| { | |
| "best_metric": 0.27047082781791687, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.04753228234175711, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00031688188227838076, | |
| "grad_norm": 0.288943886756897, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4815, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00031688188227838076, | |
| "eval_loss": 1.0073962211608887, | |
| "eval_runtime": 388.2204, | |
| "eval_samples_per_second": 13.691, | |
| "eval_steps_per_second": 3.423, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006337637645567615, | |
| "grad_norm": 0.34678709506988525, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6068, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0009506456468351422, | |
| "grad_norm": 0.34489861130714417, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5999, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.001267527529113523, | |
| "grad_norm": 0.3450894355773926, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6676, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0015844094113919036, | |
| "grad_norm": 0.31017449498176575, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6433, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0019012912936702843, | |
| "grad_norm": 0.3006879985332489, | |
| "learning_rate": 6e-05, | |
| "loss": 0.607, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0022181731759486653, | |
| "grad_norm": 0.2969091534614563, | |
| "learning_rate": 7e-05, | |
| "loss": 0.6448, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.002535055058227046, | |
| "grad_norm": 0.24432115256786346, | |
| "learning_rate": 8e-05, | |
| "loss": 0.533, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.002851936940505427, | |
| "grad_norm": 0.24627213180065155, | |
| "learning_rate": 9e-05, | |
| "loss": 0.5425, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.003168818822783807, | |
| "grad_norm": 0.24130110442638397, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5446, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.003485700705062188, | |
| "grad_norm": 0.2302355170249939, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 0.5384, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0038025825873405686, | |
| "grad_norm": 0.2180791199207306, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 0.489, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.004119464469618949, | |
| "grad_norm": 0.20865923166275024, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 0.5058, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.004436346351897331, | |
| "grad_norm": 0.22682811319828033, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 0.4016, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.004753228234175711, | |
| "grad_norm": 0.18705151975154877, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 0.466, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.005070110116454092, | |
| "grad_norm": 0.21964482963085175, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 0.4538, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.005386991998732472, | |
| "grad_norm": 0.21100586652755737, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 0.3919, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.005703873881010854, | |
| "grad_norm": 0.1652364581823349, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 0.3553, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.006020755763289234, | |
| "grad_norm": 0.17498663067817688, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 0.4155, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.006337637645567614, | |
| "grad_norm": 0.20370839536190033, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 0.4282, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0066545195278459954, | |
| "grad_norm": 0.17204949259757996, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 0.3919, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.006971401410124376, | |
| "grad_norm": 0.1701580435037613, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 0.3842, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.007288283292402757, | |
| "grad_norm": 0.1702832579612732, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 0.3708, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.007605165174681137, | |
| "grad_norm": 0.1734638214111328, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 0.4457, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.007922047056959518, | |
| "grad_norm": 0.17348387837409973, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 0.3987, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.008238928939237899, | |
| "grad_norm": 0.15915273129940033, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 0.3546, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.008555810821516279, | |
| "grad_norm": 0.17628248035907745, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 0.3845, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.008872692703794661, | |
| "grad_norm": 0.17458303272724152, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 0.3812, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.009189574586073041, | |
| "grad_norm": 0.1784602254629135, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.3656, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.009506456468351422, | |
| "grad_norm": 0.16332803666591644, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 0.3481, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.009823338350629802, | |
| "grad_norm": 0.1699024736881256, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 0.3658, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.010140220232908184, | |
| "grad_norm": 0.17193962633609772, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 0.3438, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.010457102115186565, | |
| "grad_norm": 0.17559102177619934, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 0.346, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.010773983997464945, | |
| "grad_norm": 0.1714521050453186, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 0.3722, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.011090865879743325, | |
| "grad_norm": 0.19045782089233398, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 0.3926, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.011407747762021707, | |
| "grad_norm": 0.18003332614898682, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 0.3513, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.011724629644300088, | |
| "grad_norm": 0.1978747546672821, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 0.3949, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.012041511526578468, | |
| "grad_norm": 0.19634827971458435, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 0.3866, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.012358393408856848, | |
| "grad_norm": 0.21444833278656006, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 0.4006, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.012675275291135229, | |
| "grad_norm": 0.1893952488899231, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 0.3924, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01299215717341361, | |
| "grad_norm": 0.18121342360973358, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 0.3769, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.013309039055691991, | |
| "grad_norm": 0.17903916537761688, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 0.3801, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.013625920937970371, | |
| "grad_norm": 0.17503075301647186, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 0.3669, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.013942802820248752, | |
| "grad_norm": 0.19540923833847046, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 0.3888, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.014259684702527134, | |
| "grad_norm": 0.19996744394302368, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 0.3779, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.014576566584805514, | |
| "grad_norm": 0.20995080471038818, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 0.3893, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.014893448467083894, | |
| "grad_norm": 0.21569843590259552, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 0.3882, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.015210330349362275, | |
| "grad_norm": 0.2168152928352356, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.4067, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.015527212231640657, | |
| "grad_norm": 0.2293938845396042, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 0.4036, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.015844094113919037, | |
| "grad_norm": 0.27539652585983276, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 0.4517, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015844094113919037, | |
| "eval_loss": 0.3110412061214447, | |
| "eval_runtime": 391.0829, | |
| "eval_samples_per_second": 13.59, | |
| "eval_steps_per_second": 3.398, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.016160975996197417, | |
| "grad_norm": 0.09471043199300766, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 0.1704, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.016477857878475798, | |
| "grad_norm": 0.10490579158067703, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 0.1973, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.016794739760754178, | |
| "grad_norm": 0.11275085806846619, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 0.2089, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.017111621643032558, | |
| "grad_norm": 0.09714960306882858, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 0.1874, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.017428503525310942, | |
| "grad_norm": 0.10512809455394745, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 0.1858, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.017745385407589322, | |
| "grad_norm": 0.10947681963443756, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 0.2236, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.018062267289867703, | |
| "grad_norm": 0.10696465522050858, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 0.2353, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.018379149172146083, | |
| "grad_norm": 0.12559695541858673, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 0.251, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.018696031054424463, | |
| "grad_norm": 0.1162128821015358, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 0.2488, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.019012912936702844, | |
| "grad_norm": 0.1090744361281395, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 0.2398, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.019329794818981224, | |
| "grad_norm": 0.11116234958171844, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 0.2555, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.019646676701259604, | |
| "grad_norm": 0.11902529746294022, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 0.2682, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.019963558583537985, | |
| "grad_norm": 0.10772904753684998, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 0.2352, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.02028044046581637, | |
| "grad_norm": 0.10973820835351944, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 0.2369, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.02059732234809475, | |
| "grad_norm": 0.11814217269420624, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 0.2524, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.02091420423037313, | |
| "grad_norm": 0.12398887425661087, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 0.2532, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.02123108611265151, | |
| "grad_norm": 0.12989726662635803, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.2718, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.02154796799492989, | |
| "grad_norm": 0.11946962773799896, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 0.242, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02186484987720827, | |
| "grad_norm": 0.1271256059408188, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 0.2929, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.02218173175948665, | |
| "grad_norm": 0.13123860955238342, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 0.2413, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02249861364176503, | |
| "grad_norm": 0.13517530262470245, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 0.2833, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.022815495524043414, | |
| "grad_norm": 0.14057579636573792, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 0.2952, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.023132377406321795, | |
| "grad_norm": 0.1402745395898819, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 0.3068, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.023449259288600175, | |
| "grad_norm": 0.1463419497013092, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 0.3047, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.023766141170878555, | |
| "grad_norm": 0.15095284581184387, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 0.3107, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.024083023053156936, | |
| "grad_norm": 0.1348758339881897, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 0.3007, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.024399904935435316, | |
| "grad_norm": 0.1441372185945511, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 0.2894, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.024716786817713696, | |
| "grad_norm": 0.14854104816913605, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 0.3202, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.025033668699992077, | |
| "grad_norm": 0.1389545202255249, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 0.3043, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.025350550582270457, | |
| "grad_norm": 0.15648682415485382, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 0.337, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.02566743246454884, | |
| "grad_norm": 0.14272603392601013, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 0.3011, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.02598431434682722, | |
| "grad_norm": 0.15423764288425446, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 0.3214, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.0263011962291056, | |
| "grad_norm": 0.15417426824569702, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 0.296, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.026618078111383982, | |
| "grad_norm": 0.15329402685165405, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 0.2983, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.026934959993662362, | |
| "grad_norm": 0.15941715240478516, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 0.3016, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.027251841875940742, | |
| "grad_norm": 0.15860851109027863, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.272, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.027568723758219123, | |
| "grad_norm": 0.16991178691387177, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 0.3448, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.027885605640497503, | |
| "grad_norm": 0.1755577027797699, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 0.3156, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.028202487522775887, | |
| "grad_norm": 0.1622546762228012, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 0.3063, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.028519369405054267, | |
| "grad_norm": 0.17945386469364166, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.3276, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.028836251287332648, | |
| "grad_norm": 0.18708471953868866, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 0.325, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.029153133169611028, | |
| "grad_norm": 0.1904095858335495, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 0.3476, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.029470015051889408, | |
| "grad_norm": 0.1793287843465805, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 0.3423, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02978689693416779, | |
| "grad_norm": 0.18426981568336487, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 0.321, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.03010377881644617, | |
| "grad_norm": 0.20930606126785278, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 0.3844, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.03042066069872455, | |
| "grad_norm": 0.19688951969146729, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 0.3542, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.03073754258100293, | |
| "grad_norm": 0.21573825180530548, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 0.3764, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.031054424463281313, | |
| "grad_norm": 0.22527675330638885, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 0.4116, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.031371306345559694, | |
| "grad_norm": 0.2153525948524475, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 0.3692, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.031688188227838074, | |
| "grad_norm": 0.27288129925727844, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 0.4117, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.031688188227838074, | |
| "eval_loss": 0.2861791253089905, | |
| "eval_runtime": 390.5684, | |
| "eval_samples_per_second": 13.608, | |
| "eval_steps_per_second": 3.403, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.032005070110116454, | |
| "grad_norm": 0.09543729573488235, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 0.1448, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.032321951992394835, | |
| "grad_norm": 0.12222027033567429, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 0.1824, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.032638833874673215, | |
| "grad_norm": 0.11698739975690842, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 0.1877, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.032955715756951595, | |
| "grad_norm": 0.12096019834280014, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 0.2091, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.033272597639229975, | |
| "grad_norm": 0.1101212278008461, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2147, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.033589479521508356, | |
| "grad_norm": 0.11647240072488785, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 0.2144, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.033906361403786736, | |
| "grad_norm": 0.11861401051282883, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 0.2182, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.034223243286065116, | |
| "grad_norm": 0.10909169912338257, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 0.2102, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.0345401251683435, | |
| "grad_norm": 0.11667589098215103, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 0.2165, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.034857007050621884, | |
| "grad_norm": 0.11918514221906662, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 0.233, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.035173888932900264, | |
| "grad_norm": 0.11744184792041779, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 0.2325, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.035490770815178645, | |
| "grad_norm": 0.1135401725769043, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 0.2248, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.035807652697457025, | |
| "grad_norm": 0.12545964121818542, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 0.2669, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.036124534579735405, | |
| "grad_norm": 0.12244533002376556, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 0.2485, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.036441416462013786, | |
| "grad_norm": 0.1240435466170311, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 0.259, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.036758298344292166, | |
| "grad_norm": 0.12252770364284515, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 0.2441, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.037075180226570546, | |
| "grad_norm": 0.11815881729125977, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 0.2272, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.03739206210884893, | |
| "grad_norm": 0.12181126326322556, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 0.2542, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.03770894399112731, | |
| "grad_norm": 0.13004834949970245, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 0.2644, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.03802582587340569, | |
| "grad_norm": 0.1406186819076538, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 0.2954, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03834270775568407, | |
| "grad_norm": 0.12018871307373047, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 0.2294, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.03865958963796245, | |
| "grad_norm": 0.13669250905513763, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 0.2871, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.03897647152024083, | |
| "grad_norm": 0.14799264073371887, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 0.3321, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.03929335340251921, | |
| "grad_norm": 0.12626908719539642, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.2595, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.03961023528479759, | |
| "grad_norm": 0.1423238217830658, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 0.2994, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03992711716707597, | |
| "grad_norm": 0.13488979637622833, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 0.2693, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.040243999049354356, | |
| "grad_norm": 0.14322566986083984, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 0.3032, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.04056088093163274, | |
| "grad_norm": 0.13436266779899597, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 0.2573, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.04087776281391112, | |
| "grad_norm": 0.152916818857193, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 0.3222, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0411946446961895, | |
| "grad_norm": 0.1609887182712555, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 0.3125, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.04151152657846788, | |
| "grad_norm": 0.15256395936012268, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 0.3098, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.04182840846074626, | |
| "grad_norm": 0.14366978406906128, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 0.2685, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.04214529034302464, | |
| "grad_norm": 0.1478935033082962, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 0.2959, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.04246217222530302, | |
| "grad_norm": 0.16299816966056824, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 0.295, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0427790541075814, | |
| "grad_norm": 0.15156926214694977, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 0.2783, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.04309593598985978, | |
| "grad_norm": 0.15439371764659882, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 0.2729, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.04341281787213816, | |
| "grad_norm": 0.1601949781179428, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 0.2977, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.04372969975441654, | |
| "grad_norm": 0.17758437991142273, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 0.3337, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.04404658163669492, | |
| "grad_norm": 0.1762048751115799, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 0.335, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.0443634635189733, | |
| "grad_norm": 0.17591573297977448, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 0.3153, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.04468034540125168, | |
| "grad_norm": 0.19056615233421326, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 0.3557, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.04499722728353006, | |
| "grad_norm": 0.18306946754455566, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 0.3116, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.04531410916580844, | |
| "grad_norm": 0.1650189459323883, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 0.2802, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.04563099104808683, | |
| "grad_norm": 0.1860068440437317, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 0.3471, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.04594787293036521, | |
| "grad_norm": 0.19405558705329895, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 0.3724, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.04626475481264359, | |
| "grad_norm": 0.18297216296195984, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 0.3331, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.04658163669492197, | |
| "grad_norm": 0.20970414578914642, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 0.3551, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.04689851857720035, | |
| "grad_norm": 0.21330970525741577, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 0.3632, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.04721540045947873, | |
| "grad_norm": 0.22864414751529694, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 0.3553, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.04753228234175711, | |
| "grad_norm": 0.2590240240097046, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 0.3993, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04753228234175711, | |
| "eval_loss": 0.27047082781791687, | |
| "eval_runtime": 390.4245, | |
| "eval_samples_per_second": 13.613, | |
| "eval_steps_per_second": 3.404, | |
| "step": 150 | |
| } | |
| ], | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.019046890280059e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |