| { | |
| "best_metric": 5.294073104858398, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.08583690987124463, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00042918454935622315, | |
| "grad_norm": 10628.7685546875, | |
| "learning_rate": 1e-05, | |
| "loss": 21.0635, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00042918454935622315, | |
| "eval_loss": 6.3962907791137695, | |
| "eval_runtime": 32.9445, | |
| "eval_samples_per_second": 119.109, | |
| "eval_steps_per_second": 29.777, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0008583690987124463, | |
| "grad_norm": 11912.2529296875, | |
| "learning_rate": 2e-05, | |
| "loss": 23.2765, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0012875536480686696, | |
| "grad_norm": 16264.650390625, | |
| "learning_rate": 3e-05, | |
| "loss": 23.4017, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0017167381974248926, | |
| "grad_norm": 13689.9365234375, | |
| "learning_rate": 4e-05, | |
| "loss": 23.1432, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.002145922746781116, | |
| "grad_norm": 10931.578125, | |
| "learning_rate": 5e-05, | |
| "loss": 23.3441, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.002575107296137339, | |
| "grad_norm": 23887.791015625, | |
| "learning_rate": 6e-05, | |
| "loss": 22.9707, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0030042918454935624, | |
| "grad_norm": 11153.2880859375, | |
| "learning_rate": 7e-05, | |
| "loss": 22.902, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0034334763948497852, | |
| "grad_norm": 17550.177734375, | |
| "learning_rate": 8e-05, | |
| "loss": 21.1592, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0038626609442060085, | |
| "grad_norm": 11275.908203125, | |
| "learning_rate": 9e-05, | |
| "loss": 21.7623, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.004291845493562232, | |
| "grad_norm": 19955.615234375, | |
| "learning_rate": 0.0001, | |
| "loss": 20.0114, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004721030042918455, | |
| "grad_norm": 11082.9462890625, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 20.0925, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.005150214592274678, | |
| "grad_norm": 12425.65625, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 19.7646, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005579399141630901, | |
| "grad_norm": 14407.03125, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 20.562, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.006008583690987125, | |
| "grad_norm": 12435.060546875, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 18.6204, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.006437768240343348, | |
| "grad_norm": 14993.59375, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 18.3011, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0068669527896995704, | |
| "grad_norm": 18411.9609375, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 17.7168, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.007296137339055794, | |
| "grad_norm": 19244.978515625, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 17.5028, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.007725321888412017, | |
| "grad_norm": 21387.21484375, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 17.0027, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.00815450643776824, | |
| "grad_norm": 18360.8046875, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 17.1145, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.008583690987124463, | |
| "grad_norm": 14379.1650390625, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 17.2649, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.009012875536480686, | |
| "grad_norm": 23081.27734375, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 17.4423, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.00944206008583691, | |
| "grad_norm": 29940.923828125, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 17.0243, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.009871244635193134, | |
| "grad_norm": 14467.982421875, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 17.1411, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.010300429184549357, | |
| "grad_norm": 20777.873046875, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 16.1163, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.01072961373390558, | |
| "grad_norm": 32732.693359375, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 16.489, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.011158798283261802, | |
| "grad_norm": 22580.431640625, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 17.2863, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.011587982832618025, | |
| "grad_norm": 21401.095703125, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 16.8197, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.01201716738197425, | |
| "grad_norm": 21236.1328125, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 18.0942, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.012446351931330472, | |
| "grad_norm": 21269.58203125, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 18.8153, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.012875536480686695, | |
| "grad_norm": 72275.421875, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 19.6917, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.013304721030042918, | |
| "grad_norm": 29044.935546875, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 20.4021, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.013733905579399141, | |
| "grad_norm": 46292.72265625, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 20.7917, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.014163090128755365, | |
| "grad_norm": 38369.83203125, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 22.1983, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.014592274678111588, | |
| "grad_norm": 65315.74609375, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 23.0792, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.015021459227467811, | |
| "grad_norm": 46062.6171875, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 23.4203, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.015450643776824034, | |
| "grad_norm": 52505.52734375, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 24.6427, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01587982832618026, | |
| "grad_norm": 28250.546875, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 22.3865, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01630901287553648, | |
| "grad_norm": 78103.2109375, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 22.3232, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.016738197424892704, | |
| "grad_norm": 29406.908203125, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 24.2719, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.017167381974248927, | |
| "grad_norm": 38396.484375, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 21.3504, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01759656652360515, | |
| "grad_norm": 33549.44140625, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 21.8852, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.018025751072961373, | |
| "grad_norm": 42242.23046875, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 25.3198, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.018454935622317595, | |
| "grad_norm": 45411.640625, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 23.2329, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.01888412017167382, | |
| "grad_norm": 18876.40625, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 24.4576, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.019313304721030045, | |
| "grad_norm": 47549.8671875, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 23.5283, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.019742489270386267, | |
| "grad_norm": 50706.859375, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 23.93, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.02017167381974249, | |
| "grad_norm": 31169.017578125, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 22.7424, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.020600858369098713, | |
| "grad_norm": 18738.626953125, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 22.0673, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.021030042918454936, | |
| "grad_norm": 36289.7265625, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 25.4747, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.02145922746781116, | |
| "grad_norm": 80855.2421875, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 25.9234, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02145922746781116, | |
| "eval_loss": 6.562374114990234, | |
| "eval_runtime": 33.0523, | |
| "eval_samples_per_second": 118.721, | |
| "eval_steps_per_second": 29.68, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02188841201716738, | |
| "grad_norm": 19864.068359375, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 32.4057, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.022317596566523604, | |
| "grad_norm": 43845.65625, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 34.89, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.022746781115879827, | |
| "grad_norm": 34666.484375, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 31.0985, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.02317596566523605, | |
| "grad_norm": 22230.24609375, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 30.7861, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.023605150214592276, | |
| "grad_norm": 19344.49609375, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 29.2839, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0240343347639485, | |
| "grad_norm": 26743.744140625, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 29.6186, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.024463519313304722, | |
| "grad_norm": 25800.14453125, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 28.0493, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.024892703862660945, | |
| "grad_norm": 31142.1171875, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 32.5601, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.025321888412017168, | |
| "grad_norm": 16036.462890625, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 28.4418, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.02575107296137339, | |
| "grad_norm": 15373.236328125, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 28.5077, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.026180257510729613, | |
| "grad_norm": 14576.5810546875, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 27.6621, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.026609442060085836, | |
| "grad_norm": 25582.259765625, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 27.6296, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.02703862660944206, | |
| "grad_norm": 17445.578125, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 27.2274, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.027467811158798282, | |
| "grad_norm": 22893.35546875, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 26.9392, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.027896995708154508, | |
| "grad_norm": 23648.275390625, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 27.4601, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.02832618025751073, | |
| "grad_norm": 12287.2822265625, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 27.0826, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.028755364806866954, | |
| "grad_norm": 16298.2734375, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 25.9617, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.029184549356223177, | |
| "grad_norm": 13834.390625, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 26.9404, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0296137339055794, | |
| "grad_norm": 12603.525390625, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 27.5063, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.030042918454935622, | |
| "grad_norm": 14483.2568359375, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 25.5151, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.030472103004291845, | |
| "grad_norm": 22377.703125, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 25.3501, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.030901287553648068, | |
| "grad_norm": 11517.9404296875, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 24.6406, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.03133047210300429, | |
| "grad_norm": 20101.896484375, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 24.7856, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.03175965665236052, | |
| "grad_norm": 19294.7890625, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 23.9486, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.032188841201716736, | |
| "grad_norm": 23703.109375, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 24.9903, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03261802575107296, | |
| "grad_norm": 30060.146484375, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 24.5411, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.03304721030042918, | |
| "grad_norm": 19922.330078125, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 25.7279, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.03347639484978541, | |
| "grad_norm": 17217.419921875, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 25.1845, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.03390557939914163, | |
| "grad_norm": 19183.958984375, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 25.4832, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.034334763948497854, | |
| "grad_norm": 20375.23046875, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 23.0827, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.03476394849785408, | |
| "grad_norm": 18824.501953125, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 25.0052, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0351931330472103, | |
| "grad_norm": 24895.44140625, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 23.4742, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.035622317596566526, | |
| "grad_norm": 40151.45703125, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 22.1832, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.036051502145922745, | |
| "grad_norm": 23362.876953125, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 22.6521, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.03648068669527897, | |
| "grad_norm": 27753.1640625, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 20.8269, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03690987124463519, | |
| "grad_norm": 33134.3828125, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 22.4601, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.03733905579399142, | |
| "grad_norm": 30492.09765625, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 21.7277, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.03776824034334764, | |
| "grad_norm": 19832.763671875, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 23.2241, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.03819742489270386, | |
| "grad_norm": 23929.669921875, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 20.0956, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.03862660944206009, | |
| "grad_norm": 49209.15234375, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 18.8861, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03905579399141631, | |
| "grad_norm": 30595.28515625, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 20.4827, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.039484978540772535, | |
| "grad_norm": 37350.1875, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 19.272, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.039914163090128754, | |
| "grad_norm": 50806.703125, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 18.5296, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.04034334763948498, | |
| "grad_norm": 24144.21875, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 17.9356, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0407725321888412, | |
| "grad_norm": 30585.232421875, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 18.3017, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.041201716738197426, | |
| "grad_norm": 34594.72265625, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 17.5436, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.041630901287553645, | |
| "grad_norm": 19484.173828125, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 19.3632, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.04206008583690987, | |
| "grad_norm": 33278.078125, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 17.8278, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.04248927038626609, | |
| "grad_norm": 53987.2578125, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 18.7489, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.04291845493562232, | |
| "grad_norm": 108816.4375, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 22.3774, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04291845493562232, | |
| "eval_loss": 5.826864242553711, | |
| "eval_runtime": 33.0811, | |
| "eval_samples_per_second": 118.617, | |
| "eval_steps_per_second": 29.654, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.043347639484978544, | |
| "grad_norm": 29586.859375, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 31.3516, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.04377682403433476, | |
| "grad_norm": 19528.3046875, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 28.4271, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.04420600858369099, | |
| "grad_norm": 14138.626953125, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 27.0693, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.04463519313304721, | |
| "grad_norm": 14044.6318359375, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 27.4763, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.045064377682403435, | |
| "grad_norm": 29368.255859375, | |
| "learning_rate": 5e-05, | |
| "loss": 29.0252, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.045493562231759654, | |
| "grad_norm": 160239.359375, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 26.7511, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.04592274678111588, | |
| "grad_norm": 15314.037109375, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 25.1894, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0463519313304721, | |
| "grad_norm": 26796.080078125, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 25.2649, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.046781115879828326, | |
| "grad_norm": 326209.0, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 25.0861, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.04721030042918455, | |
| "grad_norm": 18186.328125, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 24.9119, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04763948497854077, | |
| "grad_norm": 18744.583984375, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 25.3471, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.048068669527897, | |
| "grad_norm": 22808.3125, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 23.624, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.04849785407725322, | |
| "grad_norm": 19347.537109375, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 27.8759, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.048927038626609444, | |
| "grad_norm": 27354.23046875, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 23.9563, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.04935622317596566, | |
| "grad_norm": 16501.87890625, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 23.0557, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.04978540772532189, | |
| "grad_norm": 259689.109375, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 24.4779, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.05021459227467811, | |
| "grad_norm": 27897.322265625, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 22.9713, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.050643776824034335, | |
| "grad_norm": 18918.537109375, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 22.0572, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.051072961373390555, | |
| "grad_norm": 18446.068359375, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 22.0303, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.05150214592274678, | |
| "grad_norm": 14606.2001953125, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 23.1456, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.05193133047210301, | |
| "grad_norm": 22487.6484375, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 21.0624, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.05236051502145923, | |
| "grad_norm": 19732.375, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 22.8554, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.05278969957081545, | |
| "grad_norm": 23008.619140625, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 22.3963, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.05321888412017167, | |
| "grad_norm": 13546.5107421875, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 21.7482, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.0536480686695279, | |
| "grad_norm": 20482.58203125, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 22.1889, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.05407725321888412, | |
| "grad_norm": 17858.486328125, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 21.6642, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.054506437768240344, | |
| "grad_norm": 14167.671875, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 20.0264, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.054935622317596564, | |
| "grad_norm": 20402.9296875, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 21.5997, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.05536480686695279, | |
| "grad_norm": 9313.5732421875, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 21.6812, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.055793991416309016, | |
| "grad_norm": 32921.47265625, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 21.1304, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.056223175965665236, | |
| "grad_norm": 10723.96484375, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 20.2714, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.05665236051502146, | |
| "grad_norm": 18415.265625, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 19.8755, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.05708154506437768, | |
| "grad_norm": 17032.2578125, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 19.687, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.05751072961373391, | |
| "grad_norm": 14226.6083984375, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 20.3326, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.05793991416309013, | |
| "grad_norm": 16448.916015625, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 18.5379, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.05836909871244635, | |
| "grad_norm": 16763.3359375, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 18.2503, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.05879828326180257, | |
| "grad_norm": 25070.724609375, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 18.2971, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.0592274678111588, | |
| "grad_norm": 12421.0654296875, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 19.2307, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.05965665236051502, | |
| "grad_norm": 29242.57421875, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 17.8132, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.060085836909871244, | |
| "grad_norm": 14171.6220703125, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 17.8596, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06051502145922747, | |
| "grad_norm": 56702.3828125, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 17.2853, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.06094420600858369, | |
| "grad_norm": 21922.3046875, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 17.4972, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.061373390557939916, | |
| "grad_norm": 26030.953125, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 17.3916, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.061802575107296136, | |
| "grad_norm": 14679.05859375, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 18.4846, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.06223175965665236, | |
| "grad_norm": 13952.599609375, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 17.6485, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.06266094420600858, | |
| "grad_norm": 18942.3515625, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 16.6502, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.06309012875536481, | |
| "grad_norm": 14503.865234375, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 17.5226, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.06351931330472103, | |
| "grad_norm": 22509.419921875, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 15.2506, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.06394849785407725, | |
| "grad_norm": 26797.15625, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 17.5458, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.06437768240343347, | |
| "grad_norm": 31287.544921875, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 27.5095, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06437768240343347, | |
| "eval_loss": 5.46917724609375, | |
| "eval_runtime": 32.9151, | |
| "eval_samples_per_second": 119.216, | |
| "eval_steps_per_second": 29.804, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0648068669527897, | |
| "grad_norm": 25999.80078125, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 31.8518, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06523605150214593, | |
| "grad_norm": 18807.294921875, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 29.7521, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.06566523605150215, | |
| "grad_norm": 39500.671875, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 27.1969, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06609442060085836, | |
| "grad_norm": 15707.22265625, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 26.9572, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06652360515021459, | |
| "grad_norm": 31338.119140625, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 25.9306, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06695278969957082, | |
| "grad_norm": 15267.263671875, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 25.894, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.06738197424892704, | |
| "grad_norm": 22669.09375, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 28.9373, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.06781115879828326, | |
| "grad_norm": 18563.8515625, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 28.7938, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06824034334763948, | |
| "grad_norm": 20524.720703125, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 25.2427, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.06866952789699571, | |
| "grad_norm": 17444.15234375, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 23.208, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06909871244635193, | |
| "grad_norm": 16035.669921875, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 23.3063, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.06952789699570816, | |
| "grad_norm": 39131.9453125, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 27.8237, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.06995708154506437, | |
| "grad_norm": 27309.083984375, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 29.3328, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0703862660944206, | |
| "grad_norm": 18432.5390625, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 23.4278, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.07081545064377683, | |
| "grad_norm": 19710.7734375, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 22.6084, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.07124463519313305, | |
| "grad_norm": 17223.376953125, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 21.9122, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.07167381974248926, | |
| "grad_norm": 25279.587890625, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 21.6746, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.07210300429184549, | |
| "grad_norm": 33817.6171875, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 26.3172, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.07253218884120172, | |
| "grad_norm": 10719.3251953125, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 21.7262, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.07296137339055794, | |
| "grad_norm": 15343.8876953125, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 22.2938, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.07339055793991417, | |
| "grad_norm": 16401.130859375, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 21.1817, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.07381974248927038, | |
| "grad_norm": 13158.2236328125, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 21.1175, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.07424892703862661, | |
| "grad_norm": 21862.6015625, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 21.8146, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.07467811158798283, | |
| "grad_norm": 24937.91796875, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 20.9579, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.07510729613733906, | |
| "grad_norm": 32984.84375, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 20.5653, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.07553648068669527, | |
| "grad_norm": 28511.427734375, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 21.2712, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.0759656652360515, | |
| "grad_norm": 13016.1259765625, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 20.2881, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07639484978540773, | |
| "grad_norm": 17225.98046875, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 19.6842, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07682403433476395, | |
| "grad_norm": 42496.7578125, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 23.9248, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.07725321888412018, | |
| "grad_norm": 17504.453125, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 20.3548, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07768240343347639, | |
| "grad_norm": 12993.2353515625, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 19.6067, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07811158798283262, | |
| "grad_norm": 8669.2421875, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 20.2702, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07854077253218884, | |
| "grad_norm": 8531.05078125, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 19.8044, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.07896995708154507, | |
| "grad_norm": 11753.0712890625, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 19.7804, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.07939914163090128, | |
| "grad_norm": 10034.986328125, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 19.2943, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.07982832618025751, | |
| "grad_norm": 15453.01953125, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 18.6545, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.08025751072961373, | |
| "grad_norm": 19633.0078125, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 18.5271, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.08068669527896996, | |
| "grad_norm": 10748.7177734375, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 18.949, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.08111587982832617, | |
| "grad_norm": 14911.6806640625, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 17.3563, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.0815450643776824, | |
| "grad_norm": 26729.716796875, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 18.3295, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.08197424892703863, | |
| "grad_norm": 11022.7001953125, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 17.8345, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.08240343347639485, | |
| "grad_norm": 19143.1953125, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 16.9504, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.08283261802575108, | |
| "grad_norm": 16519.365234375, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 17.656, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.08326180257510729, | |
| "grad_norm": 12927.4775390625, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 17.3791, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.08369098712446352, | |
| "grad_norm": 13258.2392578125, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 17.1515, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.08412017167381974, | |
| "grad_norm": 8209.8955078125, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 14.8383, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.08454935622317597, | |
| "grad_norm": 14235.85546875, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 16.4079, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.08497854077253218, | |
| "grad_norm": 19800.390625, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 16.2311, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.08540772532188841, | |
| "grad_norm": 15021.943359375, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 15.9475, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.08583690987124463, | |
| "grad_norm": 11545.7412109375, | |
| "learning_rate": 0.0, | |
| "loss": 24.1106, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08583690987124463, | |
| "eval_loss": 5.294073104858398, | |
| "eval_runtime": 32.9183, | |
| "eval_samples_per_second": 119.204, | |
| "eval_steps_per_second": 29.801, | |
| "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": 330900386611200.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |