| { | |
| "best_metric": 1.497014045715332, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e5l510-l/checkpoint-20500", | |
| "epoch": 3.776713338246131, | |
| "eval_steps": 500, | |
| "global_step": 20500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09211495946941783, | |
| "grad_norm": 21.660480499267578, | |
| "learning_rate": 4.907885040530583e-10, | |
| "loss": 0.3974, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09211495946941783, | |
| "eval_loss": 1.5028706789016724, | |
| "eval_runtime": 74.0662, | |
| "eval_samples_per_second": 16.296, | |
| "eval_steps_per_second": 2.039, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "grad_norm": 60.26338577270508, | |
| "learning_rate": 4.815770081061165e-10, | |
| "loss": 0.462, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "eval_loss": 1.5025668144226074, | |
| "eval_runtime": 75.1835, | |
| "eval_samples_per_second": 16.054, | |
| "eval_steps_per_second": 2.008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "grad_norm": 1.4199143648147583, | |
| "learning_rate": 4.723655121591747e-10, | |
| "loss": 0.5249, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "eval_loss": 1.5022265911102295, | |
| "eval_runtime": 76.3442, | |
| "eval_samples_per_second": 15.81, | |
| "eval_steps_per_second": 1.978, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "grad_norm": 1.2705016136169434, | |
| "learning_rate": 4.6315401621223287e-10, | |
| "loss": 0.4642, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "eval_loss": 1.5019283294677734, | |
| "eval_runtime": 77.0512, | |
| "eval_samples_per_second": 15.665, | |
| "eval_steps_per_second": 1.96, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "grad_norm": 335.6837463378906, | |
| "learning_rate": 4.5394252026529114e-10, | |
| "loss": 0.5056, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "eval_loss": 1.5015931129455566, | |
| "eval_runtime": 76.8527, | |
| "eval_samples_per_second": 15.705, | |
| "eval_steps_per_second": 1.965, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "grad_norm": 402.1738586425781, | |
| "learning_rate": 4.4473102431834935e-10, | |
| "loss": 0.4604, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "eval_loss": 1.501311182975769, | |
| "eval_runtime": 76.7728, | |
| "eval_samples_per_second": 15.722, | |
| "eval_steps_per_second": 1.967, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "grad_norm": 120.70919799804688, | |
| "learning_rate": 4.3551952837140756e-10, | |
| "loss": 0.4508, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "eval_loss": 1.5010316371917725, | |
| "eval_runtime": 77.0074, | |
| "eval_samples_per_second": 15.674, | |
| "eval_steps_per_second": 1.961, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "grad_norm": 103.2845687866211, | |
| "learning_rate": 4.2630803242446577e-10, | |
| "loss": 0.4674, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "eval_loss": 1.5007821321487427, | |
| "eval_runtime": 76.7277, | |
| "eval_samples_per_second": 15.731, | |
| "eval_steps_per_second": 1.968, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "grad_norm": 464.5989074707031, | |
| "learning_rate": 4.17096536477524e-10, | |
| "loss": 0.4435, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "eval_loss": 1.500537633895874, | |
| "eval_runtime": 76.8855, | |
| "eval_samples_per_second": 15.699, | |
| "eval_steps_per_second": 1.964, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "grad_norm": 6.3550262451171875, | |
| "learning_rate": 4.078850405305822e-10, | |
| "loss": 0.4573, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "eval_loss": 1.5003185272216797, | |
| "eval_runtime": 77.2072, | |
| "eval_samples_per_second": 15.633, | |
| "eval_steps_per_second": 1.956, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "grad_norm": 0.6333007216453552, | |
| "learning_rate": 3.9867354458364045e-10, | |
| "loss": 0.4681, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "eval_loss": 1.5001204013824463, | |
| "eval_runtime": 76.9383, | |
| "eval_samples_per_second": 15.688, | |
| "eval_steps_per_second": 1.963, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "grad_norm": 238.3988800048828, | |
| "learning_rate": 3.894620486366986e-10, | |
| "loss": 0.4862, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "eval_loss": 1.4999229907989502, | |
| "eval_runtime": 76.9483, | |
| "eval_samples_per_second": 15.686, | |
| "eval_steps_per_second": 1.962, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "grad_norm": 492.2626953125, | |
| "learning_rate": 3.8025055268975687e-10, | |
| "loss": 0.4231, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "eval_loss": 1.4997295141220093, | |
| "eval_runtime": 75.1791, | |
| "eval_samples_per_second": 16.055, | |
| "eval_steps_per_second": 2.009, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "grad_norm": 3.0739829540252686, | |
| "learning_rate": 3.7103905674281503e-10, | |
| "loss": 0.3734, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "eval_loss": 1.4995641708374023, | |
| "eval_runtime": 76.6897, | |
| "eval_samples_per_second": 15.739, | |
| "eval_steps_per_second": 1.969, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "grad_norm": 142.28684997558594, | |
| "learning_rate": 3.618275607958733e-10, | |
| "loss": 0.4606, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "eval_loss": 1.4993607997894287, | |
| "eval_runtime": 75.831, | |
| "eval_samples_per_second": 15.917, | |
| "eval_steps_per_second": 1.991, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "grad_norm": 1.5455149412155151, | |
| "learning_rate": 3.526160648489315e-10, | |
| "loss": 0.467, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "eval_loss": 1.4991577863693237, | |
| "eval_runtime": 75.1339, | |
| "eval_samples_per_second": 16.065, | |
| "eval_steps_per_second": 2.01, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "grad_norm": 215.83163452148438, | |
| "learning_rate": 3.434045689019897e-10, | |
| "loss": 0.4202, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "eval_loss": 1.498979091644287, | |
| "eval_runtime": 75.4004, | |
| "eval_samples_per_second": 16.008, | |
| "eval_steps_per_second": 2.003, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "grad_norm": 2.8240511417388916, | |
| "learning_rate": 3.341930729550479e-10, | |
| "loss": 0.4866, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "eval_loss": 1.4988232851028442, | |
| "eval_runtime": 75.0613, | |
| "eval_samples_per_second": 16.08, | |
| "eval_steps_per_second": 2.012, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "grad_norm": 7.242213249206543, | |
| "learning_rate": 3.2498157700810614e-10, | |
| "loss": 0.4848, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "eval_loss": 1.4986662864685059, | |
| "eval_runtime": 75.3419, | |
| "eval_samples_per_second": 16.02, | |
| "eval_steps_per_second": 2.004, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "grad_norm": 908.7202758789062, | |
| "learning_rate": 3.1577008106116435e-10, | |
| "loss": 0.4385, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "eval_loss": 1.4985153675079346, | |
| "eval_runtime": 75.0638, | |
| "eval_samples_per_second": 16.08, | |
| "eval_steps_per_second": 2.012, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "grad_norm": 4.991658687591553, | |
| "learning_rate": 3.065585851142226e-10, | |
| "loss": 0.4781, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "eval_loss": 1.498375654220581, | |
| "eval_runtime": 75.0746, | |
| "eval_samples_per_second": 16.077, | |
| "eval_steps_per_second": 2.011, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "grad_norm": 0.4935278594493866, | |
| "learning_rate": 2.9734708916728077e-10, | |
| "loss": 0.4287, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "eval_loss": 1.4982625246047974, | |
| "eval_runtime": 75.5519, | |
| "eval_samples_per_second": 15.976, | |
| "eval_steps_per_second": 1.999, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "grad_norm": 441.09906005859375, | |
| "learning_rate": 2.8813559322033903e-10, | |
| "loss": 0.4392, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "eval_loss": 1.4981573820114136, | |
| "eval_runtime": 75.0431, | |
| "eval_samples_per_second": 16.084, | |
| "eval_steps_per_second": 2.012, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "grad_norm": 1.5746198892593384, | |
| "learning_rate": 2.789240972733972e-10, | |
| "loss": 0.4447, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "eval_loss": 1.498041033744812, | |
| "eval_runtime": 75.1277, | |
| "eval_samples_per_second": 16.066, | |
| "eval_steps_per_second": 2.01, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "grad_norm": 604.6005249023438, | |
| "learning_rate": 2.6971260132645545e-10, | |
| "loss": 0.4449, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "eval_loss": 1.497951865196228, | |
| "eval_runtime": 75.2241, | |
| "eval_samples_per_second": 16.045, | |
| "eval_steps_per_second": 2.007, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "grad_norm": 5.409970760345459, | |
| "learning_rate": 2.6050110537951366e-10, | |
| "loss": 0.4748, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "eval_loss": 1.4978498220443726, | |
| "eval_runtime": 75.3614, | |
| "eval_samples_per_second": 16.016, | |
| "eval_steps_per_second": 2.004, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "grad_norm": 17.807907104492188, | |
| "learning_rate": 2.5128960943257187e-10, | |
| "loss": 0.4935, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "eval_loss": 1.4977525472640991, | |
| "eval_runtime": 74.8827, | |
| "eval_samples_per_second": 16.119, | |
| "eval_steps_per_second": 2.016, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "grad_norm": 190.154541015625, | |
| "learning_rate": 2.420781134856301e-10, | |
| "loss": 0.4113, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "eval_loss": 1.4976625442504883, | |
| "eval_runtime": 76.3316, | |
| "eval_samples_per_second": 15.813, | |
| "eval_steps_per_second": 1.978, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "grad_norm": 3.5760951042175293, | |
| "learning_rate": 2.328666175386883e-10, | |
| "loss": 0.3934, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "eval_loss": 1.4975850582122803, | |
| "eval_runtime": 76.0918, | |
| "eval_samples_per_second": 15.862, | |
| "eval_steps_per_second": 1.984, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "grad_norm": 195.72146606445312, | |
| "learning_rate": 2.236551215917465e-10, | |
| "loss": 0.4901, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "eval_loss": 1.4975062608718872, | |
| "eval_runtime": 74.9535, | |
| "eval_samples_per_second": 16.103, | |
| "eval_steps_per_second": 2.015, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "grad_norm": 0.14779114723205566, | |
| "learning_rate": 2.1444362564480471e-10, | |
| "loss": 0.4601, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "eval_loss": 1.497436761856079, | |
| "eval_runtime": 75.9666, | |
| "eval_samples_per_second": 15.889, | |
| "eval_steps_per_second": 1.988, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "grad_norm": 205.7298583984375, | |
| "learning_rate": 2.0523212969786292e-10, | |
| "loss": 0.4224, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "eval_loss": 1.4973747730255127, | |
| "eval_runtime": 76.7648, | |
| "eval_samples_per_second": 15.723, | |
| "eval_steps_per_second": 1.967, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "grad_norm": 9.98619270324707, | |
| "learning_rate": 1.9602063375092116e-10, | |
| "loss": 0.4381, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "eval_loss": 1.497320294380188, | |
| "eval_runtime": 75.2158, | |
| "eval_samples_per_second": 16.047, | |
| "eval_steps_per_second": 2.008, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "grad_norm": 5.3734331130981445, | |
| "learning_rate": 1.8680913780397937e-10, | |
| "loss": 0.4942, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "eval_loss": 1.4972634315490723, | |
| "eval_runtime": 75.2534, | |
| "eval_samples_per_second": 16.039, | |
| "eval_steps_per_second": 2.007, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "grad_norm": 117.80599212646484, | |
| "learning_rate": 1.7759764185703758e-10, | |
| "loss": 0.3885, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "eval_loss": 1.4972138404846191, | |
| "eval_runtime": 75.5586, | |
| "eval_samples_per_second": 15.974, | |
| "eval_steps_per_second": 1.998, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "grad_norm": 81.40491485595703, | |
| "learning_rate": 1.683861459100958e-10, | |
| "loss": 0.4407, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "eval_loss": 1.4971661567687988, | |
| "eval_runtime": 75.6731, | |
| "eval_samples_per_second": 15.95, | |
| "eval_steps_per_second": 1.995, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "grad_norm": 16.30617904663086, | |
| "learning_rate": 1.59174649963154e-10, | |
| "loss": 0.4635, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "eval_loss": 1.4971280097961426, | |
| "eval_runtime": 75.728, | |
| "eval_samples_per_second": 15.939, | |
| "eval_steps_per_second": 1.994, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "grad_norm": 40.21035385131836, | |
| "learning_rate": 1.4996315401621224e-10, | |
| "loss": 0.4228, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "eval_loss": 1.4970934391021729, | |
| "eval_runtime": 76.4366, | |
| "eval_samples_per_second": 15.791, | |
| "eval_steps_per_second": 1.975, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "grad_norm": 7.447707176208496, | |
| "learning_rate": 1.4075165806927045e-10, | |
| "loss": 0.464, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "eval_loss": 1.4970619678497314, | |
| "eval_runtime": 75.781, | |
| "eval_samples_per_second": 15.927, | |
| "eval_steps_per_second": 1.993, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "grad_norm": 0.005426365882158279, | |
| "learning_rate": 1.3154016212232866e-10, | |
| "loss": 0.4746, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "eval_loss": 1.4970370531082153, | |
| "eval_runtime": 75.7317, | |
| "eval_samples_per_second": 15.938, | |
| "eval_steps_per_second": 1.994, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.776713338246131, | |
| "grad_norm": 485.1770935058594, | |
| "learning_rate": 1.223286661753869e-10, | |
| "loss": 0.4447, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.776713338246131, | |
| "eval_loss": 1.497014045715332, | |
| "eval_runtime": 76.0421, | |
| "eval_samples_per_second": 15.873, | |
| "eval_steps_per_second": 1.986, | |
| "step": 20500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 27140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 7375765717532520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |