| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "eval_steps": 500, | |
| "global_step": 113400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0368574857711792, | |
| "learning_rate": 9.900440917107584e-06, | |
| "loss": 2.4497, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.962113380432129, | |
| "learning_rate": 9.800440917107585e-06, | |
| "loss": 2.3441, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.04465913772583, | |
| "learning_rate": 9.700529100529101e-06, | |
| "loss": 2.2934, | |
| "step": 3402 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.710998773574829, | |
| "learning_rate": 9.6005291005291e-06, | |
| "loss": 2.2626, | |
| "step": 4536 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.217125177383423, | |
| "learning_rate": 9.500529100529102e-06, | |
| "loss": 2.2311, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.2306671142578125, | |
| "learning_rate": 9.400617283950619e-06, | |
| "loss": 2.1809, | |
| "step": 6804 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.223266124725342, | |
| "learning_rate": 9.300617283950618e-06, | |
| "loss": 2.1686, | |
| "step": 7938 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.7807767391204834, | |
| "learning_rate": 9.200705467372135e-06, | |
| "loss": 2.139, | |
| "step": 9072 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 6.859907150268555, | |
| "learning_rate": 9.100705467372136e-06, | |
| "loss": 2.1258, | |
| "step": 10206 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.691579818725586, | |
| "learning_rate": 9.000793650793651e-06, | |
| "loss": 2.1096, | |
| "step": 11340 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 4.632028579711914, | |
| "learning_rate": 8.90079365079365e-06, | |
| "loss": 2.1051, | |
| "step": 12474 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 5.086824417114258, | |
| "learning_rate": 8.800881834215167e-06, | |
| "loss": 2.0945, | |
| "step": 13608 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 3.4331443309783936, | |
| "learning_rate": 8.700881834215168e-06, | |
| "loss": 2.0912, | |
| "step": 14742 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.4025092124938965, | |
| "learning_rate": 8.600881834215168e-06, | |
| "loss": 2.0722, | |
| "step": 15876 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 10.316903114318848, | |
| "learning_rate": 8.500881834215169e-06, | |
| "loss": 2.0726, | |
| "step": 17010 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 5.012073040008545, | |
| "learning_rate": 8.400970017636686e-06, | |
| "loss": 2.0657, | |
| "step": 18144 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 6.108023166656494, | |
| "learning_rate": 8.301058201058203e-06, | |
| "loss": 2.0583, | |
| "step": 19278 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 3.4143471717834473, | |
| "learning_rate": 8.201058201058202e-06, | |
| "loss": 2.038, | |
| "step": 20412 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 4.766075134277344, | |
| "learning_rate": 8.101146384479719e-06, | |
| "loss": 2.0292, | |
| "step": 21546 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 2.7671144008636475, | |
| "learning_rate": 8.001146384479718e-06, | |
| "loss": 2.0344, | |
| "step": 22680 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 3.447026491165161, | |
| "learning_rate": 7.901234567901235e-06, | |
| "loss": 2.0324, | |
| "step": 23814 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 3.207721471786499, | |
| "learning_rate": 7.801234567901234e-06, | |
| "loss": 2.0119, | |
| "step": 24948 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 2.789856433868408, | |
| "learning_rate": 7.701234567901235e-06, | |
| "loss": 2.0157, | |
| "step": 26082 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 2.2711899280548096, | |
| "learning_rate": 7.601234567901235e-06, | |
| "loss": 2.0048, | |
| "step": 27216 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 1.5879839658737183, | |
| "learning_rate": 7.5013227513227514e-06, | |
| "loss": 2.0111, | |
| "step": 28350 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 4.620934963226318, | |
| "learning_rate": 7.401410934744268e-06, | |
| "loss": 1.9837, | |
| "step": 29484 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 1.8832050561904907, | |
| "learning_rate": 7.3014109347442685e-06, | |
| "loss": 1.9818, | |
| "step": 30618 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 3.1029512882232666, | |
| "learning_rate": 7.2014991181657855e-06, | |
| "loss": 1.9868, | |
| "step": 31752 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 1.3787024021148682, | |
| "learning_rate": 7.101499118165786e-06, | |
| "loss": 1.988, | |
| "step": 32886 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 4.151069164276123, | |
| "learning_rate": 7.001499118165786e-06, | |
| "loss": 1.975, | |
| "step": 34020 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 3.1992034912109375, | |
| "learning_rate": 6.901587301587303e-06, | |
| "loss": 1.9814, | |
| "step": 35154 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 5.592874526977539, | |
| "learning_rate": 6.801675485008819e-06, | |
| "loss": 1.9739, | |
| "step": 36288 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 1.8030545711517334, | |
| "learning_rate": 6.701675485008818e-06, | |
| "loss": 1.9758, | |
| "step": 37422 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 3.9878218173980713, | |
| "learning_rate": 6.601675485008819e-06, | |
| "loss": 1.9726, | |
| "step": 38556 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 3.259308099746704, | |
| "learning_rate": 6.501763668430335e-06, | |
| "loss": 1.9649, | |
| "step": 39690 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "grad_norm": 3.823164224624634, | |
| "learning_rate": 6.401763668430335e-06, | |
| "loss": 1.9547, | |
| "step": 40824 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "grad_norm": 4.652733325958252, | |
| "learning_rate": 6.301763668430335e-06, | |
| "loss": 1.9673, | |
| "step": 41958 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "grad_norm": 3.522451162338257, | |
| "learning_rate": 6.201851851851852e-06, | |
| "loss": 1.9544, | |
| "step": 43092 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "grad_norm": 4.213411331176758, | |
| "learning_rate": 6.1018518518518525e-06, | |
| "loss": 1.9601, | |
| "step": 44226 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 5.826420783996582, | |
| "learning_rate": 6.0019400352733694e-06, | |
| "loss": 1.9553, | |
| "step": 45360 | |
| }, | |
| { | |
| "epoch": 41.0, | |
| "grad_norm": 3.5787088871002197, | |
| "learning_rate": 5.902028218694886e-06, | |
| "loss": 1.9603, | |
| "step": 46494 | |
| }, | |
| { | |
| "epoch": 42.0, | |
| "grad_norm": 2.380449056625366, | |
| "learning_rate": 5.8020282186948865e-06, | |
| "loss": 1.9551, | |
| "step": 47628 | |
| }, | |
| { | |
| "epoch": 43.0, | |
| "grad_norm": 5.958215236663818, | |
| "learning_rate": 5.702028218694886e-06, | |
| "loss": 1.9515, | |
| "step": 48762 | |
| }, | |
| { | |
| "epoch": 44.0, | |
| "grad_norm": 1.0750882625579834, | |
| "learning_rate": 5.602116402116402e-06, | |
| "loss": 1.9578, | |
| "step": 49896 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 2.135838031768799, | |
| "learning_rate": 5.502116402116403e-06, | |
| "loss": 1.9412, | |
| "step": 51030 | |
| }, | |
| { | |
| "epoch": 46.0, | |
| "grad_norm": 2.9327504634857178, | |
| "learning_rate": 5.402204585537919e-06, | |
| "loss": 1.948, | |
| "step": 52164 | |
| }, | |
| { | |
| "epoch": 47.0, | |
| "grad_norm": 5.212012767791748, | |
| "learning_rate": 5.302204585537919e-06, | |
| "loss": 1.9581, | |
| "step": 53298 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 3.5629754066467285, | |
| "learning_rate": 5.202292768959436e-06, | |
| "loss": 1.9394, | |
| "step": 54432 | |
| }, | |
| { | |
| "epoch": 49.0, | |
| "grad_norm": 2.700446844100952, | |
| "learning_rate": 5.102292768959436e-06, | |
| "loss": 1.943, | |
| "step": 55566 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 5.429810523986816, | |
| "learning_rate": 5.002380952380953e-06, | |
| "loss": 1.9499, | |
| "step": 56700 | |
| }, | |
| { | |
| "epoch": 51.0, | |
| "grad_norm": 2.2767255306243896, | |
| "learning_rate": 4.902469135802469e-06, | |
| "loss": 1.933, | |
| "step": 57834 | |
| }, | |
| { | |
| "epoch": 52.0, | |
| "grad_norm": 4.359825611114502, | |
| "learning_rate": 4.802469135802469e-06, | |
| "loss": 1.9319, | |
| "step": 58968 | |
| }, | |
| { | |
| "epoch": 53.0, | |
| "grad_norm": 3.3374125957489014, | |
| "learning_rate": 4.702557319223986e-06, | |
| "loss": 1.9303, | |
| "step": 60102 | |
| }, | |
| { | |
| "epoch": 54.0, | |
| "grad_norm": 1.666292428970337, | |
| "learning_rate": 4.6025573192239865e-06, | |
| "loss": 1.9391, | |
| "step": 61236 | |
| }, | |
| { | |
| "epoch": 55.0, | |
| "grad_norm": 2.920473575592041, | |
| "learning_rate": 4.502645502645503e-06, | |
| "loss": 1.9339, | |
| "step": 62370 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "grad_norm": 3.5226597785949707, | |
| "learning_rate": 4.402645502645503e-06, | |
| "loss": 1.924, | |
| "step": 63504 | |
| }, | |
| { | |
| "epoch": 57.0, | |
| "grad_norm": 1.056983232498169, | |
| "learning_rate": 4.30273368606702e-06, | |
| "loss": 1.939, | |
| "step": 64638 | |
| }, | |
| { | |
| "epoch": 58.0, | |
| "grad_norm": 1.7839380502700806, | |
| "learning_rate": 4.20273368606702e-06, | |
| "loss": 1.9302, | |
| "step": 65772 | |
| }, | |
| { | |
| "epoch": 59.0, | |
| "grad_norm": 2.453362464904785, | |
| "learning_rate": 4.102821869488536e-06, | |
| "loss": 1.932, | |
| "step": 66906 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 1.79325532913208, | |
| "learning_rate": 4.002910052910054e-06, | |
| "loss": 1.9347, | |
| "step": 68040 | |
| }, | |
| { | |
| "epoch": 61.0, | |
| "grad_norm": 4.606123924255371, | |
| "learning_rate": 3.902910052910053e-06, | |
| "loss": 1.9163, | |
| "step": 69174 | |
| }, | |
| { | |
| "epoch": 62.0, | |
| "grad_norm": 3.7879464626312256, | |
| "learning_rate": 3.80299823633157e-06, | |
| "loss": 1.9314, | |
| "step": 70308 | |
| }, | |
| { | |
| "epoch": 63.0, | |
| "grad_norm": 3.875440835952759, | |
| "learning_rate": 3.7029982363315697e-06, | |
| "loss": 1.9364, | |
| "step": 71442 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "grad_norm": 3.8616816997528076, | |
| "learning_rate": 3.6030864197530867e-06, | |
| "loss": 1.9183, | |
| "step": 72576 | |
| }, | |
| { | |
| "epoch": 65.0, | |
| "grad_norm": 3.232619285583496, | |
| "learning_rate": 3.5030864197530868e-06, | |
| "loss": 1.9164, | |
| "step": 73710 | |
| }, | |
| { | |
| "epoch": 66.0, | |
| "grad_norm": 3.7224302291870117, | |
| "learning_rate": 3.4031746031746033e-06, | |
| "loss": 1.9225, | |
| "step": 74844 | |
| }, | |
| { | |
| "epoch": 67.0, | |
| "grad_norm": 2.774463653564453, | |
| "learning_rate": 3.303174603174604e-06, | |
| "loss": 1.9328, | |
| "step": 75978 | |
| }, | |
| { | |
| "epoch": 68.0, | |
| "grad_norm": 1.1616075038909912, | |
| "learning_rate": 3.2032627865961204e-06, | |
| "loss": 1.914, | |
| "step": 77112 | |
| }, | |
| { | |
| "epoch": 69.0, | |
| "grad_norm": 1.613097071647644, | |
| "learning_rate": 3.103350970017637e-06, | |
| "loss": 1.9147, | |
| "step": 78246 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 4.327018737792969, | |
| "learning_rate": 3.003350970017637e-06, | |
| "loss": 1.924, | |
| "step": 79380 | |
| }, | |
| { | |
| "epoch": 71.0, | |
| "grad_norm": 1.2719597816467285, | |
| "learning_rate": 2.9034391534391536e-06, | |
| "loss": 1.938, | |
| "step": 80514 | |
| }, | |
| { | |
| "epoch": 72.0, | |
| "grad_norm": 2.238760232925415, | |
| "learning_rate": 2.8034391534391537e-06, | |
| "loss": 1.9153, | |
| "step": 81648 | |
| }, | |
| { | |
| "epoch": 73.0, | |
| "grad_norm": 1.070393443107605, | |
| "learning_rate": 2.7035273368606703e-06, | |
| "loss": 1.9171, | |
| "step": 82782 | |
| }, | |
| { | |
| "epoch": 74.0, | |
| "grad_norm": 3.3792688846588135, | |
| "learning_rate": 2.603527336860671e-06, | |
| "loss": 1.9204, | |
| "step": 83916 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "grad_norm": 1.406660795211792, | |
| "learning_rate": 2.503615520282187e-06, | |
| "loss": 1.9139, | |
| "step": 85050 | |
| }, | |
| { | |
| "epoch": 76.0, | |
| "grad_norm": 1.7133662700653076, | |
| "learning_rate": 2.403615520282187e-06, | |
| "loss": 1.9238, | |
| "step": 86184 | |
| }, | |
| { | |
| "epoch": 77.0, | |
| "grad_norm": 1.5862336158752441, | |
| "learning_rate": 2.303703703703704e-06, | |
| "loss": 1.9193, | |
| "step": 87318 | |
| }, | |
| { | |
| "epoch": 78.0, | |
| "grad_norm": 2.752800226211548, | |
| "learning_rate": 2.2037918871252206e-06, | |
| "loss": 1.9147, | |
| "step": 88452 | |
| }, | |
| { | |
| "epoch": 79.0, | |
| "grad_norm": 1.8123775720596313, | |
| "learning_rate": 2.1037918871252207e-06, | |
| "loss": 1.9007, | |
| "step": 89586 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 3.6397151947021484, | |
| "learning_rate": 2.0038800705467372e-06, | |
| "loss": 1.9123, | |
| "step": 90720 | |
| }, | |
| { | |
| "epoch": 81.0, | |
| "grad_norm": 2.8387656211853027, | |
| "learning_rate": 1.9038800705467373e-06, | |
| "loss": 1.9165, | |
| "step": 91854 | |
| }, | |
| { | |
| "epoch": 82.0, | |
| "grad_norm": 0.9121940732002258, | |
| "learning_rate": 1.803968253968254e-06, | |
| "loss": 1.9179, | |
| "step": 92988 | |
| }, | |
| { | |
| "epoch": 83.0, | |
| "grad_norm": 3.447791576385498, | |
| "learning_rate": 1.7039682539682542e-06, | |
| "loss": 1.9049, | |
| "step": 94122 | |
| }, | |
| { | |
| "epoch": 84.0, | |
| "grad_norm": 2.4387011528015137, | |
| "learning_rate": 1.6040564373897708e-06, | |
| "loss": 1.9186, | |
| "step": 95256 | |
| }, | |
| { | |
| "epoch": 85.0, | |
| "grad_norm": 4.287496566772461, | |
| "learning_rate": 1.5040564373897709e-06, | |
| "loss": 1.9137, | |
| "step": 96390 | |
| }, | |
| { | |
| "epoch": 86.0, | |
| "grad_norm": 1.9334492683410645, | |
| "learning_rate": 1.4041446208112876e-06, | |
| "loss": 1.9245, | |
| "step": 97524 | |
| }, | |
| { | |
| "epoch": 87.0, | |
| "grad_norm": 3.3120267391204834, | |
| "learning_rate": 1.3042328042328042e-06, | |
| "loss": 1.9022, | |
| "step": 98658 | |
| }, | |
| { | |
| "epoch": 88.0, | |
| "grad_norm": 3.476644992828369, | |
| "learning_rate": 1.2042328042328043e-06, | |
| "loss": 1.9157, | |
| "step": 99792 | |
| }, | |
| { | |
| "epoch": 89.0, | |
| "grad_norm": 1.7216417789459229, | |
| "learning_rate": 1.104320987654321e-06, | |
| "loss": 1.9212, | |
| "step": 100926 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 1.2965023517608643, | |
| "learning_rate": 1.0043209876543212e-06, | |
| "loss": 1.9098, | |
| "step": 102060 | |
| }, | |
| { | |
| "epoch": 91.0, | |
| "grad_norm": 2.355569362640381, | |
| "learning_rate": 9.044091710758378e-07, | |
| "loss": 1.9128, | |
| "step": 103194 | |
| }, | |
| { | |
| "epoch": 92.0, | |
| "grad_norm": 1.394773006439209, | |
| "learning_rate": 8.044091710758378e-07, | |
| "loss": 1.9084, | |
| "step": 104328 | |
| }, | |
| { | |
| "epoch": 93.0, | |
| "grad_norm": 2.99224591255188, | |
| "learning_rate": 7.044973544973546e-07, | |
| "loss": 1.902, | |
| "step": 105462 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "grad_norm": 1.2547388076782227, | |
| "learning_rate": 6.044973544973545e-07, | |
| "loss": 1.9185, | |
| "step": 106596 | |
| }, | |
| { | |
| "epoch": 95.0, | |
| "grad_norm": 0.7210695147514343, | |
| "learning_rate": 5.045855379188714e-07, | |
| "loss": 1.9131, | |
| "step": 107730 | |
| }, | |
| { | |
| "epoch": 96.0, | |
| "grad_norm": 0.7888159155845642, | |
| "learning_rate": 4.0467372134038806e-07, | |
| "loss": 1.9087, | |
| "step": 108864 | |
| }, | |
| { | |
| "epoch": 97.0, | |
| "grad_norm": 0.9536716341972351, | |
| "learning_rate": 3.04673721340388e-07, | |
| "loss": 1.9029, | |
| "step": 109998 | |
| }, | |
| { | |
| "epoch": 98.0, | |
| "grad_norm": 2.2886030673980713, | |
| "learning_rate": 2.047619047619048e-07, | |
| "loss": 1.9094, | |
| "step": 111132 | |
| }, | |
| { | |
| "epoch": 99.0, | |
| "grad_norm": 2.0867316722869873, | |
| "learning_rate": 1.0476190476190476e-07, | |
| "loss": 1.9115, | |
| "step": 112266 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 0.59788578748703, | |
| "learning_rate": 4.8500881834215175e-09, | |
| "loss": 1.8996, | |
| "step": 113400 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 113400, | |
| "total_flos": 3.571154905966387e+18, | |
| "train_loss": 1.9808363689442792, | |
| "train_runtime": 84553.2908, | |
| "train_samples_per_second": 160.897, | |
| "train_steps_per_second": 1.341 | |
| } | |
| ], | |
| "logging_steps": 1134, | |
| "max_steps": 113400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 530, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.571154905966387e+18, | |
| "train_batch_size": 60, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |