| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 375, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008, | |
| "grad_norm": 6.951127529144287, | |
| "learning_rate": 2.6315789473684213e-07, | |
| "loss": 1.2947, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 6.943446636199951, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 1.2776, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 6.9143147468566895, | |
| "learning_rate": 7.894736842105263e-07, | |
| "loss": 1.272, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 6.79904317855835, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.2738, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.621460437774658, | |
| "learning_rate": 1.3157894736842106e-06, | |
| "loss": 1.2432, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.126567840576172, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 1.2115, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.056, | |
| "grad_norm": 5.147610187530518, | |
| "learning_rate": 1.8421052631578948e-06, | |
| "loss": 1.2493, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.801726818084717, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 1.2107, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 4.432796001434326, | |
| "learning_rate": 2.368421052631579e-06, | |
| "loss": 1.1829, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.735476016998291, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 1.1258, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.088, | |
| "grad_norm": 2.625633716583252, | |
| "learning_rate": 2.8947368421052634e-06, | |
| "loss": 1.1722, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.6213021278381348, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 1.1329, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.104, | |
| "grad_norm": 4.202518463134766, | |
| "learning_rate": 3.421052631578948e-06, | |
| "loss": 1.1283, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 4.3722615242004395, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 1.1362, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 4.284238338470459, | |
| "learning_rate": 3.947368421052632e-06, | |
| "loss": 1.122, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.9620132446289062, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 1.098, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.136, | |
| "grad_norm": 2.9353275299072266, | |
| "learning_rate": 4.473684210526316e-06, | |
| "loss": 1.0722, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 2.8066329956054688, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 1.0788, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.152, | |
| "grad_norm": 2.426243543624878, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0823, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.9383955001831055, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.0643, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 1.5883516073226929, | |
| "learning_rate": 5.526315789473685e-06, | |
| "loss": 1.0186, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 1.5131279230117798, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 1.0316, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.184, | |
| "grad_norm": 1.5010865926742554, | |
| "learning_rate": 6.0526315789473685e-06, | |
| "loss": 1.0223, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.1548634767532349, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.9821, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.17434561252594, | |
| "learning_rate": 6.578947368421054e-06, | |
| "loss": 0.9551, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 1.1539326906204224, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 0.9725, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 1.1245882511138916, | |
| "learning_rate": 7.1052631578947375e-06, | |
| "loss": 0.9684, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.1288598775863647, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 0.9963, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.232, | |
| "grad_norm": 1.0191954374313354, | |
| "learning_rate": 7.631578947368423e-06, | |
| "loss": 1.0004, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.9676952362060547, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 0.9658, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.248, | |
| "grad_norm": 1.1406854391098022, | |
| "learning_rate": 8.157894736842106e-06, | |
| "loss": 0.9879, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.9583864212036133, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.9472, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 0.9609050750732422, | |
| "learning_rate": 8.68421052631579e-06, | |
| "loss": 0.978, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 0.8762646913528442, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 0.9468, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.8770004510879517, | |
| "learning_rate": 9.210526315789474e-06, | |
| "loss": 0.9643, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.8456284403800964, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.9534, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.296, | |
| "grad_norm": 0.9351610541343689, | |
| "learning_rate": 9.736842105263159e-06, | |
| "loss": 0.9389, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 0.8555827736854553, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9065, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 0.8321928977966309, | |
| "learning_rate": 9.99978274148479e-06, | |
| "loss": 0.9469, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7721774578094482, | |
| "learning_rate": 9.999130984819662e-06, | |
| "loss": 0.9243, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.328, | |
| "grad_norm": 0.7928785085678101, | |
| "learning_rate": 9.998044786644492e-06, | |
| "loss": 0.9615, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.6871277093887329, | |
| "learning_rate": 9.9965242413536e-06, | |
| "loss": 0.9437, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.344, | |
| "grad_norm": 0.7224629521369934, | |
| "learning_rate": 9.994569481087552e-06, | |
| "loss": 0.9486, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 0.8585007190704346, | |
| "learning_rate": 9.992180675721671e-06, | |
| "loss": 0.9382, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.7822074294090271, | |
| "learning_rate": 9.989358032851283e-06, | |
| "loss": 0.9179, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 0.7207604646682739, | |
| "learning_rate": 9.986101797773667e-06, | |
| "loss": 0.9395, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.376, | |
| "grad_norm": 0.889811635017395, | |
| "learning_rate": 9.98241225346674e-06, | |
| "loss": 0.886, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.8683562874794006, | |
| "learning_rate": 9.978289720564471e-06, | |
| "loss": 0.8996, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.392, | |
| "grad_norm": 0.7052727341651917, | |
| "learning_rate": 9.97373455732901e-06, | |
| "loss": 0.9052, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.8796941041946411, | |
| "learning_rate": 9.968747159619556e-06, | |
| "loss": 0.9217, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 0.7054159641265869, | |
| "learning_rate": 9.963327960857962e-06, | |
| "loss": 0.9534, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.7010794281959534, | |
| "learning_rate": 9.957477431991053e-06, | |
| "loss": 0.904, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.424, | |
| "grad_norm": 0.7052696347236633, | |
| "learning_rate": 9.95119608144972e-06, | |
| "loss": 0.9093, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.6790952086448669, | |
| "learning_rate": 9.944484455104716e-06, | |
| "loss": 0.9236, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.6545783877372742, | |
| "learning_rate": 9.937343136219234e-06, | |
| "loss": 0.9295, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.6248077154159546, | |
| "learning_rate": 9.929772745398207e-06, | |
| "loss": 0.9171, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 0.6094949245452881, | |
| "learning_rate": 9.921773940534382e-06, | |
| "loss": 0.8842, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 0.7010917663574219, | |
| "learning_rate": 9.913347416751148e-06, | |
| "loss": 0.9079, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.472, | |
| "grad_norm": 0.7240394353866577, | |
| "learning_rate": 9.904493906342124e-06, | |
| "loss": 0.9433, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.71834397315979, | |
| "learning_rate": 9.895214178707516e-06, | |
| "loss": 0.9086, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.488, | |
| "grad_norm": 0.7287778854370117, | |
| "learning_rate": 9.885509040287267e-06, | |
| "loss": 0.8941, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 0.6156062483787537, | |
| "learning_rate": 9.875379334490962e-06, | |
| "loss": 0.8547, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 0.6256719827651978, | |
| "learning_rate": 9.864825941624538e-06, | |
| "loss": 0.8925, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.6274106502532959, | |
| "learning_rate": 9.853849778813777e-06, | |
| "loss": 0.9073, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.6543683409690857, | |
| "learning_rate": 9.842451799924616e-06, | |
| "loss": 0.9398, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.6689018607139587, | |
| "learning_rate": 9.830632995480243e-06, | |
| "loss": 0.9149, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.536, | |
| "grad_norm": 0.6306348443031311, | |
| "learning_rate": 9.818394392575018e-06, | |
| "loss": 0.8812, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.7672179937362671, | |
| "learning_rate": 9.805737054785223e-06, | |
| "loss": 0.9074, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.7313401103019714, | |
| "learning_rate": 9.792662082076618e-06, | |
| "loss": 0.8983, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.6449005007743835, | |
| "learning_rate": 9.779170610708872e-06, | |
| "loss": 0.95, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.568, | |
| "grad_norm": 0.7076218128204346, | |
| "learning_rate": 9.765263813136796e-06, | |
| "loss": 0.8961, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.6033435463905334, | |
| "learning_rate": 9.750942897908468e-06, | |
| "loss": 0.8845, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.584, | |
| "grad_norm": 0.6751191020011902, | |
| "learning_rate": 9.736209109560201e-06, | |
| "loss": 0.9402, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 0.6485176682472229, | |
| "learning_rate": 9.721063728508384e-06, | |
| "loss": 0.8936, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.6805006265640259, | |
| "learning_rate": 9.705508070938219e-06, | |
| "loss": 0.9238, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.7293475270271301, | |
| "learning_rate": 9.689543488689332e-06, | |
| "loss": 0.8917, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.616, | |
| "grad_norm": 0.6713281869888306, | |
| "learning_rate": 9.673171369138297e-06, | |
| "loss": 0.9051, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.7109168171882629, | |
| "learning_rate": 9.656393135078067e-06, | |
| "loss": 0.9147, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.632, | |
| "grad_norm": 0.7239603400230408, | |
| "learning_rate": 9.639210244594335e-06, | |
| "loss": 0.878, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6019280552864075, | |
| "learning_rate": 9.621624190938802e-06, | |
| "loss": 0.8919, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.7745220065116882, | |
| "learning_rate": 9.603636502399436e-06, | |
| "loss": 0.8745, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 0.8410776853561401, | |
| "learning_rate": 9.585248742167638e-06, | |
| "loss": 0.9138, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.664, | |
| "grad_norm": 0.5263102054595947, | |
| "learning_rate": 9.566462508202403e-06, | |
| "loss": 0.8924, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.9095488786697388, | |
| "learning_rate": 9.547279433091446e-06, | |
| "loss": 0.9034, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.6388502717018127, | |
| "learning_rate": 9.527701183909336e-06, | |
| "loss": 0.9433, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.5810441970825195, | |
| "learning_rate": 9.507729462072615e-06, | |
| "loss": 0.9063, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.6178363561630249, | |
| "learning_rate": 9.48736600319193e-06, | |
| "loss": 0.8911, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6862829923629761, | |
| "learning_rate": 9.466612576921223e-06, | |
| "loss": 0.8723, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.712, | |
| "grad_norm": 0.6216509938240051, | |
| "learning_rate": 9.445470986803922e-06, | |
| "loss": 0.9128, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6560526490211487, | |
| "learning_rate": 9.423943070116219e-06, | |
| "loss": 0.8976, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.728, | |
| "grad_norm": 0.6992607116699219, | |
| "learning_rate": 9.402030697707398e-06, | |
| "loss": 0.9017, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.6775327920913696, | |
| "learning_rate": 9.37973577383726e-06, | |
| "loss": 0.9186, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.5652820467948914, | |
| "learning_rate": 9.357060236010626e-06, | |
| "loss": 0.9105, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.6999420523643494, | |
| "learning_rate": 9.334006054808966e-06, | |
| "loss": 0.9206, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.6290011405944824, | |
| "learning_rate": 9.310575233719155e-06, | |
| "loss": 0.9198, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7411805987358093, | |
| "learning_rate": 9.28676980895935e-06, | |
| "loss": 0.8981, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.776, | |
| "grad_norm": 0.600430965423584, | |
| "learning_rate": 9.262591849302049e-06, | |
| "loss": 0.8867, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 0.6421351432800293, | |
| "learning_rate": 9.238043455894294e-06, | |
| "loss": 0.9053, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.7420929670333862, | |
| "learning_rate": 9.213126762075088e-06, | |
| "loss": 0.8963, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5919941663742065, | |
| "learning_rate": 9.187843933189994e-06, | |
| "loss": 0.8234, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.808, | |
| "grad_norm": 0.7523407340049744, | |
| "learning_rate": 9.162197166402957e-06, | |
| "loss": 0.9065, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.6169841885566711, | |
| "learning_rate": 9.136188690505363e-06, | |
| "loss": 0.9195, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.824, | |
| "grad_norm": 0.6386623382568359, | |
| "learning_rate": 9.109820765722357e-06, | |
| "loss": 0.8867, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.5823110342025757, | |
| "learning_rate": 9.083095683516414e-06, | |
| "loss": 0.9039, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.5956842303276062, | |
| "learning_rate": 9.056015766388205e-06, | |
| "loss": 0.9004, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.5407257676124573, | |
| "learning_rate": 9.028583367674767e-06, | |
| "loss": 0.8644, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.856, | |
| "grad_norm": 0.6211457252502441, | |
| "learning_rate": 9.00080087134498e-06, | |
| "loss": 0.9033, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.5631619095802307, | |
| "learning_rate": 8.972670691792409e-06, | |
| "loss": 0.846, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.872, | |
| "grad_norm": 0.5860468745231628, | |
| "learning_rate": 8.944195273625472e-06, | |
| "loss": 0.9196, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.5991073846817017, | |
| "learning_rate": 8.915377091454992e-06, | |
| "loss": 0.8949, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.5442657470703125, | |
| "learning_rate": 8.886218649679162e-06, | |
| "loss": 0.8848, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.49064090847969055, | |
| "learning_rate": 8.856722482265886e-06, | |
| "loss": 0.8571, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.904, | |
| "grad_norm": 0.6109375953674316, | |
| "learning_rate": 8.826891152532579e-06, | |
| "loss": 0.874, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.5844782590866089, | |
| "learning_rate": 8.796727252923403e-06, | |
| "loss": 0.8715, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.5409021377563477, | |
| "learning_rate": 8.766233404783975e-06, | |
| "loss": 0.8836, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.5358095169067383, | |
| "learning_rate": 8.735412258133562e-06, | |
| "loss": 0.8686, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.6133047938346863, | |
| "learning_rate": 8.704266491434787e-06, | |
| "loss": 0.8661, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.6640852093696594, | |
| "learning_rate": 8.672798811360863e-06, | |
| "loss": 0.8357, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.952, | |
| "grad_norm": 0.6169326901435852, | |
| "learning_rate": 8.641011952560372e-06, | |
| "loss": 0.8542, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6841595768928528, | |
| "learning_rate": 8.608908677419606e-06, | |
| "loss": 0.898, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.968, | |
| "grad_norm": 0.6243302822113037, | |
| "learning_rate": 8.576491775822527e-06, | |
| "loss": 0.8579, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.5553723573684692, | |
| "learning_rate": 8.543764064908295e-06, | |
| "loss": 0.8628, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.6372663974761963, | |
| "learning_rate": 8.510728388826464e-06, | |
| "loss": 0.8672, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.6535351872444153, | |
| "learning_rate": 8.477387618489808e-06, | |
| "loss": 0.849, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6568986773490906, | |
| "learning_rate": 8.443744651324828e-06, | |
| "loss": 0.8512, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 0.6753949522972107, | |
| "learning_rate": 8.409802411019962e-06, | |
| "loss": 0.8784, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 0.7170552015304565, | |
| "learning_rate": 8.375563847271506e-06, | |
| "loss": 0.8431, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6562941074371338, | |
| "learning_rate": 8.341031935527267e-06, | |
| "loss": 0.8489, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 0.6256612539291382, | |
| "learning_rate": 8.306209676727994e-06, | |
| "loss": 0.8586, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.7772700190544128, | |
| "learning_rate": 8.271100097046585e-06, | |
| "loss": 0.8498, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.048, | |
| "grad_norm": 0.6408899426460266, | |
| "learning_rate": 8.235706247625098e-06, | |
| "loss": 0.8249, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5903573632240295, | |
| "learning_rate": 8.200031204309604e-06, | |
| "loss": 0.8051, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 0.7947317957878113, | |
| "learning_rate": 8.16407806738288e-06, | |
| "loss": 0.8263, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.072, | |
| "grad_norm": 0.6758819818496704, | |
| "learning_rate": 8.127849961294984e-06, | |
| "loss": 0.801, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.682586669921875, | |
| "learning_rate": 8.091350034391732e-06, | |
| "loss": 0.8507, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7249560356140137, | |
| "learning_rate": 8.05458145864109e-06, | |
| "loss": 0.836, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 0.6341940760612488, | |
| "learning_rate": 8.017547429357532e-06, | |
| "loss": 0.846, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.7217475175857544, | |
| "learning_rate": 7.980251164924342e-06, | |
| "loss": 0.8489, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 0.5944041609764099, | |
| "learning_rate": 7.94269590651393e-06, | |
| "loss": 0.8149, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.6496677994728088, | |
| "learning_rate": 7.904884917806174e-06, | |
| "loss": 0.8401, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 0.604494571685791, | |
| "learning_rate": 7.866821484704777e-06, | |
| "loss": 0.8379, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 0.5416861772537231, | |
| "learning_rate": 7.828508915051724e-06, | |
| "loss": 0.817, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.144, | |
| "grad_norm": 0.5813770890235901, | |
| "learning_rate": 7.789950538339813e-06, | |
| "loss": 0.8452, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5333764553070068, | |
| "learning_rate": 7.751149705423313e-06, | |
| "loss": 0.8404, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.5556795597076416, | |
| "learning_rate": 7.712109788226763e-06, | |
| "loss": 0.8142, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.168, | |
| "grad_norm": 0.6207058429718018, | |
| "learning_rate": 7.672834179451943e-06, | |
| "loss": 0.8746, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 0.5369215607643127, | |
| "learning_rate": 7.633326292283028e-06, | |
| "loss": 0.8469, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 0.5977884531021118, | |
| "learning_rate": 7.593589560089984e-06, | |
| "loss": 0.8328, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 0.6059456467628479, | |
| "learning_rate": 7.553627436130183e-06, | |
| "loss": 0.8582, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5836285948753357, | |
| "learning_rate": 7.513443393248312e-06, | |
| "loss": 0.8558, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 0.6144264936447144, | |
| "learning_rate": 7.473040923574567e-06, | |
| "loss": 0.8549, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.626693844795227, | |
| "learning_rate": 7.432423538221179e-06, | |
| "loss": 0.8473, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 0.6432238221168518, | |
| "learning_rate": 7.391594766977277e-06, | |
| "loss": 0.8281, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 0.6119612455368042, | |
| "learning_rate": 7.350558158002154e-06, | |
| "loss": 0.8419, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.6895710229873657, | |
| "learning_rate": 7.3093172775169e-06, | |
| "loss": 0.8408, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.6269011497497559, | |
| "learning_rate": 7.2678757094945e-06, | |
| "loss": 0.8345, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 0.5723068118095398, | |
| "learning_rate": 7.226237055348369e-06, | |
| "loss": 0.8654, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.264, | |
| "grad_norm": 0.6293651461601257, | |
| "learning_rate": 7.184404933619377e-06, | |
| "loss": 0.8387, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 0.6416763067245483, | |
| "learning_rate": 7.142382979661386e-06, | |
| "loss": 0.8317, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5021880865097046, | |
| "learning_rate": 7.100174845325327e-06, | |
| "loss": 0.8281, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 0.567487895488739, | |
| "learning_rate": 7.057784198641835e-06, | |
| "loss": 0.8646, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.5684318542480469, | |
| "learning_rate": 7.015214723502496e-06, | |
| "loss": 0.8293, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 0.5071728825569153, | |
| "learning_rate": 6.972470119339692e-06, | |
| "loss": 0.8097, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.312, | |
| "grad_norm": 0.6408015489578247, | |
| "learning_rate": 6.929554100805118e-06, | |
| "loss": 0.85, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.5955577492713928, | |
| "learning_rate": 6.886470397446958e-06, | |
| "loss": 0.8343, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 0.4974721670150757, | |
| "learning_rate": 6.843222753385785e-06, | |
| "loss": 0.7749, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.336, | |
| "grad_norm": 0.6704075336456299, | |
| "learning_rate": 6.799814926989171e-06, | |
| "loss": 0.8165, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5451616644859314, | |
| "learning_rate": 6.756250690545079e-06, | |
| "loss": 0.8349, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 0.5713831782341003, | |
| "learning_rate": 6.712533829934042e-06, | |
| "loss": 0.8224, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.5288674831390381, | |
| "learning_rate": 6.6686681443001485e-06, | |
| "loss": 0.8518, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 0.6580489873886108, | |
| "learning_rate": 6.62465744572089e-06, | |
| "loss": 0.8736, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 0.5576747059822083, | |
| "learning_rate": 6.580505558875878e-06, | |
| "loss": 0.8171, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 0.6031728386878967, | |
| "learning_rate": 6.536216320714466e-06, | |
| "loss": 0.8325, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.49704813957214355, | |
| "learning_rate": 6.491793580122301e-06, | |
| "loss": 0.8348, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.5445089340209961, | |
| "learning_rate": 6.447241197586847e-06, | |
| "loss": 0.8679, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5517768263816833, | |
| "learning_rate": 6.402563044861899e-06, | |
| "loss": 0.8414, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 0.5023953318595886, | |
| "learning_rate": 6.357763004631104e-06, | |
| "loss": 0.7706, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 0.5183501243591309, | |
| "learning_rate": 6.312844970170551e-06, | |
| "loss": 0.8145, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.432, | |
| "grad_norm": 0.5883236527442932, | |
| "learning_rate": 6.267812845010431e-06, | |
| "loss": 0.8071, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5993454456329346, | |
| "learning_rate": 6.2226705425958e-06, | |
| "loss": 0.8338, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 0.5319005250930786, | |
| "learning_rate": 6.177421985946499e-06, | |
| "loss": 0.856, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.456, | |
| "grad_norm": 0.6088922619819641, | |
| "learning_rate": 6.132071107316221e-06, | |
| "loss": 0.8405, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 0.565377950668335, | |
| "learning_rate": 6.0866218478507875e-06, | |
| "loss": 0.848, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5202862024307251, | |
| "learning_rate": 6.041078157245649e-06, | |
| "loss": 0.8139, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.5979906320571899, | |
| "learning_rate": 5.995443993402647e-06, | |
| "loss": 0.8267, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.5841382145881653, | |
| "learning_rate": 5.949723322086053e-06, | |
| "loss": 0.8011, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 0.5658543705940247, | |
| "learning_rate": 5.9039201165779315e-06, | |
| "loss": 0.786, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.504, | |
| "grad_norm": 0.5966721773147583, | |
| "learning_rate": 5.858038357332851e-06, | |
| "loss": 0.8083, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 0.5970141291618347, | |
| "learning_rate": 5.812082031631966e-06, | |
| "loss": 0.8421, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.5003347992897034, | |
| "learning_rate": 5.766055133236513e-06, | |
| "loss": 0.823, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.528, | |
| "grad_norm": 0.5702365040779114, | |
| "learning_rate": 5.7199616620407325e-06, | |
| "loss": 0.8309, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.6028835773468018, | |
| "learning_rate": 5.673805623724272e-06, | |
| "loss": 0.8088, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 0.5072699189186096, | |
| "learning_rate": 5.627591029404072e-06, | |
| "loss": 0.8199, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.552, | |
| "grad_norm": 0.5302557945251465, | |
| "learning_rate": 5.581321895285787e-06, | |
| "loss": 0.8194, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.48366543650627136, | |
| "learning_rate": 5.535002242314772e-06, | |
| "loss": 0.8176, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 0.6495940089225769, | |
| "learning_rate": 5.488636095826636e-06, | |
| "loss": 0.8078, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 0.4819113314151764, | |
| "learning_rate": 5.4422274851974356e-06, | |
| "loss": 0.8397, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.5706694722175598, | |
| "learning_rate": 5.395780443493508e-06, | |
| "loss": 0.8549, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 0.5659922957420349, | |
| "learning_rate": 5.34929900712098e-06, | |
| "loss": 0.7686, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5568146705627441, | |
| "learning_rate": 5.302787215474992e-06, | |
| "loss": 0.8329, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 0.5732328295707703, | |
| "learning_rate": 5.256249110588659e-06, | |
| "loss": 0.833, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 0.5560725927352905, | |
| "learning_rate": 5.209688736781811e-06, | |
| "loss": 0.812, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.624, | |
| "grad_norm": 0.49767476320266724, | |
| "learning_rate": 5.163110140309518e-06, | |
| "loss": 0.8408, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.5641460418701172, | |
| "learning_rate": 5.116517369010467e-06, | |
| "loss": 0.8294, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.6225250363349915, | |
| "learning_rate": 5.069914471955179e-06, | |
| "loss": 0.7811, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.6480000000000001, | |
| "grad_norm": 0.5748852491378784, | |
| "learning_rate": 5.023305499094145e-06, | |
| "loss": 0.8307, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 0.5865118503570557, | |
| "learning_rate": 4.976694500905858e-06, | |
| "loss": 0.8046, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5250575542449951, | |
| "learning_rate": 4.930085528044823e-06, | |
| "loss": 0.8315, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 0.5795397162437439, | |
| "learning_rate": 4.883482630989536e-06, | |
| "loss": 0.8472, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.6399658918380737, | |
| "learning_rate": 4.8368898596904834e-06, | |
| "loss": 0.8163, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 0.5113149881362915, | |
| "learning_rate": 4.790311263218191e-06, | |
| "loss": 0.8362, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.696, | |
| "grad_norm": 0.4925471544265747, | |
| "learning_rate": 4.743750889411342e-06, | |
| "loss": 0.8364, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.5226654410362244, | |
| "learning_rate": 4.697212784525009e-06, | |
| "loss": 0.8156, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 0.49383535981178284, | |
| "learning_rate": 4.65070099287902e-06, | |
| "loss": 0.8363, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.5424585342407227, | |
| "learning_rate": 4.604219556506492e-06, | |
| "loss": 0.8252, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4512873589992523, | |
| "learning_rate": 4.557772514802564e-06, | |
| "loss": 0.8188, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 0.49828025698661804, | |
| "learning_rate": 4.511363904173366e-06, | |
| "loss": 0.835, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.744, | |
| "grad_norm": 0.5537604093551636, | |
| "learning_rate": 4.46499775768523e-06, | |
| "loss": 0.8072, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 0.509067714214325, | |
| "learning_rate": 4.418678104714214e-06, | |
| "loss": 0.8331, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.49603044986724854, | |
| "learning_rate": 4.372408970595931e-06, | |
| "loss": 0.8031, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 0.5341010689735413, | |
| "learning_rate": 4.326194376275729e-06, | |
| "loss": 0.8117, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.49610334634780884, | |
| "learning_rate": 4.280038337959268e-06, | |
| "loss": 0.8137, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 0.4939770996570587, | |
| "learning_rate": 4.2339448667634885e-06, | |
| "loss": 0.7834, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5622074007987976, | |
| "learning_rate": 4.187917968368036e-06, | |
| "loss": 0.8227, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.6220787763595581, | |
| "learning_rate": 4.141961642667152e-06, | |
| "loss": 0.8526, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 0.5713546276092529, | |
| "learning_rate": 4.09607988342207e-06, | |
| "loss": 0.8529, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.8159999999999998, | |
| "grad_norm": 0.5482434034347534, | |
| "learning_rate": 4.0502766779139485e-06, | |
| "loss": 0.8142, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.6168124079704285, | |
| "learning_rate": 4.0045560065973535e-06, | |
| "loss": 0.8127, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 0.5725027918815613, | |
| "learning_rate": 3.958921842754351e-06, | |
| "loss": 0.8405, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.5908339023590088, | |
| "learning_rate": 3.913378152149214e-06, | |
| "loss": 0.7938, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 0.5032182931900024, | |
| "learning_rate": 3.86792889268378e-06, | |
| "loss": 0.8237, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4820937514305115, | |
| "learning_rate": 3.8225780140535025e-06, | |
| "loss": 0.8157, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 0.5697328448295593, | |
| "learning_rate": 3.777329457404202e-06, | |
| "loss": 0.8471, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.5110496282577515, | |
| "learning_rate": 3.7321871549895715e-06, | |
| "loss": 0.8184, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.4492500424385071, | |
| "learning_rate": 3.68715502982945e-06, | |
| "loss": 0.8048, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.888, | |
| "grad_norm": 0.5033578872680664, | |
| "learning_rate": 3.6422369953688973e-06, | |
| "loss": 0.8462, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 0.47552290558815, | |
| "learning_rate": 3.5974369551381023e-06, | |
| "loss": 0.814, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 0.47769391536712646, | |
| "learning_rate": 3.5527588024131542e-06, | |
| "loss": 0.8076, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.912, | |
| "grad_norm": 0.47970548272132874, | |
| "learning_rate": 3.5082064198777e-06, | |
| "loss": 0.8063, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5200381875038147, | |
| "learning_rate": 3.463783679285535e-06, | |
| "loss": 0.8212, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 0.4885690212249756, | |
| "learning_rate": 3.4194944411241213e-06, | |
| "loss": 0.8087, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.936, | |
| "grad_norm": 0.523141086101532, | |
| "learning_rate": 3.3753425542791106e-06, | |
| "loss": 0.8116, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 0.5129029750823975, | |
| "learning_rate": 3.3313318556998523e-06, | |
| "loss": 0.8062, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 0.4830564260482788, | |
| "learning_rate": 3.2874661700659586e-06, | |
| "loss": 0.8428, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.4503372609615326, | |
| "learning_rate": 3.2437493094549223e-06, | |
| "loss": 0.8268, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.4484092891216278, | |
| "learning_rate": 3.200185073010831e-06, | |
| "loss": 0.8086, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 0.41616371273994446, | |
| "learning_rate": 3.1567772466142156e-06, | |
| "loss": 0.8212, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4492393732070923, | |
| "learning_rate": 3.1135296025530426e-06, | |
| "loss": 0.8217, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 0.48867785930633545, | |
| "learning_rate": 3.070445899194885e-06, | |
| "loss": 0.8394, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5146781206130981, | |
| "learning_rate": 3.0275298806603102e-06, | |
| "loss": 0.8254, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.008, | |
| "grad_norm": 0.4940868616104126, | |
| "learning_rate": 2.984785276497507e-06, | |
| "loss": 0.7948, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.5990253686904907, | |
| "learning_rate": 2.9422158013581658e-06, | |
| "loss": 0.7709, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.024, | |
| "grad_norm": 0.52504962682724, | |
| "learning_rate": 2.899825154674674e-06, | |
| "loss": 0.7659, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.47968220710754395, | |
| "learning_rate": 2.8576170203386144e-06, | |
| "loss": 0.7978, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.5218414068222046, | |
| "learning_rate": 2.8155950663806234e-06, | |
| "loss": 0.8075, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6022427082061768, | |
| "learning_rate": 2.7737629446516325e-06, | |
| "loss": 0.7967, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.056, | |
| "grad_norm": 0.45537319779396057, | |
| "learning_rate": 2.732124290505501e-06, | |
| "loss": 0.7637, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.47411370277404785, | |
| "learning_rate": 2.6906827224831024e-06, | |
| "loss": 0.7999, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.072, | |
| "grad_norm": 0.47417914867401123, | |
| "learning_rate": 2.6494418419978485e-06, | |
| "loss": 0.8002, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.47205376625061035, | |
| "learning_rate": 2.608405233022724e-06, | |
| "loss": 0.7841, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.088, | |
| "grad_norm": 0.4820186197757721, | |
| "learning_rate": 2.5675764617788233e-06, | |
| "loss": 0.7808, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 0.44489675760269165, | |
| "learning_rate": 2.526959076425434e-06, | |
| "loss": 0.7758, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.104, | |
| "grad_norm": 0.4890801012516022, | |
| "learning_rate": 2.4865566067516896e-06, | |
| "loss": 0.8111, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.4414970278739929, | |
| "learning_rate": 2.4463725638698182e-06, | |
| "loss": 0.7976, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.45259734988212585, | |
| "learning_rate": 2.406410439910017e-06, | |
| "loss": 0.7903, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.48778876662254333, | |
| "learning_rate": 2.366673707716973e-06, | |
| "loss": 0.8107, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.136, | |
| "grad_norm": 0.44795307517051697, | |
| "learning_rate": 2.327165820548059e-06, | |
| "loss": 0.7231, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 0.43515029549598694, | |
| "learning_rate": 2.287890211773238e-06, | |
| "loss": 0.7503, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.152, | |
| "grad_norm": 0.4571765661239624, | |
| "learning_rate": 2.2488502945766893e-06, | |
| "loss": 0.8206, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.4228002727031708, | |
| "learning_rate": 2.210049461660189e-06, | |
| "loss": 0.7817, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.168, | |
| "grad_norm": 0.4443894624710083, | |
| "learning_rate": 2.1714910849482777e-06, | |
| "loss": 0.7855, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4512254297733307, | |
| "learning_rate": 2.1331785152952243e-06, | |
| "loss": 0.8024, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.184, | |
| "grad_norm": 0.5297703742980957, | |
| "learning_rate": 2.0951150821938278e-06, | |
| "loss": 0.7472, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 0.46972817182540894, | |
| "learning_rate": 2.0573040934860717e-06, | |
| "loss": 0.7666, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.48370957374572754, | |
| "learning_rate": 2.0197488350756618e-06, | |
| "loss": 0.7913, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.43513721227645874, | |
| "learning_rate": 1.98245257064247e-06, | |
| "loss": 0.7855, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.216, | |
| "grad_norm": 0.44624122977256775, | |
| "learning_rate": 1.945418541358911e-06, | |
| "loss": 0.7948, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.5190390944480896, | |
| "learning_rate": 1.9086499656082685e-06, | |
| "loss": 0.7953, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.232, | |
| "grad_norm": 0.4471817910671234, | |
| "learning_rate": 1.872150038705015e-06, | |
| "loss": 0.7723, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4081578254699707, | |
| "learning_rate": 1.835921932617119e-06, | |
| "loss": 0.7863, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.248, | |
| "grad_norm": 0.4299153983592987, | |
| "learning_rate": 1.7999687956903955e-06, | |
| "loss": 0.7459, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.44286420941352844, | |
| "learning_rate": 1.7642937523749038e-06, | |
| "loss": 0.7549, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.2640000000000002, | |
| "grad_norm": 0.5079202651977539, | |
| "learning_rate": 1.7288999029534177e-06, | |
| "loss": 0.7945, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.5463809967041016, | |
| "learning_rate": 1.6937903232720076e-06, | |
| "loss": 0.8004, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 0.4327748715877533, | |
| "learning_rate": 1.6589680644727347e-06, | |
| "loss": 0.7723, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 0.4358852505683899, | |
| "learning_rate": 1.6244361527284953e-06, | |
| "loss": 0.7721, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.296, | |
| "grad_norm": 0.41679710149765015, | |
| "learning_rate": 1.5901975889800387e-06, | |
| "loss": 0.7891, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4202753007411957, | |
| "learning_rate": 1.556255348675174e-06, | |
| "loss": 0.7585, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.312, | |
| "grad_norm": 0.4579218029975891, | |
| "learning_rate": 1.522612381510195e-06, | |
| "loss": 0.7672, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.47101694345474243, | |
| "learning_rate": 1.489271611173538e-06, | |
| "loss": 0.7861, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.328, | |
| "grad_norm": 0.4235668182373047, | |
| "learning_rate": 1.4562359350917054e-06, | |
| "loss": 0.791, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 0.44197484850883484, | |
| "learning_rate": 1.423508224177474e-06, | |
| "loss": 0.7861, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.344, | |
| "grad_norm": 0.43885377049446106, | |
| "learning_rate": 1.3910913225803946e-06, | |
| "loss": 0.7965, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.4231855273246765, | |
| "learning_rate": 1.35898804743963e-06, | |
| "loss": 0.7815, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.48017239570617676, | |
| "learning_rate": 1.3272011886391368e-06, | |
| "loss": 0.8028, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.46980807185173035, | |
| "learning_rate": 1.295733508565213e-06, | |
| "loss": 0.7692, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.376, | |
| "grad_norm": 0.45487645268440247, | |
| "learning_rate": 1.2645877418664394e-06, | |
| "loss": 0.7546, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 0.44050565361976624, | |
| "learning_rate": 1.2337665952160266e-06, | |
| "loss": 0.798, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.392, | |
| "grad_norm": 0.4430494010448456, | |
| "learning_rate": 1.2032727470765982e-06, | |
| "loss": 0.7849, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.4752819836139679, | |
| "learning_rate": 1.1731088474674235e-06, | |
| "loss": 0.8045, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.408, | |
| "grad_norm": 0.4466467499732971, | |
| "learning_rate": 1.1432775177341165e-06, | |
| "loss": 0.7841, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.4148922562599182, | |
| "learning_rate": 1.11378135032084e-06, | |
| "loss": 0.7993, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.424, | |
| "grad_norm": 0.44224417209625244, | |
| "learning_rate": 1.08462290854501e-06, | |
| "loss": 0.7655, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.42925554513931274, | |
| "learning_rate": 1.0558047263745297e-06, | |
| "loss": 0.7607, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.4702661633491516, | |
| "learning_rate": 1.0273293082075914e-06, | |
| "loss": 0.7425, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.44665732979774475, | |
| "learning_rate": 9.991991286550207e-07, | |
| "loss": 0.8086, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.456, | |
| "grad_norm": 0.44039228558540344, | |
| "learning_rate": 9.71416632325235e-07, | |
| "loss": 0.7751, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 0.4385325610637665, | |
| "learning_rate": 9.439842336117954e-07, | |
| "loss": 0.8141, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.472, | |
| "grad_norm": 0.43111181259155273, | |
| "learning_rate": 9.169043164835867e-07, | |
| "loss": 0.762, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.4057021737098694, | |
| "learning_rate": 8.901792342776439e-07, | |
| "loss": 0.8327, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.488, | |
| "grad_norm": 0.4096888303756714, | |
| "learning_rate": 8.638113094946382e-07, | |
| "loss": 0.8251, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4527164101600647, | |
| "learning_rate": 8.378028335970451e-07, | |
| "loss": 0.7605, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.504, | |
| "grad_norm": 0.4072956144809723, | |
| "learning_rate": 8.121560668100065e-07, | |
| "loss": 0.7836, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.4347890615463257, | |
| "learning_rate": 7.868732379249122e-07, | |
| "loss": 0.7888, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.44405418634414673, | |
| "learning_rate": 7.619565441057075e-07, | |
| "loss": 0.7476, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 0.4290863573551178, | |
| "learning_rate": 7.37408150697953e-07, | |
| "loss": 0.776, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.536, | |
| "grad_norm": 0.4163723289966583, | |
| "learning_rate": 7.132301910406503e-07, | |
| "loss": 0.7841, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.39156779646873474, | |
| "learning_rate": 6.894247662808456e-07, | |
| "loss": 0.8141, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.552, | |
| "grad_norm": 0.3718986511230469, | |
| "learning_rate": 6.659939451910341e-07, | |
| "loss": 0.8007, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.39346760511398315, | |
| "learning_rate": 6.429397639893758e-07, | |
| "loss": 0.7884, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.568, | |
| "grad_norm": 0.38905784487724304, | |
| "learning_rate": 6.202642261627411e-07, | |
| "loss": 0.7697, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 0.4061625301837921, | |
| "learning_rate": 5.979693022926025e-07, | |
| "loss": 0.7715, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.584, | |
| "grad_norm": 0.4345610737800598, | |
| "learning_rate": 5.760569298837825e-07, | |
| "loss": 0.785, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.3941546678543091, | |
| "learning_rate": 5.54529013196079e-07, | |
| "loss": 0.8025, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.36712539196014404, | |
| "learning_rate": 5.333874230787772e-07, | |
| "loss": 0.8007, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.3812052607536316, | |
| "learning_rate": 5.126339968080696e-07, | |
| "loss": 0.7889, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.616, | |
| "grad_norm": 0.4085177481174469, | |
| "learning_rate": 4.922705379273862e-07, | |
| "loss": 0.7641, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.42239484190940857, | |
| "learning_rate": 4.7229881609066387e-07, | |
| "loss": 0.7979, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.632, | |
| "grad_norm": 0.4121035635471344, | |
| "learning_rate": 4.5272056690855494e-07, | |
| "loss": 0.7858, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.38635125756263733, | |
| "learning_rate": 4.335374917975982e-07, | |
| "loss": 0.8095, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.648, | |
| "grad_norm": 0.38656213879585266, | |
| "learning_rate": 4.147512578323615e-07, | |
| "loss": 0.7669, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.41909682750701904, | |
| "learning_rate": 3.9636349760056427e-07, | |
| "loss": 0.7788, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.664, | |
| "grad_norm": 0.4400654137134552, | |
| "learning_rate": 3.783758090611983e-07, | |
| "loss": 0.7382, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 0.3858467638492584, | |
| "learning_rate": 3.6078975540566716e-07, | |
| "loss": 0.767, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.36701512336730957, | |
| "learning_rate": 3.4360686492193263e-07, | |
| "loss": 0.7928, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.38759228587150574, | |
| "learning_rate": 3.268286308617041e-07, | |
| "loss": 0.8181, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.6959999999999997, | |
| "grad_norm": 0.38279959559440613, | |
| "learning_rate": 3.104565113106689e-07, | |
| "loss": 0.7943, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.4579544961452484, | |
| "learning_rate": 2.9449192906178205e-07, | |
| "loss": 0.7605, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.7119999999999997, | |
| "grad_norm": 0.40908747911453247, | |
| "learning_rate": 2.789362714916172e-07, | |
| "loss": 0.8095, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.37257906794548035, | |
| "learning_rate": 2.6379089043980064e-07, | |
| "loss": 0.7988, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.7279999999999998, | |
| "grad_norm": 0.431416779756546, | |
| "learning_rate": 2.4905710209153224e-07, | |
| "loss": 0.7819, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.4172951579093933, | |
| "learning_rate": 2.3473618686320477e-07, | |
| "loss": 0.7597, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.7439999999999998, | |
| "grad_norm": 0.41569289565086365, | |
| "learning_rate": 2.208293892911284e-07, | |
| "loss": 0.7797, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.42253097891807556, | |
| "learning_rate": 2.0733791792338197e-07, | |
| "loss": 0.787, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.36844950914382935, | |
| "learning_rate": 1.9426294521477874e-07, | |
| "loss": 0.8194, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.4011365473270416, | |
| "learning_rate": 1.8160560742498223e-07, | |
| "loss": 0.7881, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.776, | |
| "grad_norm": 0.44703608751296997, | |
| "learning_rate": 1.6936700451975818e-07, | |
| "loss": 0.7567, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.39994892477989197, | |
| "learning_rate": 1.5754820007538473e-07, | |
| "loss": 0.7683, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.792, | |
| "grad_norm": 0.4645727872848511, | |
| "learning_rate": 1.461502211862237e-07, | |
| "loss": 0.7094, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.3852178156375885, | |
| "learning_rate": 1.3517405837546404e-07, | |
| "loss": 0.8263, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.808, | |
| "grad_norm": 0.4058782160282135, | |
| "learning_rate": 1.2462066550903818e-07, | |
| "loss": 0.7728, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.4285104274749756, | |
| "learning_rate": 1.1449095971273305e-07, | |
| "loss": 0.8048, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 2.824, | |
| "grad_norm": 0.39162999391555786, | |
| "learning_rate": 1.0478582129248516e-07, | |
| "loss": 0.8227, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.39359840750694275, | |
| "learning_rate": 9.550609365787888e-08, | |
| "loss": 0.7782, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.4036370813846588, | |
| "learning_rate": 8.66525832488535e-08, | |
| "loss": 0.7619, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.42931556701660156, | |
| "learning_rate": 7.822605946561923e-08, | |
| "loss": 0.7714, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 2.856, | |
| "grad_norm": 0.4127103090286255, | |
| "learning_rate": 7.022725460179459e-08, | |
| "loss": 0.7252, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 0.3982616364955902, | |
| "learning_rate": 6.265686378076729e-08, | |
| "loss": 0.8099, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 2.872, | |
| "grad_norm": 0.4003240168094635, | |
| "learning_rate": 5.5515544895284324e-08, | |
| "loss": 0.7717, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3641519844532013, | |
| "learning_rate": 4.880391855028088e-08, | |
| "loss": 0.7815, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.888, | |
| "grad_norm": 0.3692556619644165, | |
| "learning_rate": 4.252256800894694e-08, | |
| "loss": 0.8232, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.3871900141239166, | |
| "learning_rate": 3.6672039142039426e-08, | |
| "loss": 0.7802, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 2.904, | |
| "grad_norm": 0.39646491408348083, | |
| "learning_rate": 3.125284038044407e-08, | |
| "loss": 0.7824, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 0.40282607078552246, | |
| "learning_rate": 2.6265442670991293e-08, | |
| "loss": 0.7715, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.4330500662326813, | |
| "learning_rate": 2.1710279435530058e-08, | |
| "loss": 0.7895, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.3595670461654663, | |
| "learning_rate": 1.7587746533260786e-08, | |
| "loss": 0.8252, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 2.936, | |
| "grad_norm": 0.41389673948287964, | |
| "learning_rate": 1.3898202226333424e-08, | |
| "loss": 0.7914, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.42384371161460876, | |
| "learning_rate": 1.0641967148716236e-08, | |
| "loss": 0.775, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 2.952, | |
| "grad_norm": 0.4105876088142395, | |
| "learning_rate": 7.819324278328099e-09, | |
| "loss": 0.8268, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.4283929467201233, | |
| "learning_rate": 5.430518912448169e-09, | |
| "loss": 0.785, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.968, | |
| "grad_norm": 0.4095473289489746, | |
| "learning_rate": 3.4757586464001513e-09, | |
| "loss": 0.7683, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.38301780819892883, | |
| "learning_rate": 1.9552133555084117e-09, | |
| "loss": 0.7837, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.984, | |
| "grad_norm": 0.4291246831417084, | |
| "learning_rate": 8.690151803386615e-10, | |
| "loss": 0.7784, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "grad_norm": 0.4616396725177765, | |
| "learning_rate": 2.1725851521103847e-10, | |
| "loss": 0.753, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4092578887939453, | |
| "learning_rate": 0.0, | |
| "loss": 0.7939, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 375, | |
| "total_flos": 372577161248768.0, | |
| "train_loss": 0.8553513642946879, | |
| "train_runtime": 20932.817, | |
| "train_samples_per_second": 1.72, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 375, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 372577161248768.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |