| { | |
| "best_metric": 6.290847301483154, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-50", | |
| "epoch": 0.009231479344564966, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 4.615739672282483e-05, | |
| "grad_norm": 22445.78125, | |
| "learning_rate": 1e-05, | |
| "loss": 25.784, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 4.615739672282483e-05, | |
| "eval_loss": 6.5567145347595215, | |
| "eval_runtime": 116.9938, | |
| "eval_samples_per_second": 311.888, | |
| "eval_steps_per_second": 77.978, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 9.231479344564966e-05, | |
| "grad_norm": 12910.5078125, | |
| "learning_rate": 2e-05, | |
| "loss": 27.8226, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0001384721901684745, | |
| "grad_norm": 17848.23046875, | |
| "learning_rate": 3e-05, | |
| "loss": 27.2563, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00018462958689129932, | |
| "grad_norm": 12322.72265625, | |
| "learning_rate": 4e-05, | |
| "loss": 27.3904, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00023078698361412417, | |
| "grad_norm": 10741.056640625, | |
| "learning_rate": 5e-05, | |
| "loss": 28.2983, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.000276944380336949, | |
| "grad_norm": 13121.9970703125, | |
| "learning_rate": 6e-05, | |
| "loss": 26.237, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00032310177705977385, | |
| "grad_norm": 9293.3037109375, | |
| "learning_rate": 7e-05, | |
| "loss": 25.5588, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00036925917378259865, | |
| "grad_norm": 10365.8857421875, | |
| "learning_rate": 8e-05, | |
| "loss": 26.4811, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0004154165705054235, | |
| "grad_norm": 12928.04296875, | |
| "learning_rate": 9e-05, | |
| "loss": 25.8373, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00046157396722824835, | |
| "grad_norm": 14451.408203125, | |
| "learning_rate": 0.0001, | |
| "loss": 26.4935, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0005077313639510731, | |
| "grad_norm": 19713.658203125, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 25.1195, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.000553888760673898, | |
| "grad_norm": 15381.8095703125, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 24.9361, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0006000461573967228, | |
| "grad_norm": 24253.5546875, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 25.9728, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0006462035541195477, | |
| "grad_norm": 14783.53125, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 23.7536, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0006923609508423724, | |
| "grad_norm": 15313.6494140625, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 25.9306, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0007385183475651973, | |
| "grad_norm": 22333.41015625, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 22.6163, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0007846757442880221, | |
| "grad_norm": 20572.08984375, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 23.9031, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.000830833141010847, | |
| "grad_norm": 19698.630859375, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 23.5744, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0008769905377336718, | |
| "grad_norm": 13384.6240234375, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 23.5693, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0009231479344564967, | |
| "grad_norm": 26537.767578125, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 23.3416, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0009693053311793214, | |
| "grad_norm": 19650.4921875, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 23.3517, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0010154627279021463, | |
| "grad_norm": 18629.486328125, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 20.4999, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0010616201246249712, | |
| "grad_norm": 22304.865234375, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 22.6462, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.001107777521347796, | |
| "grad_norm": 14506.77734375, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 23.4764, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0011539349180706207, | |
| "grad_norm": 14228.169921875, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 23.762, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0012000923147934457, | |
| "grad_norm": 30771.2109375, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 22.8363, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0012462497115162704, | |
| "grad_norm": 56593.390625, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 21.8793, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0012924071082390954, | |
| "grad_norm": 20859.037109375, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 22.4923, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0013385645049619201, | |
| "grad_norm": 27690.1875, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 21.5724, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0013847219016847449, | |
| "grad_norm": 16696.064453125, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 19.0704, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0014308792984075698, | |
| "grad_norm": 31018.873046875, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 20.8645, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0014770366951303946, | |
| "grad_norm": 31504.15625, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 20.111, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0015231940918532195, | |
| "grad_norm": 32905.55859375, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 21.8384, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0015693514885760443, | |
| "grad_norm": 25131.814453125, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 23.6569, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0016155088852988692, | |
| "grad_norm": 22027.71484375, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 22.2301, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.001661666282021694, | |
| "grad_norm": 27400.876953125, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 21.5273, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0017078236787445187, | |
| "grad_norm": 24637.80859375, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 22.5958, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0017539810754673437, | |
| "grad_norm": 20805.537109375, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 21.467, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0018001384721901684, | |
| "grad_norm": 36721.640625, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 23.575, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0018462958689129934, | |
| "grad_norm": 24539.470703125, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 23.2596, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0018924532656358181, | |
| "grad_norm": 32397.0703125, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 20.8464, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0019386106623586429, | |
| "grad_norm": 22713.24609375, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 19.6454, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0019847680590814676, | |
| "grad_norm": 15757.3203125, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 21.6769, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0020309254558042926, | |
| "grad_norm": 23368.83203125, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 20.7625, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0020770828525271175, | |
| "grad_norm": 22798.123046875, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 20.2746, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0021232402492499425, | |
| "grad_norm": 27159.142578125, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 23.0228, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.002169397645972767, | |
| "grad_norm": 43737.3984375, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 20.1612, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.002215555042695592, | |
| "grad_norm": 38735.39453125, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 22.1198, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.002261712439418417, | |
| "grad_norm": 24909.6484375, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 23.5771, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0023078698361412415, | |
| "grad_norm": 46400.6328125, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 24.5076, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0023078698361412415, | |
| "eval_loss": 6.290847301483154, | |
| "eval_runtime": 117.0237, | |
| "eval_samples_per_second": 311.809, | |
| "eval_steps_per_second": 77.959, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0023540272328640664, | |
| "grad_norm": 13528.9306640625, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 29.2787, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0024001846295868914, | |
| "grad_norm": 14640.6796875, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 29.954, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0024463420263097163, | |
| "grad_norm": 19368.150390625, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 28.7966, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.002492499423032541, | |
| "grad_norm": 17936.0078125, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 28.7941, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.002538656819755366, | |
| "grad_norm": 23861.60546875, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 29.3609, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.002584814216478191, | |
| "grad_norm": 15433.3095703125, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 27.1609, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0026309716132010153, | |
| "grad_norm": 33403.921875, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 29.2413, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0026771290099238403, | |
| "grad_norm": 24673.494140625, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 27.6385, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0027232864066466652, | |
| "grad_norm": 14325.19140625, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 29.6801, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0027694438033694898, | |
| "grad_norm": 17081.51171875, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 28.7645, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0028156012000923147, | |
| "grad_norm": 29413.771484375, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 28.8996, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0028617585968151397, | |
| "grad_norm": 16099.470703125, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 28.155, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0029079159935379646, | |
| "grad_norm": 22059.09765625, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 29.3335, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.002954073390260789, | |
| "grad_norm": 17140.3203125, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 27.0672, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.003000230786983614, | |
| "grad_norm": 28841.115234375, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 28.5047, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.003046388183706439, | |
| "grad_norm": 18518.37109375, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 27.5791, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0030925455804292636, | |
| "grad_norm": 19345.005859375, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 28.3685, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.0031387029771520886, | |
| "grad_norm": 29759.44921875, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 26.2761, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0031848603738749135, | |
| "grad_norm": 18335.62109375, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 26.6666, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0032310177705977385, | |
| "grad_norm": 30440.87890625, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 26.5062, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.003277175167320563, | |
| "grad_norm": 19112.134765625, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 25.7603, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.003323332564043388, | |
| "grad_norm": 24852.76171875, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 27.9856, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.003369489960766213, | |
| "grad_norm": 46107.21484375, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 25.6559, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.0034156473574890375, | |
| "grad_norm": 24512.013671875, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 27.834, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.0034618047542118624, | |
| "grad_norm": 29530.43359375, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 26.7404, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0035079621509346874, | |
| "grad_norm": 27868.212890625, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 25.2077, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.003554119547657512, | |
| "grad_norm": 18155.494140625, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 26.284, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.003600276944380337, | |
| "grad_norm": 21550.98046875, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 25.5477, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.003646434341103162, | |
| "grad_norm": 31559.13671875, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 25.403, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0036925917378259868, | |
| "grad_norm": 18903.1328125, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 23.5741, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0037387491345488113, | |
| "grad_norm": 48414.99609375, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 24.3138, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0037849065312716363, | |
| "grad_norm": 21682.453125, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 23.5924, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.0038310639279944612, | |
| "grad_norm": 21422.771484375, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 23.3001, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0038772213247172858, | |
| "grad_norm": 22182.619140625, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 23.6324, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.003923378721440111, | |
| "grad_norm": 24638.53515625, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 24.0123, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.003969536118162935, | |
| "grad_norm": 42181.08203125, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 24.0208, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.004015693514885761, | |
| "grad_norm": 17223.33203125, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 23.8072, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.004061850911608585, | |
| "grad_norm": 37419.80078125, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 21.9256, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.00410800830833141, | |
| "grad_norm": 18106.880859375, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 22.0732, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.004154165705054235, | |
| "grad_norm": 27022.27734375, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 23.5882, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.00420032310177706, | |
| "grad_norm": 24592.55859375, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 22.5416, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.004246480498499885, | |
| "grad_norm": 26019.220703125, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 22.1786, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.0042926378952227095, | |
| "grad_norm": 19021.7421875, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 21.6818, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.004338795291945534, | |
| "grad_norm": 44170.796875, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 23.2813, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0043849526886683594, | |
| "grad_norm": 22554.763671875, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 24.2983, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.004431110085391184, | |
| "grad_norm": 37566.48046875, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 22.8671, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0044772674821140085, | |
| "grad_norm": 46563.9453125, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 22.6329, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.004523424878836834, | |
| "grad_norm": 11167.861328125, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 20.1984, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.004569582275559658, | |
| "grad_norm": 65252.296875, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 23.824, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.004615739672282483, | |
| "grad_norm": 43028.50390625, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 21.3231, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.004615739672282483, | |
| "eval_loss": 6.653355121612549, | |
| "eval_runtime": 116.8995, | |
| "eval_samples_per_second": 312.14, | |
| "eval_steps_per_second": 78.041, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.004661897069005308, | |
| "grad_norm": 13000.0380859375, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 32.0844, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.004708054465728133, | |
| "grad_norm": 12532.7470703125, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 31.0504, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.004754211862450957, | |
| "grad_norm": 21590.57421875, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 30.8403, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.004800369259173783, | |
| "grad_norm": 16138.33203125, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 29.4348, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.004846526655896607, | |
| "grad_norm": 12663.6875, | |
| "learning_rate": 5e-05, | |
| "loss": 29.8385, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.004892684052619433, | |
| "grad_norm": 17485.61328125, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 29.5445, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.004938841449342257, | |
| "grad_norm": 12701.091796875, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 28.8904, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.004984998846065082, | |
| "grad_norm": 17760.87109375, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 29.5446, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.005031156242787907, | |
| "grad_norm": 15029.7021484375, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 28.3419, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.005077313639510732, | |
| "grad_norm": 17356.201171875, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 29.5726, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.005123471036233556, | |
| "grad_norm": 18930.1796875, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 30.3205, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.005169628432956382, | |
| "grad_norm": 33717.44140625, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 29.4096, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.005215785829679206, | |
| "grad_norm": 12564.3232421875, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 27.909, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.005261943226402031, | |
| "grad_norm": 15485.4404296875, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 27.2055, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.005308100623124856, | |
| "grad_norm": 14564.55859375, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 27.1391, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0053542580198476806, | |
| "grad_norm": 28163.482421875, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 27.5595, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.005400415416570505, | |
| "grad_norm": 15327.2060546875, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 27.7242, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.0054465728132933305, | |
| "grad_norm": 12743.4599609375, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 29.3176, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.005492730210016155, | |
| "grad_norm": 17723.337890625, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 27.6727, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0055388876067389795, | |
| "grad_norm": 11900.8056640625, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 27.0558, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.005585045003461805, | |
| "grad_norm": 30333.73828125, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 25.7016, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.0056312024001846294, | |
| "grad_norm": 14563.4892578125, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 24.8602, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.005677359796907455, | |
| "grad_norm": 18867.26953125, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 25.5498, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.005723517193630279, | |
| "grad_norm": 17261.439453125, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 24.7577, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.005769674590353104, | |
| "grad_norm": 14557.5341796875, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 24.8685, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.005815831987075929, | |
| "grad_norm": 55429.828125, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 24.7386, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.005861989383798754, | |
| "grad_norm": 37720.05859375, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 25.7687, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.005908146780521578, | |
| "grad_norm": 13499.6328125, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 25.2399, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.005954304177244404, | |
| "grad_norm": 21841.525390625, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 24.0868, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.006000461573967228, | |
| "grad_norm": 18137.44921875, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 25.1236, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.006046618970690053, | |
| "grad_norm": 18170.18359375, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 24.8144, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.006092776367412878, | |
| "grad_norm": 12616.4169921875, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 26.8521, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.006138933764135703, | |
| "grad_norm": 14475.0224609375, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 24.1187, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.006185091160858527, | |
| "grad_norm": 13354.6005859375, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 24.8671, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.006231248557581353, | |
| "grad_norm": 34306.61328125, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 24.901, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.006277405954304177, | |
| "grad_norm": 14065.1328125, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 23.7815, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.006323563351027002, | |
| "grad_norm": 13215.8115234375, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 22.6957, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.006369720747749827, | |
| "grad_norm": 32365.15234375, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 22.3192, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.006415878144472652, | |
| "grad_norm": 21075.822265625, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 23.3173, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.006462035541195477, | |
| "grad_norm": 17104.109375, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 25.3204, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.0065081929379183015, | |
| "grad_norm": 23697.595703125, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 22.6917, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.006554350334641126, | |
| "grad_norm": 20324.15625, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 22.6145, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.006600507731363951, | |
| "grad_norm": 20542.833984375, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 22.4909, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.006646665128086776, | |
| "grad_norm": 43104.8125, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 24.1033, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0066928225248096005, | |
| "grad_norm": 16736.55078125, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 23.1356, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.006738979921532426, | |
| "grad_norm": 26106.3359375, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 22.9239, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.00678513731825525, | |
| "grad_norm": 16372.2353515625, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 22.2321, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.006831294714978075, | |
| "grad_norm": 27215.10546875, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 22.7279, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.0068774521117009, | |
| "grad_norm": 16365.025390625, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 23.7236, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.006923609508423725, | |
| "grad_norm": 39346.94140625, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 19.913, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.006923609508423725, | |
| "eval_loss": 6.5524163246154785, | |
| "eval_runtime": 117.0961, | |
| "eval_samples_per_second": 311.616, | |
| "eval_steps_per_second": 77.91, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.006969766905146549, | |
| "grad_norm": 13133.1728515625, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 30.7831, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.007015924301869375, | |
| "grad_norm": 14437.8154296875, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 32.2855, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.007062081698592199, | |
| "grad_norm": 12261.015625, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 30.5466, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.007108239095315024, | |
| "grad_norm": 14204.9453125, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 29.4702, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.007154396492037849, | |
| "grad_norm": 35599.56640625, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 30.0853, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.007200553888760674, | |
| "grad_norm": 15908.6923828125, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 30.2363, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.007246711285483499, | |
| "grad_norm": 21062.623046875, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 29.9957, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.007292868682206324, | |
| "grad_norm": 13865.328125, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 29.3194, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.007339026078929148, | |
| "grad_norm": 18418.775390625, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 30.0034, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.0073851834756519736, | |
| "grad_norm": 17927.546875, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 27.3433, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.007431340872374798, | |
| "grad_norm": 16150.69140625, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 29.3834, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.007477498269097623, | |
| "grad_norm": 53190.5546875, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 28.7297, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.007523655665820448, | |
| "grad_norm": 13157.083984375, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 29.4415, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0075698130625432725, | |
| "grad_norm": 18151.259765625, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 26.6588, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.007615970459266097, | |
| "grad_norm": 14146.330078125, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 28.4469, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0076621278559889225, | |
| "grad_norm": 54733.3046875, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 26.0468, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.007708285252711747, | |
| "grad_norm": 18280.759765625, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 29.2003, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.0077544426494345715, | |
| "grad_norm": 22410.99609375, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 27.8878, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.007800600046157397, | |
| "grad_norm": 11329.37109375, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 26.4854, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.007846757442880221, | |
| "grad_norm": 20874.708984375, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 26.5411, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.007892914839603046, | |
| "grad_norm": 19707.91796875, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 26.1978, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.00793907223632587, | |
| "grad_norm": 78664.015625, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 26.7962, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.007985229633048697, | |
| "grad_norm": 17584.62890625, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 25.9352, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.008031387029771521, | |
| "grad_norm": 19179.896484375, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 26.8773, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.008077544426494346, | |
| "grad_norm": 17171.498046875, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 25.3397, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.00812370182321717, | |
| "grad_norm": 17196.4609375, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 25.0458, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.008169859219939995, | |
| "grad_norm": 20215.6796875, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 26.9744, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.00821601661666282, | |
| "grad_norm": 18055.56640625, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 25.2058, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.008262174013385646, | |
| "grad_norm": 17127.599609375, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 23.8571, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.00830833141010847, | |
| "grad_norm": 21535.5078125, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 26.9367, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.008354488806831295, | |
| "grad_norm": 20788.958984375, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 22.6921, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.00840064620355412, | |
| "grad_norm": 17802.25, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 25.2687, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.008446803600276944, | |
| "grad_norm": 15224.77734375, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 25.1923, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.00849296099699977, | |
| "grad_norm": 25008.404296875, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 23.9257, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.008539118393722595, | |
| "grad_norm": 19298.392578125, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 23.0269, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.008585275790445419, | |
| "grad_norm": 16348.4912109375, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 23.2064, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.008631433187168244, | |
| "grad_norm": 32499.119140625, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 23.963, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.008677590583891068, | |
| "grad_norm": 19804.919921875, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 23.3042, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.008723747980613893, | |
| "grad_norm": 13096.90625, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 23.024, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.008769905377336719, | |
| "grad_norm": 13534.07421875, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 24.1031, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.008816062774059543, | |
| "grad_norm": 9464.7373046875, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 22.4278, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.008862220170782368, | |
| "grad_norm": 15782.4638671875, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 22.7746, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.008908377567505192, | |
| "grad_norm": 19455.8515625, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 23.3504, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.008954534964228017, | |
| "grad_norm": 25072.66796875, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 21.0983, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.009000692360950842, | |
| "grad_norm": 17993.9296875, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 23.3598, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.009046849757673668, | |
| "grad_norm": 23557.642578125, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 22.9573, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.009093007154396492, | |
| "grad_norm": 19041.326171875, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 23.1464, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.009139164551119317, | |
| "grad_norm": 26971.62890625, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 20.0905, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.009185321947842141, | |
| "grad_norm": 14283.2109375, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 21.5863, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.009231479344564966, | |
| "grad_norm": 17224.2734375, | |
| "learning_rate": 0.0, | |
| "loss": 20.5303, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.009231479344564966, | |
| "eval_loss": 6.4810471534729, | |
| "eval_runtime": 117.0685, | |
| "eval_samples_per_second": 311.689, | |
| "eval_steps_per_second": 77.929, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 330900386611200.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |