| { | |
| "best_metric": 0.13024483621120453, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 1.0695187165775402, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0071301247771836, | |
| "grad_norm": 4.8426337242126465, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2642, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0071301247771836, | |
| "eval_loss": 3.368955373764038, | |
| "eval_runtime": 17.4805, | |
| "eval_samples_per_second": 13.501, | |
| "eval_steps_per_second": 3.375, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0142602495543672, | |
| "grad_norm": 4.2209672927856445, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4353, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0213903743315508, | |
| "grad_norm": 5.881971836090088, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4871, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0285204991087344, | |
| "grad_norm": 4.883190155029297, | |
| "learning_rate": 4e-05, | |
| "loss": 2.3509, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.035650623885918005, | |
| "grad_norm": 3.3302810192108154, | |
| "learning_rate": 5e-05, | |
| "loss": 2.1371, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0427807486631016, | |
| "grad_norm": 3.647418737411499, | |
| "learning_rate": 6e-05, | |
| "loss": 1.785, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.049910873440285206, | |
| "grad_norm": 6.729278564453125, | |
| "learning_rate": 7e-05, | |
| "loss": 1.7518, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0570409982174688, | |
| "grad_norm": 7.104023456573486, | |
| "learning_rate": 8e-05, | |
| "loss": 1.846, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06417112299465241, | |
| "grad_norm": 5.479939937591553, | |
| "learning_rate": 9e-05, | |
| "loss": 1.7269, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07130124777183601, | |
| "grad_norm": 4.303497314453125, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1691, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0784313725490196, | |
| "grad_norm": 8.257266998291016, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 1.5674, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "grad_norm": 4.5696797370910645, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 1.3183, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.09269162210338681, | |
| "grad_norm": 5.195154190063477, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 1.0984, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09982174688057041, | |
| "grad_norm": 6.245325565338135, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 1.3965, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10695187165775401, | |
| "grad_norm": 6.468475341796875, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 1.2189, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1140819964349376, | |
| "grad_norm": 7.070935249328613, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 1.2871, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 5.198769569396973, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 1.1191, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12834224598930483, | |
| "grad_norm": 4.482051849365234, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.8275, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1354723707664884, | |
| "grad_norm": 4.659496784210205, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.6453, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.14260249554367202, | |
| "grad_norm": 10.78939151763916, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 2.0706, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1497326203208556, | |
| "grad_norm": 7.3973164558410645, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 1.4084, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1568627450980392, | |
| "grad_norm": 15.801050186157227, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 1.3198, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.16399286987522282, | |
| "grad_norm": 13.212759017944336, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 1.3635, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "grad_norm": 8.838081359863281, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 1.4169, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.17825311942959002, | |
| "grad_norm": 8.929149627685547, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 1.7153, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.18538324420677363, | |
| "grad_norm": 6.735871315002441, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 1.1534, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1925133689839572, | |
| "grad_norm": 3.1337316036224365, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.8431, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19964349376114082, | |
| "grad_norm": 4.328713417053223, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.917, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.20677361853832443, | |
| "grad_norm": 17.148948669433594, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.551, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.21390374331550802, | |
| "grad_norm": 16.10226058959961, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 1.9147, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.22103386809269163, | |
| "grad_norm": 8.083459854125977, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 1.429, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2281639928698752, | |
| "grad_norm": 18.02008819580078, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 2.2392, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 9.912565231323242, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 1.2989, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 7.664438247680664, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.9611, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24955436720142601, | |
| "grad_norm": 7.64049768447876, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.8516, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "grad_norm": 17.550363540649414, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 2.9945, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2638146167557932, | |
| "grad_norm": 13.104034423828125, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 2.3501, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2709447415329768, | |
| "grad_norm": 9.25308895111084, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 1.6601, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.27807486631016043, | |
| "grad_norm": 4.239262104034424, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.8481, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.28520499108734404, | |
| "grad_norm": 2.5438952445983887, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.6431, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.29233511586452765, | |
| "grad_norm": 2.9467761516571045, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.5759, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2994652406417112, | |
| "grad_norm": 2.582361936569214, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.555, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.3065953654188948, | |
| "grad_norm": 3.7745120525360107, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.6859, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3137254901960784, | |
| "grad_norm": 2.906780481338501, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.5011, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.32085561497326204, | |
| "grad_norm": 2.0319948196411133, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.3182, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.32798573975044565, | |
| "grad_norm": 4.595136642456055, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.5804, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.33511586452762926, | |
| "grad_norm": 3.3042995929718018, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.5566, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "grad_norm": 2.7046265602111816, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.4686, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3493761140819964, | |
| "grad_norm": 2.852973222732544, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.206, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.35650623885918004, | |
| "grad_norm": 3.929203987121582, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.2773, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35650623885918004, | |
| "eval_loss": 0.6693209409713745, | |
| "eval_runtime": 17.8065, | |
| "eval_samples_per_second": 13.254, | |
| "eval_steps_per_second": 3.313, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.8368263244628906, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 0.2142, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.37076648841354726, | |
| "grad_norm": 7.45966911315918, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 0.2947, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3778966131907308, | |
| "grad_norm": 3.3990819454193115, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 0.258, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3850267379679144, | |
| "grad_norm": 2.759211778640747, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.3852, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.39215686274509803, | |
| "grad_norm": 4.810303688049316, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.5276, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.39928698752228164, | |
| "grad_norm": 5.013471603393555, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.3654, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.40641711229946526, | |
| "grad_norm": 4.822093486785889, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.4552, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.41354723707664887, | |
| "grad_norm": 4.117964744567871, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.6252, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4206773618538324, | |
| "grad_norm": 3.389159679412842, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.3813, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "grad_norm": 4.948807239532471, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.4031, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.43493761140819964, | |
| "grad_norm": 2.4192276000976562, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.2109, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.44206773618538325, | |
| "grad_norm": 2.610354423522949, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.313, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.44919786096256686, | |
| "grad_norm": 5.960036754608154, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.5946, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4563279857397504, | |
| "grad_norm": 3.539583921432495, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.4566, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.46345811051693403, | |
| "grad_norm": 4.794179916381836, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.4307, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 5.651267051696777, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.6152, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.47771836007130125, | |
| "grad_norm": 8.529621124267578, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.6235, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 3.700084924697876, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.3216, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4919786096256685, | |
| "grad_norm": 3.2698822021484375, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.443, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.49910873440285203, | |
| "grad_norm": 5.935876846313477, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.515, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5062388591800356, | |
| "grad_norm": 13.412459373474121, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 1.974, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "grad_norm": 10.322650909423828, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 1.6109, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5204991087344029, | |
| "grad_norm": 8.35332202911377, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 1.229, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5276292335115864, | |
| "grad_norm": 5.663819313049316, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.7897, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5347593582887701, | |
| "grad_norm": 4.571802139282227, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.4717, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5418894830659536, | |
| "grad_norm": 2.7498021125793457, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.4664, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5490196078431373, | |
| "grad_norm": 2.3304195404052734, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.3486, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5561497326203209, | |
| "grad_norm": 4.024349212646484, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.447, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5632798573975044, | |
| "grad_norm": 5.358171463012695, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.3599, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5704099821746881, | |
| "grad_norm": 1.9654799699783325, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.2669, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5775401069518716, | |
| "grad_norm": 1.602853536605835, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.2146, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5846702317290553, | |
| "grad_norm": 3.970834255218506, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.3121, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5918003565062389, | |
| "grad_norm": 1.8870638608932495, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.1878, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "grad_norm": 5.286268711090088, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.1967, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 5.626857757568359, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.1777, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6131907308377896, | |
| "grad_norm": 3.088507890701294, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.1277, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6203208556149733, | |
| "grad_norm": 2.9896271228790283, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.3302, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6274509803921569, | |
| "grad_norm": 4.0551629066467285, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.3202, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6345811051693404, | |
| "grad_norm": 2.5737249851226807, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.2139, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6417112299465241, | |
| "grad_norm": 2.4582886695861816, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.2197, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6488413547237076, | |
| "grad_norm": 1.8644245862960815, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.1425, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6559714795008913, | |
| "grad_norm": 3.0225026607513428, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.1253, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6631016042780749, | |
| "grad_norm": 3.376275062561035, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.2162, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6702317290552585, | |
| "grad_norm": 3.654719591140747, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.2584, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6773618538324421, | |
| "grad_norm": 1.9528889656066895, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.1685, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "grad_norm": 1.6919232606887817, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.0871, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6916221033868093, | |
| "grad_norm": 1.6383646726608276, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.2077, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6987522281639929, | |
| "grad_norm": 2.661104440689087, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.2855, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 2.2017295360565186, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.2031, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7130124777183601, | |
| "grad_norm": 2.1253581047058105, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.2846, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7130124777183601, | |
| "eval_loss": 0.3400123119354248, | |
| "eval_runtime": 17.7933, | |
| "eval_samples_per_second": 13.263, | |
| "eval_steps_per_second": 3.316, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7201426024955436, | |
| "grad_norm": 2.9856224060058594, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.2371, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 4.570830821990967, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.3743, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7344028520499108, | |
| "grad_norm": 4.89229679107666, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.1655, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7415329768270945, | |
| "grad_norm": 2.2233779430389404, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.2178, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7486631016042781, | |
| "grad_norm": 4.377749443054199, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4931, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7557932263814616, | |
| "grad_norm": 7.703902721405029, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.7411, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7629233511586453, | |
| "grad_norm": 5.32340145111084, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.6452, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "grad_norm": 3.3561203479766846, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.5094, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7771836007130125, | |
| "grad_norm": 4.73153829574585, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.4919, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 3.925480365753174, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.3558, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7914438502673797, | |
| "grad_norm": 1.5577930212020874, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.2809, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7985739750445633, | |
| "grad_norm": 1.836594820022583, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.2073, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.8057040998217468, | |
| "grad_norm": 3.001798391342163, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.1924, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8128342245989305, | |
| "grad_norm": 2.8092217445373535, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.1937, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.8199643493761141, | |
| "grad_norm": 1.406787633895874, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.2105, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8270944741532977, | |
| "grad_norm": 2.4974722862243652, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.2551, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8342245989304813, | |
| "grad_norm": 3.2987828254699707, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.2149, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8413547237076648, | |
| "grad_norm": 2.0800249576568604, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.1475, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 1.760972261428833, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.2447, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "grad_norm": 1.7634668350219727, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.2162, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8627450980392157, | |
| "grad_norm": 3.028109073638916, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.2006, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8698752228163993, | |
| "grad_norm": 2.379579544067383, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.1466, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8770053475935828, | |
| "grad_norm": 1.5792676210403442, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.1942, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.8841354723707665, | |
| "grad_norm": 4.645526885986328, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.2629, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8912655971479501, | |
| "grad_norm": 1.0095664262771606, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.0914, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8983957219251337, | |
| "grad_norm": 1.4276849031448364, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.1227, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.9055258467023173, | |
| "grad_norm": 0.8317928314208984, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.0315, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.9126559714795008, | |
| "grad_norm": 1.818306803703308, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.2228, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.9197860962566845, | |
| "grad_norm": 4.03300666809082, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.2389, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.9269162210338681, | |
| "grad_norm": 3.8486204147338867, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.2066, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9340463458110517, | |
| "grad_norm": 4.065092086791992, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.2141, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 1.6100399494171143, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.1784, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.948306595365419, | |
| "grad_norm": 2.6661133766174316, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.3567, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9554367201426025, | |
| "grad_norm": 4.373043537139893, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.2701, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9625668449197861, | |
| "grad_norm": 2.485142707824707, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.2768, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 2.9201934337615967, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.2867, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9768270944741533, | |
| "grad_norm": 2.0322253704071045, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.1691, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.983957219251337, | |
| "grad_norm": 3.244508743286133, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.2272, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9910873440285205, | |
| "grad_norm": 3.3805246353149414, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.2056, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9982174688057041, | |
| "grad_norm": 2.5575058460235596, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.3891, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0053475935828877, | |
| "grad_norm": 4.942562580108643, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.5803, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.0124777183600713, | |
| "grad_norm": 2.94242262840271, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.2574, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.0196078431372548, | |
| "grad_norm": 2.386302947998047, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.2749, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.0267379679144386, | |
| "grad_norm": 2.1066482067108154, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.2647, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.0338680926916222, | |
| "grad_norm": 1.3415483236312866, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.2325, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0409982174688057, | |
| "grad_norm": 1.8504228591918945, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.2718, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0481283422459893, | |
| "grad_norm": 2.384942054748535, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.211, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0552584670231728, | |
| "grad_norm": 3.9082236289978027, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.2135, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0623885918003566, | |
| "grad_norm": 2.6325087547302246, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.1609, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0695187165775402, | |
| "grad_norm": 1.9559813737869263, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.1264, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0695187165775402, | |
| "eval_loss": 0.13024483621120453, | |
| "eval_runtime": 17.7654, | |
| "eval_samples_per_second": 13.284, | |
| "eval_steps_per_second": 3.321, | |
| "step": 150 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2608847261453517e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |