| { | |
| "best_metric": 0.4866587221622467, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l57-l/checkpoint-21500", | |
| "epoch": 2.34641638225256, | |
| "eval_steps": 500, | |
| "global_step": 22000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05332764505119454, | |
| "grad_norm": 414.2552795410156, | |
| "learning_rate": 4.911120591581342e-07, | |
| "loss": 0.3809, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05332764505119454, | |
| "eval_loss": 1.1109352111816406, | |
| "eval_runtime": 63.6734, | |
| "eval_samples_per_second": 15.501, | |
| "eval_steps_per_second": 1.947, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "grad_norm": 45.880027770996094, | |
| "learning_rate": 4.822241183162685e-07, | |
| "loss": 0.2649, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10665529010238908, | |
| "eval_loss": 0.9588962197303772, | |
| "eval_runtime": 62.6825, | |
| "eval_samples_per_second": 15.746, | |
| "eval_steps_per_second": 1.978, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "grad_norm": 432.9915466308594, | |
| "learning_rate": 4.733361774744027e-07, | |
| "loss": 0.2104, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1599829351535836, | |
| "eval_loss": 0.90110844373703, | |
| "eval_runtime": 62.6757, | |
| "eval_samples_per_second": 15.748, | |
| "eval_steps_per_second": 1.978, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "grad_norm": 0.014208819717168808, | |
| "learning_rate": 4.6444823663253695e-07, | |
| "loss": 0.1849, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21331058020477817, | |
| "eval_loss": 0.8570474982261658, | |
| "eval_runtime": 63.7296, | |
| "eval_samples_per_second": 15.487, | |
| "eval_steps_per_second": 1.946, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "grad_norm": 174.73324584960938, | |
| "learning_rate": 4.5556029579067116e-07, | |
| "loss": 0.2056, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2666382252559727, | |
| "eval_loss": 0.7895939350128174, | |
| "eval_runtime": 63.8387, | |
| "eval_samples_per_second": 15.461, | |
| "eval_steps_per_second": 1.942, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "grad_norm": 9.852073823424234e-09, | |
| "learning_rate": 4.4667235494880547e-07, | |
| "loss": 0.159, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3199658703071672, | |
| "eval_loss": 0.7647623419761658, | |
| "eval_runtime": 63.9155, | |
| "eval_samples_per_second": 15.442, | |
| "eval_steps_per_second": 1.94, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "grad_norm": 0.0025215104687958956, | |
| "learning_rate": 4.377844141069397e-07, | |
| "loss": 0.1696, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.37329351535836175, | |
| "eval_loss": 0.7638036012649536, | |
| "eval_runtime": 63.8003, | |
| "eval_samples_per_second": 15.47, | |
| "eval_steps_per_second": 1.944, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "grad_norm": 2.4611830711364746, | |
| "learning_rate": 4.2889647326507393e-07, | |
| "loss": 0.1125, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.42662116040955633, | |
| "eval_loss": 0.7486374378204346, | |
| "eval_runtime": 63.9847, | |
| "eval_samples_per_second": 15.426, | |
| "eval_steps_per_second": 1.938, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "grad_norm": 5.230295658111572, | |
| "learning_rate": 4.2000853242320814e-07, | |
| "loss": 0.1322, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.47994880546075086, | |
| "eval_loss": 0.7251659035682678, | |
| "eval_runtime": 64.4782, | |
| "eval_samples_per_second": 15.308, | |
| "eval_steps_per_second": 1.923, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "grad_norm": 3.821002974291332e-05, | |
| "learning_rate": 4.1112059158134245e-07, | |
| "loss": 0.1633, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5332764505119454, | |
| "eval_loss": 0.7035414576530457, | |
| "eval_runtime": 62.6186, | |
| "eval_samples_per_second": 15.762, | |
| "eval_steps_per_second": 1.98, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "grad_norm": 5.916939244343666e-06, | |
| "learning_rate": 4.0223265073947665e-07, | |
| "loss": 0.125, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.58660409556314, | |
| "eval_loss": 0.6792380809783936, | |
| "eval_runtime": 65.0551, | |
| "eval_samples_per_second": 15.172, | |
| "eval_steps_per_second": 1.906, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "grad_norm": 92.8386459350586, | |
| "learning_rate": 3.933447098976109e-07, | |
| "loss": 0.1297, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6399317406143344, | |
| "eval_loss": 0.6549482345581055, | |
| "eval_runtime": 64.0827, | |
| "eval_samples_per_second": 15.402, | |
| "eval_steps_per_second": 1.935, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "grad_norm": 0.00014717792510055006, | |
| "learning_rate": 3.8445676905574517e-07, | |
| "loss": 0.1114, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.693259385665529, | |
| "eval_loss": 0.6223254799842834, | |
| "eval_runtime": 64.89, | |
| "eval_samples_per_second": 15.21, | |
| "eval_steps_per_second": 1.911, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "grad_norm": 4.789559397977428e-07, | |
| "learning_rate": 3.755688282138794e-07, | |
| "loss": 0.144, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7465870307167235, | |
| "eval_loss": 0.6204637289047241, | |
| "eval_runtime": 62.7674, | |
| "eval_samples_per_second": 15.725, | |
| "eval_steps_per_second": 1.976, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "grad_norm": 1.7078508138656616, | |
| "learning_rate": 3.6668088737201363e-07, | |
| "loss": 0.1399, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7999146757679181, | |
| "eval_loss": 0.6074371933937073, | |
| "eval_runtime": 63.4694, | |
| "eval_samples_per_second": 15.551, | |
| "eval_steps_per_second": 1.954, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "grad_norm": 870.1183471679688, | |
| "learning_rate": 3.5779294653014783e-07, | |
| "loss": 0.141, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8532423208191127, | |
| "eval_loss": 0.5950366854667664, | |
| "eval_runtime": 62.915, | |
| "eval_samples_per_second": 15.688, | |
| "eval_steps_per_second": 1.971, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "grad_norm": 5.781071013188921e-05, | |
| "learning_rate": 3.4890500568828214e-07, | |
| "loss": 0.1291, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9065699658703071, | |
| "eval_loss": 0.5800224542617798, | |
| "eval_runtime": 62.3336, | |
| "eval_samples_per_second": 15.834, | |
| "eval_steps_per_second": 1.989, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "grad_norm": 1.3669992685317993, | |
| "learning_rate": 3.4001706484641635e-07, | |
| "loss": 0.1246, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9598976109215017, | |
| "eval_loss": 0.5702486038208008, | |
| "eval_runtime": 64.055, | |
| "eval_samples_per_second": 15.409, | |
| "eval_steps_per_second": 1.936, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "grad_norm": 644.68115234375, | |
| "learning_rate": 3.311291240045506e-07, | |
| "loss": 0.1257, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0132252559726962, | |
| "eval_loss": 0.5807780623435974, | |
| "eval_runtime": 62.5312, | |
| "eval_samples_per_second": 15.784, | |
| "eval_steps_per_second": 1.983, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "grad_norm": 27.854568481445312, | |
| "learning_rate": 3.2224118316268486e-07, | |
| "loss": 0.0336, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0665529010238908, | |
| "eval_loss": 0.5713164210319519, | |
| "eval_runtime": 62.5622, | |
| "eval_samples_per_second": 15.776, | |
| "eval_steps_per_second": 1.982, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "grad_norm": 0.00045933027286082506, | |
| "learning_rate": 3.133532423208191e-07, | |
| "loss": 0.0663, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1198805460750854, | |
| "eval_loss": 0.5601483583450317, | |
| "eval_runtime": 63.4675, | |
| "eval_samples_per_second": 15.551, | |
| "eval_steps_per_second": 1.954, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "grad_norm": 0.14330387115478516, | |
| "learning_rate": 3.044653014789533e-07, | |
| "loss": 0.078, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.17320819112628, | |
| "eval_loss": 0.5580261945724487, | |
| "eval_runtime": 63.3927, | |
| "eval_samples_per_second": 15.57, | |
| "eval_steps_per_second": 1.956, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "grad_norm": 4.066005518388316e-11, | |
| "learning_rate": 2.955773606370876e-07, | |
| "loss": 0.1118, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2265358361774745, | |
| "eval_loss": 0.5491370558738708, | |
| "eval_runtime": 63.7293, | |
| "eval_samples_per_second": 15.487, | |
| "eval_steps_per_second": 1.946, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "grad_norm": 7.519358769059181e-05, | |
| "learning_rate": 2.8668941979522184e-07, | |
| "loss": 0.0534, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2798634812286689, | |
| "eval_loss": 0.5478764772415161, | |
| "eval_runtime": 62.7547, | |
| "eval_samples_per_second": 15.728, | |
| "eval_steps_per_second": 1.976, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "grad_norm": 8.400617446113756e-08, | |
| "learning_rate": 2.778014789533561e-07, | |
| "loss": 0.0507, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.3331911262798635, | |
| "eval_loss": 0.576600968837738, | |
| "eval_runtime": 64.301, | |
| "eval_samples_per_second": 15.35, | |
| "eval_steps_per_second": 1.928, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "grad_norm": 2.5385464130084356e-12, | |
| "learning_rate": 2.689135381114903e-07, | |
| "loss": 0.0784, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.386518771331058, | |
| "eval_loss": 0.5729931592941284, | |
| "eval_runtime": 63.652, | |
| "eval_samples_per_second": 15.506, | |
| "eval_steps_per_second": 1.948, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "grad_norm": 0.08409273624420166, | |
| "learning_rate": 2.600255972696246e-07, | |
| "loss": 0.079, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.4398464163822526, | |
| "eval_loss": 0.5694039463996887, | |
| "eval_runtime": 62.8247, | |
| "eval_samples_per_second": 15.71, | |
| "eval_steps_per_second": 1.974, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "grad_norm": 3.7334253022436314e-08, | |
| "learning_rate": 2.511376564277588e-07, | |
| "loss": 0.0507, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.493174061433447, | |
| "eval_loss": 0.5713907480239868, | |
| "eval_runtime": 62.3577, | |
| "eval_samples_per_second": 15.828, | |
| "eval_steps_per_second": 1.989, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "grad_norm": 1.5219894647598267, | |
| "learning_rate": 2.422497155858931e-07, | |
| "loss": 0.0774, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5465017064846416, | |
| "eval_loss": 0.5674533843994141, | |
| "eval_runtime": 62.3753, | |
| "eval_samples_per_second": 15.824, | |
| "eval_steps_per_second": 1.988, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "grad_norm": 879.1193237304688, | |
| "learning_rate": 2.333617747440273e-07, | |
| "loss": 0.0376, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5998293515358362, | |
| "eval_loss": 0.561253547668457, | |
| "eval_runtime": 62.4788, | |
| "eval_samples_per_second": 15.797, | |
| "eval_steps_per_second": 1.985, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "grad_norm": 0.0006221308722160757, | |
| "learning_rate": 2.2447383390216154e-07, | |
| "loss": 0.0929, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.6531569965870307, | |
| "eval_loss": 0.5395999550819397, | |
| "eval_runtime": 62.5651, | |
| "eval_samples_per_second": 15.776, | |
| "eval_steps_per_second": 1.982, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "grad_norm": 37.85109329223633, | |
| "learning_rate": 2.155858930602958e-07, | |
| "loss": 0.0442, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7064846416382253, | |
| "eval_loss": 0.5307884216308594, | |
| "eval_runtime": 62.4681, | |
| "eval_samples_per_second": 15.8, | |
| "eval_steps_per_second": 1.985, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "grad_norm": 9.600337885773413e-11, | |
| "learning_rate": 2.0669795221843002e-07, | |
| "loss": 0.0593, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.75981228668942, | |
| "eval_loss": 0.5230685472488403, | |
| "eval_runtime": 63.8124, | |
| "eval_samples_per_second": 15.467, | |
| "eval_steps_per_second": 1.943, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "grad_norm": 9.067448940129619e-13, | |
| "learning_rate": 1.9781001137656428e-07, | |
| "loss": 0.0578, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8131399317406145, | |
| "eval_loss": 0.5200281739234924, | |
| "eval_runtime": 62.3767, | |
| "eval_samples_per_second": 15.823, | |
| "eval_steps_per_second": 1.988, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8664675767918089, | |
| "grad_norm": 0.00843018852174282, | |
| "learning_rate": 1.889220705346985e-07, | |
| "loss": 0.0928, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.8664675767918089, | |
| "eval_loss": 0.509168803691864, | |
| "eval_runtime": 62.4108, | |
| "eval_samples_per_second": 15.815, | |
| "eval_steps_per_second": 1.987, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.9197952218430034, | |
| "grad_norm": 4.264631314754297e-09, | |
| "learning_rate": 1.8003412969283277e-07, | |
| "loss": 0.0633, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.9197952218430034, | |
| "eval_loss": 0.4954444468021393, | |
| "eval_runtime": 63.3639, | |
| "eval_samples_per_second": 15.577, | |
| "eval_steps_per_second": 1.957, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.9731228668941978, | |
| "grad_norm": 459.5137939453125, | |
| "learning_rate": 1.7114618885096697e-07, | |
| "loss": 0.067, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.9731228668941978, | |
| "eval_loss": 0.49844062328338623, | |
| "eval_runtime": 63.6573, | |
| "eval_samples_per_second": 15.505, | |
| "eval_steps_per_second": 1.948, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.0264505119453924, | |
| "grad_norm": 0.27590689063072205, | |
| "learning_rate": 1.6225824800910123e-07, | |
| "loss": 0.0525, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.0264505119453924, | |
| "eval_loss": 0.4999145269393921, | |
| "eval_runtime": 63.1727, | |
| "eval_samples_per_second": 15.624, | |
| "eval_steps_per_second": 1.963, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.079778156996587, | |
| "grad_norm": 1.349623744317796e-06, | |
| "learning_rate": 1.533703071672355e-07, | |
| "loss": 0.0301, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.079778156996587, | |
| "eval_loss": 0.4926412105560303, | |
| "eval_runtime": 63.4421, | |
| "eval_samples_per_second": 15.557, | |
| "eval_steps_per_second": 1.955, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.1331058020477816, | |
| "grad_norm": 4.928138878312893e-06, | |
| "learning_rate": 1.4448236632536972e-07, | |
| "loss": 0.0408, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.1331058020477816, | |
| "eval_loss": 0.4872625172138214, | |
| "eval_runtime": 63.2047, | |
| "eval_samples_per_second": 15.616, | |
| "eval_steps_per_second": 1.962, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.186433447098976, | |
| "grad_norm": 0.003605826525017619, | |
| "learning_rate": 1.3559442548350398e-07, | |
| "loss": 0.0489, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.186433447098976, | |
| "eval_loss": 0.48834049701690674, | |
| "eval_runtime": 63.5831, | |
| "eval_samples_per_second": 15.523, | |
| "eval_steps_per_second": 1.95, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.2397610921501707, | |
| "grad_norm": 0.0027391298208385706, | |
| "learning_rate": 1.267064846416382e-07, | |
| "loss": 0.0354, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.2397610921501707, | |
| "eval_loss": 0.4887964129447937, | |
| "eval_runtime": 63.0462, | |
| "eval_samples_per_second": 15.655, | |
| "eval_steps_per_second": 1.967, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.2930887372013653, | |
| "grad_norm": 1.5921540352792363e-07, | |
| "learning_rate": 1.1781854379977247e-07, | |
| "loss": 0.0667, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.2930887372013653, | |
| "eval_loss": 0.4866587221622467, | |
| "eval_runtime": 63.4143, | |
| "eval_samples_per_second": 15.564, | |
| "eval_steps_per_second": 1.955, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.34641638225256, | |
| "grad_norm": 3.954861771404694e-09, | |
| "learning_rate": 1.0893060295790671e-07, | |
| "loss": 0.0315, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.34641638225256, | |
| "eval_loss": 0.48805058002471924, | |
| "eval_runtime": 63.4731, | |
| "eval_samples_per_second": 15.55, | |
| "eval_steps_per_second": 1.954, | |
| "step": 22000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 28128, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7915403244422520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |