albert-tiny-spanish-finetuned-ner / trainer_state.json
sdonoso's picture
best albert-tiny model finetuned on ner
9a0d740
{
"best_metric": 0.1745937019586563,
"best_model_checkpoint": "/home/sdonoso/data/all_results/ner-c/albert-tiny/epochs_4_bs_16_lr_5e-5/checkpoint-1400",
"epoch": 4.0,
"global_step": 2084,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"eval_accuracy": 0.863771172477427,
"eval_f1": 0.24746022618363045,
"eval_loss": 0.4690728187561035,
"eval_precision": 0.3093697579678888,
"eval_recall": 0.20619709311611564,
"eval_runtime": 1.4686,
"eval_samples_per_second": 1304.616,
"eval_steps_per_second": 81.709,
"step": 50
},
{
"epoch": 0.19,
"eval_accuracy": 0.9084722862980293,
"eval_f1": 0.5024801196756161,
"eval_loss": 0.3399835526943207,
"eval_precision": 0.4954968944099379,
"eval_recall": 0.5096629931320875,
"eval_runtime": 1.4666,
"eval_samples_per_second": 1306.423,
"eval_steps_per_second": 81.822,
"step": 100
},
{
"epoch": 0.29,
"eval_accuracy": 0.9171060436301325,
"eval_f1": 0.5566067486597288,
"eval_loss": 0.29001280665397644,
"eval_precision": 0.5495874202086253,
"eval_recall": 0.5638076984507268,
"eval_runtime": 1.5018,
"eval_samples_per_second": 1275.795,
"eval_steps_per_second": 79.904,
"step": 150
},
{
"epoch": 0.38,
"eval_accuracy": 0.9215053054768338,
"eval_f1": 0.5868562353475006,
"eval_loss": 0.26263922452926636,
"eval_precision": 0.5573111174949728,
"eval_recall": 0.6197093116115636,
"eval_runtime": 1.4723,
"eval_samples_per_second": 1301.379,
"eval_steps_per_second": 81.506,
"step": 200
},
{
"epoch": 0.48,
"eval_accuracy": 0.9255750346009358,
"eval_f1": 0.6203032300593275,
"eval_loss": 0.25319042801856995,
"eval_precision": 0.6406808510638298,
"eval_recall": 0.6011819198211148,
"eval_runtime": 1.469,
"eval_samples_per_second": 1304.326,
"eval_steps_per_second": 81.691,
"step": 250
},
{
"epoch": 0.58,
"eval_accuracy": 0.9309793712515653,
"eval_f1": 0.6298792914249882,
"eval_loss": 0.22614283859729767,
"eval_precision": 0.6184392796675389,
"eval_recall": 0.6417505190864079,
"eval_runtime": 1.4706,
"eval_samples_per_second": 1302.847,
"eval_steps_per_second": 81.598,
"step": 300
},
{
"epoch": 0.67,
"eval_accuracy": 0.9334838199433204,
"eval_f1": 0.6383647798742139,
"eval_loss": 0.21767833828926086,
"eval_precision": 0.6285802755844558,
"eval_recall": 0.6484587126657083,
"eval_runtime": 1.47,
"eval_samples_per_second": 1303.418,
"eval_steps_per_second": 81.634,
"step": 350
},
{
"epoch": 0.77,
"eval_accuracy": 0.9345383246556384,
"eval_f1": 0.6418154877953678,
"eval_loss": 0.21528242528438568,
"eval_precision": 0.6270953977445901,
"eval_recall": 0.657243251876697,
"eval_runtime": 1.5017,
"eval_samples_per_second": 1275.883,
"eval_steps_per_second": 79.909,
"step": 400
},
{
"epoch": 0.86,
"eval_accuracy": 0.9356422592763461,
"eval_f1": 0.6500153704272977,
"eval_loss": 0.21207766234874725,
"eval_precision": 0.626425714708932,
"eval_recall": 0.6754512058776553,
"eval_runtime": 1.4752,
"eval_samples_per_second": 1298.836,
"eval_steps_per_second": 81.347,
"step": 450
},
{
"epoch": 0.96,
"learning_rate": 3.807581573896353e-05,
"loss": 0.2732,
"step": 500
},
{
"epoch": 0.96,
"eval_accuracy": 0.9400415211230475,
"eval_f1": 0.6775403935982847,
"eval_loss": 0.19518034160137177,
"eval_precision": 0.6507796410709032,
"eval_recall": 0.7065963903529787,
"eval_runtime": 1.4979,
"eval_samples_per_second": 1279.083,
"eval_steps_per_second": 80.11,
"step": 500
},
{
"epoch": 1.06,
"eval_accuracy": 0.9415738482831345,
"eval_f1": 0.68428873074216,
"eval_loss": 0.18678122758865356,
"eval_precision": 0.6703953417100827,
"eval_recall": 0.6987701645104616,
"eval_runtime": 1.473,
"eval_samples_per_second": 1300.753,
"eval_steps_per_second": 81.467,
"step": 550
},
{
"epoch": 1.15,
"eval_accuracy": 0.9421670071838133,
"eval_f1": 0.6916773778920309,
"eval_loss": 0.19731050729751587,
"eval_precision": 0.6958138031356069,
"eval_recall": 0.6875898418782942,
"eval_runtime": 1.4718,
"eval_samples_per_second": 1301.838,
"eval_steps_per_second": 81.535,
"step": 600
},
{
"epoch": 1.25,
"eval_accuracy": 0.9407500164766361,
"eval_f1": 0.6780254777070065,
"eval_loss": 0.1952008754014969,
"eval_precision": 0.6759803143356088,
"eval_recall": 0.6800830538252676,
"eval_runtime": 1.5038,
"eval_samples_per_second": 1274.147,
"eval_steps_per_second": 79.8,
"step": 650
},
{
"epoch": 1.34,
"eval_accuracy": 0.9399097080340078,
"eval_f1": 0.689172175007869,
"eval_loss": 0.19703227281570435,
"eval_precision": 0.679230649914689,
"eval_recall": 0.6994090400894426,
"eval_runtime": 1.4697,
"eval_samples_per_second": 1303.698,
"eval_steps_per_second": 81.651,
"step": 700
},
{
"epoch": 1.44,
"eval_accuracy": 0.9425789230870626,
"eval_f1": 0.697266881028939,
"eval_loss": 0.19310028851032257,
"eval_precision": 0.7018935102767438,
"eval_recall": 0.6927008465101422,
"eval_runtime": 1.4705,
"eval_samples_per_second": 1302.989,
"eval_steps_per_second": 81.607,
"step": 750
},
{
"epoch": 1.54,
"eval_accuracy": 0.9420351940947737,
"eval_f1": 0.689247483623582,
"eval_loss": 0.1976134181022644,
"eval_precision": 0.6894677960684034,
"eval_recall": 0.6890273119310014,
"eval_runtime": 1.4714,
"eval_samples_per_second": 1302.156,
"eval_steps_per_second": 81.555,
"step": 800
},
{
"epoch": 1.63,
"eval_accuracy": 0.9442265867000593,
"eval_f1": 0.6969575339166734,
"eval_loss": 0.18812265992164612,
"eval_precision": 0.7006132989025178,
"eval_recall": 0.6933397220891231,
"eval_runtime": 1.4736,
"eval_samples_per_second": 1300.193,
"eval_steps_per_second": 81.432,
"step": 850
},
{
"epoch": 1.73,
"eval_accuracy": 0.9432050352600013,
"eval_f1": 0.6990461049284578,
"eval_loss": 0.18850077688694,
"eval_precision": 0.69583794904257,
"eval_recall": 0.702283980194857,
"eval_runtime": 1.5056,
"eval_samples_per_second": 1272.624,
"eval_steps_per_second": 79.705,
"step": 900
},
{
"epoch": 1.82,
"eval_accuracy": 0.9444078296974889,
"eval_f1": 0.7056041617403641,
"eval_loss": 0.18260571360588074,
"eval_precision": 0.6965452847805789,
"eval_recall": 0.7149017728797317,
"eval_runtime": 1.4756,
"eval_samples_per_second": 1298.431,
"eval_steps_per_second": 81.321,
"step": 950
},
{
"epoch": 1.92,
"learning_rate": 2.607965451055662e-05,
"loss": 0.1235,
"step": 1000
},
{
"epoch": 1.92,
"eval_accuracy": 0.9451163250510776,
"eval_f1": 0.7066519546027743,
"eval_loss": 0.18071572482585907,
"eval_precision": 0.6975260619262487,
"eval_recall": 0.7160198051429484,
"eval_runtime": 1.4701,
"eval_samples_per_second": 1303.293,
"eval_steps_per_second": 81.626,
"step": 1000
},
{
"epoch": 2.02,
"eval_accuracy": 0.9455282409543267,
"eval_f1": 0.7122369446609509,
"eval_loss": 0.17818446457386017,
"eval_precision": 0.6955396559598113,
"eval_recall": 0.7297556300910397,
"eval_runtime": 1.4738,
"eval_samples_per_second": 1300.036,
"eval_steps_per_second": 81.422,
"step": 1050
},
{
"epoch": 2.11,
"eval_accuracy": 0.9441277268832795,
"eval_f1": 0.7015257469802925,
"eval_loss": 0.18732576072216034,
"eval_precision": 0.6980863514154674,
"eval_recall": 0.7049992014055263,
"eval_runtime": 1.4745,
"eval_samples_per_second": 1299.388,
"eval_steps_per_second": 81.381,
"step": 1100
},
{
"epoch": 2.21,
"eval_accuracy": 0.9422493903644632,
"eval_f1": 0.6907414521705724,
"eval_loss": 0.19080577790737152,
"eval_precision": 0.6655315368670418,
"eval_recall": 0.7179364318798914,
"eval_runtime": 1.5005,
"eval_samples_per_second": 1276.934,
"eval_steps_per_second": 79.975,
"step": 1150
},
{
"epoch": 2.3,
"eval_accuracy": 0.945956633493706,
"eval_f1": 0.712525347059741,
"eval_loss": 0.181381955742836,
"eval_precision": 0.696235329980186,
"eval_recall": 0.7295959111962945,
"eval_runtime": 1.4739,
"eval_samples_per_second": 1299.963,
"eval_steps_per_second": 81.417,
"step": 1200
},
{
"epoch": 2.4,
"eval_accuracy": 0.9461873063995254,
"eval_f1": 0.7136230091232412,
"eval_loss": 0.17689131200313568,
"eval_precision": 0.6915929866626704,
"eval_recall": 0.7371026992493211,
"eval_runtime": 1.472,
"eval_samples_per_second": 1301.631,
"eval_steps_per_second": 81.522,
"step": 1250
},
{
"epoch": 2.5,
"eval_accuracy": 0.9468463718447242,
"eval_f1": 0.7252625760088447,
"eval_loss": 0.1800052374601364,
"eval_precision": 0.7172758512964699,
"eval_recall": 0.7334291646701805,
"eval_runtime": 1.4724,
"eval_samples_per_second": 1301.261,
"eval_steps_per_second": 81.499,
"step": 1300
},
{
"epoch": 2.59,
"eval_accuracy": 0.9469617082976339,
"eval_f1": 0.7245991396167384,
"eval_loss": 0.17777691781520844,
"eval_precision": 0.7099938687921521,
"eval_recall": 0.7398179204599904,
"eval_runtime": 1.4757,
"eval_samples_per_second": 1298.39,
"eval_steps_per_second": 81.319,
"step": 1350
},
{
"epoch": 2.69,
"eval_accuracy": 0.9473241942924933,
"eval_f1": 0.7241245438310427,
"eval_loss": 0.1745937019586563,
"eval_precision": 0.7045935327893623,
"eval_recall": 0.7447692061970931,
"eval_runtime": 1.5088,
"eval_samples_per_second": 1269.882,
"eval_steps_per_second": 79.533,
"step": 1400
},
{
"epoch": 2.78,
"eval_accuracy": 0.9480985961906018,
"eval_f1": 0.7267085687007716,
"eval_loss": 0.17996077239513397,
"eval_precision": 0.724001268230818,
"eval_recall": 0.7294361923015493,
"eval_runtime": 1.4708,
"eval_samples_per_second": 1302.711,
"eval_steps_per_second": 81.589,
"step": 1450
},
{
"epoch": 2.88,
"learning_rate": 1.4083493282149713e-05,
"loss": 0.083,
"step": 1500
},
{
"epoch": 2.88,
"eval_accuracy": 0.9467475120279444,
"eval_f1": 0.714475847490801,
"eval_loss": 0.1799446940422058,
"eval_precision": 0.7007063882063882,
"eval_recall": 0.7287973167225683,
"eval_runtime": 1.4712,
"eval_samples_per_second": 1302.299,
"eval_steps_per_second": 81.564,
"step": 1500
},
{
"epoch": 2.98,
"eval_accuracy": 0.9470440914782838,
"eval_f1": 0.7189552956257009,
"eval_loss": 0.18403704464435577,
"eval_precision": 0.7212666773830574,
"eval_recall": 0.7166586807219294,
"eval_runtime": 1.4732,
"eval_samples_per_second": 1300.543,
"eval_steps_per_second": 81.454,
"step": 1550
},
{
"epoch": 3.07,
"eval_accuracy": 0.9481150728267317,
"eval_f1": 0.7218129404228059,
"eval_loss": 0.18616726994514465,
"eval_precision": 0.7237835233659868,
"eval_recall": 0.7198530586168344,
"eval_runtime": 1.4726,
"eval_samples_per_second": 1301.059,
"eval_steps_per_second": 81.486,
"step": 1600
},
{
"epoch": 3.17,
"eval_accuracy": 0.9484446055493311,
"eval_f1": 0.7236070954129874,
"eval_loss": 0.18031327426433563,
"eval_precision": 0.7083843329253366,
"eval_recall": 0.7394984826704999,
"eval_runtime": 1.5062,
"eval_samples_per_second": 1272.051,
"eval_steps_per_second": 79.669,
"step": 1650
},
{
"epoch": 3.26,
"eval_accuracy": 0.9472582877479734,
"eval_f1": 0.7165758509042092,
"eval_loss": 0.1847413182258606,
"eval_precision": 0.7086847860043737,
"eval_recall": 0.7246446254591918,
"eval_runtime": 1.4724,
"eval_samples_per_second": 1301.279,
"eval_steps_per_second": 81.5,
"step": 1700
},
{
"epoch": 3.36,
"eval_accuracy": 0.9482139326435115,
"eval_f1": 0.7222266150075118,
"eval_loss": 0.18235260248184204,
"eval_precision": 0.7151581584716568,
"eval_recall": 0.7294361923015493,
"eval_runtime": 1.4742,
"eval_samples_per_second": 1299.703,
"eval_steps_per_second": 81.401,
"step": 1750
},
{
"epoch": 3.45,
"eval_accuracy": 0.9486917550912806,
"eval_f1": 0.7286566227244193,
"eval_loss": 0.17917561531066895,
"eval_precision": 0.7161807805028536,
"eval_recall": 0.7415748283021881,
"eval_runtime": 1.4683,
"eval_samples_per_second": 1304.936,
"eval_steps_per_second": 81.729,
"step": 1800
},
{
"epoch": 3.55,
"eval_accuracy": 0.9481480260989916,
"eval_f1": 0.7228443887540136,
"eval_loss": 0.1824788898229599,
"eval_precision": 0.7091272280270436,
"eval_recall": 0.7371026992493211,
"eval_runtime": 1.471,
"eval_samples_per_second": 1302.492,
"eval_steps_per_second": 81.576,
"step": 1850
},
{
"epoch": 3.65,
"eval_accuracy": 0.9487411849996704,
"eval_f1": 0.7265369649805447,
"eval_loss": 0.17811530828475952,
"eval_precision": 0.708453483077857,
"eval_recall": 0.7455678006708194,
"eval_runtime": 1.5062,
"eval_samples_per_second": 1272.042,
"eval_steps_per_second": 79.669,
"step": 1900
},
{
"epoch": 3.74,
"eval_accuracy": 0.9482963158241613,
"eval_f1": 0.7255009823182711,
"eval_loss": 0.18192069232463837,
"eval_precision": 0.7141089108910891,
"eval_recall": 0.7372624181440665,
"eval_runtime": 1.4716,
"eval_samples_per_second": 1301.966,
"eval_steps_per_second": 81.543,
"step": 1950
},
{
"epoch": 3.84,
"learning_rate": 2.0873320537428026e-06,
"loss": 0.0639,
"step": 2000
},
{
"epoch": 3.84,
"eval_accuracy": 0.9487576616358004,
"eval_f1": 0.7273722057214319,
"eval_loss": 0.17967551946640015,
"eval_precision": 0.7122302158273381,
"eval_recall": 0.7431720172496407,
"eval_runtime": 1.4718,
"eval_samples_per_second": 1301.808,
"eval_steps_per_second": 81.533,
"step": 2000
},
{
"epoch": 3.93,
"eval_accuracy": 0.9482468859157714,
"eval_f1": 0.7234075810863619,
"eval_loss": 0.18143223226070404,
"eval_precision": 0.7082950719314356,
"eval_recall": 0.7391790448810094,
"eval_runtime": 1.4715,
"eval_samples_per_second": 1302.101,
"eval_steps_per_second": 81.551,
"step": 2050
},
{
"epoch": 4.0,
"step": 2084,
"total_flos": 20529237663360.0,
"train_loss": 0.1328041532904539,
"train_runtime": 90.5157,
"train_samples_per_second": 367.848,
"train_steps_per_second": 23.024
}
],
"max_steps": 2084,
"num_train_epochs": 4,
"total_flos": 20529237663360.0,
"trial_name": null,
"trial_params": null
}