Swansen1993/RepositoryBertModels

1---2# Model Performances and overview metrics will be posted here for all trained models. This model card is used as a overview3# of the metrics of the different models. The tag is the model name, the corresponfing models can be found in the Swansen1993 folder.4# Model performances are from the evaluation phase/ test set5language:6- en7#base model was always the same8base_model:9- google-bert/bert-base-uncased10pipeline_tag: text-classification11 12tags:13- v1_lr2e-5_nowarmup_nodecay14epochs: 315average loss in each epoch:16- Average loss in epoch 1: 0.880517- Average loss in epoch 2: 0.383518- Average loss in epoch 3: 0.253719metrics:20- overall accuracy: 0,87621- irony22- precision: 0,9423- recall: 0,8324- f1-score: 0,8825- sample size: 13126- multipolarity27- precision: 0,7128- recall: 0,8929- f1-score: 0,7930- sample size: 15031- negation32- precision: 0,7233- recall: 0,5534- f1-score: 0,6335- sample size: 15036- negative37- precision: 0,8338- recall: 0,9239- f1-score: 0,8740- sample size: 22541- neutral42- precision: 0,8343- recall: 0,7344- f1-score: 0,7845- sample size: 15046- positive47- precision: 0,8948- recall: 0,9049- f1-score: 0,8950- sample size: 22551- sarcastic52- precision: 0,9753- recall: 0,9754- f1-score: 0,9755- sample size: 45056 57tags v2:58- v2_lr2e-5_nowarmup_0.01weight_decay59epochs v2: 360average loss in each epoch v2:61- Average loss in epoch 1: 0.946362- Average loss in epoch 2: 0.413363- Average loss in epoch 3: 0.268464metrics v2:65- overall accuracy: 0,86266- irony67- precision: 0,9568- recall: 0,7969- f1-score: 0,8670- sample size: 13171- multipolarity72- precision: 0,7373- recall: 0,9174- f1-score: 0,8175- sample size: 15076- negation77- precision: 0,6778- recall: 0,4979- f1-score: 0,5680- sample size: 15081- negative82- precision: 0,8283- recall: 0,9284- f1-score: 0,8785- sample size: 22586- neutral87- precision: 0,8188- recall: 0,7589- f1-score: 0,7890- sample size: 15091- positive92- precision: 0,9393- recall: 0,9094- f1-score: 0,9195- sample size: 22596- sarcastic97- precision: 0,9598- recall: 0,9899- f1-score: 0,96100- sample size: 450101 102tags v3:103- v3_lr2e-5_100warmupsteps_0.01weight_decay104epochs v3: 3105average loss in each epoch v3:106- Average loss in epoch 1: 1.0157107- Average loss in epoch 2: 0.4125108- Average loss in epoch 3: 0.2722109metrics v3:110- overall accuracy: 0,867111- irony112- precision: 0,97113- recall: 0,77114- f1-score: 0,86115- sample size: 131116- multipolarity117- precision: 0,72118- recall: 0,91119- f1-score: 0,80120- sample size: 150121- negation122- precision: 0,73123- recall: 0,56124- f1-score: 0,63125- sample size: 150126- negative127- precision: 0,83128- recall: 0,94129- f1-score: 0,88130- sample size: 225131- neutral132- precision: 0,84133- recall: 0,72134- f1-score: 0,78135- sample size: 150136- positive137- precision: 0,90138- recall: 0,89139- f1-score: 0,90140- sample size: 225141- sarcastic142- precision: 0,95143- recall: 0,98144- f1-score: 0,97145- sample size: 450146 147tags v4:148- v4_lr2e-5_100warmupsteps_nodecay149epochs v4: 3150average loss in each epoch v4:151- Average loss in epoch 1: 1.0900152- Average loss in epoch 2: 0.4439153- Average loss in epoch 3: 0.2898154metrics v4:155- overall accuracy: 0,860156- irony157- precision: 0,96158- recall: 0,79159- f1-score: 0,87160- sample size: 131161- multipolarity162- precision: 0,72163- recall: 0,92164- f1-score: 0,81165- sample size: 150166- negation167- precision: 0,70168- recall: 0,47169- f1-score: 0,56170- sample size: 150171- negative172- precision: 0,81173- recall: 0,94174- f1-score: 0,87175- sample size: 225176- neutral177- precision: 0,79178- recall: 0,73179- f1-score: 0,76180- sample size: 150181- positive182- precision: 0,92183- recall: 0,89184- f1-score: 0,90185- sample size: 225186- sarcastic187- precision: 0,95188- recall: 0,98189- f1-score: 0,96190- sample size: 450191 192tags v5:193- v5_lr4e-5_100warmupsteps_decay0.015194epochs v5: 3195average loss in each epoch v5:196- Average loss in epoch 1: 0.9703197- Average loss in epoch 2: 0.3887198- Average loss in epoch 3: 0.2168199metrics v5:200- overall accuracy: 0,875201- irony202- precision: 0,95203- recall: 0,82204- f1-score: 0,88205- sample size: 131206- multipolarity207- precision: 0,75208- recall: 0,89209- f1-score: 0,82210- sample size: 150211- negation212- precision: 0,777213- recall: 0,59214- f1-score: 0,67215- sample size: 150216- negative217- precision: 0,85218- recall: 0,95219- f1-score: 0,90220- sample size: 225221- neutral222- precision: 0,82223- recall: 0,76224- f1-score: 0,79225- sample size: 150226- positive227- precision: 0,88228- recall: 0,89229- f1-score: 0,89230- sample size: 225231- sarcastic232- precision: 0,96233- recall: 0,97234- f1-score: 0,96235- sample size: 450236 237tags v6:238- v6_lr2e-5_150warmupsteps_decay0.01239epochs v6: 5240average loss in each epoch v6:241- Average loss in epoch 1: 1.1274242- Average loss in epoch 2: 0.4649243- Average loss in epoch 3: 0.2806244- Average loss in epoch 4: 0.2009245- Average loss in epoch 5: 0.1389246metrics v6:247- overall accuracy: 0,876248- irony249- precision: 0,97250- recall: 0,82251- f1-score: 0,89252- sample size: 131253- multipolarity254- precision: 0,73255- recall: 0,89256- f1-score: 0,80257- sample size: 150258- negation259- precision: 0,73260- recall: 0,59261- f1-score: 0,65262- sample size: 150263- negative264- precision: 0,85265- recall: 0,94266- f1-score: 0,89267- sample size: 225268- neutral269- precision: 0,84270- recall: 0,75271- f1-score: 0,79272- sample size: 150273- positive274- precision: 0,91275- recall: 0,90276- f1-score: 0,91277- sample size: 225278- sarcastic279- precision: 0,96280- recall: 0,98281- f1-score: 0,97282- sample size: 450283---