| { | |
| "best_metric": 2.083789825439453, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 1.1111111111111112, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005555555555555556, | |
| "grad_norm": 4.495181560516357, | |
| "learning_rate": 1e-05, | |
| "loss": 3.101, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005555555555555556, | |
| "eval_loss": 2.971418619155884, | |
| "eval_runtime": 5.61, | |
| "eval_samples_per_second": 54.189, | |
| "eval_steps_per_second": 13.547, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011111111111111112, | |
| "grad_norm": 4.676938533782959, | |
| "learning_rate": 2e-05, | |
| "loss": 2.9841, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.016666666666666666, | |
| "grad_norm": 4.591571807861328, | |
| "learning_rate": 3e-05, | |
| "loss": 2.8521, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.022222222222222223, | |
| "grad_norm": 4.44553279876709, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8005, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 4.705663204193115, | |
| "learning_rate": 5e-05, | |
| "loss": 2.6049, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03333333333333333, | |
| "grad_norm": 6.448979377746582, | |
| "learning_rate": 6e-05, | |
| "loss": 2.5441, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03888888888888889, | |
| "grad_norm": 4.940516948699951, | |
| "learning_rate": 7e-05, | |
| "loss": 2.6022, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 8.077047348022461, | |
| "learning_rate": 8e-05, | |
| "loss": 2.3294, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 9.639278411865234, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4294, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 7.703786849975586, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3853, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06111111111111111, | |
| "grad_norm": 4.978358268737793, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 2.3943, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06666666666666667, | |
| "grad_norm": 6.4761810302734375, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 2.4724, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07222222222222222, | |
| "grad_norm": 5.821319580078125, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 2.3078, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07777777777777778, | |
| "grad_norm": 3.4821279048919678, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 2.3868, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 3.1514482498168945, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 2.2606, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 3.100623607635498, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 2.3348, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09444444444444444, | |
| "grad_norm": 3.1726274490356445, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 2.1337, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.2123072147369385, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 2.3387, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10555555555555556, | |
| "grad_norm": 3.102133274078369, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 2.2055, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 2.907766342163086, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 2.1471, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11666666666666667, | |
| "grad_norm": 2.9859018325805664, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 2.2325, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12222222222222222, | |
| "grad_norm": 3.149094820022583, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 2.3846, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12777777777777777, | |
| "grad_norm": 3.331679105758667, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 2.229, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 3.0126988887786865, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 2.1966, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 3.1853792667388916, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 2.1256, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14444444444444443, | |
| "grad_norm": 2.80102801322937, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 2.1837, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.3799564838409424, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 2.3204, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15555555555555556, | |
| "grad_norm": 3.184105396270752, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 2.0423, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.16111111111111112, | |
| "grad_norm": 3.0086214542388916, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.324, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 2.793081760406494, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 2.1956, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17222222222222222, | |
| "grad_norm": 2.8765249252319336, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 2.2785, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 3.536093235015869, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 2.1477, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18333333333333332, | |
| "grad_norm": 3.01804256439209, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 2.1674, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.18888888888888888, | |
| "grad_norm": 2.828728437423706, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 2.3468, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 2.8197877407073975, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 2.3459, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.400102376937866, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 2.2952, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.20555555555555555, | |
| "grad_norm": 3.0070595741271973, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 2.2848, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2111111111111111, | |
| "grad_norm": 2.8950300216674805, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 2.1191, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.21666666666666667, | |
| "grad_norm": 2.8364169597625732, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 2.1494, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 2.7214770317077637, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 2.3524, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22777777777777777, | |
| "grad_norm": 2.8703582286834717, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 2.2382, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.23333333333333334, | |
| "grad_norm": 2.726114273071289, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 2.1445, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2388888888888889, | |
| "grad_norm": 2.6854159832000732, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 2.2542, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.24444444444444444, | |
| "grad_norm": 2.903628349304199, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 2.1134, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 3.1313066482543945, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 2.3135, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.25555555555555554, | |
| "grad_norm": 2.6870551109313965, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 2.3362, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.2611111111111111, | |
| "grad_norm": 2.718916177749634, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 2.1098, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 2.618040084838867, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.1239, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2722222222222222, | |
| "grad_norm": 2.5252163410186768, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 2.1402, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 2.4961345195770264, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 2.0671, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "eval_loss": 2.1783969402313232, | |
| "eval_runtime": 5.3061, | |
| "eval_samples_per_second": 57.293, | |
| "eval_steps_per_second": 14.323, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2833333333333333, | |
| "grad_norm": 2.598252296447754, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 2.2901, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.28888888888888886, | |
| "grad_norm": 2.652029037475586, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 2.285, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.29444444444444445, | |
| "grad_norm": 2.5119314193725586, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 2.2247, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.6834304332733154, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 2.0835, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 2.6870522499084473, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 2.244, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 2.425607681274414, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 2.1662, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.31666666666666665, | |
| "grad_norm": 2.5925562381744385, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 2.3357, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.32222222222222224, | |
| "grad_norm": 2.4682435989379883, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 2.0191, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3277777777777778, | |
| "grad_norm": 2.5179603099823, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 2.2036, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 2.506409168243408, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 2.0915, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3388888888888889, | |
| "grad_norm": 2.6131038665771484, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 2.228, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.34444444444444444, | |
| "grad_norm": 2.6575188636779785, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 2.3363, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.7023885250091553, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 2.0732, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.679713487625122, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 2.0837, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 2.5868608951568604, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 2.126, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.36666666666666664, | |
| "grad_norm": 2.431974411010742, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 2.075, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.37222222222222223, | |
| "grad_norm": 2.66987681388855, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.011, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.37777777777777777, | |
| "grad_norm": 2.872534990310669, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 2.2826, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.38333333333333336, | |
| "grad_norm": 2.536738872528076, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 2.2653, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 2.6215434074401855, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 2.0538, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.39444444444444443, | |
| "grad_norm": 2.6606903076171875, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 2.2091, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.857592821121216, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 2.2052, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.40555555555555556, | |
| "grad_norm": 2.7153890132904053, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 2.1232, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4111111111111111, | |
| "grad_norm": 2.545321226119995, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 1.9781, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 2.9500622749328613, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 2.1488, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4222222222222222, | |
| "grad_norm": 2.9101979732513428, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 2.3642, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.42777777777777776, | |
| "grad_norm": 2.7557241916656494, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 2.2467, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.43333333333333335, | |
| "grad_norm": 2.8893096446990967, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 2.1667, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4388888888888889, | |
| "grad_norm": 2.5453028678894043, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 2.048, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 2.5639288425445557, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 2.1292, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.5094361305236816, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 2.2029, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.45555555555555555, | |
| "grad_norm": 2.5174753665924072, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 2.0341, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.46111111111111114, | |
| "grad_norm": 2.728111743927002, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 2.2932, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4666666666666667, | |
| "grad_norm": 2.6332783699035645, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 2.2324, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 2.5300662517547607, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 2.119, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4777777777777778, | |
| "grad_norm": 2.7975356578826904, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.03, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.48333333333333334, | |
| "grad_norm": 2.5400803089141846, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 2.2483, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 2.830970525741577, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 2.1029, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.49444444444444446, | |
| "grad_norm": 3.057438611984253, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 2.3245, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.6523685455322266, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 2.0426, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5055555555555555, | |
| "grad_norm": 3.5712132453918457, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 2.4352, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5111111111111111, | |
| "grad_norm": 4.198061943054199, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 2.0641, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5166666666666667, | |
| "grad_norm": 3.0330898761749268, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 2.1259, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5222222222222223, | |
| "grad_norm": 2.8275649547576904, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 2.2233, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5277777777777778, | |
| "grad_norm": 2.2473394870758057, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 1.9539, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 2.5870609283447266, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 2.1022, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5388888888888889, | |
| "grad_norm": 2.9209086894989014, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 2.0656, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5444444444444444, | |
| "grad_norm": 3.0494041442871094, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 2.1193, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.0358989238739014, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.2129, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 2.6447176933288574, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 2.1513, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "eval_loss": 2.1226329803466797, | |
| "eval_runtime": 5.3168, | |
| "eval_samples_per_second": 57.178, | |
| "eval_steps_per_second": 14.294, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5611111111111111, | |
| "grad_norm": 2.384932041168213, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 2.2277, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5666666666666667, | |
| "grad_norm": 2.800978422164917, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 2.0029, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5722222222222222, | |
| "grad_norm": 2.901449680328369, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 2.0475, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 2.8509185314178467, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 2.1435, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5833333333333334, | |
| "grad_norm": 2.8648464679718018, | |
| "learning_rate": 5e-05, | |
| "loss": 2.0758, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5888888888888889, | |
| "grad_norm": 2.626039743423462, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 2.0858, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5944444444444444, | |
| "grad_norm": 2.3633663654327393, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 1.9879, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.501535415649414, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 2.2116, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6055555555555555, | |
| "grad_norm": 2.617394208908081, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 2.138, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6111111111111112, | |
| "grad_norm": 2.579564094543457, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 2.1465, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6166666666666667, | |
| "grad_norm": 2.570713520050049, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 2.09, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 2.512786626815796, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 2.1812, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6277777777777778, | |
| "grad_norm": 2.4086341857910156, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 2.0978, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6333333333333333, | |
| "grad_norm": 2.4698939323425293, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 2.1408, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6388888888888888, | |
| "grad_norm": 2.347113847732544, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 1.9826, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6444444444444445, | |
| "grad_norm": 2.5002825260162354, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 1.9889, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 2.45943546295166, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 2.2029, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6555555555555556, | |
| "grad_norm": 2.230510711669922, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 2.0398, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6611111111111111, | |
| "grad_norm": 2.3624422550201416, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 1.999, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 2.53621506690979, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 2.2093, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6722222222222223, | |
| "grad_norm": 2.5943751335144043, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 2.1875, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6777777777777778, | |
| "grad_norm": 2.359910011291504, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 2.0942, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6833333333333333, | |
| "grad_norm": 2.4836995601654053, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 1.994, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6888888888888889, | |
| "grad_norm": 2.694197416305542, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.2861, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 2.2708804607391357, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 2.0382, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.550246238708496, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 2.1292, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7055555555555556, | |
| "grad_norm": 2.4213714599609375, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 2.1918, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 2.4564764499664307, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 2.0905, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7166666666666667, | |
| "grad_norm": 2.3561134338378906, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 2.1168, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7222222222222222, | |
| "grad_norm": 2.6135506629943848, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 2.1918, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7277777777777777, | |
| "grad_norm": 2.4894702434539795, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 2.1104, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7333333333333333, | |
| "grad_norm": 2.3881235122680664, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 2.0762, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7388888888888889, | |
| "grad_norm": 2.4601657390594482, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 2.0198, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7444444444444445, | |
| "grad_norm": 2.5906982421875, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 2.1932, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.688217878341675, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 2.1464, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 3.826042413711548, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 2.1686, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7611111111111111, | |
| "grad_norm": 4.0038299560546875, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 2.1037, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7666666666666667, | |
| "grad_norm": 3.4031991958618164, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 2.0544, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7722222222222223, | |
| "grad_norm": 3.369370698928833, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 1.9373, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 2.829681158065796, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 1.946, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7833333333333333, | |
| "grad_norm": 2.5274407863616943, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 2.1871, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7888888888888889, | |
| "grad_norm": 2.421165704727173, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 2.0364, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7944444444444444, | |
| "grad_norm": 2.4030814170837402, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.17, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.948323965072632, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 2.2331, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8055555555555556, | |
| "grad_norm": 2.916417121887207, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 2.0981, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8111111111111111, | |
| "grad_norm": 2.9353768825531006, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 2.069, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.8166666666666667, | |
| "grad_norm": 2.745278835296631, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.0316, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8222222222222222, | |
| "grad_norm": 3.0665552616119385, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 1.9733, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8277777777777777, | |
| "grad_norm": 2.384312391281128, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.0964, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 2.5258023738861084, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.2636, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "eval_loss": 2.093482494354248, | |
| "eval_runtime": 5.3128, | |
| "eval_samples_per_second": 57.22, | |
| "eval_steps_per_second": 14.305, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8388888888888889, | |
| "grad_norm": 2.3465325832366943, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 2.0406, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 2.519423723220825, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 1.968, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.4950923919677734, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 2.1471, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8555555555555555, | |
| "grad_norm": 2.3640804290771484, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 1.9919, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8611111111111112, | |
| "grad_norm": 2.518543243408203, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 2.1513, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8666666666666667, | |
| "grad_norm": 2.504188060760498, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 2.0638, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8722222222222222, | |
| "grad_norm": 2.3272154331207275, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 1.9518, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8777777777777778, | |
| "grad_norm": 2.5546417236328125, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 2.059, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8833333333333333, | |
| "grad_norm": 2.6515941619873047, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 2.0171, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 2.4597809314727783, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 1.9185, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8944444444444445, | |
| "grad_norm": 2.5070078372955322, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 2.0902, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.5468130111694336, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.9904, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9055555555555556, | |
| "grad_norm": 2.375117301940918, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 2.0713, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9111111111111111, | |
| "grad_norm": 2.35435152053833, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 1.9669, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9166666666666666, | |
| "grad_norm": 2.3441059589385986, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 2.0499, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9222222222222223, | |
| "grad_norm": 2.365280866622925, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 2.0398, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9277777777777778, | |
| "grad_norm": 2.42503023147583, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 2.1046, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 2.4105420112609863, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 1.9839, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9388888888888889, | |
| "grad_norm": 2.3553500175476074, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 2.0738, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9444444444444444, | |
| "grad_norm": 2.450026750564575, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 1.9891, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.6159915924072266, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 2.0233, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9555555555555556, | |
| "grad_norm": 2.5502679347991943, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 2.2083, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9611111111111111, | |
| "grad_norm": 2.5647997856140137, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 2.2597, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9666666666666667, | |
| "grad_norm": 2.8217949867248535, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 2.058, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 2.727836847305298, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 2.0708, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 2.5638632774353027, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 2.0893, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9833333333333333, | |
| "grad_norm": 2.7340164184570312, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 2.1115, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9888888888888889, | |
| "grad_norm": 2.7915878295898438, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 2.0715, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.9944444444444445, | |
| "grad_norm": 3.0439765453338623, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 2.1619, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.057664394378662, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 1.9983, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0055555555555555, | |
| "grad_norm": 2.75246524810791, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.9903, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.011111111111111, | |
| "grad_norm": 2.7853286266326904, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 1.9104, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.0166666666666666, | |
| "grad_norm": 2.938938856124878, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 1.8358, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.0222222222222221, | |
| "grad_norm": 2.8226351737976074, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 1.6926, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.0277777777777777, | |
| "grad_norm": 2.8052713871002197, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 1.7155, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.0333333333333334, | |
| "grad_norm": 2.8661975860595703, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 1.8953, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.038888888888889, | |
| "grad_norm": 2.903599262237549, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 1.8219, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.0444444444444445, | |
| "grad_norm": 2.703796148300171, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 1.8375, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.4769649505615234, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 1.9221, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.0555555555555556, | |
| "grad_norm": 2.5778965950012207, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 1.8365, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0611111111111111, | |
| "grad_norm": 2.750174045562744, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 1.8767, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 2.631240129470825, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 1.7292, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.0722222222222222, | |
| "grad_norm": 2.6291799545288086, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 1.8598, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.0777777777777777, | |
| "grad_norm": 2.477027416229248, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 1.7905, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.0833333333333333, | |
| "grad_norm": 2.389650344848633, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 1.699, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.0888888888888888, | |
| "grad_norm": 2.764362335205078, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 1.6669, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.0944444444444446, | |
| "grad_norm": 2.358754873275757, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 1.7975, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 2.6524245738983154, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 1.7758, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.1055555555555556, | |
| "grad_norm": 2.298368453979492, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 1.771, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 2.4102272987365723, | |
| "learning_rate": 0.0, | |
| "loss": 1.8416, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "eval_loss": 2.083789825439453, | |
| "eval_runtime": 5.3127, | |
| "eval_samples_per_second": 57.221, | |
| "eval_steps_per_second": 14.305, | |
| "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": 1.54569648635904e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |