| { | |
| "best_metric": 0.7762666344642639, | |
| "best_model_checkpoint": "./output/clip-finetuned-csu-p14-336-e6l56-l/checkpoint-21000", | |
| "epoch": 3.868828297715549, | |
| "eval_steps": 500, | |
| "global_step": 21000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09211495946941783, | |
| "grad_norm": 0.49671515822410583, | |
| "learning_rate": 4.923237533775486e-06, | |
| "loss": 0.2159, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09211495946941783, | |
| "eval_loss": 1.4818799495697021, | |
| "eval_runtime": 72.2844, | |
| "eval_samples_per_second": 16.698, | |
| "eval_steps_per_second": 2.089, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "grad_norm": 0.0015633106231689453, | |
| "learning_rate": 4.846475067550971e-06, | |
| "loss": 0.2648, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.18422991893883567, | |
| "eval_loss": 1.4225908517837524, | |
| "eval_runtime": 73.6999, | |
| "eval_samples_per_second": 16.377, | |
| "eval_steps_per_second": 2.049, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "grad_norm": 0.047471947968006134, | |
| "learning_rate": 4.769712601326456e-06, | |
| "loss": 0.328, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2763448784082535, | |
| "eval_loss": 1.412400245666504, | |
| "eval_runtime": 74.9915, | |
| "eval_samples_per_second": 16.095, | |
| "eval_steps_per_second": 2.014, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "grad_norm": 3.9284778208248383e-10, | |
| "learning_rate": 4.692950135101941e-06, | |
| "loss": 0.1813, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36845983787767134, | |
| "eval_loss": 1.2844351530075073, | |
| "eval_runtime": 74.7649, | |
| "eval_samples_per_second": 16.144, | |
| "eval_steps_per_second": 2.02, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "grad_norm": 514.6055297851562, | |
| "learning_rate": 4.616187668877426e-06, | |
| "loss": 0.4836, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.46057479734708917, | |
| "eval_loss": 1.301351547241211, | |
| "eval_runtime": 74.7994, | |
| "eval_samples_per_second": 16.136, | |
| "eval_steps_per_second": 2.019, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "grad_norm": 127.98405456542969, | |
| "learning_rate": 4.5394252026529115e-06, | |
| "loss": 0.2803, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.552689756816507, | |
| "eval_loss": 1.1920161247253418, | |
| "eval_runtime": 75.5447, | |
| "eval_samples_per_second": 15.977, | |
| "eval_steps_per_second": 1.999, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "grad_norm": 0.048586923629045486, | |
| "learning_rate": 4.462662736428396e-06, | |
| "loss": 0.2056, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6448047162859248, | |
| "eval_loss": 1.1428992748260498, | |
| "eval_runtime": 75.7202, | |
| "eval_samples_per_second": 15.94, | |
| "eval_steps_per_second": 1.994, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "grad_norm": 228.63064575195312, | |
| "learning_rate": 4.385900270203882e-06, | |
| "loss": 0.2174, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7369196757553427, | |
| "eval_loss": 1.1957736015319824, | |
| "eval_runtime": 74.1933, | |
| "eval_samples_per_second": 16.268, | |
| "eval_steps_per_second": 2.035, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "grad_norm": 226.2799530029297, | |
| "learning_rate": 4.3091378039793666e-06, | |
| "loss": 0.2236, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8290346352247605, | |
| "eval_loss": 1.0961644649505615, | |
| "eval_runtime": 74.6968, | |
| "eval_samples_per_second": 16.159, | |
| "eval_steps_per_second": 2.022, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "grad_norm": 8.084862640478363e-12, | |
| "learning_rate": 4.232375337754851e-06, | |
| "loss": 0.1966, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9211495946941783, | |
| "eval_loss": 1.1026136875152588, | |
| "eval_runtime": 75.566, | |
| "eval_samples_per_second": 15.973, | |
| "eval_steps_per_second": 1.998, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "grad_norm": 1.592839794284373e-06, | |
| "learning_rate": 4.155612871530337e-06, | |
| "loss": 0.2547, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.013264554163596, | |
| "eval_loss": 1.0125067234039307, | |
| "eval_runtime": 74.4786, | |
| "eval_samples_per_second": 16.206, | |
| "eval_steps_per_second": 2.027, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "grad_norm": 0.06377054750919342, | |
| "learning_rate": 4.078850405305822e-06, | |
| "loss": 0.1057, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.105379513633014, | |
| "eval_loss": 1.1245251893997192, | |
| "eval_runtime": 74.102, | |
| "eval_samples_per_second": 16.288, | |
| "eval_steps_per_second": 2.038, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "grad_norm": 0.00903751328587532, | |
| "learning_rate": 4.002087939081307e-06, | |
| "loss": 0.1085, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1974944731024317, | |
| "eval_loss": 1.1834831237792969, | |
| "eval_runtime": 74.4532, | |
| "eval_samples_per_second": 16.212, | |
| "eval_steps_per_second": 2.028, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "grad_norm": 5.0311258631419165e-11, | |
| "learning_rate": 3.925325472856792e-06, | |
| "loss": 0.1174, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.2896094325718497, | |
| "eval_loss": 1.1334316730499268, | |
| "eval_runtime": 74.1756, | |
| "eval_samples_per_second": 16.272, | |
| "eval_steps_per_second": 2.036, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "grad_norm": 3.813414917175351e-09, | |
| "learning_rate": 3.848563006632278e-06, | |
| "loss": 0.122, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3817243920412676, | |
| "eval_loss": 1.1058012247085571, | |
| "eval_runtime": 74.0285, | |
| "eval_samples_per_second": 16.305, | |
| "eval_steps_per_second": 2.04, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "grad_norm": 4.6230165935412515e-06, | |
| "learning_rate": 3.7718005404077624e-06, | |
| "loss": 0.1653, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.4738393515106853, | |
| "eval_loss": 1.8118412494659424, | |
| "eval_runtime": 73.9465, | |
| "eval_samples_per_second": 16.323, | |
| "eval_steps_per_second": 2.042, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "grad_norm": 6.103370031596e-10, | |
| "learning_rate": 3.6950380741832476e-06, | |
| "loss": 0.1144, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.565954310980103, | |
| "eval_loss": 1.0845656394958496, | |
| "eval_runtime": 73.8897, | |
| "eval_samples_per_second": 16.335, | |
| "eval_steps_per_second": 2.044, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "grad_norm": 0.023263949900865555, | |
| "learning_rate": 3.6182756079587327e-06, | |
| "loss": 0.1866, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.658069270449521, | |
| "eval_loss": 1.2360613346099854, | |
| "eval_runtime": 73.8786, | |
| "eval_samples_per_second": 16.338, | |
| "eval_steps_per_second": 2.044, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "grad_norm": 1.1146563672070897e-09, | |
| "learning_rate": 3.541513141734218e-06, | |
| "loss": 0.1385, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.750184229918939, | |
| "eval_loss": 1.0680217742919922, | |
| "eval_runtime": 73.8064, | |
| "eval_samples_per_second": 16.354, | |
| "eval_steps_per_second": 2.046, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "grad_norm": 0.0801646038889885, | |
| "learning_rate": 3.464750675509703e-06, | |
| "loss": 0.1178, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.8422991893883567, | |
| "eval_loss": 0.9592632055282593, | |
| "eval_runtime": 73.9232, | |
| "eval_samples_per_second": 16.328, | |
| "eval_steps_per_second": 2.043, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "grad_norm": 8.68252456598384e-08, | |
| "learning_rate": 3.3879882092851883e-06, | |
| "loss": 0.1506, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.9344141488577744, | |
| "eval_loss": 1.0038937330245972, | |
| "eval_runtime": 74.3344, | |
| "eval_samples_per_second": 16.237, | |
| "eval_steps_per_second": 2.031, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "grad_norm": 1.1008462763584248e-07, | |
| "learning_rate": 3.311225743060673e-06, | |
| "loss": 0.0938, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.026529108327192, | |
| "eval_loss": 0.931797206401825, | |
| "eval_runtime": 74.2035, | |
| "eval_samples_per_second": 16.266, | |
| "eval_steps_per_second": 2.035, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "grad_norm": 7.901914614194538e-06, | |
| "learning_rate": 3.234463276836158e-06, | |
| "loss": 0.093, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.1186440677966103, | |
| "eval_loss": 0.9992227554321289, | |
| "eval_runtime": 74.2516, | |
| "eval_samples_per_second": 16.256, | |
| "eval_steps_per_second": 2.034, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "grad_norm": 0.04380282014608383, | |
| "learning_rate": 3.1577008106116434e-06, | |
| "loss": 0.0877, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.210759027266028, | |
| "eval_loss": 0.9989096522331238, | |
| "eval_runtime": 75.2761, | |
| "eval_samples_per_second": 16.034, | |
| "eval_steps_per_second": 2.006, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "grad_norm": 65.85383605957031, | |
| "learning_rate": 3.080938344387129e-06, | |
| "loss": 0.1102, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3028739867354457, | |
| "eval_loss": 0.9193492531776428, | |
| "eval_runtime": 74.3653, | |
| "eval_samples_per_second": 16.231, | |
| "eval_steps_per_second": 2.031, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "grad_norm": 0.0006995272124186158, | |
| "learning_rate": 3.004175878162614e-06, | |
| "loss": 0.0916, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.3949889462048635, | |
| "eval_loss": 0.9769579172134399, | |
| "eval_runtime": 74.2536, | |
| "eval_samples_per_second": 16.255, | |
| "eval_steps_per_second": 2.034, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "grad_norm": 3.747671507881023e-05, | |
| "learning_rate": 2.9274134119380993e-06, | |
| "loss": 0.0665, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.4871039056742816, | |
| "eval_loss": 0.982865035533905, | |
| "eval_runtime": 75.5772, | |
| "eval_samples_per_second": 15.97, | |
| "eval_steps_per_second": 1.998, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "grad_norm": 4.0381178855896, | |
| "learning_rate": 2.850650945713584e-06, | |
| "loss": 0.0545, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.5792188651436994, | |
| "eval_loss": 0.9746910929679871, | |
| "eval_runtime": 74.7109, | |
| "eval_samples_per_second": 16.156, | |
| "eval_steps_per_second": 2.021, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "grad_norm": 7.164443260876396e-09, | |
| "learning_rate": 2.7738884794890693e-06, | |
| "loss": 0.1023, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.671333824613117, | |
| "eval_loss": 0.9872701168060303, | |
| "eval_runtime": 73.8157, | |
| "eval_samples_per_second": 16.352, | |
| "eval_steps_per_second": 2.046, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "grad_norm": 1.5922936302814605e-09, | |
| "learning_rate": 2.6971260132645544e-06, | |
| "loss": 0.1057, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.7634487840825352, | |
| "eval_loss": 0.946104109287262, | |
| "eval_runtime": 73.9037, | |
| "eval_samples_per_second": 16.332, | |
| "eval_steps_per_second": 2.043, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "grad_norm": 8.37097544059695e-11, | |
| "learning_rate": 2.6203635470400396e-06, | |
| "loss": 0.087, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.855563743551953, | |
| "eval_loss": 0.980368435382843, | |
| "eval_runtime": 73.8687, | |
| "eval_samples_per_second": 16.34, | |
| "eval_steps_per_second": 2.044, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "grad_norm": 1.7481942176818848, | |
| "learning_rate": 2.543601080815525e-06, | |
| "loss": 0.046, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.9476787030213707, | |
| "eval_loss": 0.9453437328338623, | |
| "eval_runtime": 73.8853, | |
| "eval_samples_per_second": 16.336, | |
| "eval_steps_per_second": 2.044, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "grad_norm": 2.1040523923848697e-12, | |
| "learning_rate": 2.46683861459101e-06, | |
| "loss": 0.0914, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.0397936624907884, | |
| "eval_loss": 0.9565027952194214, | |
| "eval_runtime": 74.0088, | |
| "eval_samples_per_second": 16.309, | |
| "eval_steps_per_second": 2.04, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "grad_norm": 1.3199319415235777e-10, | |
| "learning_rate": 2.390076148366495e-06, | |
| "loss": 0.0486, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.131908621960206, | |
| "eval_loss": 0.962292492389679, | |
| "eval_runtime": 73.9793, | |
| "eval_samples_per_second": 16.315, | |
| "eval_steps_per_second": 2.041, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "grad_norm": 5.488172246259637e-05, | |
| "learning_rate": 2.31331368214198e-06, | |
| "loss": 0.1024, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.2240235814296243, | |
| "eval_loss": 0.8799718022346497, | |
| "eval_runtime": 73.833, | |
| "eval_samples_per_second": 16.348, | |
| "eval_steps_per_second": 2.045, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "grad_norm": 1.8391082927393665e-12, | |
| "learning_rate": 2.236551215917465e-06, | |
| "loss": 0.0641, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.316138540899042, | |
| "eval_loss": 0.8668168187141418, | |
| "eval_runtime": 74.1002, | |
| "eval_samples_per_second": 16.289, | |
| "eval_steps_per_second": 2.038, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "grad_norm": 4.760212224397264e-10, | |
| "learning_rate": 2.1597887496929502e-06, | |
| "loss": 0.0982, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.4082535003684598, | |
| "eval_loss": 0.8307517766952515, | |
| "eval_runtime": 74.2129, | |
| "eval_samples_per_second": 16.264, | |
| "eval_steps_per_second": 2.035, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "grad_norm": 0.00010056859900942072, | |
| "learning_rate": 2.0830262834684354e-06, | |
| "loss": 0.0256, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.500368459837878, | |
| "eval_loss": 0.7900328636169434, | |
| "eval_runtime": 73.8942, | |
| "eval_samples_per_second": 16.334, | |
| "eval_steps_per_second": 2.043, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "grad_norm": 3.8125975265757006e-07, | |
| "learning_rate": 2.0062638172439206e-06, | |
| "loss": 0.0684, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.5924834193072956, | |
| "eval_loss": 0.807513415813446, | |
| "eval_runtime": 75.2826, | |
| "eval_samples_per_second": 16.033, | |
| "eval_steps_per_second": 2.006, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "grad_norm": 2.2643770742464785e-14, | |
| "learning_rate": 1.9295013510194058e-06, | |
| "loss": 0.0463, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.6845983787767134, | |
| "eval_loss": 0.7804784178733826, | |
| "eval_runtime": 74.4513, | |
| "eval_samples_per_second": 16.212, | |
| "eval_steps_per_second": 2.028, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.776713338246131, | |
| "grad_norm": 48.942562103271484, | |
| "learning_rate": 1.8527388847948907e-06, | |
| "loss": 0.0536, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.776713338246131, | |
| "eval_loss": 0.7935602068901062, | |
| "eval_runtime": 74.0045, | |
| "eval_samples_per_second": 16.31, | |
| "eval_steps_per_second": 2.04, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.868828297715549, | |
| "grad_norm": 1.9078814916806408e-11, | |
| "learning_rate": 1.775976418570376e-06, | |
| "loss": 0.0901, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.868828297715549, | |
| "eval_loss": 0.7762666344642639, | |
| "eval_runtime": 75.3556, | |
| "eval_samples_per_second": 16.017, | |
| "eval_steps_per_second": 2.004, | |
| "step": 21000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32568, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "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": 7555644893162520.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |