SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/google_t5
/t5_small_twitter
/trainer_state.json
| { | |
| "best_metric": 0.49757373332977295, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_t5/t5_small_twitter/checkpoint-200", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 816, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.9986526370048523, | |
| "learning_rate": 0.0004938725490196079, | |
| "loss": 0.6031, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.9707651734352112, | |
| "learning_rate": 0.0004877450980392157, | |
| "loss": 0.5921, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.3216536045074463, | |
| "learning_rate": 0.00048161764705882356, | |
| "loss": 0.5945, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.8769760131835938, | |
| "learning_rate": 0.00047549019607843134, | |
| "loss": 0.5244, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 1.3491202592849731, | |
| "learning_rate": 0.0004693627450980392, | |
| "loss": 0.4779, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.7564338235294118, | |
| "eval_f1_macro": 0.7177028105371891, | |
| "eval_f1_micro": 0.7564338235294118, | |
| "eval_loss": 0.5150883793830872, | |
| "eval_runtime": 0.9275, | |
| "eval_samples_per_second": 1173.066, | |
| "eval_steps_per_second": 36.658, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.29738450050354, | |
| "learning_rate": 0.0004632352941176471, | |
| "loss": 0.4934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 3.7542002201080322, | |
| "learning_rate": 0.0004571078431372549, | |
| "loss": 0.5007, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.893402099609375, | |
| "learning_rate": 0.0004509803921568628, | |
| "loss": 0.4738, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.0344367027282715, | |
| "learning_rate": 0.00044485294117647056, | |
| "loss": 0.4594, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.5279059410095215, | |
| "learning_rate": 0.00043872549019607844, | |
| "loss": 0.5123, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.7527573529411765, | |
| "eval_f1_macro": 0.6986562303405088, | |
| "eval_f1_micro": 0.7527573529411765, | |
| "eval_loss": 0.5059758424758911, | |
| "eval_runtime": 0.8718, | |
| "eval_samples_per_second": 1248.006, | |
| "eval_steps_per_second": 39.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.4078208208084106, | |
| "learning_rate": 0.0004325980392156863, | |
| "loss": 0.516, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.224971055984497, | |
| "learning_rate": 0.0004264705882352941, | |
| "loss": 0.5001, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.39370858669281, | |
| "learning_rate": 0.000420343137254902, | |
| "loss": 0.4835, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.9512380361557007, | |
| "learning_rate": 0.0004142156862745098, | |
| "loss": 0.5218, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.359387755393982, | |
| "learning_rate": 0.00040808823529411766, | |
| "loss": 0.4617, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.7270220588235294, | |
| "eval_f1_macro": 0.614936097953347, | |
| "eval_f1_micro": 0.7270220588235294, | |
| "eval_loss": 0.5286622047424316, | |
| "eval_runtime": 0.9283, | |
| "eval_samples_per_second": 1172.017, | |
| "eval_steps_per_second": 36.626, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1093153953552246, | |
| "learning_rate": 0.0004019607843137255, | |
| "loss": 0.4961, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.4549243450164795, | |
| "learning_rate": 0.0003958333333333333, | |
| "loss": 0.5351, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 1.2358920574188232, | |
| "learning_rate": 0.0003897058823529412, | |
| "loss": 0.5129, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.4391891956329346, | |
| "learning_rate": 0.00038357843137254904, | |
| "loss": 0.4945, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.3604400157928467, | |
| "learning_rate": 0.0003774509803921569, | |
| "loss": 0.4942, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.7444852941176471, | |
| "eval_f1_macro": 0.6956024668695554, | |
| "eval_f1_micro": 0.7444852941176471, | |
| "eval_loss": 0.49757373332977295, | |
| "eval_runtime": 0.9302, | |
| "eval_samples_per_second": 1169.596, | |
| "eval_steps_per_second": 36.55, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.0586875677108765, | |
| "learning_rate": 0.0003713235294117647, | |
| "loss": 0.5041, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 5.99719762802124, | |
| "learning_rate": 0.00036519607843137254, | |
| "loss": 0.5088, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.8691691160202026, | |
| "learning_rate": 0.0003590686274509804, | |
| "loss": 0.4577, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.8101117610931396, | |
| "learning_rate": 0.00035294117647058826, | |
| "loss": 0.4562, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.3150193691253662, | |
| "learning_rate": 0.0003468137254901961, | |
| "loss": 0.4783, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.7573529411764706, | |
| "eval_f1_macro": 0.7124355755253073, | |
| "eval_f1_micro": 0.7573529411764706, | |
| "eval_loss": 0.49782079458236694, | |
| "eval_runtime": 0.8783, | |
| "eval_samples_per_second": 1238.797, | |
| "eval_steps_per_second": 38.712, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.0855648517608643, | |
| "learning_rate": 0.0003406862745098039, | |
| "loss": 0.4206, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.4224942922592163, | |
| "learning_rate": 0.00033455882352941176, | |
| "loss": 0.532, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.6312692165374756, | |
| "learning_rate": 0.0003284313725490196, | |
| "loss": 0.465, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.7025282382965088, | |
| "learning_rate": 0.0003223039215686275, | |
| "loss": 0.4146, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.1280652284622192, | |
| "learning_rate": 0.0003161764705882353, | |
| "loss": 0.4369, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_accuracy": 0.7601102941176471, | |
| "eval_f1_macro": 0.7124360123342397, | |
| "eval_f1_micro": 0.7601102941176471, | |
| "eval_loss": 0.5052167773246765, | |
| "eval_runtime": 0.8783, | |
| "eval_samples_per_second": 1238.751, | |
| "eval_steps_per_second": 38.711, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 1.285010814666748, | |
| "learning_rate": 0.00031004901960784314, | |
| "loss": 0.448, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.8714994788169861, | |
| "learning_rate": 0.00030392156862745097, | |
| "loss": 0.4204, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.7784018516540527, | |
| "learning_rate": 0.0002977941176470588, | |
| "loss": 0.4939, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.4992843866348267, | |
| "learning_rate": 0.0002916666666666667, | |
| "loss": 0.4632, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.7776957750320435, | |
| "learning_rate": 0.0002855392156862745, | |
| "loss": 0.439, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_accuracy": 0.7564338235294118, | |
| "eval_f1_macro": 0.7223909884910498, | |
| "eval_f1_micro": 0.7564338235294118, | |
| "eval_loss": 0.5092098116874695, | |
| "eval_runtime": 0.9311, | |
| "eval_samples_per_second": 1168.532, | |
| "eval_steps_per_second": 36.517, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.2894947528839111, | |
| "learning_rate": 0.00027941176470588236, | |
| "loss": 0.4339, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.262589693069458, | |
| "learning_rate": 0.0002732843137254902, | |
| "loss": 0.3804, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.4415295124053955, | |
| "learning_rate": 0.000267156862745098, | |
| "loss": 0.5278, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.8975765109062195, | |
| "learning_rate": 0.0002610294117647059, | |
| "loss": 0.4198, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 1.8955622911453247, | |
| "learning_rate": 0.00025490196078431374, | |
| "loss": 0.4417, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_accuracy": 0.7545955882352942, | |
| "eval_f1_macro": 0.6808490301741896, | |
| "eval_f1_micro": 0.7545955882352942, | |
| "eval_loss": 0.5228371620178223, | |
| "eval_runtime": 0.8846, | |
| "eval_samples_per_second": 1229.999, | |
| "eval_steps_per_second": 38.437, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 1.5438319444656372, | |
| "learning_rate": 0.00024877450980392157, | |
| "loss": 0.4897, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 2.117488384246826, | |
| "learning_rate": 0.0002426470588235294, | |
| "loss": 0.4292, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 1.4254924058914185, | |
| "learning_rate": 0.00023651960784313726, | |
| "loss": 0.4201, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 1.8626124858856201, | |
| "learning_rate": 0.0002303921568627451, | |
| "loss": 0.4013, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 1.8252973556518555, | |
| "learning_rate": 0.00022426470588235296, | |
| "loss": 0.47, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_accuracy": 0.7693014705882353, | |
| "eval_f1_macro": 0.7234537896394413, | |
| "eval_f1_micro": 0.7693014705882353, | |
| "eval_loss": 0.5087092518806458, | |
| "eval_runtime": 0.9392, | |
| "eval_samples_per_second": 1158.396, | |
| "eval_steps_per_second": 36.2, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 1.173530101776123, | |
| "learning_rate": 0.0002181372549019608, | |
| "loss": 0.4267, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 1.2974258661270142, | |
| "learning_rate": 0.00021200980392156862, | |
| "loss": 0.429, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.577986717224121, | |
| "learning_rate": 0.00020588235294117645, | |
| "loss": 0.3808, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.0223047733306885, | |
| "learning_rate": 0.00019975490196078434, | |
| "loss": 0.4225, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.9677762985229492, | |
| "learning_rate": 0.00019362745098039217, | |
| "loss": 0.4415, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "eval_accuracy": 0.7647058823529411, | |
| "eval_f1_macro": 0.7262032085561497, | |
| "eval_f1_micro": 0.7647058823529411, | |
| "eval_loss": 0.510553240776062, | |
| "eval_runtime": 0.9392, | |
| "eval_samples_per_second": 1158.407, | |
| "eval_steps_per_second": 36.2, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.1366698741912842, | |
| "learning_rate": 0.0001875, | |
| "loss": 0.4622, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 1.5312350988388062, | |
| "learning_rate": 0.00018137254901960784, | |
| "loss": 0.4507, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.2124278545379639, | |
| "learning_rate": 0.00017524509803921567, | |
| "loss": 0.4495, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 1.3845272064208984, | |
| "learning_rate": 0.00016911764705882356, | |
| "loss": 0.4161, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 1.2308210134506226, | |
| "learning_rate": 0.0001629901960784314, | |
| "loss": 0.4297, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_accuracy": 0.7628676470588235, | |
| "eval_f1_macro": 0.7291254940711462, | |
| "eval_f1_micro": 0.7628676470588235, | |
| "eval_loss": 0.5022936463356018, | |
| "eval_runtime": 0.9417, | |
| "eval_samples_per_second": 1155.3, | |
| "eval_steps_per_second": 36.103, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 1.803513765335083, | |
| "learning_rate": 0.00015686274509803922, | |
| "loss": 0.4231, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.3074411153793335, | |
| "learning_rate": 0.00015073529411764705, | |
| "loss": 0.3723, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 1.8602393865585327, | |
| "learning_rate": 0.0001446078431372549, | |
| "loss": 0.4003, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 1.486925721168518, | |
| "learning_rate": 0.00013848039215686274, | |
| "loss": 0.3796, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 1.5388500690460205, | |
| "learning_rate": 0.0001323529411764706, | |
| "loss": 0.4366, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "eval_accuracy": 0.7555147058823529, | |
| "eval_f1_macro": 0.7127019171206659, | |
| "eval_f1_micro": 0.7555147058823529, | |
| "eval_loss": 0.5225424766540527, | |
| "eval_runtime": 0.9394, | |
| "eval_samples_per_second": 1158.157, | |
| "eval_steps_per_second": 36.192, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.38102388381958, | |
| "learning_rate": 0.00012622549019607844, | |
| "loss": 0.4028, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 1.0879701375961304, | |
| "learning_rate": 0.00012009803921568628, | |
| "loss": 0.3619, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.56364107131958, | |
| "learning_rate": 0.00011397058823529411, | |
| "loss": 0.3883, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 1.285326600074768, | |
| "learning_rate": 0.00010784313725490197, | |
| "loss": 0.4058, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 1.1956150531768799, | |
| "learning_rate": 0.0001017156862745098, | |
| "loss": 0.3623, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "eval_accuracy": 0.7582720588235294, | |
| "eval_f1_macro": 0.7157047718685389, | |
| "eval_f1_micro": 0.7582720588235294, | |
| "eval_loss": 0.522619366645813, | |
| "eval_runtime": 0.9391, | |
| "eval_samples_per_second": 1158.576, | |
| "eval_steps_per_second": 36.206, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 1.5235141515731812, | |
| "learning_rate": 9.558823529411764e-05, | |
| "loss": 0.4147, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 1.760062575340271, | |
| "learning_rate": 8.94607843137255e-05, | |
| "loss": 0.4255, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.636400580406189, | |
| "learning_rate": 8.333333333333333e-05, | |
| "loss": 0.3687, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 1.4576902389526367, | |
| "learning_rate": 7.720588235294118e-05, | |
| "loss": 0.3764, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 1.5375081300735474, | |
| "learning_rate": 7.107843137254902e-05, | |
| "loss": 0.3337, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "eval_accuracy": 0.7573529411764706, | |
| "eval_f1_macro": 0.7143845832587554, | |
| "eval_f1_micro": 0.7573529411764706, | |
| "eval_loss": 0.5312902927398682, | |
| "eval_runtime": 0.9369, | |
| "eval_samples_per_second": 1161.294, | |
| "eval_steps_per_second": 36.29, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.7449744939804077, | |
| "learning_rate": 6.495098039215687e-05, | |
| "loss": 0.3967, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 1.1091108322143555, | |
| "learning_rate": 5.882352941176471e-05, | |
| "loss": 0.3403, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 1.650840163230896, | |
| "learning_rate": 5.2696078431372553e-05, | |
| "loss": 0.3709, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.3335641622543335, | |
| "learning_rate": 4.656862745098039e-05, | |
| "loss": 0.3852, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 1.7249501943588257, | |
| "learning_rate": 4.044117647058824e-05, | |
| "loss": 0.4158, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "eval_accuracy": 0.7601102941176471, | |
| "eval_f1_macro": 0.7210790741531978, | |
| "eval_f1_micro": 0.7601102941176471, | |
| "eval_loss": 0.5384809374809265, | |
| "eval_runtime": 0.9393, | |
| "eval_samples_per_second": 1158.291, | |
| "eval_steps_per_second": 36.197, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 1.5633273124694824, | |
| "learning_rate": 3.4313725490196084e-05, | |
| "loss": 0.3583, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 1.3377939462661743, | |
| "learning_rate": 2.8186274509803924e-05, | |
| "loss": 0.3841, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 1.5654672384262085, | |
| "learning_rate": 2.2058823529411766e-05, | |
| "loss": 0.3559, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.3034334182739258, | |
| "learning_rate": 1.593137254901961e-05, | |
| "loss": 0.3877, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 1.5207332372665405, | |
| "learning_rate": 9.803921568627451e-06, | |
| "loss": 0.4003, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "eval_accuracy": 0.7545955882352942, | |
| "eval_f1_macro": 0.7058253459511188, | |
| "eval_f1_micro": 0.7545955882352942, | |
| "eval_loss": 0.5349971055984497, | |
| "eval_runtime": 0.9377, | |
| "eval_samples_per_second": 1160.255, | |
| "eval_steps_per_second": 36.258, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 2.046905279159546, | |
| "learning_rate": 3.6764705882352942e-06, | |
| "loss": 0.3939, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 816, | |
| "total_flos": 888799123472384.0, | |
| "train_loss": 0.44335933526357013, | |
| "train_runtime": 92.5, | |
| "train_samples_per_second": 282.162, | |
| "train_steps_per_second": 8.822 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 816, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 888799123472384.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |