| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7036059806508356, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003518029903254178, | |
| "grad_norm": 1.260473608970642, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9049, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007036059806508356, | |
| "grad_norm": 1.2573540210723877, | |
| "learning_rate": 9.949748743718594e-05, | |
| "loss": 2.9229, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010554089709762533, | |
| "grad_norm": 1.1971232891082764, | |
| "learning_rate": 9.899497487437186e-05, | |
| "loss": 2.7224, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014072119613016711, | |
| "grad_norm": 1.3032786846160889, | |
| "learning_rate": 9.84924623115578e-05, | |
| "loss": 2.5683, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01759014951627089, | |
| "grad_norm": 1.358597755432129, | |
| "learning_rate": 9.798994974874372e-05, | |
| "loss": 2.4848, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021108179419525065, | |
| "grad_norm": 1.439229130744934, | |
| "learning_rate": 9.748743718592965e-05, | |
| "loss": 2.2625, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.024626209322779244, | |
| "grad_norm": 1.313475251197815, | |
| "learning_rate": 9.698492462311559e-05, | |
| "loss": 2.0213, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.028144239226033423, | |
| "grad_norm": 1.3972655534744263, | |
| "learning_rate": 9.64824120603015e-05, | |
| "loss": 2.0339, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0316622691292876, | |
| "grad_norm": 1.2032045125961304, | |
| "learning_rate": 9.597989949748745e-05, | |
| "loss": 1.7561, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03518029903254178, | |
| "grad_norm": 1.1419575214385986, | |
| "learning_rate": 9.547738693467337e-05, | |
| "loss": 1.6908, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03869832893579595, | |
| "grad_norm": 1.0084314346313477, | |
| "learning_rate": 9.49748743718593e-05, | |
| "loss": 1.4594, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04221635883905013, | |
| "grad_norm": 0.8950707912445068, | |
| "learning_rate": 9.447236180904523e-05, | |
| "loss": 1.4036, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04573438874230431, | |
| "grad_norm": 0.735814094543457, | |
| "learning_rate": 9.396984924623115e-05, | |
| "loss": 1.3852, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04925241864555849, | |
| "grad_norm": 0.7151777148246765, | |
| "learning_rate": 9.34673366834171e-05, | |
| "loss": 1.4866, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.052770448548812667, | |
| "grad_norm": 0.6819817423820496, | |
| "learning_rate": 9.296482412060302e-05, | |
| "loss": 1.342, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.056288478452066845, | |
| "grad_norm": 0.6794399619102478, | |
| "learning_rate": 9.246231155778895e-05, | |
| "loss": 1.3458, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05980650835532102, | |
| "grad_norm": 0.7077609896659851, | |
| "learning_rate": 9.195979899497488e-05, | |
| "loss": 1.3624, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0633245382585752, | |
| "grad_norm": 0.6987395286560059, | |
| "learning_rate": 9.14572864321608e-05, | |
| "loss": 1.2826, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06684256816182937, | |
| "grad_norm": 0.7677096724510193, | |
| "learning_rate": 9.095477386934675e-05, | |
| "loss": 1.3643, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07036059806508356, | |
| "grad_norm": 0.7091855406761169, | |
| "learning_rate": 9.045226130653267e-05, | |
| "loss": 1.2746, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07387862796833773, | |
| "grad_norm": 0.7494800090789795, | |
| "learning_rate": 8.99497487437186e-05, | |
| "loss": 1.3333, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0773966578715919, | |
| "grad_norm": 0.7067513465881348, | |
| "learning_rate": 8.944723618090453e-05, | |
| "loss": 1.1819, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08091468777484609, | |
| "grad_norm": 0.7177988290786743, | |
| "learning_rate": 8.894472361809045e-05, | |
| "loss": 1.248, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08443271767810026, | |
| "grad_norm": 0.8215416669845581, | |
| "learning_rate": 8.84422110552764e-05, | |
| "loss": 1.2657, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08795074758135445, | |
| "grad_norm": 0.7781690955162048, | |
| "learning_rate": 8.793969849246232e-05, | |
| "loss": 1.2697, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09146877748460862, | |
| "grad_norm": 0.7476558685302734, | |
| "learning_rate": 8.743718592964825e-05, | |
| "loss": 1.1899, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09498680738786279, | |
| "grad_norm": 0.7552306056022644, | |
| "learning_rate": 8.693467336683418e-05, | |
| "loss": 1.2209, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09850483729111698, | |
| "grad_norm": 0.8574095368385315, | |
| "learning_rate": 8.64321608040201e-05, | |
| "loss": 1.2418, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10202286719437115, | |
| "grad_norm": 0.8529037833213806, | |
| "learning_rate": 8.592964824120603e-05, | |
| "loss": 1.2028, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10554089709762533, | |
| "grad_norm": 0.9519858360290527, | |
| "learning_rate": 8.542713567839196e-05, | |
| "loss": 1.2501, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1090589270008795, | |
| "grad_norm": 0.929736852645874, | |
| "learning_rate": 8.49246231155779e-05, | |
| "loss": 1.162, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11257695690413369, | |
| "grad_norm": 1.0004149675369263, | |
| "learning_rate": 8.442211055276383e-05, | |
| "loss": 1.297, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.11609498680738786, | |
| "grad_norm": 0.9739089608192444, | |
| "learning_rate": 8.391959798994975e-05, | |
| "loss": 1.2161, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11961301671064203, | |
| "grad_norm": 1.010574460029602, | |
| "learning_rate": 8.341708542713568e-05, | |
| "loss": 1.2653, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12313104661389622, | |
| "grad_norm": 1.1182818412780762, | |
| "learning_rate": 8.291457286432161e-05, | |
| "loss": 1.2406, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1266490765171504, | |
| "grad_norm": 1.0420522689819336, | |
| "learning_rate": 8.241206030150754e-05, | |
| "loss": 1.1295, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13016710642040458, | |
| "grad_norm": 0.9873694777488708, | |
| "learning_rate": 8.190954773869348e-05, | |
| "loss": 1.1819, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.13368513632365875, | |
| "grad_norm": 0.9985349774360657, | |
| "learning_rate": 8.14070351758794e-05, | |
| "loss": 1.0707, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13720316622691292, | |
| "grad_norm": 0.9374362826347351, | |
| "learning_rate": 8.090452261306533e-05, | |
| "loss": 1.1942, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.14072119613016712, | |
| "grad_norm": 0.9439392685890198, | |
| "learning_rate": 8.040201005025126e-05, | |
| "loss": 1.249, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1442392260334213, | |
| "grad_norm": 0.7446234822273254, | |
| "learning_rate": 7.989949748743719e-05, | |
| "loss": 1.2198, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.14775725593667546, | |
| "grad_norm": 0.6614280939102173, | |
| "learning_rate": 7.939698492462313e-05, | |
| "loss": 1.147, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.15127528583992964, | |
| "grad_norm": 0.7048504948616028, | |
| "learning_rate": 7.889447236180904e-05, | |
| "loss": 1.2024, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1547933157431838, | |
| "grad_norm": 0.6736462116241455, | |
| "learning_rate": 7.839195979899498e-05, | |
| "loss": 1.1622, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.158311345646438, | |
| "grad_norm": 0.651723325252533, | |
| "learning_rate": 7.788944723618091e-05, | |
| "loss": 1.115, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16182937554969218, | |
| "grad_norm": 0.6364433169364929, | |
| "learning_rate": 7.738693467336684e-05, | |
| "loss": 1.1094, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.16534740545294635, | |
| "grad_norm": 0.6916666626930237, | |
| "learning_rate": 7.688442211055277e-05, | |
| "loss": 1.0492, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16886543535620052, | |
| "grad_norm": 0.6645711064338684, | |
| "learning_rate": 7.638190954773869e-05, | |
| "loss": 1.1187, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1723834652594547, | |
| "grad_norm": 0.6615894436836243, | |
| "learning_rate": 7.587939698492463e-05, | |
| "loss": 1.0937, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.1759014951627089, | |
| "grad_norm": 0.6639358997344971, | |
| "learning_rate": 7.537688442211056e-05, | |
| "loss": 1.1598, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17941952506596306, | |
| "grad_norm": 0.7072437405586243, | |
| "learning_rate": 7.487437185929649e-05, | |
| "loss": 1.0835, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.18293755496921724, | |
| "grad_norm": 0.6130443811416626, | |
| "learning_rate": 7.437185929648241e-05, | |
| "loss": 1.1116, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.1864555848724714, | |
| "grad_norm": 0.6709645986557007, | |
| "learning_rate": 7.386934673366834e-05, | |
| "loss": 1.0814, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18997361477572558, | |
| "grad_norm": 0.724520742893219, | |
| "learning_rate": 7.336683417085427e-05, | |
| "loss": 1.1467, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.19349164467897978, | |
| "grad_norm": 0.7026971578598022, | |
| "learning_rate": 7.28643216080402e-05, | |
| "loss": 1.1671, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.19700967458223395, | |
| "grad_norm": 0.6167840957641602, | |
| "learning_rate": 7.236180904522614e-05, | |
| "loss": 1.0342, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.20052770448548812, | |
| "grad_norm": 0.6177359819412231, | |
| "learning_rate": 7.185929648241206e-05, | |
| "loss": 0.9947, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.2040457343887423, | |
| "grad_norm": 0.6578599810600281, | |
| "learning_rate": 7.135678391959799e-05, | |
| "loss": 1.083, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2075637642919965, | |
| "grad_norm": 0.6553827524185181, | |
| "learning_rate": 7.085427135678392e-05, | |
| "loss": 1.1451, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.21108179419525067, | |
| "grad_norm": 0.6147003769874573, | |
| "learning_rate": 7.035175879396985e-05, | |
| "loss": 1.1026, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21459982409850484, | |
| "grad_norm": 0.6656669974327087, | |
| "learning_rate": 6.984924623115579e-05, | |
| "loss": 1.0638, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.218117854001759, | |
| "grad_norm": 0.6732162237167358, | |
| "learning_rate": 6.93467336683417e-05, | |
| "loss": 1.1639, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.22163588390501318, | |
| "grad_norm": 0.6583305597305298, | |
| "learning_rate": 6.884422110552764e-05, | |
| "loss": 1.0694, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.22515391380826738, | |
| "grad_norm": 0.573226809501648, | |
| "learning_rate": 6.834170854271357e-05, | |
| "loss": 1.0237, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.22867194371152155, | |
| "grad_norm": 0.6894761919975281, | |
| "learning_rate": 6.78391959798995e-05, | |
| "loss": 1.0884, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.23218997361477572, | |
| "grad_norm": 0.6620854735374451, | |
| "learning_rate": 6.733668341708544e-05, | |
| "loss": 1.0515, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2357080035180299, | |
| "grad_norm": 0.695426344871521, | |
| "learning_rate": 6.683417085427135e-05, | |
| "loss": 1.1149, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.23922603342128407, | |
| "grad_norm": 0.6343066096305847, | |
| "learning_rate": 6.633165829145729e-05, | |
| "loss": 1.0736, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.24274406332453827, | |
| "grad_norm": 0.6478216052055359, | |
| "learning_rate": 6.582914572864322e-05, | |
| "loss": 1.0576, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.24626209322779244, | |
| "grad_norm": 0.7264822125434875, | |
| "learning_rate": 6.532663316582915e-05, | |
| "loss": 1.16, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2497801231310466, | |
| "grad_norm": 0.7927188277244568, | |
| "learning_rate": 6.482412060301508e-05, | |
| "loss": 1.0784, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2532981530343008, | |
| "grad_norm": 0.6734123826026917, | |
| "learning_rate": 6.4321608040201e-05, | |
| "loss": 1.1155, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.256816182937555, | |
| "grad_norm": 0.6928442120552063, | |
| "learning_rate": 6.381909547738694e-05, | |
| "loss": 1.1, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.26033421284080915, | |
| "grad_norm": 0.6205620765686035, | |
| "learning_rate": 6.331658291457287e-05, | |
| "loss": 1.0557, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2638522427440633, | |
| "grad_norm": 0.6895455718040466, | |
| "learning_rate": 6.28140703517588e-05, | |
| "loss": 1.0893, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2673702726473175, | |
| "grad_norm": 0.7075064778327942, | |
| "learning_rate": 6.231155778894473e-05, | |
| "loss": 1.0281, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.27088830255057167, | |
| "grad_norm": 0.8777890801429749, | |
| "learning_rate": 6.180904522613065e-05, | |
| "loss": 1.1, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.27440633245382584, | |
| "grad_norm": 0.7415732145309448, | |
| "learning_rate": 6.130653266331658e-05, | |
| "loss": 1.1266, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.27792436235708, | |
| "grad_norm": 0.6941065192222595, | |
| "learning_rate": 6.080402010050251e-05, | |
| "loss": 1.1073, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.28144239226033424, | |
| "grad_norm": 0.713752269744873, | |
| "learning_rate": 6.030150753768844e-05, | |
| "loss": 1.0299, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2849604221635884, | |
| "grad_norm": 0.672386884689331, | |
| "learning_rate": 5.979899497487438e-05, | |
| "loss": 1.1285, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2884784520668426, | |
| "grad_norm": 0.6600875854492188, | |
| "learning_rate": 5.929648241206031e-05, | |
| "loss": 1.0618, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.29199648197009676, | |
| "grad_norm": 0.7304966449737549, | |
| "learning_rate": 5.879396984924623e-05, | |
| "loss": 1.068, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2955145118733509, | |
| "grad_norm": 0.7191479206085205, | |
| "learning_rate": 5.829145728643216e-05, | |
| "loss": 1.0915, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2990325417766051, | |
| "grad_norm": 0.6817315220832825, | |
| "learning_rate": 5.778894472361809e-05, | |
| "loss": 1.0081, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30255057167985927, | |
| "grad_norm": 0.7097010016441345, | |
| "learning_rate": 5.728643216080403e-05, | |
| "loss": 1.0442, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.30606860158311344, | |
| "grad_norm": 0.7585952281951904, | |
| "learning_rate": 5.6783919597989955e-05, | |
| "loss": 1.0238, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3095866314863676, | |
| "grad_norm": 0.7607995271682739, | |
| "learning_rate": 5.628140703517588e-05, | |
| "loss": 1.0959, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3131046613896218, | |
| "grad_norm": 0.67258220911026, | |
| "learning_rate": 5.577889447236181e-05, | |
| "loss": 0.9929, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.316622691292876, | |
| "grad_norm": 0.75568026304245, | |
| "learning_rate": 5.527638190954774e-05, | |
| "loss": 1.105, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3201407211961302, | |
| "grad_norm": 0.8852781057357788, | |
| "learning_rate": 5.477386934673368e-05, | |
| "loss": 1.083, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.32365875109938436, | |
| "grad_norm": 0.6639973521232605, | |
| "learning_rate": 5.4271356783919604e-05, | |
| "loss": 1.073, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.32717678100263853, | |
| "grad_norm": 0.7528688311576843, | |
| "learning_rate": 5.376884422110553e-05, | |
| "loss": 1.0957, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3306948109058927, | |
| "grad_norm": 0.7375084757804871, | |
| "learning_rate": 5.3266331658291455e-05, | |
| "loss": 1.0804, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.33421284080914687, | |
| "grad_norm": 0.8116129040718079, | |
| "learning_rate": 5.276381909547739e-05, | |
| "loss": 1.0797, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.33773087071240104, | |
| "grad_norm": 0.7964279055595398, | |
| "learning_rate": 5.226130653266332e-05, | |
| "loss": 1.1213, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3412489006156552, | |
| "grad_norm": 0.765575110912323, | |
| "learning_rate": 5.175879396984925e-05, | |
| "loss": 1.0384, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3447669305189094, | |
| "grad_norm": 0.6614196300506592, | |
| "learning_rate": 5.125628140703518e-05, | |
| "loss": 1.0332, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3482849604221636, | |
| "grad_norm": 0.7407499551773071, | |
| "learning_rate": 5.0753768844221104e-05, | |
| "loss": 1.0688, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3518029903254178, | |
| "grad_norm": 0.8672274947166443, | |
| "learning_rate": 5.0251256281407036e-05, | |
| "loss": 1.0742, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35532102022867196, | |
| "grad_norm": 0.6899972558021545, | |
| "learning_rate": 4.974874371859297e-05, | |
| "loss": 0.9776, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.35883905013192613, | |
| "grad_norm": 0.7466877698898315, | |
| "learning_rate": 4.92462311557789e-05, | |
| "loss": 1.0293, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3623570800351803, | |
| "grad_norm": 0.7986593842506409, | |
| "learning_rate": 4.874371859296483e-05, | |
| "loss": 1.0399, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3658751099384345, | |
| "grad_norm": 0.6813223958015442, | |
| "learning_rate": 4.824120603015075e-05, | |
| "loss": 1.063, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.36939313984168864, | |
| "grad_norm": 0.7377122044563293, | |
| "learning_rate": 4.7738693467336685e-05, | |
| "loss": 0.9959, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3729111697449428, | |
| "grad_norm": 0.7429965138435364, | |
| "learning_rate": 4.723618090452262e-05, | |
| "loss": 1.0617, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.376429199648197, | |
| "grad_norm": 0.8200985193252563, | |
| "learning_rate": 4.673366834170855e-05, | |
| "loss": 1.069, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.37994722955145116, | |
| "grad_norm": 0.734062910079956, | |
| "learning_rate": 4.6231155778894475e-05, | |
| "loss": 1.1513, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3834652594547054, | |
| "grad_norm": 0.8677653670310974, | |
| "learning_rate": 4.57286432160804e-05, | |
| "loss": 1.1646, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.38698328935795956, | |
| "grad_norm": 0.7318121790885925, | |
| "learning_rate": 4.522613065326633e-05, | |
| "loss": 1.0443, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.39050131926121373, | |
| "grad_norm": 0.8211216330528259, | |
| "learning_rate": 4.4723618090452266e-05, | |
| "loss": 1.1295, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3940193491644679, | |
| "grad_norm": 0.6949535608291626, | |
| "learning_rate": 4.42211055276382e-05, | |
| "loss": 1.0175, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3975373790677221, | |
| "grad_norm": 0.7230639457702637, | |
| "learning_rate": 4.3718592964824124e-05, | |
| "loss": 1.0341, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.40105540897097625, | |
| "grad_norm": 0.793847918510437, | |
| "learning_rate": 4.321608040201005e-05, | |
| "loss": 1.0576, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4045734388742304, | |
| "grad_norm": 0.7108281850814819, | |
| "learning_rate": 4.271356783919598e-05, | |
| "loss": 1.0636, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.4080914687774846, | |
| "grad_norm": 0.7297809720039368, | |
| "learning_rate": 4.2211055276381914e-05, | |
| "loss": 1.0821, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.41160949868073876, | |
| "grad_norm": 0.6856512427330017, | |
| "learning_rate": 4.170854271356784e-05, | |
| "loss": 0.9826, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.415127528583993, | |
| "grad_norm": 0.7112051248550415, | |
| "learning_rate": 4.120603015075377e-05, | |
| "loss": 1.0463, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.41864555848724716, | |
| "grad_norm": 0.6769644021987915, | |
| "learning_rate": 4.07035175879397e-05, | |
| "loss": 1.0091, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.42216358839050133, | |
| "grad_norm": 0.7250102758407593, | |
| "learning_rate": 4.020100502512563e-05, | |
| "loss": 1.0686, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4256816182937555, | |
| "grad_norm": 0.7410470843315125, | |
| "learning_rate": 3.969849246231156e-05, | |
| "loss": 1.0755, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4291996481970097, | |
| "grad_norm": 0.7236255407333374, | |
| "learning_rate": 3.919597989949749e-05, | |
| "loss": 1.0721, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.43271767810026385, | |
| "grad_norm": 0.7625666856765747, | |
| "learning_rate": 3.869346733668342e-05, | |
| "loss": 0.966, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.436235708003518, | |
| "grad_norm": 0.7245182394981384, | |
| "learning_rate": 3.8190954773869346e-05, | |
| "loss": 1.0801, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4397537379067722, | |
| "grad_norm": 0.7869658470153809, | |
| "learning_rate": 3.768844221105528e-05, | |
| "loss": 1.022, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.44327176781002636, | |
| "grad_norm": 0.7516188621520996, | |
| "learning_rate": 3.7185929648241204e-05, | |
| "loss": 1.0499, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4467897977132806, | |
| "grad_norm": 0.7964783906936646, | |
| "learning_rate": 3.668341708542714e-05, | |
| "loss": 1.0321, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.45030782761653476, | |
| "grad_norm": 0.8271761536598206, | |
| "learning_rate": 3.618090452261307e-05, | |
| "loss": 1.0488, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.45382585751978893, | |
| "grad_norm": 0.69193434715271, | |
| "learning_rate": 3.5678391959798995e-05, | |
| "loss": 0.9999, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4573438874230431, | |
| "grad_norm": 0.7824375033378601, | |
| "learning_rate": 3.517587939698493e-05, | |
| "loss": 1.0199, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4608619173262973, | |
| "grad_norm": 0.7616211771965027, | |
| "learning_rate": 3.467336683417085e-05, | |
| "loss": 0.9752, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.46437994722955145, | |
| "grad_norm": 0.7464612126350403, | |
| "learning_rate": 3.4170854271356785e-05, | |
| "loss": 0.9756, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4678979771328056, | |
| "grad_norm": 0.7916256189346313, | |
| "learning_rate": 3.366834170854272e-05, | |
| "loss": 1.1048, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4714160070360598, | |
| "grad_norm": 0.7534184455871582, | |
| "learning_rate": 3.3165829145728643e-05, | |
| "loss": 0.9938, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.47493403693931396, | |
| "grad_norm": 0.6909853219985962, | |
| "learning_rate": 3.2663316582914576e-05, | |
| "loss": 0.9762, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.47845206684256814, | |
| "grad_norm": 0.7753147482872009, | |
| "learning_rate": 3.21608040201005e-05, | |
| "loss": 1.105, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.48197009674582236, | |
| "grad_norm": 0.7884505391120911, | |
| "learning_rate": 3.1658291457286434e-05, | |
| "loss": 1.0783, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.48548812664907653, | |
| "grad_norm": 0.7265881896018982, | |
| "learning_rate": 3.1155778894472366e-05, | |
| "loss": 0.9864, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4890061565523307, | |
| "grad_norm": 0.7939391732215881, | |
| "learning_rate": 3.065326633165829e-05, | |
| "loss": 1.1004, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4925241864555849, | |
| "grad_norm": 0.739389955997467, | |
| "learning_rate": 3.015075376884422e-05, | |
| "loss": 1.0617, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.49604221635883905, | |
| "grad_norm": 0.8098007440567017, | |
| "learning_rate": 2.9648241206030153e-05, | |
| "loss": 1.0949, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4995602462620932, | |
| "grad_norm": 0.8120628595352173, | |
| "learning_rate": 2.914572864321608e-05, | |
| "loss": 1.0509, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5030782761653474, | |
| "grad_norm": 0.8424797654151917, | |
| "learning_rate": 2.8643216080402015e-05, | |
| "loss": 1.1095, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5065963060686016, | |
| "grad_norm": 0.7528412938117981, | |
| "learning_rate": 2.814070351758794e-05, | |
| "loss": 0.9922, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5101143359718557, | |
| "grad_norm": 0.7280577421188354, | |
| "learning_rate": 2.763819095477387e-05, | |
| "loss": 1.0284, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.51363236587511, | |
| "grad_norm": 0.7835600972175598, | |
| "learning_rate": 2.7135678391959802e-05, | |
| "loss": 1.0266, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5171503957783641, | |
| "grad_norm": 0.7442212104797363, | |
| "learning_rate": 2.6633165829145728e-05, | |
| "loss": 1.0308, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5206684256816183, | |
| "grad_norm": 0.7954034209251404, | |
| "learning_rate": 2.613065326633166e-05, | |
| "loss": 1.1187, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5241864555848724, | |
| "grad_norm": 0.6770613193511963, | |
| "learning_rate": 2.562814070351759e-05, | |
| "loss": 0.9689, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5277044854881267, | |
| "grad_norm": 0.7365478277206421, | |
| "learning_rate": 2.5125628140703518e-05, | |
| "loss": 0.9841, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5312225153913809, | |
| "grad_norm": 0.7578640580177307, | |
| "learning_rate": 2.462311557788945e-05, | |
| "loss": 1.041, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.534740545294635, | |
| "grad_norm": 0.7007668614387512, | |
| "learning_rate": 2.4120603015075376e-05, | |
| "loss": 1.0407, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5382585751978892, | |
| "grad_norm": 0.7602474689483643, | |
| "learning_rate": 2.361809045226131e-05, | |
| "loss": 0.9901, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5417766051011433, | |
| "grad_norm": 0.8691968321800232, | |
| "learning_rate": 2.3115577889447238e-05, | |
| "loss": 1.0138, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5452946350043976, | |
| "grad_norm": 0.7328104376792908, | |
| "learning_rate": 2.2613065326633167e-05, | |
| "loss": 1.0706, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5488126649076517, | |
| "grad_norm": 0.7762755751609802, | |
| "learning_rate": 2.21105527638191e-05, | |
| "loss": 1.0248, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5523306948109059, | |
| "grad_norm": 0.854016900062561, | |
| "learning_rate": 2.1608040201005025e-05, | |
| "loss": 1.0046, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.55584872471416, | |
| "grad_norm": 0.7478740215301514, | |
| "learning_rate": 2.1105527638190957e-05, | |
| "loss": 1.0434, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5593667546174143, | |
| "grad_norm": 0.8043814301490784, | |
| "learning_rate": 2.0603015075376886e-05, | |
| "loss": 1.0655, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5628847845206685, | |
| "grad_norm": 0.8493765592575073, | |
| "learning_rate": 2.0100502512562815e-05, | |
| "loss": 1.0395, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5664028144239226, | |
| "grad_norm": 0.783991813659668, | |
| "learning_rate": 1.9597989949748744e-05, | |
| "loss": 0.9494, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5699208443271768, | |
| "grad_norm": 0.7609344124794006, | |
| "learning_rate": 1.9095477386934673e-05, | |
| "loss": 1.055, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5734388742304309, | |
| "grad_norm": 0.766476035118103, | |
| "learning_rate": 1.8592964824120602e-05, | |
| "loss": 1.0782, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5769569041336852, | |
| "grad_norm": 0.7780715227127075, | |
| "learning_rate": 1.8090452261306535e-05, | |
| "loss": 0.9793, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5804749340369393, | |
| "grad_norm": 0.7344515919685364, | |
| "learning_rate": 1.7587939698492464e-05, | |
| "loss": 1.0129, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5839929639401935, | |
| "grad_norm": 0.7865444421768188, | |
| "learning_rate": 1.7085427135678393e-05, | |
| "loss": 1.0503, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5875109938434476, | |
| "grad_norm": 0.8012449741363525, | |
| "learning_rate": 1.6582914572864322e-05, | |
| "loss": 1.0298, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5910290237467019, | |
| "grad_norm": 0.8140902519226074, | |
| "learning_rate": 1.608040201005025e-05, | |
| "loss": 1.1027, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.594547053649956, | |
| "grad_norm": 0.9053994417190552, | |
| "learning_rate": 1.5577889447236183e-05, | |
| "loss": 1.0591, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5980650835532102, | |
| "grad_norm": 0.7367292642593384, | |
| "learning_rate": 1.507537688442211e-05, | |
| "loss": 1.0475, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6015831134564644, | |
| "grad_norm": 0.8504379391670227, | |
| "learning_rate": 1.457286432160804e-05, | |
| "loss": 0.9989, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.6051011433597185, | |
| "grad_norm": 0.7499436736106873, | |
| "learning_rate": 1.407035175879397e-05, | |
| "loss": 1.0329, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6086191732629728, | |
| "grad_norm": 0.8187640309333801, | |
| "learning_rate": 1.3567839195979901e-05, | |
| "loss": 1.0425, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6121372031662269, | |
| "grad_norm": 0.7070643305778503, | |
| "learning_rate": 1.306532663316583e-05, | |
| "loss": 0.9766, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6156552330694811, | |
| "grad_norm": 0.8162341713905334, | |
| "learning_rate": 1.2562814070351759e-05, | |
| "loss": 0.9974, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6191732629727352, | |
| "grad_norm": 0.7759721875190735, | |
| "learning_rate": 1.2060301507537688e-05, | |
| "loss": 1.0475, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6226912928759895, | |
| "grad_norm": 0.7885333299636841, | |
| "learning_rate": 1.1557788944723619e-05, | |
| "loss": 1.0531, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6262093227792436, | |
| "grad_norm": 0.7671830654144287, | |
| "learning_rate": 1.105527638190955e-05, | |
| "loss": 0.9974, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6297273526824978, | |
| "grad_norm": 0.7737442851066589, | |
| "learning_rate": 1.0552763819095479e-05, | |
| "loss": 1.0145, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.633245382585752, | |
| "grad_norm": 0.8488346338272095, | |
| "learning_rate": 1.0050251256281408e-05, | |
| "loss": 1.024, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6367634124890061, | |
| "grad_norm": 0.7485771775245667, | |
| "learning_rate": 9.547738693467337e-06, | |
| "loss": 1.0519, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6402814423922604, | |
| "grad_norm": 0.8044915795326233, | |
| "learning_rate": 9.045226130653267e-06, | |
| "loss": 0.977, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6437994722955145, | |
| "grad_norm": 0.8789907693862915, | |
| "learning_rate": 8.542713567839196e-06, | |
| "loss": 1.0284, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6473175021987687, | |
| "grad_norm": 0.7542572617530823, | |
| "learning_rate": 8.040201005025125e-06, | |
| "loss": 1.0125, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6508355321020228, | |
| "grad_norm": 0.7793267965316772, | |
| "learning_rate": 7.537688442211055e-06, | |
| "loss": 1.0383, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6543535620052771, | |
| "grad_norm": 0.774917721748352, | |
| "learning_rate": 7.035175879396985e-06, | |
| "loss": 1.0392, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6578715919085312, | |
| "grad_norm": 0.8436054587364197, | |
| "learning_rate": 6.532663316582915e-06, | |
| "loss": 1.0772, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6613896218117854, | |
| "grad_norm": 0.7968306541442871, | |
| "learning_rate": 6.030150753768844e-06, | |
| "loss": 1.0723, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6649076517150396, | |
| "grad_norm": 0.8724409341812134, | |
| "learning_rate": 5.527638190954775e-06, | |
| "loss": 1.0429, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6684256816182937, | |
| "grad_norm": 0.9110769033432007, | |
| "learning_rate": 5.025125628140704e-06, | |
| "loss": 1.0439, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.671943711521548, | |
| "grad_norm": 0.8945828080177307, | |
| "learning_rate": 4.522613065326634e-06, | |
| "loss": 1.0797, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6754617414248021, | |
| "grad_norm": 0.8030518889427185, | |
| "learning_rate": 4.020100502512563e-06, | |
| "loss": 1.0457, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6789797713280563, | |
| "grad_norm": 0.8692275285720825, | |
| "learning_rate": 3.5175879396984926e-06, | |
| "loss": 1.0895, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6824978012313104, | |
| "grad_norm": 0.7445128560066223, | |
| "learning_rate": 3.015075376884422e-06, | |
| "loss": 0.9904, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6860158311345647, | |
| "grad_norm": 0.7754868865013123, | |
| "learning_rate": 2.512562814070352e-06, | |
| "loss": 1.0576, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6895338610378188, | |
| "grad_norm": 0.8235899806022644, | |
| "learning_rate": 2.0100502512562813e-06, | |
| "loss": 0.9928, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.693051890941073, | |
| "grad_norm": 0.8219490051269531, | |
| "learning_rate": 1.507537688442211e-06, | |
| "loss": 1.0847, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6965699208443272, | |
| "grad_norm": 0.7800722122192383, | |
| "learning_rate": 1.0050251256281407e-06, | |
| "loss": 1.0303, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7000879507475813, | |
| "grad_norm": 0.8147994875907898, | |
| "learning_rate": 5.025125628140703e-07, | |
| "loss": 1.045, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7036059806508356, | |
| "grad_norm": 0.7462975978851318, | |
| "learning_rate": 0.0, | |
| "loss": 0.9956, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 1.871454142739251e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |