|
{ |
|
"best_metric": 0.8607594936708861, |
|
"best_model_checkpoint": "beit-base-patch16-224-fold2/checkpoint-77", |
|
"epoch": 85.71428571428571, |
|
"eval_steps": 500, |
|
"global_step": 300, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.8571428571428571, |
|
"eval_accuracy": 0.6075949367088608, |
|
"eval_loss": 0.678494930267334, |
|
"eval_runtime": 1.1189, |
|
"eval_samples_per_second": 70.604, |
|
"eval_steps_per_second": 2.681, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.6329113924050633, |
|
"eval_loss": 0.6621109843254089, |
|
"eval_runtime": 1.1177, |
|
"eval_samples_per_second": 70.683, |
|
"eval_steps_per_second": 2.684, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"grad_norm": 6.013461112976074, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.7083, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 2.857142857142857, |
|
"eval_accuracy": 0.5822784810126582, |
|
"eval_loss": 0.6433596014976501, |
|
"eval_runtime": 1.1574, |
|
"eval_samples_per_second": 68.257, |
|
"eval_steps_per_second": 2.592, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.569620253164557, |
|
"eval_loss": 0.6708337068557739, |
|
"eval_runtime": 1.1991, |
|
"eval_samples_per_second": 65.882, |
|
"eval_steps_per_second": 2.502, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 4.857142857142857, |
|
"eval_accuracy": 0.6075949367088608, |
|
"eval_loss": 0.6701093316078186, |
|
"eval_runtime": 1.2358, |
|
"eval_samples_per_second": 63.928, |
|
"eval_steps_per_second": 2.428, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 5.714285714285714, |
|
"grad_norm": 14.194457054138184, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.6009, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5949367088607594, |
|
"eval_loss": 0.7957927584648132, |
|
"eval_runtime": 1.2868, |
|
"eval_samples_per_second": 61.391, |
|
"eval_steps_per_second": 2.331, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 6.857142857142857, |
|
"eval_accuracy": 0.6455696202531646, |
|
"eval_loss": 0.5951915979385376, |
|
"eval_runtime": 1.2473, |
|
"eval_samples_per_second": 63.338, |
|
"eval_steps_per_second": 2.405, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.6962025316455697, |
|
"eval_loss": 0.8007984757423401, |
|
"eval_runtime": 1.2233, |
|
"eval_samples_per_second": 64.58, |
|
"eval_steps_per_second": 2.452, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 8.571428571428571, |
|
"grad_norm": 9.016133308410645, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5315, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 8.857142857142858, |
|
"eval_accuracy": 0.6329113924050633, |
|
"eval_loss": 0.8903242945671082, |
|
"eval_runtime": 1.2441, |
|
"eval_samples_per_second": 63.497, |
|
"eval_steps_per_second": 2.411, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.6708860759493671, |
|
"eval_loss": 0.7070391178131104, |
|
"eval_runtime": 1.269, |
|
"eval_samples_per_second": 62.255, |
|
"eval_steps_per_second": 2.364, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 10.857142857142858, |
|
"eval_accuracy": 0.759493670886076, |
|
"eval_loss": 0.5331233739852905, |
|
"eval_runtime": 1.2407, |
|
"eval_samples_per_second": 63.672, |
|
"eval_steps_per_second": 2.418, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 11.428571428571429, |
|
"grad_norm": 4.008541584014893, |
|
"learning_rate": 4.814814814814815e-05, |
|
"loss": 0.5756, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.5307104587554932, |
|
"eval_runtime": 1.2262, |
|
"eval_samples_per_second": 64.427, |
|
"eval_steps_per_second": 2.447, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 12.857142857142858, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.5069800615310669, |
|
"eval_runtime": 1.286, |
|
"eval_samples_per_second": 61.429, |
|
"eval_steps_per_second": 2.333, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.7215189873417721, |
|
"eval_loss": 0.6117033958435059, |
|
"eval_runtime": 1.239, |
|
"eval_samples_per_second": 63.763, |
|
"eval_steps_per_second": 2.421, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 14.285714285714286, |
|
"grad_norm": 13.79133129119873, |
|
"learning_rate": 4.62962962962963e-05, |
|
"loss": 0.4519, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 14.857142857142858, |
|
"eval_accuracy": 0.7468354430379747, |
|
"eval_loss": 0.4667030870914459, |
|
"eval_runtime": 1.2442, |
|
"eval_samples_per_second": 63.495, |
|
"eval_steps_per_second": 2.411, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.41506943106651306, |
|
"eval_runtime": 1.2516, |
|
"eval_samples_per_second": 63.12, |
|
"eval_steps_per_second": 2.397, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 16.857142857142858, |
|
"eval_accuracy": 0.7721518987341772, |
|
"eval_loss": 0.4434549808502197, |
|
"eval_runtime": 1.2483, |
|
"eval_samples_per_second": 63.284, |
|
"eval_steps_per_second": 2.403, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 17.142857142857142, |
|
"grad_norm": 4.975181579589844, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 0.3821, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.41142967343330383, |
|
"eval_runtime": 1.2652, |
|
"eval_samples_per_second": 62.442, |
|
"eval_steps_per_second": 2.371, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 18.857142857142858, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.4066586494445801, |
|
"eval_runtime": 1.24, |
|
"eval_samples_per_second": 63.712, |
|
"eval_steps_per_second": 2.419, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 5.205602645874023, |
|
"learning_rate": 4.259259259259259e-05, |
|
"loss": 0.328, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.4459022879600525, |
|
"eval_runtime": 1.2373, |
|
"eval_samples_per_second": 63.847, |
|
"eval_steps_per_second": 2.425, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 20.857142857142858, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.38586243987083435, |
|
"eval_runtime": 1.2783, |
|
"eval_samples_per_second": 61.799, |
|
"eval_steps_per_second": 2.347, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 22.0, |
|
"eval_accuracy": 0.8607594936708861, |
|
"eval_loss": 0.3405309021472931, |
|
"eval_runtime": 1.2673, |
|
"eval_samples_per_second": 62.336, |
|
"eval_steps_per_second": 2.367, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"grad_norm": 9.780171394348145, |
|
"learning_rate": 4.074074074074074e-05, |
|
"loss": 0.3344, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 22.857142857142858, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.3702424168586731, |
|
"eval_runtime": 1.2485, |
|
"eval_samples_per_second": 63.277, |
|
"eval_steps_per_second": 2.403, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 24.0, |
|
"eval_accuracy": 0.7848101265822784, |
|
"eval_loss": 0.4351886808872223, |
|
"eval_runtime": 1.2427, |
|
"eval_samples_per_second": 63.569, |
|
"eval_steps_per_second": 2.414, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 24.857142857142858, |
|
"eval_accuracy": 0.7341772151898734, |
|
"eval_loss": 0.6776776313781738, |
|
"eval_runtime": 1.2553, |
|
"eval_samples_per_second": 62.931, |
|
"eval_steps_per_second": 2.39, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 25.714285714285715, |
|
"grad_norm": 5.561969757080078, |
|
"learning_rate": 3.888888888888889e-05, |
|
"loss": 0.2747, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 26.0, |
|
"eval_accuracy": 0.7974683544303798, |
|
"eval_loss": 0.5707988739013672, |
|
"eval_runtime": 1.2633, |
|
"eval_samples_per_second": 62.533, |
|
"eval_steps_per_second": 2.375, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 26.857142857142858, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.4431754946708679, |
|
"eval_runtime": 1.2698, |
|
"eval_samples_per_second": 62.215, |
|
"eval_steps_per_second": 2.363, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 28.0, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.37358519434928894, |
|
"eval_runtime": 1.2475, |
|
"eval_samples_per_second": 63.326, |
|
"eval_steps_per_second": 2.405, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 28.571428571428573, |
|
"grad_norm": 4.9703569412231445, |
|
"learning_rate": 3.7037037037037037e-05, |
|
"loss": 0.2634, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 28.857142857142858, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.39375337958335876, |
|
"eval_runtime": 1.2572, |
|
"eval_samples_per_second": 62.84, |
|
"eval_steps_per_second": 2.386, |
|
"step": 101 |
|
}, |
|
{ |
|
"epoch": 30.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.44604796171188354, |
|
"eval_runtime": 1.2495, |
|
"eval_samples_per_second": 63.226, |
|
"eval_steps_per_second": 2.401, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 30.857142857142858, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.4381825029850006, |
|
"eval_runtime": 1.2425, |
|
"eval_samples_per_second": 63.58, |
|
"eval_steps_per_second": 2.414, |
|
"step": 108 |
|
}, |
|
{ |
|
"epoch": 31.428571428571427, |
|
"grad_norm": 5.964456081390381, |
|
"learning_rate": 3.518518518518519e-05, |
|
"loss": 0.2306, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 32.0, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.5574378967285156, |
|
"eval_runtime": 1.2448, |
|
"eval_samples_per_second": 63.462, |
|
"eval_steps_per_second": 2.41, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 32.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.3862878680229187, |
|
"eval_runtime": 1.2372, |
|
"eval_samples_per_second": 63.854, |
|
"eval_steps_per_second": 2.425, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 34.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.4390255808830261, |
|
"eval_runtime": 1.2617, |
|
"eval_samples_per_second": 62.614, |
|
"eval_steps_per_second": 2.378, |
|
"step": 119 |
|
}, |
|
{ |
|
"epoch": 34.285714285714285, |
|
"grad_norm": 3.8251123428344727, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.2214, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 34.857142857142854, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.48393675684928894, |
|
"eval_runtime": 1.2419, |
|
"eval_samples_per_second": 63.611, |
|
"eval_steps_per_second": 2.416, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 36.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.4523105323314667, |
|
"eval_runtime": 1.2521, |
|
"eval_samples_per_second": 63.092, |
|
"eval_steps_per_second": 2.396, |
|
"step": 126 |
|
}, |
|
{ |
|
"epoch": 36.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.40220630168914795, |
|
"eval_runtime": 1.2399, |
|
"eval_samples_per_second": 63.715, |
|
"eval_steps_per_second": 2.42, |
|
"step": 129 |
|
}, |
|
{ |
|
"epoch": 37.142857142857146, |
|
"grad_norm": 5.482424736022949, |
|
"learning_rate": 3.148148148148148e-05, |
|
"loss": 0.1945, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 38.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.4408385455608368, |
|
"eval_runtime": 1.2496, |
|
"eval_samples_per_second": 63.22, |
|
"eval_steps_per_second": 2.401, |
|
"step": 133 |
|
}, |
|
{ |
|
"epoch": 38.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.39884084463119507, |
|
"eval_runtime": 1.2583, |
|
"eval_samples_per_second": 62.782, |
|
"eval_steps_per_second": 2.384, |
|
"step": 136 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"grad_norm": 5.258836269378662, |
|
"learning_rate": 2.962962962962963e-05, |
|
"loss": 0.1863, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 40.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.4467245042324066, |
|
"eval_runtime": 1.2362, |
|
"eval_samples_per_second": 63.907, |
|
"eval_steps_per_second": 2.427, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 40.857142857142854, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.478800505399704, |
|
"eval_runtime": 1.2478, |
|
"eval_samples_per_second": 63.31, |
|
"eval_steps_per_second": 2.404, |
|
"step": 143 |
|
}, |
|
{ |
|
"epoch": 42.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.4749128222465515, |
|
"eval_runtime": 1.2505, |
|
"eval_samples_per_second": 63.175, |
|
"eval_steps_per_second": 2.399, |
|
"step": 147 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"grad_norm": 4.518224716186523, |
|
"learning_rate": 2.777777777777778e-05, |
|
"loss": 0.1718, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 42.857142857142854, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.47265732288360596, |
|
"eval_runtime": 1.2572, |
|
"eval_samples_per_second": 62.836, |
|
"eval_steps_per_second": 2.386, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 44.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.4632214307785034, |
|
"eval_runtime": 1.2427, |
|
"eval_samples_per_second": 63.57, |
|
"eval_steps_per_second": 2.414, |
|
"step": 154 |
|
}, |
|
{ |
|
"epoch": 44.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.45606979727745056, |
|
"eval_runtime": 1.2354, |
|
"eval_samples_per_second": 63.948, |
|
"eval_steps_per_second": 2.428, |
|
"step": 157 |
|
}, |
|
{ |
|
"epoch": 45.714285714285715, |
|
"grad_norm": 6.867657661437988, |
|
"learning_rate": 2.5925925925925925e-05, |
|
"loss": 0.1535, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 46.0, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.5113399028778076, |
|
"eval_runtime": 1.238, |
|
"eval_samples_per_second": 63.815, |
|
"eval_steps_per_second": 2.423, |
|
"step": 161 |
|
}, |
|
{ |
|
"epoch": 46.857142857142854, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.6505078673362732, |
|
"eval_runtime": 1.274, |
|
"eval_samples_per_second": 62.008, |
|
"eval_steps_per_second": 2.355, |
|
"step": 164 |
|
}, |
|
{ |
|
"epoch": 48.0, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.5612393021583557, |
|
"eval_runtime": 1.2503, |
|
"eval_samples_per_second": 63.184, |
|
"eval_steps_per_second": 2.399, |
|
"step": 168 |
|
}, |
|
{ |
|
"epoch": 48.57142857142857, |
|
"grad_norm": 2.030773401260376, |
|
"learning_rate": 2.4074074074074074e-05, |
|
"loss": 0.1454, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 48.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6824959516525269, |
|
"eval_runtime": 1.2393, |
|
"eval_samples_per_second": 63.745, |
|
"eval_steps_per_second": 2.421, |
|
"step": 171 |
|
}, |
|
{ |
|
"epoch": 50.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7959554195404053, |
|
"eval_runtime": 1.2446, |
|
"eval_samples_per_second": 63.472, |
|
"eval_steps_per_second": 2.41, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 50.857142857142854, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.5914857387542725, |
|
"eval_runtime": 1.2425, |
|
"eval_samples_per_second": 63.581, |
|
"eval_steps_per_second": 2.414, |
|
"step": 178 |
|
}, |
|
{ |
|
"epoch": 51.42857142857143, |
|
"grad_norm": 5.444547176361084, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.1327, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 52.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6199972629547119, |
|
"eval_runtime": 1.2836, |
|
"eval_samples_per_second": 61.544, |
|
"eval_steps_per_second": 2.337, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 52.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.49771010875701904, |
|
"eval_runtime": 1.254, |
|
"eval_samples_per_second": 63.001, |
|
"eval_steps_per_second": 2.392, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 54.0, |
|
"eval_accuracy": 0.8607594936708861, |
|
"eval_loss": 0.6180350184440613, |
|
"eval_runtime": 1.2463, |
|
"eval_samples_per_second": 63.389, |
|
"eval_steps_per_second": 2.407, |
|
"step": 189 |
|
}, |
|
{ |
|
"epoch": 54.285714285714285, |
|
"grad_norm": 6.831031799316406, |
|
"learning_rate": 2.037037037037037e-05, |
|
"loss": 0.1491, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 54.857142857142854, |
|
"eval_accuracy": 0.8607594936708861, |
|
"eval_loss": 0.6473775506019592, |
|
"eval_runtime": 1.244, |
|
"eval_samples_per_second": 63.505, |
|
"eval_steps_per_second": 2.412, |
|
"step": 192 |
|
}, |
|
{ |
|
"epoch": 56.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.5886236429214478, |
|
"eval_runtime": 1.2691, |
|
"eval_samples_per_second": 62.249, |
|
"eval_steps_per_second": 2.364, |
|
"step": 196 |
|
}, |
|
{ |
|
"epoch": 56.857142857142854, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.6742523312568665, |
|
"eval_runtime": 1.2454, |
|
"eval_samples_per_second": 63.433, |
|
"eval_steps_per_second": 2.409, |
|
"step": 199 |
|
}, |
|
{ |
|
"epoch": 57.142857142857146, |
|
"grad_norm": 6.931390762329102, |
|
"learning_rate": 1.8518518518518518e-05, |
|
"loss": 0.1666, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 58.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6475793123245239, |
|
"eval_runtime": 1.2506, |
|
"eval_samples_per_second": 63.17, |
|
"eval_steps_per_second": 2.399, |
|
"step": 203 |
|
}, |
|
{ |
|
"epoch": 58.857142857142854, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.6483187079429626, |
|
"eval_runtime": 1.244, |
|
"eval_samples_per_second": 63.507, |
|
"eval_steps_per_second": 2.412, |
|
"step": 206 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"grad_norm": 2.6429920196533203, |
|
"learning_rate": 1.6666666666666667e-05, |
|
"loss": 0.1219, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 60.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7215976119041443, |
|
"eval_runtime": 1.257, |
|
"eval_samples_per_second": 62.847, |
|
"eval_steps_per_second": 2.387, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 60.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6541433334350586, |
|
"eval_runtime": 1.255, |
|
"eval_samples_per_second": 62.949, |
|
"eval_steps_per_second": 2.39, |
|
"step": 213 |
|
}, |
|
{ |
|
"epoch": 62.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6635527014732361, |
|
"eval_runtime": 1.2585, |
|
"eval_samples_per_second": 62.774, |
|
"eval_steps_per_second": 2.384, |
|
"step": 217 |
|
}, |
|
{ |
|
"epoch": 62.857142857142854, |
|
"grad_norm": 2.814028739929199, |
|
"learning_rate": 1.4814814814814815e-05, |
|
"loss": 0.1339, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 62.857142857142854, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6707887053489685, |
|
"eval_runtime": 1.262, |
|
"eval_samples_per_second": 62.601, |
|
"eval_steps_per_second": 2.377, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 64.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.6735276579856873, |
|
"eval_runtime": 1.2462, |
|
"eval_samples_per_second": 63.393, |
|
"eval_steps_per_second": 2.407, |
|
"step": 224 |
|
}, |
|
{ |
|
"epoch": 64.85714285714286, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7030315399169922, |
|
"eval_runtime": 1.2489, |
|
"eval_samples_per_second": 63.256, |
|
"eval_steps_per_second": 2.402, |
|
"step": 227 |
|
}, |
|
{ |
|
"epoch": 65.71428571428571, |
|
"grad_norm": 5.088428020477295, |
|
"learning_rate": 1.2962962962962962e-05, |
|
"loss": 0.1227, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 66.0, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.6779046654701233, |
|
"eval_runtime": 1.2468, |
|
"eval_samples_per_second": 63.364, |
|
"eval_steps_per_second": 2.406, |
|
"step": 231 |
|
}, |
|
{ |
|
"epoch": 66.85714285714286, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7091224193572998, |
|
"eval_runtime": 1.2843, |
|
"eval_samples_per_second": 61.512, |
|
"eval_steps_per_second": 2.336, |
|
"step": 234 |
|
}, |
|
{ |
|
"epoch": 68.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6858163475990295, |
|
"eval_runtime": 1.2712, |
|
"eval_samples_per_second": 62.148, |
|
"eval_steps_per_second": 2.36, |
|
"step": 238 |
|
}, |
|
{ |
|
"epoch": 68.57142857142857, |
|
"grad_norm": 7.267394065856934, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.1316, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 68.85714285714286, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6668309569358826, |
|
"eval_runtime": 1.2603, |
|
"eval_samples_per_second": 62.681, |
|
"eval_steps_per_second": 2.38, |
|
"step": 241 |
|
}, |
|
{ |
|
"epoch": 70.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.6490625739097595, |
|
"eval_runtime": 1.2412, |
|
"eval_samples_per_second": 63.648, |
|
"eval_steps_per_second": 2.417, |
|
"step": 245 |
|
}, |
|
{ |
|
"epoch": 70.85714285714286, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.7164068222045898, |
|
"eval_runtime": 1.2338, |
|
"eval_samples_per_second": 64.032, |
|
"eval_steps_per_second": 2.432, |
|
"step": 248 |
|
}, |
|
{ |
|
"epoch": 71.42857142857143, |
|
"grad_norm": 4.775552749633789, |
|
"learning_rate": 9.259259259259259e-06, |
|
"loss": 0.1124, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 72.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.8063094019889832, |
|
"eval_runtime": 1.2431, |
|
"eval_samples_per_second": 63.551, |
|
"eval_steps_per_second": 2.413, |
|
"step": 252 |
|
}, |
|
{ |
|
"epoch": 72.85714285714286, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.7436769604682922, |
|
"eval_runtime": 1.2512, |
|
"eval_samples_per_second": 63.141, |
|
"eval_steps_per_second": 2.398, |
|
"step": 255 |
|
}, |
|
{ |
|
"epoch": 74.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.8528115153312683, |
|
"eval_runtime": 1.2483, |
|
"eval_samples_per_second": 63.285, |
|
"eval_steps_per_second": 2.403, |
|
"step": 259 |
|
}, |
|
{ |
|
"epoch": 74.28571428571429, |
|
"grad_norm": 3.7733418941497803, |
|
"learning_rate": 7.4074074074074075e-06, |
|
"loss": 0.1036, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 74.85714285714286, |
|
"eval_accuracy": 0.810126582278481, |
|
"eval_loss": 0.9348064661026001, |
|
"eval_runtime": 1.2578, |
|
"eval_samples_per_second": 62.809, |
|
"eval_steps_per_second": 2.385, |
|
"step": 262 |
|
}, |
|
{ |
|
"epoch": 76.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.8077890872955322, |
|
"eval_runtime": 1.2522, |
|
"eval_samples_per_second": 63.09, |
|
"eval_steps_per_second": 2.396, |
|
"step": 266 |
|
}, |
|
{ |
|
"epoch": 76.85714285714286, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.7697485685348511, |
|
"eval_runtime": 1.231, |
|
"eval_samples_per_second": 64.173, |
|
"eval_steps_per_second": 2.437, |
|
"step": 269 |
|
}, |
|
{ |
|
"epoch": 77.14285714285714, |
|
"grad_norm": 6.019745826721191, |
|
"learning_rate": 5.555555555555556e-06, |
|
"loss": 0.1057, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 78.0, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.8040142059326172, |
|
"eval_runtime": 1.2413, |
|
"eval_samples_per_second": 63.641, |
|
"eval_steps_per_second": 2.417, |
|
"step": 273 |
|
}, |
|
{ |
|
"epoch": 78.85714285714286, |
|
"eval_accuracy": 0.8481012658227848, |
|
"eval_loss": 0.8197168111801147, |
|
"eval_runtime": 1.2521, |
|
"eval_samples_per_second": 63.095, |
|
"eval_steps_per_second": 2.396, |
|
"step": 276 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"grad_norm": 5.121321201324463, |
|
"learning_rate": 3.7037037037037037e-06, |
|
"loss": 0.099, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 80.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.8256491422653198, |
|
"eval_runtime": 1.2518, |
|
"eval_samples_per_second": 63.109, |
|
"eval_steps_per_second": 2.397, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 80.85714285714286, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.8057142496109009, |
|
"eval_runtime": 1.2389, |
|
"eval_samples_per_second": 63.765, |
|
"eval_steps_per_second": 2.421, |
|
"step": 283 |
|
}, |
|
{ |
|
"epoch": 82.0, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7797062397003174, |
|
"eval_runtime": 1.2603, |
|
"eval_samples_per_second": 62.684, |
|
"eval_steps_per_second": 2.38, |
|
"step": 287 |
|
}, |
|
{ |
|
"epoch": 82.85714285714286, |
|
"grad_norm": 4.244006156921387, |
|
"learning_rate": 1.8518518518518519e-06, |
|
"loss": 0.0927, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 82.85714285714286, |
|
"eval_accuracy": 0.8354430379746836, |
|
"eval_loss": 0.7806990742683411, |
|
"eval_runtime": 1.2548, |
|
"eval_samples_per_second": 62.959, |
|
"eval_steps_per_second": 2.391, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 84.0, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.7956676483154297, |
|
"eval_runtime": 1.2427, |
|
"eval_samples_per_second": 63.572, |
|
"eval_steps_per_second": 2.414, |
|
"step": 294 |
|
}, |
|
{ |
|
"epoch": 84.85714285714286, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.8031200766563416, |
|
"eval_runtime": 1.241, |
|
"eval_samples_per_second": 63.659, |
|
"eval_steps_per_second": 2.417, |
|
"step": 297 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"grad_norm": 3.5336740016937256, |
|
"learning_rate": 0.0, |
|
"loss": 0.0995, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"eval_accuracy": 0.8227848101265823, |
|
"eval_loss": 0.806117594242096, |
|
"eval_runtime": 1.2702, |
|
"eval_samples_per_second": 62.195, |
|
"eval_steps_per_second": 2.362, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"step": 300, |
|
"total_flos": 2.9349165326823014e+18, |
|
"train_loss": 0.24419225533803304, |
|
"train_runtime": 1773.1424, |
|
"train_samples_per_second": 24.927, |
|
"train_steps_per_second": 0.169 |
|
}, |
|
{ |
|
"epoch": 85.71428571428571, |
|
"eval_accuracy": 0.8607594936708861, |
|
"eval_loss": 0.3405309021472931, |
|
"eval_runtime": 1.2249, |
|
"eval_samples_per_second": 64.497, |
|
"eval_steps_per_second": 2.449, |
|
"step": 300 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 300, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 100, |
|
"save_steps": 500, |
|
"total_flos": 2.9349165326823014e+18, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|