| { | |
| "best_metric": 1.2759346961975098, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.3401360544217687, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0017006802721088435, | |
| "grad_norm": 25.771764755249023, | |
| "learning_rate": 1e-05, | |
| "loss": 5.4753, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0017006802721088435, | |
| "eval_loss": 1.5546903610229492, | |
| "eval_runtime": 11.5445, | |
| "eval_samples_per_second": 85.755, | |
| "eval_steps_per_second": 21.482, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003401360544217687, | |
| "grad_norm": 24.665054321289062, | |
| "learning_rate": 2e-05, | |
| "loss": 5.4607, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00510204081632653, | |
| "grad_norm": 25.340229034423828, | |
| "learning_rate": 3e-05, | |
| "loss": 5.9128, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.006802721088435374, | |
| "grad_norm": 23.048992156982422, | |
| "learning_rate": 4e-05, | |
| "loss": 5.8313, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.008503401360544218, | |
| "grad_norm": 24.37986183166504, | |
| "learning_rate": 5e-05, | |
| "loss": 5.9566, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01020408163265306, | |
| "grad_norm": 25.966943740844727, | |
| "learning_rate": 6e-05, | |
| "loss": 6.003, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.011904761904761904, | |
| "grad_norm": 23.275394439697266, | |
| "learning_rate": 7e-05, | |
| "loss": 5.9538, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.013605442176870748, | |
| "grad_norm": 25.879196166992188, | |
| "learning_rate": 8e-05, | |
| "loss": 5.6283, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.015306122448979591, | |
| "grad_norm": 25.57696533203125, | |
| "learning_rate": 9e-05, | |
| "loss": 6.1644, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.017006802721088437, | |
| "grad_norm": 29.063772201538086, | |
| "learning_rate": 0.0001, | |
| "loss": 6.1406, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01870748299319728, | |
| "grad_norm": 28.84071922302246, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 6.0843, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 35.08443832397461, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 6.3408, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.022108843537414966, | |
| "grad_norm": 39.56977844238281, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 6.089, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.023809523809523808, | |
| "grad_norm": 36.34089660644531, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 6.1436, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.025510204081632654, | |
| "grad_norm": 35.21949768066406, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 6.0946, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.027210884353741496, | |
| "grad_norm": 30.47332191467285, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 6.2502, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02891156462585034, | |
| "grad_norm": 34.45054244995117, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 5.9068, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.030612244897959183, | |
| "grad_norm": 34.17898941040039, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 6.2611, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03231292517006803, | |
| "grad_norm": 31.242021560668945, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 6.0134, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.034013605442176874, | |
| "grad_norm": 54.89870834350586, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 5.9536, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03571428571428571, | |
| "grad_norm": 34.17111587524414, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 6.356, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03741496598639456, | |
| "grad_norm": 35.953392028808594, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 6.1854, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0391156462585034, | |
| "grad_norm": 38.30813217163086, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 6.25, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 37.83503341674805, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 5.8835, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04251700680272109, | |
| "grad_norm": 48.75051498413086, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 6.0714, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04421768707482993, | |
| "grad_norm": 46.86067581176758, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 5.8296, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04591836734693878, | |
| "grad_norm": 47.48677062988281, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 5.7537, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 63.90276336669922, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 6.0631, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.04931972789115646, | |
| "grad_norm": 68.13435363769531, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 5.9166, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05102040816326531, | |
| "grad_norm": 37.53318405151367, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 6.2247, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05272108843537415, | |
| "grad_norm": 47.694053649902344, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 5.825, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05442176870748299, | |
| "grad_norm": 34.079559326171875, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 5.877, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05612244897959184, | |
| "grad_norm": 40.28605270385742, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 5.8585, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05782312925170068, | |
| "grad_norm": 42.48761749267578, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 5.7666, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.05952380952380952, | |
| "grad_norm": 40.06511306762695, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 5.785, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 36.0867805480957, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 5.9405, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06292517006802721, | |
| "grad_norm": 49.06697463989258, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 5.9568, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06462585034013606, | |
| "grad_norm": 111.22212219238281, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 5.5869, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0663265306122449, | |
| "grad_norm": 35.14386749267578, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 5.8226, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.06802721088435375, | |
| "grad_norm": 36.978759765625, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 5.4815, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06972789115646258, | |
| "grad_norm": 35.120521545410156, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 5.7292, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07142857142857142, | |
| "grad_norm": 32.75640106201172, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 5.1914, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07312925170068027, | |
| "grad_norm": 39.890159606933594, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 5.6042, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07482993197278912, | |
| "grad_norm": 37.68291473388672, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 5.796, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.07653061224489796, | |
| "grad_norm": 43.04936981201172, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 5.4306, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0782312925170068, | |
| "grad_norm": 36.302406311035156, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 5.0626, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.07993197278911565, | |
| "grad_norm": 42.774600982666016, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 5.2214, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 35.731285095214844, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 5.2327, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 40.49732971191406, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 5.1229, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08503401360544217, | |
| "grad_norm": 42.47202682495117, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 5.0709, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08503401360544217, | |
| "eval_loss": 1.416332483291626, | |
| "eval_runtime": 11.5669, | |
| "eval_samples_per_second": 85.589, | |
| "eval_steps_per_second": 21.44, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08673469387755102, | |
| "grad_norm": 29.37403678894043, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 4.9712, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.08843537414965986, | |
| "grad_norm": 30.86339569091797, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 5.3793, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09013605442176871, | |
| "grad_norm": 30.918533325195312, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 5.5769, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09183673469387756, | |
| "grad_norm": 23.81951141357422, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 5.1939, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0935374149659864, | |
| "grad_norm": 20.515365600585938, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 5.5364, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 21.413488388061523, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 5.4537, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.09693877551020408, | |
| "grad_norm": 19.256364822387695, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 5.5546, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.09863945578231292, | |
| "grad_norm": 16.272668838500977, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 5.8678, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10034013605442177, | |
| "grad_norm": 15.786639213562012, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 6.0125, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 16.961414337158203, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 5.7725, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10374149659863946, | |
| "grad_norm": 15.298925399780273, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 5.7823, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1054421768707483, | |
| "grad_norm": 15.63038158416748, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 5.6047, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.10714285714285714, | |
| "grad_norm": 17.067108154296875, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 5.6644, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.10884353741496598, | |
| "grad_norm": 21.148035049438477, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.8186, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11054421768707483, | |
| "grad_norm": 17.97235107421875, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 5.7798, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11224489795918367, | |
| "grad_norm": 15.511758804321289, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 5.7978, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.11394557823129252, | |
| "grad_norm": 15.689850807189941, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.5338, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.11564625850340136, | |
| "grad_norm": 16.96357536315918, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 5.5503, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.11734693877551021, | |
| "grad_norm": 18.011661529541016, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 5.8305, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.11904761904761904, | |
| "grad_norm": 17.602231979370117, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 5.5255, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12074829931972789, | |
| "grad_norm": 16.966541290283203, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 5.6748, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 17.173913955688477, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 5.8363, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.12414965986394558, | |
| "grad_norm": 19.994428634643555, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 5.4537, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.12585034013605442, | |
| "grad_norm": 17.963293075561523, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 5.8565, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.12755102040816327, | |
| "grad_norm": 18.77520751953125, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 5.6593, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1292517006802721, | |
| "grad_norm": 20.010395050048828, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 5.5668, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13095238095238096, | |
| "grad_norm": 20.018394470214844, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 5.5457, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.1326530612244898, | |
| "grad_norm": 16.6497859954834, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 5.5911, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.13435374149659865, | |
| "grad_norm": 19.21900177001953, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 5.6037, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.1360544217687075, | |
| "grad_norm": 19.526813507080078, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 5.3774, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1377551020408163, | |
| "grad_norm": 17.894052505493164, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 5.4939, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.13945578231292516, | |
| "grad_norm": 19.774980545043945, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 5.1935, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.141156462585034, | |
| "grad_norm": 26.265398025512695, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 5.3791, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 48.09366989135742, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 5.2917, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.1445578231292517, | |
| "grad_norm": 23.147611618041992, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 5.2825, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.14625850340136054, | |
| "grad_norm": 33.764434814453125, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 5.1781, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.14795918367346939, | |
| "grad_norm": 28.532094955444336, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 5.3634, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.14965986394557823, | |
| "grad_norm": 31.085493087768555, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 5.1369, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.15136054421768708, | |
| "grad_norm": 30.25196647644043, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 5.2544, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.15306122448979592, | |
| "grad_norm": 28.777196884155273, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 5.1256, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15476190476190477, | |
| "grad_norm": 35.92982482910156, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 5.0517, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1564625850340136, | |
| "grad_norm": 39.41917419433594, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 5.3004, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.15816326530612246, | |
| "grad_norm": 27.697032928466797, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 4.9328, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.1598639455782313, | |
| "grad_norm": 42.105812072753906, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 5.0553, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.16156462585034015, | |
| "grad_norm": 35.63790512084961, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 4.7895, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 28.941566467285156, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 4.7431, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.1649659863945578, | |
| "grad_norm": 28.25985336303711, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 4.7258, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 31.44864845275879, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 4.9412, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1683673469387755, | |
| "grad_norm": 34.225006103515625, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 4.9528, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.17006802721088435, | |
| "grad_norm": 43.27703094482422, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 4.8265, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.17006802721088435, | |
| "eval_loss": 1.3221343755722046, | |
| "eval_runtime": 11.4708, | |
| "eval_samples_per_second": 86.306, | |
| "eval_steps_per_second": 21.62, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1717687074829932, | |
| "grad_norm": 32.49237823486328, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 5.034, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.17346938775510204, | |
| "grad_norm": 26.757333755493164, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 5.1461, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.17517006802721088, | |
| "grad_norm": 21.87533950805664, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 5.2722, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.17687074829931973, | |
| "grad_norm": 22.367048263549805, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 5.4894, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.17857142857142858, | |
| "grad_norm": 21.383411407470703, | |
| "learning_rate": 5e-05, | |
| "loss": 5.4114, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.18027210884353742, | |
| "grad_norm": 22.514957427978516, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 5.4015, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.18197278911564627, | |
| "grad_norm": 23.984758377075195, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 5.2921, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 21.759788513183594, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 5.4827, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.18537414965986396, | |
| "grad_norm": 20.15304183959961, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 5.5669, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1870748299319728, | |
| "grad_norm": 20.60021209716797, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 5.299, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18877551020408162, | |
| "grad_norm": 24.306785583496094, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 5.4923, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 19.58234405517578, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 5.5558, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.1921768707482993, | |
| "grad_norm": 19.0377140045166, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 5.3701, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.19387755102040816, | |
| "grad_norm": 20.250268936157227, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 5.4652, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.195578231292517, | |
| "grad_norm": 20.23456573486328, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 5.5539, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.19727891156462585, | |
| "grad_norm": 27.842456817626953, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 5.6446, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.1989795918367347, | |
| "grad_norm": 20.882328033447266, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 5.5002, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.20068027210884354, | |
| "grad_norm": 18.743694305419922, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 5.6031, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.20238095238095238, | |
| "grad_norm": 18.00966453552246, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 5.3068, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 16.10498046875, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 5.3238, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.20578231292517007, | |
| "grad_norm": 18.90121841430664, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 5.4182, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.20748299319727892, | |
| "grad_norm": 16.165494918823242, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 5.3903, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.20918367346938777, | |
| "grad_norm": 18.660297393798828, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 5.4441, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2108843537414966, | |
| "grad_norm": 16.453760147094727, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 5.5855, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.21258503401360543, | |
| "grad_norm": 15.383659362792969, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 5.3182, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.21428571428571427, | |
| "grad_norm": 18.764667510986328, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 5.3291, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.21598639455782312, | |
| "grad_norm": 16.44414710998535, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 5.3767, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.21768707482993196, | |
| "grad_norm": 14.29847240447998, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 5.415, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.2193877551020408, | |
| "grad_norm": 15.01552963256836, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 5.1709, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.22108843537414966, | |
| "grad_norm": 16.77971649169922, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 5.2718, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2227891156462585, | |
| "grad_norm": 16.29701805114746, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 5.1363, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 16.925735473632812, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 5.1654, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2261904761904762, | |
| "grad_norm": 15.25202465057373, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 5.308, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.22789115646258504, | |
| "grad_norm": 33.06045913696289, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 5.4922, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.22959183673469388, | |
| "grad_norm": 18.302505493164062, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 5.333, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.23129251700680273, | |
| "grad_norm": 16.267621994018555, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 5.3612, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.23299319727891157, | |
| "grad_norm": 16.986467361450195, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 5.0567, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.23469387755102042, | |
| "grad_norm": 15.164219856262207, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 4.6008, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.23639455782312926, | |
| "grad_norm": 21.55219841003418, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 4.8607, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 20.613637924194336, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 5.1134, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23979591836734693, | |
| "grad_norm": 37.14155578613281, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 4.7279, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.24149659863945577, | |
| "grad_norm": 15.671425819396973, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 5.0258, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.24319727891156462, | |
| "grad_norm": 20.68080711364746, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 4.6693, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 26.475250244140625, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 4.8914, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.2465986394557823, | |
| "grad_norm": 21.820087432861328, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 4.8816, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.24829931972789115, | |
| "grad_norm": 20.836872100830078, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 4.8017, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 22.175506591796875, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 4.8856, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.25170068027210885, | |
| "grad_norm": 30.824220657348633, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 4.717, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.2534013605442177, | |
| "grad_norm": 30.54734230041504, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 4.4062, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "grad_norm": 27.12770652770996, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 4.5298, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.25510204081632654, | |
| "eval_loss": 1.2868432998657227, | |
| "eval_runtime": 11.5532, | |
| "eval_samples_per_second": 85.69, | |
| "eval_steps_per_second": 21.466, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2568027210884354, | |
| "grad_norm": 18.7049503326416, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 4.9375, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.2585034013605442, | |
| "grad_norm": 18.078922271728516, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 5.0545, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2602040816326531, | |
| "grad_norm": 19.154966354370117, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 5.3526, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.2619047619047619, | |
| "grad_norm": 18.028196334838867, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 5.1987, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.26360544217687076, | |
| "grad_norm": 15.923282623291016, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 5.4463, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 18.457412719726562, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 5.2983, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.26700680272108845, | |
| "grad_norm": 17.998483657836914, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 5.4181, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.2687074829931973, | |
| "grad_norm": 16.7733154296875, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 5.3638, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.27040816326530615, | |
| "grad_norm": 13.827960968017578, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 5.4928, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 14.617654800415039, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 5.291, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27380952380952384, | |
| "grad_norm": 14.159627914428711, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 5.4403, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2755102040816326, | |
| "grad_norm": 13.921249389648438, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 5.5992, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.27721088435374147, | |
| "grad_norm": 15.430550575256348, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 5.6775, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2789115646258503, | |
| "grad_norm": 14.329670906066895, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 5.2063, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.28061224489795916, | |
| "grad_norm": 15.070326805114746, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 5.3976, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.282312925170068, | |
| "grad_norm": 14.749815940856934, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 5.6894, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.28401360544217685, | |
| "grad_norm": 16.868249893188477, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 5.2744, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 14.273650169372559, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 5.398, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.28741496598639454, | |
| "grad_norm": 13.507621765136719, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 5.3162, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.2891156462585034, | |
| "grad_norm": 19.16738510131836, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 5.4874, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.29081632653061223, | |
| "grad_norm": 18.393007278442383, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 5.359, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.2925170068027211, | |
| "grad_norm": 15.776576042175293, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 5.3738, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.2942176870748299, | |
| "grad_norm": 14.64944076538086, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 5.4677, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.29591836734693877, | |
| "grad_norm": 18.698768615722656, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 5.5293, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2976190476190476, | |
| "grad_norm": 12.899678230285645, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 5.4947, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.29931972789115646, | |
| "grad_norm": 18.404571533203125, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 5.4355, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3010204081632653, | |
| "grad_norm": 18.381723403930664, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 5.3453, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.30272108843537415, | |
| "grad_norm": 14.628609657287598, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 5.4148, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.304421768707483, | |
| "grad_norm": 16.81824493408203, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 5.2486, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 14.485462188720703, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 5.248, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3078231292517007, | |
| "grad_norm": 16.301626205444336, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 5.4492, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.30952380952380953, | |
| "grad_norm": 15.432479858398438, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 5.1182, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3112244897959184, | |
| "grad_norm": 14.198790550231934, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 4.9904, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3129251700680272, | |
| "grad_norm": 15.504638671875, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 5.143, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.31462585034013607, | |
| "grad_norm": 18.87157440185547, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 5.0935, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3163265306122449, | |
| "grad_norm": 17.14332389831543, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 5.1447, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.31802721088435376, | |
| "grad_norm": 16.795663833618164, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 5.1939, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3197278911564626, | |
| "grad_norm": 20.62322998046875, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 4.822, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.32142857142857145, | |
| "grad_norm": 20.075130462646484, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 5.0069, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.3231292517006803, | |
| "grad_norm": 15.727253913879395, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 4.909, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32482993197278914, | |
| "grad_norm": 20.30358123779297, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 4.9415, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 19.216651916503906, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 4.3706, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3282312925170068, | |
| "grad_norm": 19.906362533569336, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 4.622, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.3299319727891156, | |
| "grad_norm": 21.7601261138916, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 4.6673, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.33163265306122447, | |
| "grad_norm": 20.11528968811035, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 4.4147, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 19.714672088623047, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 4.5006, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.33503401360544216, | |
| "grad_norm": 21.023738861083984, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 4.504, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.336734693877551, | |
| "grad_norm": 27.90838050842285, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 4.5986, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.33843537414965985, | |
| "grad_norm": 25.331344604492188, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 4.8014, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.3401360544217687, | |
| "grad_norm": 28.763505935668945, | |
| "learning_rate": 0.0, | |
| "loss": 4.4791, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3401360544217687, | |
| "eval_loss": 1.2759346961975098, | |
| "eval_runtime": 11.5149, | |
| "eval_samples_per_second": 85.975, | |
| "eval_steps_per_second": 21.537, | |
| "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": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4976948111409152e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |