refusal-classifier

This model is a fine-tuned version of bert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3054
  • F1: 1.0
  • Roc Auc: 1.0
  • Accuracy: 1.0

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: 1e-05
  • train_batch_size: 384
  • eval_batch_size: 384
  • seed: 42
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 2
  • label_smoothing_factor: 0.05

Training results

Training Loss Epoch Step Validation Loss F1 Roc Auc Accuracy
No log 0.0076 100 2.4712 0.1582 0.7093 0.2062
No log 0.0151 200 2.0531 0.4984 0.9014 0.5231
No log 0.0227 300 1.3834 0.6824 0.9524 0.6938
No log 0.0302 400 1.0466 0.7584 0.9684 0.7623
1.7761 0.0378 500 0.8311 0.8162 0.9822 0.8177
1.7761 0.0454 600 0.7246 0.8474 0.9878 0.8469
1.7761 0.0529 700 0.6630 0.8680 0.9908 0.8677
1.7761 0.0605 800 0.6067 0.8865 0.9933 0.8862
1.7761 0.0681 900 0.5693 0.8948 0.9948 0.8946
0.6613 0.0756 1000 0.5578 0.8993 0.9957 0.8992
0.6613 0.0832 1100 0.5338 0.9103 0.9963 0.9108
0.6613 0.0907 1200 0.5117 0.9154 0.9971 0.9154
0.6613 0.0983 1300 0.5015 0.9253 0.9972 0.9254
0.6613 0.1059 1400 0.4856 0.9277 0.9977 0.9277
0.5142 0.1134 1500 0.4666 0.9448 0.9980 0.9446
0.5142 0.1210 1600 0.4659 0.9391 0.9981 0.9392
0.5142 0.1285 1700 0.4543 0.9399 0.9984 0.94
0.5142 0.1361 1800 0.4557 0.9392 0.9985 0.9392
0.5142 0.1437 1900 0.4472 0.9454 0.9985 0.9454
0.4593 0.1512 2000 0.4354 0.9523 0.9987 0.9523
0.4593 0.1588 2100 0.4317 0.9507 0.9989 0.9508
0.4593 0.1664 2200 0.4296 0.9479 0.9990 0.9477
0.4593 0.1739 2300 0.4335 0.9479 0.9989 0.9477
0.4593 0.1815 2400 0.4237 0.9525 0.9991 0.9523
0.4301 0.1890 2500 0.4158 0.9592 0.9991 0.9592
0.4301 0.1966 2600 0.4071 0.9537 0.9993 0.9538
0.4301 0.2042 2700 0.4111 0.9555 0.9993 0.9554
0.4301 0.2117 2800 0.3970 0.9623 0.9994 0.9623
0.4301 0.2193 2900 0.3992 0.9559 0.9994 0.9562
0.4091 0.2268 3000 0.3954 0.9631 0.9994 0.9631
0.4091 0.2344 3100 0.3868 0.9676 0.9996 0.9677
0.4091 0.2420 3200 0.3947 0.9615 0.9995 0.9615
0.4091 0.2495 3300 0.3935 0.9653 0.9994 0.9654
0.4091 0.2571 3400 0.3843 0.9661 0.9995 0.9662
0.3959 0.2647 3500 0.3811 0.9653 0.9995 0.9654
0.3959 0.2722 3600 0.3813 0.9654 0.9996 0.9654
0.3959 0.2798 3700 0.3870 0.9646 0.9995 0.9646
0.3959 0.2873 3800 0.3768 0.9692 0.9996 0.9692
0.3959 0.2949 3900 0.3825 0.9669 0.9996 0.9669
0.3846 0.3025 4000 0.3741 0.9677 0.9996 0.9677
0.3846 0.3100 4100 0.3739 0.9716 0.9996 0.9715
0.3846 0.3176 4200 0.3645 0.9769 0.9997 0.9769
0.3846 0.3251 4300 0.3651 0.9739 0.9997 0.9738
0.3846 0.3327 4400 0.3644 0.9761 0.9997 0.9762
0.3752 0.3403 4500 0.3682 0.9753 0.9997 0.9754
0.3752 0.3478 4600 0.3634 0.9731 0.9998 0.9731
0.3752 0.3554 4700 0.3577 0.9777 0.9998 0.9777
0.3752 0.3629 4800 0.3575 0.9777 0.9998 0.9777
0.3752 0.3705 4900 0.3605 0.9785 0.9998 0.9785
0.3685 0.3781 5000 0.3569 0.9762 0.9998 0.9762
0.3685 0.3856 5100 0.3546 0.9769 0.9999 0.9769
0.3685 0.3932 5200 0.3580 0.9792 0.9998 0.9792
0.3685 0.4008 5300 0.3596 0.9738 0.9998 0.9738
0.3685 0.4083 5400 0.3554 0.9777 0.9999 0.9777
0.3616 0.4159 5500 0.3491 0.9839 0.9999 0.9838
0.3616 0.4234 5600 0.3487 0.9807 0.9999 0.9808
0.3616 0.4310 5700 0.3481 0.9831 0.9999 0.9831
0.3616 0.4386 5800 0.3437 0.9839 0.9999 0.9838
0.3616 0.4461 5900 0.3467 0.9823 0.9999 0.9823
0.3556 0.4537 6000 0.3411 0.9823 0.9999 0.9823
0.3556 0.4612 6100 0.3422 0.9846 0.9999 0.9846
0.3556 0.4688 6200 0.3482 0.9823 0.9999 0.9823
0.3556 0.4764 6300 0.3406 0.9839 0.9999 0.9838
0.3556 0.4839 6400 0.3368 0.9854 0.9999 0.9854
0.3513 0.4915 6500 0.3416 0.9800 0.9999 0.98
0.3513 0.4991 6600 0.3382 0.9862 0.9999 0.9862
0.3513 0.5066 6700 0.3361 0.9869 1.0000 0.9869
0.3513 0.5142 6800 0.3329 0.9854 1.0000 0.9854
0.3513 0.5217 6900 0.3303 0.9877 1.0000 0.9877
0.3487 0.5293 7000 0.3393 0.9838 1.0000 0.9838
0.3487 0.5369 7100 0.3298 0.9892 1.0000 0.9892
0.3487 0.5444 7200 0.3319 0.9831 1.0000 0.9831
0.3487 0.5520 7300 0.3365 0.9815 1.0000 0.9815
0.3487 0.5595 7400 0.3331 0.9908 1.0000 0.9908
0.3456 0.5671 7500 0.3406 0.9838 0.9999 0.9838
0.3456 0.5747 7600 0.3320 0.9877 1.0000 0.9877
0.3456 0.5822 7700 0.3341 0.9877 1.0000 0.9877
0.3456 0.5898 7800 0.3296 0.9885 1.0000 0.9885
0.3456 0.5974 7900 0.3319 0.9884 1.0000 0.9885
0.3419 0.6049 8000 0.3282 0.9908 1.0000 0.9908
0.3419 0.6125 8100 0.3272 0.9908 1.0000 0.9908
0.3419 0.6200 8200 0.3244 0.9923 1.0000 0.9923
0.3419 0.6276 8300 0.3304 0.9860 1.0000 0.9862
0.3419 0.6352 8400 0.3301 0.9877 1.0000 0.9877
0.3392 0.6427 8500 0.3240 0.9923 1.0000 0.9923
0.3392 0.6503 8600 0.3229 0.9908 1.0000 0.9908
0.3392 0.6578 8700 0.3247 0.9923 1.0000 0.9923
0.3392 0.6654 8800 0.3254 0.9900 1.0000 0.99
0.3392 0.6730 8900 0.3248 0.9900 1.0000 0.99
0.3374 0.6805 9000 0.3235 0.9900 1.0000 0.99
0.3374 0.6881 9100 0.3250 0.9915 1.0000 0.9915
0.3374 0.6957 9200 0.3232 0.9938 1.0000 0.9938
0.3374 0.7032 9300 0.3236 0.9915 1.0000 0.9915
0.3374 0.7108 9400 0.3237 0.9908 1.0000 0.9908
0.3352 0.7183 9500 0.3241 0.9908 1.0000 0.9908
0.3352 0.7259 9600 0.3160 0.9954 1.0000 0.9954
0.3352 0.7335 9700 0.3185 0.9938 1.0000 0.9938
0.3352 0.7410 9800 0.3187 0.9939 1.0000 0.9938
0.3352 0.7486 9900 0.3247 0.9892 1.0000 0.9892
0.3333 0.7561 10000 0.3174 0.9939 1.0000 0.9938
0.3333 0.7637 10100 0.3207 0.9938 1.0000 0.9938
0.3333 0.7713 10200 0.3237 0.9923 1.0000 0.9923
0.3333 0.7788 10300 0.3197 0.9938 1.0000 0.9938
0.3333 0.7864 10400 0.3194 0.9931 1.0000 0.9931
0.3309 0.7940 10500 0.3209 0.9923 1.0000 0.9923
0.3309 0.8015 10600 0.3178 0.9946 1.0000 0.9946
0.3309 0.8091 10700 0.3159 0.9946 1.0000 0.9946
0.3309 0.8166 10800 0.3138 0.9954 1.0000 0.9954
0.3309 0.8242 10900 0.3142 0.9954 1.0000 0.9954
0.3298 0.8318 11000 0.3131 0.9954 1.0000 0.9954
0.3298 0.8393 11100 0.3171 0.9939 1.0000 0.9938
0.3298 0.8469 11200 0.3167 0.9946 1.0000 0.9946
0.3298 0.8544 11300 0.3182 0.9938 1.0000 0.9938
0.3298 0.8620 11400 0.3184 0.9923 1.0000 0.9923
0.3287 0.8696 11500 0.3173 0.9962 1.0000 0.9962
0.3287 0.8771 11600 0.3195 0.9946 1.0000 0.9946
0.3287 0.8847 11700 0.3142 0.9962 1.0000 0.9962
0.3287 0.8922 11800 0.3148 0.9946 1.0000 0.9946
0.3287 0.8998 11900 0.3139 0.9962 1.0000 0.9962
0.3265 0.9074 12000 0.3162 0.9954 1.0000 0.9954
0.3265 0.9149 12100 0.3129 0.9962 1.0000 0.9962
0.3265 0.9225 12200 0.3176 0.9931 1.0000 0.9931
0.3265 0.9301 12300 0.3221 0.9931 0.9999 0.9931
0.3265 0.9376 12400 0.3168 0.9962 1.0000 0.9962
0.3257 0.9452 12500 0.3129 0.9969 1.0000 0.9969
0.3257 0.9527 12600 0.3140 0.9954 1.0000 0.9954
0.3257 0.9603 12700 0.3160 0.9954 1.0000 0.9954
0.3257 0.9679 12800 0.3133 0.9954 1.0000 0.9954
0.3257 0.9754 12900 0.3116 0.9977 1.0000 0.9977
0.3242 0.9830 13000 0.3091 0.9985 1.0000 0.9985
0.3242 0.9905 13100 0.3124 0.9969 1.0000 0.9969
0.3242 0.9981 13200 0.3148 0.9946 1.0000 0.9946
0.3242 1.0057 13300 0.3191 0.9946 1.0000 0.9946
0.3242 1.0132 13400 0.3135 0.9946 1.0000 0.9946
0.3213 1.0208 13500 0.3191 0.9915 1.0000 0.9915
0.3213 1.0284 13600 0.3145 0.9939 1.0000 0.9938
0.3213 1.0359 13700 0.3135 0.9946 1.0000 0.9946
0.3213 1.0435 13800 0.3126 0.9931 1.0 0.9931
0.3213 1.0510 13900 0.3125 0.9962 1.0000 0.9962
0.32 1.0586 14000 0.3129 0.9961 1.0000 0.9962
0.32 1.0662 14100 0.3099 0.9969 1.0000 0.9969
0.32 1.0737 14200 0.3113 0.9977 1.0000 0.9977
0.32 1.0813 14300 0.3111 0.9969 1.0000 0.9969
0.32 1.0888 14400 0.3074 0.9985 1.0 0.9985
0.3189 1.0964 14500 0.3088 0.9969 1.0000 0.9969
0.3189 1.1040 14600 0.3087 0.9977 1.0 0.9977
0.3189 1.1115 14700 0.3110 0.9962 1.0000 0.9962
0.3189 1.1191 14800 0.3097 0.9961 1.0000 0.9962
0.3189 1.1267 14900 0.3086 0.9977 1.0 0.9977
0.3183 1.1342 15000 0.3087 0.9962 1.0000 0.9962
0.3183 1.1418 15100 0.3133 0.9946 1.0000 0.9946
0.3183 1.1493 15200 0.3098 0.9969 1.0 0.9969
0.3183 1.1569 15300 0.3111 0.9962 1.0000 0.9962
0.3183 1.1645 15400 0.3149 0.9946 1.0000 0.9946
0.3172 1.1720 15500 0.3143 0.9954 1.0000 0.9954
0.3172 1.1796 15600 0.3090 0.9969 1.0 0.9969
0.3172 1.1871 15700 0.3074 0.9977 1.0000 0.9977
0.3172 1.1947 15800 0.3116 0.9969 1.0000 0.9969
0.3172 1.2023 15900 0.3113 0.9969 1.0000 0.9969
0.3163 1.2098 16000 0.3101 0.9977 1.0 0.9977
0.3163 1.2174 16100 0.3081 0.9977 1.0000 0.9977
0.3163 1.2250 16200 0.3163 0.9938 0.9999 0.9938
0.3163 1.2325 16300 0.3141 0.9946 1.0000 0.9946
0.3163 1.2401 16400 0.3134 0.9946 1.0000 0.9946
0.3157 1.2476 16500 0.3131 0.9954 1.0000 0.9954
0.3157 1.2552 16600 0.3168 0.9946 1.0000 0.9946
0.3157 1.2628 16700 0.3112 0.9962 1.0000 0.9962
0.3157 1.2703 16800 0.3158 0.9962 1.0000 0.9962
0.3157 1.2779 16900 0.3118 0.9969 1.0000 0.9969
0.3156 1.2854 17000 0.3115 0.9969 1.0000 0.9969
0.3156 1.2930 17100 0.3091 0.9969 1.0000 0.9969
0.3156 1.3006 17200 0.3087 0.9969 1.0000 0.9969
0.3156 1.3081 17300 0.3087 0.9985 1.0000 0.9985
0.3156 1.3157 17400 0.3095 0.9969 1.0000 0.9969
0.315 1.3233 17500 0.3075 0.9985 1.0000 0.9985
0.315 1.3308 17600 0.3087 0.9969 1.0000 0.9969
0.315 1.3384 17700 0.3095 0.9985 1.0000 0.9985
0.315 1.3459 17800 0.3082 0.9977 1.0 0.9977
0.315 1.3535 17900 0.3074 0.9977 1.0 0.9977
0.3145 1.3611 18000 0.3086 0.9977 1.0000 0.9977
0.3145 1.3686 18100 0.3069 0.9969 1.0000 0.9969
0.3145 1.3762 18200 0.3092 0.9977 1.0000 0.9977
0.3145 1.3837 18300 0.3081 0.9969 1.0000 0.9969
0.3145 1.3913 18400 0.3129 0.9954 1.0000 0.9954
0.314 1.3989 18500 0.3086 0.9977 1.0 0.9977
0.314 1.4064 18600 0.3093 0.9962 1.0000 0.9962
0.314 1.4140 18700 0.3085 0.9969 1.0000 0.9969
0.314 1.4216 18800 0.3081 0.9977 1.0 0.9977
0.314 1.4291 18900 0.3100 0.9962 1.0 0.9962
0.3138 1.4367 19000 0.3067 0.9985 1.0 0.9985
0.3138 1.4442 19100 0.3080 0.9969 1.0 0.9969
0.3138 1.4518 19200 0.3080 0.9977 1.0000 0.9977
0.3138 1.4594 19300 0.3116 0.9954 1.0000 0.9954
0.3138 1.4669 19400 0.3117 0.9946 1.0000 0.9946
0.313 1.4745 19500 0.3106 0.9962 1.0000 0.9962
0.313 1.4820 19600 0.3091 0.9969 1.0000 0.9969
0.313 1.4896 19700 0.3086 0.9977 1.0 0.9977
0.313 1.4972 19800 0.3071 0.9977 1.0 0.9977
0.313 1.5047 19900 0.3081 0.9977 1.0000 0.9977
0.3131 1.5123 20000 0.3090 0.9961 1.0000 0.9962
0.3131 1.5198 20100 0.3071 0.9977 1.0000 0.9977
0.3131 1.5274 20200 0.3089 0.9969 1.0000 0.9969
0.3131 1.5350 20300 0.3092 0.9969 1.0000 0.9969
0.3131 1.5425 20400 0.3064 0.9985 1.0 0.9985
0.3123 1.5501 20500 0.3113 0.9961 1.0000 0.9962
0.3123 1.5577 20600 0.3096 0.9954 1.0000 0.9954
0.3123 1.5652 20700 0.3078 0.9977 1.0000 0.9977
0.3123 1.5728 20800 0.3093 0.9969 1.0000 0.9969
0.3123 1.5803 20900 0.3077 0.9985 1.0000 0.9985
0.3119 1.5879 21000 0.3073 0.9977 1.0000 0.9977
0.3119 1.5955 21100 0.3053 0.9992 1.0 0.9992
0.3119 1.6030 21200 0.3066 0.9985 1.0 0.9985
0.3119 1.6106 21300 0.3085 0.9985 1.0 0.9985
0.3119 1.6181 21400 0.3082 0.9969 1.0000 0.9969
0.3114 1.6257 21500 0.3079 0.9977 1.0000 0.9977
0.3114 1.6333 21600 0.3063 0.9977 1.0 0.9977
0.3114 1.6408 21700 0.3065 0.9985 1.0 0.9985
0.3114 1.6484 21800 0.3083 0.9969 1.0 0.9969
0.3114 1.6560 21900 0.3066 0.9977 1.0 0.9977
0.3118 1.6635 22000 0.3071 0.9977 1.0 0.9977
0.3118 1.6711 22100 0.3075 0.9977 1.0 0.9977
0.3118 1.6786 22200 0.3076 0.9969 1.0 0.9969
0.3118 1.6862 22300 0.3070 0.9977 1.0 0.9977
0.3118 1.6938 22400 0.3069 0.9985 1.0 0.9985
0.3113 1.7013 22500 0.3078 0.9969 1.0 0.9969
0.3113 1.7089 22600 0.3063 0.9985 1.0 0.9985
0.3113 1.7164 22700 0.3057 0.9992 1.0 0.9992
0.3113 1.7240 22800 0.3070 0.9977 1.0 0.9977
0.3113 1.7316 22900 0.3067 0.9977 1.0 0.9977
0.3107 1.7391 23000 0.3057 0.9985 1.0 0.9985
0.3107 1.7467 23100 0.3061 0.9992 1.0 0.9992
0.3107 1.7543 23200 0.3059 0.9992 1.0 0.9992
0.3107 1.7618 23300 0.3055 0.9992 1.0 0.9992
0.3107 1.7694 23400 0.3054 1.0 1.0 1.0
0.3114 1.7769 23500 0.3055 0.9985 1.0 0.9985
0.3114 1.7845 23600 0.3067 0.9985 1.0 0.9985
0.3114 1.7921 23700 0.3063 0.9985 1.0 0.9985
0.3114 1.7996 23800 0.3053 0.9992 1.0 0.9992
0.3114 1.8072 23900 0.3055 0.9985 1.0 0.9985
0.3105 1.8147 24000 0.3057 0.9992 1.0 0.9992
0.3105 1.8223 24100 0.3053 0.9992 1.0 0.9992
0.3105 1.8299 24200 0.3065 0.9992 1.0 0.9992
0.3105 1.8374 24300 0.3064 0.9977 1.0 0.9977
0.3105 1.8450 24400 0.3066 0.9985 1.0 0.9985
0.3106 1.8526 24500 0.3070 0.9977 1.0000 0.9977
0.3106 1.8601 24600 0.3054 0.9992 1.0 0.9992
0.3106 1.8677 24700 0.3062 0.9985 1.0 0.9985
0.3106 1.8752 24800 0.3060 0.9992 1.0 0.9992
0.3106 1.8828 24900 0.3063 0.9992 1.0000 0.9992
0.3101 1.8904 25000 0.3058 0.9992 1.0 0.9992
0.3101 1.8979 25100 0.3062 0.9985 1.0 0.9985
0.3101 1.9055 25200 0.3061 0.9992 1.0000 0.9992
0.3101 1.9130 25300 0.3064 0.9985 1.0000 0.9985
0.3101 1.9206 25400 0.3057 0.9992 1.0 0.9992
0.3102 1.9282 25500 0.3060 0.9992 1.0 0.9992
0.3102 1.9357 25600 0.3059 0.9985 1.0 0.9985
0.3102 1.9433 25700 0.3053 0.9992 1.0 0.9992
0.3102 1.9509 25800 0.3049 1.0 1.0 1.0
0.3102 1.9584 25900 0.3053 0.9992 1.0 0.9992
0.3098 1.9660 26000 0.3052 0.9992 1.0 0.9992
0.3098 1.9735 26100 0.3051 1.0 1.0 1.0
0.3098 1.9811 26200 0.3052 0.9992 1.0 0.9992
0.3098 1.9887 26300 0.3051 0.9992 1.0 0.9992
0.3098 1.9962 26400 0.3051 0.9992 1.0 0.9992

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.1.0+cu118
  • Datasets 3.1.0
  • Tokenizers 0.20.3
Downloads last month
13
Safetensors
Model size
109M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for avrecum/refusal-classifier

Finetuned
(2305)
this model