| { | |
| "best_metric": 0.8666148781776428, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e4l58-l/checkpoint-20000", | |
| "epoch": 3.6845983787767134, | |
| "eval_steps": 500, | |
| "global_step": 20000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09211495946941783, | |
| "grad_norm": 20.04043960571289, | |
| "learning_rate": 4.8848563006632274e-08, | |
| "loss": 0.3758, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09211495946941783, | |
| "eval_loss": 1.4184927940368652, | |
| "eval_runtime": 74.3562, | |
| "eval_samples_per_second": 16.233, | |
| "eval_steps_per_second": 2.031, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "grad_norm": 74.25025939941406, | |
| "learning_rate": 4.7697126013264556e-08, | |
| "loss": 0.4103, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "eval_loss": 1.3500770330429077, | |
| "eval_runtime": 75.7013, | |
| "eval_samples_per_second": 15.944, | |
| "eval_steps_per_second": 1.995, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "grad_norm": 0.5102410912513733, | |
| "learning_rate": 4.6545689019896826e-08, | |
| "loss": 0.433, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "eval_loss": 1.2885024547576904, | |
| "eval_runtime": 75.197, | |
| "eval_samples_per_second": 16.051, | |
| "eval_steps_per_second": 2.008, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "grad_norm": 0.1667678952217102, | |
| "learning_rate": 4.539425202652911e-08, | |
| "loss": 0.3424, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "eval_loss": 1.239119052886963, | |
| "eval_runtime": 75.387, | |
| "eval_samples_per_second": 16.011, | |
| "eval_steps_per_second": 2.003, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "grad_norm": 343.8175048828125, | |
| "learning_rate": 4.4242815033161385e-08, | |
| "loss": 0.3645, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "eval_loss": 1.1902339458465576, | |
| "eval_runtime": 75.9899, | |
| "eval_samples_per_second": 15.884, | |
| "eval_steps_per_second": 1.987, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "grad_norm": 305.5626525878906, | |
| "learning_rate": 4.309137803979366e-08, | |
| "loss": 0.3172, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "eval_loss": 1.1506118774414062, | |
| "eval_runtime": 74.8655, | |
| "eval_samples_per_second": 16.122, | |
| "eval_steps_per_second": 2.017, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "grad_norm": 65.59048461914062, | |
| "learning_rate": 4.193994104642594e-08, | |
| "loss": 0.2751, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "eval_loss": 1.1169357299804688, | |
| "eval_runtime": 76.2936, | |
| "eval_samples_per_second": 15.82, | |
| "eval_steps_per_second": 1.979, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "grad_norm": 18.106201171875, | |
| "learning_rate": 4.078850405305821e-08, | |
| "loss": 0.2919, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "eval_loss": 1.0920671224594116, | |
| "eval_runtime": 74.9101, | |
| "eval_samples_per_second": 16.113, | |
| "eval_steps_per_second": 2.016, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "grad_norm": 506.40570068359375, | |
| "learning_rate": 3.9637067059690496e-08, | |
| "loss": 0.2583, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "eval_loss": 1.0721209049224854, | |
| "eval_runtime": 75.8053, | |
| "eval_samples_per_second": 15.922, | |
| "eval_steps_per_second": 1.992, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "grad_norm": 1.1508910655975342, | |
| "learning_rate": 3.848563006632277e-08, | |
| "loss": 0.2679, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "eval_loss": 1.0519349575042725, | |
| "eval_runtime": 75.226, | |
| "eval_samples_per_second": 16.045, | |
| "eval_steps_per_second": 2.007, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "grad_norm": 0.1393290013074875, | |
| "learning_rate": 3.733419307295505e-08, | |
| "loss": 0.2472, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "eval_loss": 1.0355629920959473, | |
| "eval_runtime": 74.6535, | |
| "eval_samples_per_second": 16.168, | |
| "eval_steps_per_second": 2.023, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "grad_norm": 200.34664916992188, | |
| "learning_rate": 3.6182756079587324e-08, | |
| "loss": 0.26, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "eval_loss": 1.0176944732666016, | |
| "eval_runtime": 74.4374, | |
| "eval_samples_per_second": 16.215, | |
| "eval_steps_per_second": 2.029, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "grad_norm": 431.1804504394531, | |
| "learning_rate": 3.50313190862196e-08, | |
| "loss": 0.2153, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "eval_loss": 1.0044586658477783, | |
| "eval_runtime": 74.3247, | |
| "eval_samples_per_second": 16.24, | |
| "eval_steps_per_second": 2.032, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "grad_norm": 0.6204941272735596, | |
| "learning_rate": 3.387988209285188e-08, | |
| "loss": 0.1791, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "eval_loss": 0.9926707148551941, | |
| "eval_runtime": 75.7317, | |
| "eval_samples_per_second": 15.938, | |
| "eval_steps_per_second": 1.994, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "grad_norm": 52.27273178100586, | |
| "learning_rate": 3.272844509948415e-08, | |
| "loss": 0.2082, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "eval_loss": 0.9803809523582458, | |
| "eval_runtime": 74.9936, | |
| "eval_samples_per_second": 16.095, | |
| "eval_steps_per_second": 2.014, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "grad_norm": 0.0976635068655014, | |
| "learning_rate": 3.1577008106116435e-08, | |
| "loss": 0.196, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "eval_loss": 0.9711655378341675, | |
| "eval_runtime": 74.3537, | |
| "eval_samples_per_second": 16.233, | |
| "eval_steps_per_second": 2.031, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "grad_norm": 46.64162826538086, | |
| "learning_rate": 3.042557111274871e-08, | |
| "loss": 0.1946, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "eval_loss": 0.9621181488037109, | |
| "eval_runtime": 75.7205, | |
| "eval_samples_per_second": 15.94, | |
| "eval_steps_per_second": 1.994, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "grad_norm": 0.05690698325634003, | |
| "learning_rate": 2.927413411938099e-08, | |
| "loss": 0.2422, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "eval_loss": 0.9536784291267395, | |
| "eval_runtime": 75.1745, | |
| "eval_samples_per_second": 16.056, | |
| "eval_steps_per_second": 2.009, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "grad_norm": 0.08470064401626587, | |
| "learning_rate": 2.8122697126013263e-08, | |
| "loss": 0.2106, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "eval_loss": 0.9457550644874573, | |
| "eval_runtime": 74.5062, | |
| "eval_samples_per_second": 16.2, | |
| "eval_steps_per_second": 2.027, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "grad_norm": 1.5745967626571655, | |
| "learning_rate": 2.697126013264554e-08, | |
| "loss": 0.1801, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "eval_loss": 0.9393072724342346, | |
| "eval_runtime": 75.9432, | |
| "eval_samples_per_second": 15.893, | |
| "eval_steps_per_second": 1.988, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "grad_norm": 0.06661002337932587, | |
| "learning_rate": 2.5819823139277818e-08, | |
| "loss": 0.2117, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "eval_loss": 0.9308408498764038, | |
| "eval_runtime": 75.0811, | |
| "eval_samples_per_second": 16.076, | |
| "eval_steps_per_second": 2.011, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "grad_norm": 0.236691415309906, | |
| "learning_rate": 2.4668386145910094e-08, | |
| "loss": 0.2061, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "eval_loss": 0.9237484335899353, | |
| "eval_runtime": 74.6134, | |
| "eval_samples_per_second": 16.177, | |
| "eval_steps_per_second": 2.024, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "grad_norm": 8.831416130065918, | |
| "learning_rate": 2.3516949152542374e-08, | |
| "loss": 0.1878, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "eval_loss": 0.9166750311851501, | |
| "eval_runtime": 75.6886, | |
| "eval_samples_per_second": 15.947, | |
| "eval_steps_per_second": 1.995, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "grad_norm": 0.787421464920044, | |
| "learning_rate": 2.236551215917465e-08, | |
| "loss": 0.1655, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "eval_loss": 0.9109458327293396, | |
| "eval_runtime": 74.9866, | |
| "eval_samples_per_second": 16.096, | |
| "eval_steps_per_second": 2.014, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "grad_norm": 535.4310913085938, | |
| "learning_rate": 2.1214075165806926e-08, | |
| "loss": 0.1946, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "eval_loss": 0.9070548415184021, | |
| "eval_runtime": 74.3796, | |
| "eval_samples_per_second": 16.228, | |
| "eval_steps_per_second": 2.03, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "grad_norm": 0.8420014381408691, | |
| "learning_rate": 2.0062638172439202e-08, | |
| "loss": 0.1882, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "eval_loss": 0.9021016359329224, | |
| "eval_runtime": 75.8812, | |
| "eval_samples_per_second": 15.906, | |
| "eval_steps_per_second": 1.99, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "grad_norm": 1.4041686058044434, | |
| "learning_rate": 1.891120117907148e-08, | |
| "loss": 0.1871, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "eval_loss": 0.8960274457931519, | |
| "eval_runtime": 75.2283, | |
| "eval_samples_per_second": 16.044, | |
| "eval_steps_per_second": 2.007, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "grad_norm": 327.2938537597656, | |
| "learning_rate": 1.7759764185703757e-08, | |
| "loss": 0.1419, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "eval_loss": 0.8913142085075378, | |
| "eval_runtime": 74.4503, | |
| "eval_samples_per_second": 16.212, | |
| "eval_steps_per_second": 2.028, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "grad_norm": 0.01779676228761673, | |
| "learning_rate": 1.6608327192336037e-08, | |
| "loss": 0.1431, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "eval_loss": 0.8879275321960449, | |
| "eval_runtime": 75.8636, | |
| "eval_samples_per_second": 15.91, | |
| "eval_steps_per_second": 1.99, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "grad_norm": 1.6428073644638062, | |
| "learning_rate": 1.5456890198968313e-08, | |
| "loss": 0.1811, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "eval_loss": 0.8848333954811096, | |
| "eval_runtime": 75.1911, | |
| "eval_samples_per_second": 16.052, | |
| "eval_steps_per_second": 2.008, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "grad_norm": 0.0006026208866387606, | |
| "learning_rate": 1.430545320560059e-08, | |
| "loss": 0.1694, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "eval_loss": 0.8826532959938049, | |
| "eval_runtime": 74.6044, | |
| "eval_samples_per_second": 16.179, | |
| "eval_steps_per_second": 2.024, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "grad_norm": 226.09982299804688, | |
| "learning_rate": 1.3154016212232866e-08, | |
| "loss": 0.1718, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "eval_loss": 0.8798421621322632, | |
| "eval_runtime": 76.0825, | |
| "eval_samples_per_second": 15.864, | |
| "eval_steps_per_second": 1.985, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "grad_norm": 5.1134706154698506e-05, | |
| "learning_rate": 1.2002579218865142e-08, | |
| "loss": 0.153, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "eval_loss": 0.8776962757110596, | |
| "eval_runtime": 75.3205, | |
| "eval_samples_per_second": 16.025, | |
| "eval_steps_per_second": 2.005, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "grad_norm": 0.017252299934625626, | |
| "learning_rate": 1.085114222549742e-08, | |
| "loss": 0.1715, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "eval_loss": 0.8758523464202881, | |
| "eval_runtime": 74.4544, | |
| "eval_samples_per_second": 16.211, | |
| "eval_steps_per_second": 2.028, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "grad_norm": 68.8683853149414, | |
| "learning_rate": 9.699705232129698e-09, | |
| "loss": 0.1558, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "eval_loss": 0.874191403388977, | |
| "eval_runtime": 75.7598, | |
| "eval_samples_per_second": 15.932, | |
| "eval_steps_per_second": 1.993, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "grad_norm": 0.791266918182373, | |
| "learning_rate": 8.548268238761974e-09, | |
| "loss": 0.1384, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "eval_loss": 0.8715030550956726, | |
| "eval_runtime": 74.7336, | |
| "eval_samples_per_second": 16.151, | |
| "eval_steps_per_second": 2.021, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "grad_norm": 6.072669982910156, | |
| "learning_rate": 7.396831245394252e-09, | |
| "loss": 0.1788, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "eval_loss": 0.8695403933525085, | |
| "eval_runtime": 75.2877, | |
| "eval_samples_per_second": 16.032, | |
| "eval_steps_per_second": 2.006, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "grad_norm": 0.43182799220085144, | |
| "learning_rate": 6.2453942520265294e-09, | |
| "loss": 0.1668, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "eval_loss": 0.8684694766998291, | |
| "eval_runtime": 75.7597, | |
| "eval_samples_per_second": 15.932, | |
| "eval_steps_per_second": 1.993, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "grad_norm": 0.01970759592950344, | |
| "learning_rate": 5.093957258658806e-09, | |
| "loss": 0.1697, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "eval_loss": 0.8674086928367615, | |
| "eval_runtime": 74.5942, | |
| "eval_samples_per_second": 16.181, | |
| "eval_steps_per_second": 2.024, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "grad_norm": 0.0002275915176142007, | |
| "learning_rate": 3.942520265291083e-09, | |
| "loss": 0.1764, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "eval_loss": 0.8666148781776428, | |
| "eval_runtime": 74.3994, | |
| "eval_samples_per_second": 16.223, | |
| "eval_steps_per_second": 2.03, | |
| "step": 20000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 21712, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 7195886541902520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |