| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.988, | |
| "eval_steps": 500, | |
| "global_step": 249, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012, | |
| "grad_norm": 6.98012638092041, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 1.2874, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 7.2280497550964355, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.3379, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.036, | |
| "grad_norm": 6.49622106552124, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.2253, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.694257736206055, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 1.2606, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.282245635986328, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.2299, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 4.9895453453063965, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.1999, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.084, | |
| "grad_norm": 4.745405673980713, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.2123, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.9413177967071533, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.2145, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.108, | |
| "grad_norm": 2.6902289390563965, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.1699, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.5415780544281006, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.1446, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.132, | |
| "grad_norm": 4.769739627838135, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.1619, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 4.757033824920654, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.1448, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.156, | |
| "grad_norm": 4.72186803817749, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.1301, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 3.7590582370758057, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.1112, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 3.620077133178711, | |
| "learning_rate": 6e-06, | |
| "loss": 1.1117, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 3.142825126647949, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.0957, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.204, | |
| "grad_norm": 2.2360799312591553, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.0306, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 1.9428048133850098, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.069, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.228, | |
| "grad_norm": 1.7547845840454102, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 1.0052, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.657777190208435, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.0323, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.252, | |
| "grad_norm": 1.529606819152832, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.0235, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 1.401602029800415, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.0052, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.276, | |
| "grad_norm": 1.3033355474472046, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 0.9629, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.35280442237854, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9827, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.1778582334518433, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9396, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 1.0478476285934448, | |
| "learning_rate": 9.999508258797876e-06, | |
| "loss": 0.9898, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.324, | |
| "grad_norm": 1.2109336853027344, | |
| "learning_rate": 9.998033131915266e-06, | |
| "loss": 0.9925, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 1.2328909635543823, | |
| "learning_rate": 9.995574909504434e-06, | |
| "loss": 0.9736, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.348, | |
| "grad_norm": 1.0876905918121338, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.946, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.9405947327613831, | |
| "learning_rate": 9.987711305469232e-06, | |
| "loss": 0.9901, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.372, | |
| "grad_norm": 0.8621402382850647, | |
| "learning_rate": 9.982307470588097e-06, | |
| "loss": 0.9884, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.858891487121582, | |
| "learning_rate": 9.975923633360985e-06, | |
| "loss": 0.9645, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.396, | |
| "grad_norm": 0.9253868460655212, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.9543, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 0.8706124424934387, | |
| "learning_rate": 9.960221167098124e-06, | |
| "loss": 0.9672, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.731913149356842, | |
| "learning_rate": 9.950905626682229e-06, | |
| "loss": 0.9399, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.7874249815940857, | |
| "learning_rate": 9.940616260552545e-06, | |
| "loss": 0.9548, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.444, | |
| "grad_norm": 0.8237317800521851, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.959, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 0.812666118144989, | |
| "learning_rate": 9.917124337830242e-06, | |
| "loss": 0.9314, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.468, | |
| "grad_norm": 0.681806743144989, | |
| "learning_rate": 9.903926402016153e-06, | |
| "loss": 0.9564, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.7917366623878479, | |
| "learning_rate": 9.889763881136439e-06, | |
| "loss": 0.958, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.492, | |
| "grad_norm": 0.8405086398124695, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.9135, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.504, | |
| "grad_norm": 0.7085656523704529, | |
| "learning_rate": 9.858556416234755e-06, | |
| "loss": 0.9224, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.516, | |
| "grad_norm": 0.6934413909912109, | |
| "learning_rate": 9.841517610611309e-06, | |
| "loss": 0.9147, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.8959766626358032, | |
| "learning_rate": 9.82352649551188e-06, | |
| "loss": 0.9471, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.643950879573822, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.9136, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.552, | |
| "grad_norm": 0.6084837913513184, | |
| "learning_rate": 9.784701678661045e-06, | |
| "loss": 0.9142, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.564, | |
| "grad_norm": 0.7732555866241455, | |
| "learning_rate": 9.763875613614482e-06, | |
| "loss": 0.887, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.617849588394165, | |
| "learning_rate": 9.742112510999516e-06, | |
| "loss": 0.9138, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.588, | |
| "grad_norm": 0.8307777643203735, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.9335, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.711719810962677, | |
| "learning_rate": 9.69579249943714e-06, | |
| "loss": 0.9457, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.612, | |
| "grad_norm": 0.6059475541114807, | |
| "learning_rate": 9.671244701472999e-06, | |
| "loss": 0.9146, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.6742075085639954, | |
| "learning_rate": 9.645778086114892e-06, | |
| "loss": 0.9178, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.636, | |
| "grad_norm": 0.6479842662811279, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.908, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.648, | |
| "grad_norm": 0.6552858948707581, | |
| "learning_rate": 9.592108619734107e-06, | |
| "loss": 0.953, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.637143075466156, | |
| "learning_rate": 9.563916325306595e-06, | |
| "loss": 0.8592, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.649308979511261, | |
| "learning_rate": 9.534826324599002e-06, | |
| "loss": 0.9176, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.684, | |
| "grad_norm": 0.9127522706985474, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.9613, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.696, | |
| "grad_norm": 0.6097024083137512, | |
| "learning_rate": 9.473976267396831e-06, | |
| "loss": 0.9182, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.708, | |
| "grad_norm": 0.7733173370361328, | |
| "learning_rate": 9.442228179894362e-06, | |
| "loss": 0.9183, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6656049489974976, | |
| "learning_rate": 9.409606321741776e-06, | |
| "loss": 0.9708, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.732, | |
| "grad_norm": 0.6555663347244263, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.9074, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.744, | |
| "grad_norm": 0.6316003203392029, | |
| "learning_rate": 9.341767130510529e-06, | |
| "loss": 0.8991, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.756, | |
| "grad_norm": 0.6999967694282532, | |
| "learning_rate": 9.306563141162046e-06, | |
| "loss": 0.8979, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.6841825842857361, | |
| "learning_rate": 9.270512065999139e-06, | |
| "loss": 0.8945, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.6653499603271484, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.9179, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.792, | |
| "grad_norm": 0.6969425082206726, | |
| "learning_rate": 9.195897187932513e-06, | |
| "loss": 0.8887, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.804, | |
| "grad_norm": 0.6159932017326355, | |
| "learning_rate": 9.157348061512728e-06, | |
| "loss": 0.8882, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.6578161120414734, | |
| "learning_rate": 9.117981199359575e-06, | |
| "loss": 0.9022, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.828, | |
| "grad_norm": 0.5723689794540405, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.8962, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.670845627784729, | |
| "learning_rate": 9.036825400468814e-06, | |
| "loss": 0.8789, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.852, | |
| "grad_norm": 0.6901213526725769, | |
| "learning_rate": 8.995052426791247e-06, | |
| "loss": 0.8723, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.5967451333999634, | |
| "learning_rate": 8.952493640360518e-06, | |
| "loss": 0.9267, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.876, | |
| "grad_norm": 0.6454184651374817, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.8762, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.888, | |
| "grad_norm": 0.6845690011978149, | |
| "learning_rate": 8.865052266813686e-06, | |
| "loss": 0.9348, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.616240918636322, | |
| "learning_rate": 8.820186879108038e-06, | |
| "loss": 0.9024, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.660294234752655, | |
| "learning_rate": 8.77457007408708e-06, | |
| "loss": 0.8927, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.924, | |
| "grad_norm": 0.6262758374214172, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.9414, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.936, | |
| "grad_norm": 0.6183608174324036, | |
| "learning_rate": 8.681118248795548e-06, | |
| "loss": 0.9135, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.948, | |
| "grad_norm": 0.6649749875068665, | |
| "learning_rate": 8.633301610170136e-06, | |
| "loss": 0.9005, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6269786357879639, | |
| "learning_rate": 8.584770313904138e-06, | |
| "loss": 0.8882, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.972, | |
| "grad_norm": 0.5906413197517395, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.8829, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.984, | |
| "grad_norm": 0.581932544708252, | |
| "learning_rate": 8.485602070884118e-06, | |
| "loss": 0.9358, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.996, | |
| "grad_norm": 0.689324676990509, | |
| "learning_rate": 8.43498463017451e-06, | |
| "loss": 0.9051, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.0778684616088867, | |
| "learning_rate": 8.383691540076372e-06, | |
| "loss": 1.5002, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.7571648359298706, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.826, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 0.8757691979408264, | |
| "learning_rate": 8.279118899309121e-06, | |
| "loss": 0.9041, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.044, | |
| "grad_norm": 0.5550661683082581, | |
| "learning_rate": 8.22585991771044e-06, | |
| "loss": 0.8642, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.713104784488678, | |
| "learning_rate": 8.171966420818227e-06, | |
| "loss": 0.8225, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.068, | |
| "grad_norm": 0.6943645477294922, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.9152, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.5287421941757202, | |
| "learning_rate": 8.062318406519751e-06, | |
| "loss": 0.7907, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.092, | |
| "grad_norm": 0.6482039093971252, | |
| "learning_rate": 8.00658545649203e-06, | |
| "loss": 0.9164, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.6448190808296204, | |
| "learning_rate": 7.950261121685642e-06, | |
| "loss": 0.8928, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.116, | |
| "grad_norm": 0.5866568684577942, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.8455, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 0.5920262932777405, | |
| "learning_rate": 7.835882727074779e-06, | |
| "loss": 0.7902, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 0.6534630656242371, | |
| "learning_rate": 7.777851165098012e-06, | |
| "loss": 0.917, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6771339774131775, | |
| "learning_rate": 7.719273209572745e-06, | |
| "loss": 0.94, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.164, | |
| "grad_norm": 0.6356242299079895, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.7936, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 0.5693090558052063, | |
| "learning_rate": 7.600524311394873e-06, | |
| "loss": 0.8666, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.188, | |
| "grad_norm": 0.7795962691307068, | |
| "learning_rate": 7.540376726232648e-06, | |
| "loss": 0.8397, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.6267092227935791, | |
| "learning_rate": 7.4797294579083405e-06, | |
| "loss": 0.8552, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.212, | |
| "grad_norm": 0.657116174697876, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.8907, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 0.5222663879394531, | |
| "learning_rate": 7.3569836841299905e-06, | |
| "loss": 0.8099, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.236, | |
| "grad_norm": 0.5663891434669495, | |
| "learning_rate": 7.294909322337689e-06, | |
| "loss": 0.7832, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.5922184586524963, | |
| "learning_rate": 7.232383559957815e-06, | |
| "loss": 0.9434, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.5749194025993347, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.8623, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 0.6411036252975464, | |
| "learning_rate": 7.106027114194856e-06, | |
| "loss": 0.8438, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.284, | |
| "grad_norm": 0.6471683979034424, | |
| "learning_rate": 7.042221284679982e-06, | |
| "loss": 0.8188, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.6250800490379333, | |
| "learning_rate": 6.978013757425295e-06, | |
| "loss": 0.9154, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.308, | |
| "grad_norm": 0.6909330487251282, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.8314, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.5431961417198181, | |
| "learning_rate": 6.848444203803476e-06, | |
| "loss": 0.823, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.332, | |
| "grad_norm": 0.6324174404144287, | |
| "learning_rate": 6.783107663311566e-06, | |
| "loss": 0.8475, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5820900201797485, | |
| "learning_rate": 6.717420391817306e-06, | |
| "loss": 0.8092, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.3559999999999999, | |
| "grad_norm": 0.5760152339935303, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.8584, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 0.6144810914993286, | |
| "learning_rate": 6.585045404088442e-06, | |
| "loss": 0.7915, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.5397148728370667, | |
| "learning_rate": 6.518383725548074e-06, | |
| "loss": 0.9026, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5049560070037842, | |
| "learning_rate": 6.451423386272312e-06, | |
| "loss": 0.84, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.404, | |
| "grad_norm": 0.51853346824646, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.7698, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 0.5880471467971802, | |
| "learning_rate": 6.3166594651218235e-06, | |
| "loss": 0.9143, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.428, | |
| "grad_norm": 0.612112283706665, | |
| "learning_rate": 6.248882390836135e-06, | |
| "loss": 0.8552, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.46327412128448486, | |
| "learning_rate": 6.180859665779173e-06, | |
| "loss": 0.808, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.452, | |
| "grad_norm": 0.6470435857772827, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.9172, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 0.526380181312561, | |
| "learning_rate": 6.04413082836085e-06, | |
| "loss": 0.8035, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.476, | |
| "grad_norm": 0.508842945098877, | |
| "learning_rate": 5.975451610080643e-06, | |
| "loss": 0.7913, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.5358572602272034, | |
| "learning_rate": 5.906580523901493e-06, | |
| "loss": 0.8627, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.573665201663971, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.8117, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 0.55750972032547, | |
| "learning_rate": 5.768316969722651e-06, | |
| "loss": 0.7876, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.524, | |
| "grad_norm": 0.4913860261440277, | |
| "learning_rate": 5.698951697677498e-06, | |
| "loss": 0.7882, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.682537853717804, | |
| "learning_rate": 5.629448944293128e-06, | |
| "loss": 0.8077, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.548, | |
| "grad_norm": 0.6094300150871277, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.9296, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.47651195526123047, | |
| "learning_rate": 5.490085701647805e-06, | |
| "loss": 0.8909, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.572, | |
| "grad_norm": 0.5251836180686951, | |
| "learning_rate": 5.420252624646238e-06, | |
| "loss": 0.8243, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.6414013504981995, | |
| "learning_rate": 5.350336885432337e-06, | |
| "loss": 0.9343, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.596, | |
| "grad_norm": 0.5134192109107971, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.8446, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 0.4695882499217987, | |
| "learning_rate": 5.210312442641327e-06, | |
| "loss": 0.825, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.5449735522270203, | |
| "learning_rate": 5.140231281379345e-06, | |
| "loss": 0.906, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.4647942781448364, | |
| "learning_rate": 5.070122537117812e-06, | |
| "loss": 0.7803, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.6440000000000001, | |
| "grad_norm": 0.49307671189308167, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8701, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 0.45072510838508606, | |
| "learning_rate": 4.92987746288219e-06, | |
| "loss": 0.7693, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.6680000000000001, | |
| "grad_norm": 0.6335264444351196, | |
| "learning_rate": 4.859768718620656e-06, | |
| "loss": 0.9239, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.5734366178512573, | |
| "learning_rate": 4.789687557358676e-06, | |
| "loss": 0.8102, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.692, | |
| "grad_norm": 0.5393184423446655, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.8653, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 0.6001816391944885, | |
| "learning_rate": 4.649663114567663e-06, | |
| "loss": 0.9439, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.716, | |
| "grad_norm": 0.5139857530593872, | |
| "learning_rate": 4.579747375353763e-06, | |
| "loss": 0.8445, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4925212264060974, | |
| "learning_rate": 4.509914298352197e-06, | |
| "loss": 0.7356, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.5880061388015747, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.9677, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 0.6019755601882935, | |
| "learning_rate": 4.3705510557068746e-06, | |
| "loss": 0.8576, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.764, | |
| "grad_norm": 0.47299277782440186, | |
| "learning_rate": 4.3010483023225045e-06, | |
| "loss": 0.8079, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.48285701870918274, | |
| "learning_rate": 4.231683030277349e-06, | |
| "loss": 0.8219, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.788, | |
| "grad_norm": 0.5679399967193604, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.8778, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.5238530039787292, | |
| "learning_rate": 4.0934194760985095e-06, | |
| "loss": 0.8278, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.812, | |
| "grad_norm": 0.5001329779624939, | |
| "learning_rate": 4.02454838991936e-06, | |
| "loss": 0.8468, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.4934837520122528, | |
| "learning_rate": 3.955869171639151e-06, | |
| "loss": 0.8676, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.8359999999999999, | |
| "grad_norm": 0.4868137836456299, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.6974, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 0.5343292951583862, | |
| "learning_rate": 3.81914033422083e-06, | |
| "loss": 0.9395, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.8599999999999999, | |
| "grad_norm": 0.4892309308052063, | |
| "learning_rate": 3.751117609163865e-06, | |
| "loss": 0.7526, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.5625379085540771, | |
| "learning_rate": 3.683340534878176e-06, | |
| "loss": 0.8781, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.884, | |
| "grad_norm": 0.5415952801704407, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.8064, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 0.5701708197593689, | |
| "learning_rate": 3.5485766137276894e-06, | |
| "loss": 0.8821, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.908, | |
| "grad_norm": 0.5524731874465942, | |
| "learning_rate": 3.4816162744519266e-06, | |
| "loss": 0.8881, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.48896679282188416, | |
| "learning_rate": 3.4149545959115604e-06, | |
| "loss": 0.8135, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.932, | |
| "grad_norm": 0.4823067784309387, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.7966, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 0.5454317927360535, | |
| "learning_rate": 3.2825796081826943e-06, | |
| "loss": 0.8471, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.956, | |
| "grad_norm": 0.5044663548469543, | |
| "learning_rate": 3.216892336688435e-06, | |
| "loss": 0.8319, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.500585675239563, | |
| "learning_rate": 3.1515557961965254e-06, | |
| "loss": 0.9108, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.4640965163707733, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.8201, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 0.5118164420127869, | |
| "learning_rate": 3.021986242574707e-06, | |
| "loss": 0.8368, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.004, | |
| "grad_norm": 1.088891625404358, | |
| "learning_rate": 2.95777871532002e-06, | |
| "loss": 1.361, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.47330203652381897, | |
| "learning_rate": 2.893972885805148e-06, | |
| "loss": 0.7807, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.028, | |
| "grad_norm": 0.5352301597595215, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.7914, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.580504298210144, | |
| "learning_rate": 2.7676164400421864e-06, | |
| "loss": 0.8962, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.052, | |
| "grad_norm": 0.5661837458610535, | |
| "learning_rate": 2.705090677662311e-06, | |
| "loss": 0.7885, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.45867210626602173, | |
| "learning_rate": 2.6430163158700116e-06, | |
| "loss": 0.6953, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.076, | |
| "grad_norm": 0.5421217679977417, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.8564, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.088, | |
| "grad_norm": 0.5122601389884949, | |
| "learning_rate": 2.520270542091663e-06, | |
| "loss": 0.7943, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.5720393657684326, | |
| "learning_rate": 2.4596232737673544e-06, | |
| "loss": 0.7948, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5577499270439148, | |
| "learning_rate": 2.3994756886051267e-06, | |
| "loss": 0.8502, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.124, | |
| "grad_norm": 0.5329925417900085, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.7862, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.136, | |
| "grad_norm": 0.5803031921386719, | |
| "learning_rate": 2.280726790427258e-06, | |
| "loss": 0.8887, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.148, | |
| "grad_norm": 0.45233362913131714, | |
| "learning_rate": 2.2221488349019903e-06, | |
| "loss": 0.813, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.4876217842102051, | |
| "learning_rate": 2.1641172729252206e-06, | |
| "loss": 0.7921, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.172, | |
| "grad_norm": 0.5214042663574219, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.8221, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.184, | |
| "grad_norm": 0.5316967368125916, | |
| "learning_rate": 2.04973887831436e-06, | |
| "loss": 0.7736, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.196, | |
| "grad_norm": 0.4661819338798523, | |
| "learning_rate": 1.9934145435079705e-06, | |
| "loss": 0.8011, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.46781396865844727, | |
| "learning_rate": 1.9376815934802496e-06, | |
| "loss": 0.8939, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.47688379883766174, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.7347, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.232, | |
| "grad_norm": 0.4810101389884949, | |
| "learning_rate": 1.8280335791817733e-06, | |
| "loss": 0.8875, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.2439999999999998, | |
| "grad_norm": 0.4670160114765167, | |
| "learning_rate": 1.7741400822895633e-06, | |
| "loss": 0.7385, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.4591081738471985, | |
| "learning_rate": 1.7208811006908798e-06, | |
| "loss": 0.8053, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.268, | |
| "grad_norm": 0.44285574555397034, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.8803, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 0.4147845208644867, | |
| "learning_rate": 1.6163084599236278e-06, | |
| "loss": 0.7583, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.292, | |
| "grad_norm": 0.4563939869403839, | |
| "learning_rate": 1.5650153698254916e-06, | |
| "loss": 0.8772, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.46312814950942993, | |
| "learning_rate": 1.514397929115884e-06, | |
| "loss": 0.72, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.316, | |
| "grad_norm": 0.5045288801193237, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.8059, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.328, | |
| "grad_norm": 0.4324779510498047, | |
| "learning_rate": 1.4152296860958641e-06, | |
| "loss": 0.7998, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.4602769613265991, | |
| "learning_rate": 1.3666983898298659e-06, | |
| "loss": 0.7974, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.40926551818847656, | |
| "learning_rate": 1.3188817512044544e-06, | |
| "loss": 0.7687, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.364, | |
| "grad_norm": 0.4703604280948639, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.9035, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.376, | |
| "grad_norm": 0.4227948784828186, | |
| "learning_rate": 1.225429925912921e-06, | |
| "loss": 0.7542, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.388, | |
| "grad_norm": 0.4166393280029297, | |
| "learning_rate": 1.1798131208919628e-06, | |
| "loss": 0.8088, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.38077840209007263, | |
| "learning_rate": 1.134947733186315e-06, | |
| "loss": 0.6584, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.412, | |
| "grad_norm": 0.42249420285224915, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.8569, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.424, | |
| "grad_norm": 0.49399280548095703, | |
| "learning_rate": 1.047506359639483e-06, | |
| "loss": 0.8836, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.436, | |
| "grad_norm": 0.4305580258369446, | |
| "learning_rate": 1.004947573208756e-06, | |
| "loss": 0.7372, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.49399375915527344, | |
| "learning_rate": 9.631745995311881e-07, | |
| "loss": 0.7824, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.3984535336494446, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.7465, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.472, | |
| "grad_norm": 0.36112886667251587, | |
| "learning_rate": 8.820188006404268e-07, | |
| "loss": 0.694, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.484, | |
| "grad_norm": 0.45247265696525574, | |
| "learning_rate": 8.426519384872733e-07, | |
| "loss": 0.8708, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.38195085525512695, | |
| "learning_rate": 8.041028120674894e-07, | |
| "loss": 0.7514, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.508, | |
| "grad_norm": 0.4380353093147278, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.8477, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.42108631134033203, | |
| "learning_rate": 7.294879340008632e-07, | |
| "loss": 0.8247, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.532, | |
| "grad_norm": 0.4263460636138916, | |
| "learning_rate": 6.934368588379553e-07, | |
| "loss": 0.8271, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.48352324962615967, | |
| "learning_rate": 6.582328694894729e-07, | |
| "loss": 0.8894, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.556, | |
| "grad_norm": 0.37158602476119995, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.7195, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.568, | |
| "grad_norm": 0.3880118131637573, | |
| "learning_rate": 5.903936782582253e-07, | |
| "loss": 0.7907, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.40234553813934326, | |
| "learning_rate": 5.577718201056392e-07, | |
| "loss": 0.8158, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.42842331528663635, | |
| "learning_rate": 5.260237326031698e-07, | |
| "loss": 0.7774, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.604, | |
| "grad_norm": 0.4132654070854187, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.7719, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.616, | |
| "grad_norm": 0.4403928518295288, | |
| "learning_rate": 4.651736754009972e-07, | |
| "loss": 0.7448, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.628, | |
| "grad_norm": 0.4745451509952545, | |
| "learning_rate": 4.3608367469340553e-07, | |
| "loss": 0.9587, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3818591833114624, | |
| "learning_rate": 4.078913802658946e-07, | |
| "loss": 0.7279, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.652, | |
| "grad_norm": 0.4163016974925995, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.7954, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.664, | |
| "grad_norm": 0.37762609124183655, | |
| "learning_rate": 3.542219138851094e-07, | |
| "loss": 0.8308, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.676, | |
| "grad_norm": 0.4249165654182434, | |
| "learning_rate": 3.287552985270015e-07, | |
| "loss": 0.7808, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.43331316113471985, | |
| "learning_rate": 3.0420750056286195e-07, | |
| "loss": 0.8398, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.38995715975761414, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.707, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.7119999999999997, | |
| "grad_norm": 0.43745648860931396, | |
| "learning_rate": 2.5788748900048676e-07, | |
| "loss": 0.8001, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.724, | |
| "grad_norm": 0.3968169093132019, | |
| "learning_rate": 2.3612438638551837e-07, | |
| "loss": 0.8116, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.43742409348487854, | |
| "learning_rate": 2.152983213389559e-07, | |
| "loss": 0.776, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.748, | |
| "grad_norm": 0.39760884642601013, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.8247, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.40744730830192566, | |
| "learning_rate": 1.7647350448812105e-07, | |
| "loss": 0.8348, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.7720000000000002, | |
| "grad_norm": 0.36414623260498047, | |
| "learning_rate": 1.5848238938869332e-07, | |
| "loss": 0.7256, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.42025619745254517, | |
| "learning_rate": 1.4144358376524504e-07, | |
| "loss": 0.8586, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.7960000000000003, | |
| "grad_norm": 0.4019131362438202, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.7834, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.808, | |
| "grad_norm": 0.43940842151641846, | |
| "learning_rate": 1.10236118863562e-07, | |
| "loss": 0.8294, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.4242531657218933, | |
| "learning_rate": 9.607359798384785e-08, | |
| "loss": 0.7621, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.4360555112361908, | |
| "learning_rate": 8.287566216975795e-08, | |
| "loss": 0.8652, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.844, | |
| "grad_norm": 0.38368791341781616, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.7203, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.856, | |
| "grad_norm": 0.4965880215167999, | |
| "learning_rate": 5.938373944745612e-08, | |
| "loss": 0.8733, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.868, | |
| "grad_norm": 0.43186068534851074, | |
| "learning_rate": 4.909437331777178e-08, | |
| "loss": 0.7493, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3746281862258911, | |
| "learning_rate": 3.977883290187667e-08, | |
| "loss": 0.7482, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.892, | |
| "grad_norm": 0.42593878507614136, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.8974, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.904, | |
| "grad_norm": 0.3988572657108307, | |
| "learning_rate": 2.4076366639015914e-08, | |
| "loss": 0.7592, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.916, | |
| "grad_norm": 0.39131683111190796, | |
| "learning_rate": 1.769252941190458e-08, | |
| "loss": 0.7935, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.39934051036834717, | |
| "learning_rate": 1.2288694530769862e-08, | |
| "loss": 0.78, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.44333237409591675, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.8765, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.952, | |
| "grad_norm": 0.37822088599205017, | |
| "learning_rate": 4.4250904955656095e-09, | |
| "loss": 0.807, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.964, | |
| "grad_norm": 0.41814127564430237, | |
| "learning_rate": 1.9668680847356735e-09, | |
| "loss": 0.8633, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.36398470401763916, | |
| "learning_rate": 4.91741202124918e-10, | |
| "loss": 0.7087, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.988, | |
| "grad_norm": 0.47774454951286316, | |
| "learning_rate": 0.0, | |
| "loss": 0.9268, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.988, | |
| "step": 249, | |
| "total_flos": 248634998849536.0, | |
| "train_loss": 0.883572997817074, | |
| "train_runtime": 14151.8701, | |
| "train_samples_per_second": 1.696, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 249, | |
| "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": 248634998849536.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |