490966797, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 11.4603, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.44655581947743467, | |
| "grad_norm": 57.1037712097168, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 10.9868, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.45605700712589076, | |
| "grad_norm": 67.33989715576172, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 10.9432, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4655581947743468, | |
| "grad_norm": 75.16851043701172, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 11.4707, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4750593824228028, | |
| "grad_norm": 74.13951110839844, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 11.5781, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4750593824228028, | |
| "eval_loss": 2.5815207958221436, | |
| "eval_runtime": 2.1835, | |
| "eval_samples_per_second": 81.519, | |
| "eval_steps_per_second": 20.609, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4845605700712589, | |
| "grad_norm": 73.47399139404297, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 11.8677, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.49406175771971494, | |
| "grad_norm": 75.97662353515625, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 12.5136, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.503562945368171, | |
| "grad_norm": 25.05925178527832, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 10.7167, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5130641330166271, | |
| "grad_norm": 30.213838577270508, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 10.791, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5225653206650831, | |
| "grad_norm": 28.441303253173828, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 10.4234, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5320665083135392, | |
| "grad_norm": 28.828868865966797, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 10.6945, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5415676959619953, | |
| "grad_norm": 28.981857299804688, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 10.5335, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5510688836104513, | |
| "grad_norm": 29.173961639404297, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 9.9706, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5605700712589073, | |
| "grad_norm": 27.359298706054688, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.3842, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5700712589073634, | |
| "grad_norm": 32.77239227294922, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 9.0034, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5795724465558195, | |
| "grad_norm": 32.115116119384766, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 9.7039, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5890736342042755, | |
| "grad_norm": 35.204383850097656, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 9.9224, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5985748218527316, | |
| "grad_norm": 39.001319885253906, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 9.2109, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6080760095011877, | |
| "grad_norm": 38.51883316040039, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 9.3607, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6175771971496437, | |
| "grad_norm": 40.05404281616211, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 9.3722, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6270783847980997, | |
| "grad_norm": 38.425113677978516, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 9.362, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6365795724465558, | |
| "grad_norm": 42.36941909790039, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.4244, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6460807600950119, | |
| "grad_norm": 43.2212028503418, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 9.9641, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6555819477434679, | |
| "grad_norm": 40.110958099365234, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 9.7888, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.665083135391924, | |
| "grad_norm": 51.71419143676758, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 10.4713, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6745843230403801, | |
| "grad_norm": 49.315128326416016, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 10.473, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.684085510688836, | |
| "grad_norm": 43.141693115234375, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 11.1162, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6935866983372921, | |
| "grad_norm": 40.67599105834961, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.3638, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7030878859857482, | |
| "grad_norm": 40.67022705078125, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 10.1749, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7125890736342043, | |
| "grad_norm": 47.472530364990234, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 10.528, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7220902612826603, | |
| "grad_norm": 58.274471282958984, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 11.7439, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7315914489311164, | |
| "grad_norm": 53.202579498291016, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 11.8825, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7410926365795725, | |
| "grad_norm": 74.2247543334961, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 11.4543, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7505938242280285, | |
| "grad_norm": 23.56029510498047, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 10.1815, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7600950118764845, | |
| "grad_norm": 25.578887939453125, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 10.0859, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7695961995249406, | |
| "grad_norm": 29.133159637451172, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 10.1942, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7790973871733967, | |
| "grad_norm": 26.358640670776367, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 10.2081, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7885985748218527, | |
| "grad_norm": 28.30341148376465, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 10.021, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7980997624703088, | |
| "grad_norm": 30.32122802734375, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 10.3293, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8076009501187649, | |
| "grad_norm": 32.282203674316406, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 9.6215, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8171021377672208, | |
| "grad_norm": 35.80141067504883, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 9.51, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8266033254156769, | |
| "grad_norm": 37.250389099121094, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 9.1855, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.836104513064133, | |
| "grad_norm": 30.585859298706055, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 9.3132, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8456057007125891, | |
| "grad_norm": 37.24182891845703, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 9.5975, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8551068883610451, | |
| "grad_norm": 38.571170806884766, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 9.6558, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8646080760095012, | |
| "grad_norm": 37.505733489990234, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 8.9013, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8741092636579573, | |
| "grad_norm": 50.142913818359375, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 9.4002, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8836104513064132, | |
| "grad_norm": 49.40633010864258, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 9.5774, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8931116389548693, | |
| "grad_norm": 54.912620544433594, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 9.8763, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9026128266033254, | |
| "grad_norm": 47.32883834838867, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 10.2501, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9121140142517815, | |
| "grad_norm": 63.21529769897461, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 10.962, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9216152019002375, | |
| "grad_norm": 64.93472290039062, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 10.2987, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9311163895486936, | |
| "grad_norm": 80.15961456298828, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 11.2168, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9406175771971497, | |
| "grad_norm": 69.57113647460938, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 10.4517, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "grad_norm": 66.7508544921875, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 10.3117, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "eval_loss": 2.4882469177246094, | |
| "eval_runtime": 2.183, | |
| "eval_samples_per_second": 81.54, | |
| "eval_steps_per_second": 20.614, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9596199524940617, | |
| "grad_norm": 60.58907699584961, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 10.7966, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9691211401425178, | |
| "grad_norm": 72.10790252685547, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 11.0507, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9786223277909739, | |
| "grad_norm": 81.67748260498047, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 11.0962, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9881235154394299, | |
| "grad_norm": 96.13243865966797, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 10.8405, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.997624703087886, | |
| "grad_norm": 45.95366287231445, | |
| "learning_rate": 5e-05, | |
| "loss": 9.4617, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.007125890736342, | |
| "grad_norm": 74.89264678955078, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 10.7786, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0166270783847982, | |
| "grad_norm": 34.36727523803711, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 10.6115, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0261282660332542, | |
| "grad_norm": 34.2283935546875, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 9.8407, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.03562945368171, | |
| "grad_norm": 33.08857345581055, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 9.727, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.0451306413301662, | |
| "grad_norm": 36.63699722290039, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 10.1565, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0546318289786223, | |
| "grad_norm": 39.32516098022461, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 10.0876, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0641330166270784, | |
| "grad_norm": 51.5899658203125, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 9.2029, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0736342042755345, | |
| "grad_norm": 43.274932861328125, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 9.0035, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0831353919239906, | |
| "grad_norm": 42.604061126708984, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 9.1692, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0926365795724466, | |
| "grad_norm": 41.310462951660156, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 9.1696, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1021377672209025, | |
| "grad_norm": 43.11902618408203, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 9.2453, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1116389548693586, | |
| "grad_norm": 39.28797912597656, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 9.2359, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1211401425178147, | |
| "grad_norm": 39.07838439941406, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 8.8084, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1306413301662708, | |
| "grad_norm": 42.56367111206055, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.8215, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1401425178147269, | |
| "grad_norm": 52.60874938964844, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 9.3862, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.149643705463183, | |
| "grad_norm": 59.03802490234375, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 10.0051, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.159144893111639, | |
| "grad_norm": 53.51556396484375, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 9.7166, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.168646080760095, | |
| "grad_norm": 61.38289260864258, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 9.8521, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.178147268408551, | |
| "grad_norm": 59.38209533691406, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 9.8615, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.187648456057007, | |
| "grad_norm": 69.93914031982422, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 10.217, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1971496437054632, | |
| "grad_norm": 53.72102737426758, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 10.1848, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2066508313539193, | |
| "grad_norm": 54.664764404296875, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 10.483, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2161520190023754, | |
| "grad_norm": 60.76587677001953, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 10.0884, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2256532066508314, | |
| "grad_norm": 67.71343994140625, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 10.7442, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2351543942992875, | |
| "grad_norm": 69.5775375366211, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 11.2854, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2446555819477434, | |
| "grad_norm": 88.2365951538086, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 10.9993, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2541567695961995, | |
| "grad_norm": 40.71464538574219, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 10.1865, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2636579572446556, | |
| "grad_norm": 27.133167266845703, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 10.0801, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2731591448931117, | |
| "grad_norm": 27.42853355407715, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 9.7753, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2826603325415677, | |
| "grad_norm": 29.747236251831055, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 9.9648, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2921615201900236, | |
| "grad_norm": 28.10453224182129, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 9.6276, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3016627078384797, | |
| "grad_norm": 32.8948860168457, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 9.6534, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3111638954869358, | |
| "grad_norm": 36.85212707519531, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 9.1545, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3206650831353919, | |
| "grad_norm": 42.10161590576172, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 8.7032, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.330166270783848, | |
| "grad_norm": 33.20311737060547, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 9.1932, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.339667458432304, | |
| "grad_norm": 40.20091247558594, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 8.9803, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3491686460807601, | |
| "grad_norm": 52.291587829589844, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 9.1159, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3586698337292162, | |
| "grad_norm": 43.315677642822266, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 9.0998, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3681710213776723, | |
| "grad_norm": 33.63870620727539, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 8.7301, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3776722090261282, | |
| "grad_norm": 43.19037628173828, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 9.222, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3871733966745843, | |
| "grad_norm": 49.447837829589844, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 9.0844, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.3966745843230404, | |
| "grad_norm": 52.0371208190918, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 10.1305, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4061757719714965, | |
| "grad_norm": 40.418521881103516, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 9.1222, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4156769596199525, | |
| "grad_norm": 56.46171188354492, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 9.7529, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4251781472684084, | |
| "grad_norm": 39.67538833618164, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 9.8505, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4251781472684084, | |
| "eval_loss": 2.4354910850524902, | |
| "eval_runtime": 2.1805, | |
| "eval_samples_per_second": 81.632, | |
| "eval_steps_per_second": 20.637, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4346793349168645, | |
| "grad_norm": 48.840999603271484, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 10.2448, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4441805225653206, | |
| "grad_norm": 61.02261734008789, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 11.0807, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4536817102137767, | |
| "grad_norm": 70.39476776123047, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 9.2582, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4631828978622328, | |
| "grad_norm": 57.5440673828125, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 9.9702, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4726840855106889, | |
| "grad_norm": 61.25320053100586, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 10.2709, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.482185273159145, | |
| "grad_norm": 55.396263122558594, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 10.4431, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.491686460807601, | |
| "grad_norm": 57.45940017700195, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 10.6606, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5011876484560571, | |
| "grad_norm": 45.15793228149414, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 10.8481, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5106888361045132, | |
| "grad_norm": 27.951114654541016, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 10.3779, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.520190023752969, | |
| "grad_norm": 27.518173217773438, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 10.0454, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5296912114014252, | |
| "grad_norm": 33.57125473022461, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 9.7147, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5391923990498813, | |
| "grad_norm": 43.2706184387207, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 9.4158, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5486935866983373, | |
| "grad_norm": 33.121360778808594, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 10.0536, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5581947743467932, | |
| "grad_norm": 36.35709762573242, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 9.1358, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.5676959619952493, | |
| "grad_norm": 34.599037170410156, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 8.8221, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5771971496437054, | |
| "grad_norm": 34.87372970581055, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 8.6934, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.5866983372921615, | |
| "grad_norm": 37.04258346557617, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 9.4313, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.5961995249406176, | |
| "grad_norm": 39.22598648071289, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 8.9076, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6057007125890737, | |
| "grad_norm": 37.508792877197266, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 8.7888, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6152019002375297, | |
| "grad_norm": 33.696937561035156, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 8.6522, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6247030878859858, | |
| "grad_norm": 33.85466384887695, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 8.8511, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.634204275534442, | |
| "grad_norm": 37.68962860107422, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 9.1598, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.643705463182898, | |
| "grad_norm": 41.454708099365234, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 9.8324, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6532066508313539, | |
| "grad_norm": 48.535736083984375, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 9.7711, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.66270783847981, | |
| "grad_norm": 48.91841506958008, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 9.8255, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.672209026128266, | |
| "grad_norm": 42.789329528808594, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 10.4297, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6817102137767221, | |
| "grad_norm": 45.49766159057617, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 10.3094, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.691211401425178, | |
| "grad_norm": 48.995635986328125, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.0495, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.700712589073634, | |
| "grad_norm": 48.69024658203125, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 10.4346, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.7102137767220902, | |
| "grad_norm": 52.096458435058594, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 10.4964, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7197149643705463, | |
| "grad_norm": 54.455482482910156, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 10.1291, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7292161520190024, | |
| "grad_norm": 48.2535400390625, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 10.4835, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7387173396674585, | |
| "grad_norm": 64.61412048339844, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 10.5077, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7482185273159145, | |
| "grad_norm": 41.12872314453125, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 10.383, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.7577197149643706, | |
| "grad_norm": 23.937122344970703, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.1728, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7672209026128267, | |
| "grad_norm": 35.46590805053711, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 9.9774, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7767220902612828, | |
| "grad_norm": 23.37141990661621, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 9.7724, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.7862232779097387, | |
| "grad_norm": 27.467708587646484, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 10.0875, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.7957244655581948, | |
| "grad_norm": 29.722068786621094, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 10.2215, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8052256532066508, | |
| "grad_norm": 36.735897064208984, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 9.78, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.814726840855107, | |
| "grad_norm": 40.1739616394043, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 8.8518, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8242280285035628, | |
| "grad_norm": 44.420196533203125, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 8.7969, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.833729216152019, | |
| "grad_norm": 40.424556732177734, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 9.2837, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.843230403800475, | |
| "grad_norm": 38.00553512573242, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 9.3804, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.852731591448931, | |
| "grad_norm": 36.450809478759766, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 8.6721, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8622327790973872, | |
| "grad_norm": 39.79310607910156, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 9.2604, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8717339667458432, | |
| "grad_norm": 42.907955169677734, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 8.9399, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8812351543942993, | |
| "grad_norm": 39.591670989990234, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 9.4672, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.8907363420427554, | |
| "grad_norm": 43.547306060791016, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 9.1594, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.9002375296912115, | |
| "grad_norm": 51.04115676879883, | |
| "learning_rate": 0.0, | |
| "loss": 9.6482, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9002375296912115, | |
| "eval_loss": 2.4192488193511963, | |
| "eval_runtime": 2.1809, | |
| "eval_samples_per_second": 81.618, | |
| "eval_steps_per_second": 20.634, | |
| "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 | |
| } | |