CoolFace
Modelpublic

asun17904/multiberts-seed_2-step_2000k_stereoset_classifieronly

sourceHugging Faceapache-2.0updated 4y agoView on Hugging Face
0likes9downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

multiberts-seed2-step2000kstereosetclassifieronly

This model is a fine-tuned version of google/multiberts-seed_2-step_2000k on the stereoset dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.6815
  • —Accuracy: 0.5691
  • —Tp: 0.3077
  • —Tn: 0.2614
  • —Fp: 0.2410
  • —Fn: 0.1900

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • —learning_rate: 5e-05
  • —trainbatchsize: 64
  • —evalbatchsize: 64
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —num_epochs: 50

Training results

Training LossEpochStepValidation LossAccuracyTpTnFpFn
0.71420.43200.68670.54320.34070.20250.29980.1570
0.71730.85400.68690.53380.26450.26920.23310.2331
0.70041.28600.68670.54470.30850.23630.26610.1892
0.70471.7800.68710.54320.29980.24330.25900.1978
0.69442.131000.69070.51180.13500.37680.12560.3626
0.68852.551200.68670.53920.29430.24490.25750.2033
0.70542.981400.68750.52830.24960.27860.22370.2480
0.69073.41600.68720.52980.25200.27790.22450.2457
0.69933.831800.68660.54000.31790.22210.28020.1797
0.70324.262000.68900.52980.22680.30300.19940.2708
0.70154.682200.68790.53300.25900.27390.22840.2386
0.69695.112400.68650.54790.34460.20330.29910.1531
0.6955.532600.68570.54080.30850.23230.27000.1892
0.69435.962800.68670.52830.25590.27240.23000.2418
0.70086.383000.69020.51730.10130.41600.08630.3964
0.70376.813200.68590.53380.28490.24880.25350.2127
0.69677.233400.68610.55570.37830.17740.32500.1193
0.69227.663600.68560.53770.28180.25590.24650.2159
0.69518.093800.68870.51880.12170.39720.10520.3760
0.68938.514000.68600.54240.24410.29830.20410.2535
0.69928.944200.68570.53850.27320.26530.23700.2245
0.68219.364400.68540.55100.32650.22450.27790.1711
0.70069.794600.68550.53610.29040.24570.25670.2072
0.693410.214800.68640.53920.25820.28100.22140.2394
0.693510.645000.68940.52040.15380.36660.13580.3438
0.696111.065200.68650.54160.33590.20570.29670.1617
0.692511.495400.68730.53920.24100.29830.20410.2567
0.697211.915600.68750.52040.19390.32650.17580.3038
0.693512.345800.68470.55180.35240.19940.30300.1452
0.684712.776000.68840.51650.10750.40890.09340.3901
0.691213.196200.68680.54320.24570.29750.20490.2520
0.695713.626400.68720.52280.18840.33440.16800.3093
0.703614.046600.68530.55490.38460.17030.33200.1130
0.694814.476800.68520.55100.36030.19070.31160.1374
0.698114.897000.68600.53850.26370.27470.22760.2339
0.69515.327200.68440.55810.32500.23310.26920.1727
0.68815.747400.68370.56120.37830.18290.31950.1193
0.700916.177600.68690.52900.13500.39400.10830.3626
0.68716.67800.68680.53220.13970.39250.10990.3579
0.687617.028000.68440.55340.33200.22140.28100.1656
0.694117.458200.68460.56120.30060.26060.24180.1970
0.697217.878400.68350.56360.36730.19620.30610.1303
0.691918.38600.68360.55650.33360.22290.27940.1641
0.686118.728800.68290.56360.35400.20960.29280.1436
0.70119.159000.68550.53300.19310.33990.16250.3046
0.689819.579200.68600.53220.19700.33520.16720.3006
0.690520.09400.68510.55810.29360.26450.23780.2041
0.685820.439600.68480.55570.28960.26610.23630.2080
0.68920.859800.68490.54790.21190.33590.16640.2857
0.701621.2810000.68300.56510.34070.22450.27790.1570
0.68621.710200.68290.56750.34620.22140.28100.1515
0.690822.1310400.68390.54400.22610.31790.18450.2716
0.687122.5510600.68350.56280.28180.28100.22140.2159
0.702922.9810800.68300.56830.31000.25820.24410.1876
0.690623.411000.68280.56670.32890.23780.26450.1688
0.686423.8311200.68290.56120.36730.19390.30850.1303
0.691824.2611400.68330.56590.30140.26450.23780.1962
0.693824.6811600.68340.56280.33280.23000.27240.1648
0.686425.1111800.68380.55650.25120.30530.19700.2465
0.69825.5312000.68290.56750.29980.26770.23470.1978
0.70225.9612200.68240.56040.34690.21350.28890.1507
0.699626.3812400.68230.55970.39170.16800.33440.1060
0.694626.8112600.68270.56590.28810.27790.22450.2096
0.690827.2312800.68310.56360.27160.29200.21040.2261
0.700927.6613000.68290.56590.33280.23310.26920.1648
0.688528.0913200.68290.56990.31950.25040.25200.1782
0.685228.5113400.68270.56910.30060.26840.23390.1970
0.687928.9413600.68240.57060.29830.27240.23000.1994
0.684829.3613800.68240.56750.27630.29120.21110.2214
0.685729.7914000.68200.56510.33360.23160.27080.1641
0.690930.2114200.68190.56280.34930.21350.28890.1484
0.686530.6414400.68190.55970.34690.21270.28960.1507
0.696231.0614600.68160.56440.35160.21270.28960.1460
0.695431.4914800.68170.56990.33670.23310.26920.1609
0.681531.9115000.68170.56830.33280.23550.26690.1648
0.69232.3415200.68180.57220.31870.25350.24880.1790
0.690732.7715400.68130.56510.34220.22290.27940.1554
0.693633.1915600.68180.56590.29430.27160.23080.2033
0.696533.6215800.68250.56120.25670.30460.19780.2410
0.681134.0416000.68220.56440.28100.28340.21900.2166
0.692634.4716200.68200.56510.29360.27160.23080.2041
0.684334.8916400.68170.55890.33520.22370.27860.1625
0.690235.3216600.68180.57300.32970.24330.25900.1680
0.686835.7416800.68220.56590.29590.27000.23230.2017
0.682536.1717000.68270.55420.25040.30380.19860.2473
0.688836.617200.68280.55490.25270.30220.20020.2449
0.683537.0217400.68240.56510.29750.26770.23470.2002
0.691737.4517600.68200.56440.33990.22450.27790.1578
0.6937.8717800.68240.56990.30930.26060.24180.1884
0.68438.318000.68220.57300.31870.25430.24800.1790
0.681938.7218200.68200.57540.33200.24330.25900.1656
0.692439.1518400.68290.55100.23940.31160.19070.2582
0.686839.5718600.68340.55260.21110.34140.16090.2865
0.684240.018800.68360.55180.20090.35090.15150.2967
0.688340.4319000.68260.55650.26220.29430.20800.2355
0.678940.8519200.68250.55490.25900.29590.20640.2386
0.699241.2819400.68210.56990.31160.25820.24410.1860
0.682741.719600.68220.56040.29750.26300.23940.2002
0.688442.1319800.68170.57850.34070.23780.26450.1570
0.690242.5520000.68180.56360.29980.26370.23860.1978
0.685442.9820200.68170.56990.30610.26370.23860.1915
0.684543.420400.68150.57930.33050.24880.25350.1672
0.691243.8320600.68130.56830.34070.22760.27470.1570
0.682344.2620800.68140.57460.31630.25820.24410.1813
0.67844.6821000.68140.57060.30770.26300.23940.1900
0.685745.1121200.68130.57380.31400.25980.24250.1837
0.687445.5321400.68130.57690.33120.24570.25670.1664
0.686445.9621600.68140.57610.31790.25820.24410.1797
0.686546.3821800.68150.57220.31480.25750.24490.1829
0.684346.8122000.68150.56830.30460.26370.23860.1931
0.689947.2322200.68160.56440.30300.26140.24100.1947
0.688647.6622400.68160.56510.29590.26920.23310.2017
0.689748.0922600.68160.56510.29980.26530.23700.1978
0.684748.5122800.68160.56510.30060.26450.23780.1970
0.688348.9423000.68150.56990.30610.26370.23860.1915
0.691349.3623200.68150.57060.30930.26140.24100.1884
0.684949.7923400.68150.56910.30770.26140.24100.1900

Framework versions

  • —Transformers 4.26.1
  • —Pytorch 1.13.1
  • —Datasets 2.10.1
  • —Tokenizers 0.13.2