metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - crows_pairs
metrics:
  - accuracy
model-index:
  - name: t5-small_crows_pairs_finetuned
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: crows_pairs
          type: crows_pairs
          config: crows_pairs
          split: test
          args: crows_pairs
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.5827814569536424

t5-small_crows_pairs_finetuned

This model is a fine-tuned version of t5-small on the crows_pairs dataset. It achieves the following results on the evaluation set:

Loss: 0.5333
Accuracy: 0.5828
Tp: 0.4967
Tn: 0.0861
Fp: 0.4106
Fn: 0.0066

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Tp	Tn	Fp	Fn
4.1732	1.05	20	2.2158	0.8245	0.4404	0.3841	0.1126	0.0629
0.6547	2.11	40	0.4472	0.5033	0.5033	0.0	0.4967	0.0
0.4173	3.16	60	0.3662	0.5033	0.5033	0.0	0.4967	0.0
0.3822	4.21	80	0.3529	0.5033	0.5033	0.0	0.4967	0.0
0.394	5.26	100	0.3584	0.5033	0.5033	0.0	0.4967	0.0
0.36	6.32	120	0.3446	0.5033	0.5033	0.0	0.4967	0.0
0.3472	7.37	140	0.3410	0.5033	0.5033	0.0	0.4967	0.0
0.3241	8.42	160	0.3369	0.5033	0.5033	0.0	0.4967	0.0
0.3247	9.47	180	0.3252	0.5033	0.5033	0.0	0.4967	0.0
0.3045	10.53	200	0.3152	0.5033	0.5033	0.0	0.4967	0.0
0.2863	11.58	220	0.3034	0.5232	0.5033	0.0199	0.4768	0.0
0.2462	12.63	240	0.2958	0.5232	0.5033	0.0199	0.4768	0.0
0.2151	13.68	260	0.2988	0.5232	0.5033	0.0199	0.4768	0.0
0.2323	14.74	280	0.2964	0.5397	0.5033	0.0364	0.4603	0.0
0.1951	15.79	300	0.2994	0.5430	0.5033	0.0397	0.4570	0.0
0.1848	16.84	320	0.3049	0.5530	0.5033	0.0497	0.4470	0.0
0.1875	17.89	340	0.3117	0.5596	0.5033	0.0563	0.4404	0.0
0.1646	18.95	360	0.3291	0.5596	0.5033	0.0563	0.4404	0.0
0.1622	20.0	380	0.3341	0.5563	0.5033	0.0530	0.4437	0.0
0.1511	21.05	400	0.3415	0.5596	0.5033	0.0563	0.4404	0.0
0.1497	22.11	420	0.3556	0.5629	0.5033	0.0596	0.4371	0.0
0.1195	23.16	440	0.3717	0.5563	0.5033	0.0530	0.4437	0.0
0.1319	24.21	460	0.3753	0.5662	0.5033	0.0629	0.4338	0.0
0.1099	25.26	480	0.3812	0.5728	0.5033	0.0695	0.4272	0.0
0.1078	26.32	500	0.3879	0.5695	0.5033	0.0662	0.4305	0.0
0.091	27.37	520	0.4009	0.5762	0.5	0.0762	0.4205	0.0033
0.0939	28.42	540	0.4204	0.5795	0.5	0.0795	0.4172	0.0033
0.1011	29.47	560	0.4295	0.5828	0.5	0.0828	0.4139	0.0033
0.0871	30.53	580	0.4336	0.5762	0.5	0.0762	0.4205	0.0033
0.0782	31.58	600	0.4421	0.5762	0.5	0.0762	0.4205	0.0033
0.0842	32.63	620	0.4455	0.5762	0.5	0.0762	0.4205	0.0033
0.0779	33.68	640	0.4588	0.5828	0.4967	0.0861	0.4106	0.0066
0.0684	34.74	660	0.4728	0.5795	0.4967	0.0828	0.4139	0.0066
0.0719	35.79	680	0.4747	0.5828	0.4967	0.0861	0.4106	0.0066
0.0754	36.84	700	0.4822	0.5795	0.4967	0.0828	0.4139	0.0066
0.0622	37.89	720	0.4887	0.5795	0.4967	0.0828	0.4139	0.0066
0.0604	38.95	740	0.4963	0.5795	0.4967	0.0828	0.4139	0.0066
0.0592	40.0	760	0.5034	0.5828	0.4967	0.0861	0.4106	0.0066
0.0621	41.05	780	0.5099	0.5795	0.4967	0.0828	0.4139	0.0066
0.0548	42.11	800	0.5151	0.5861	0.4967	0.0894	0.4073	0.0066
0.0653	43.16	820	0.5197	0.5795	0.4967	0.0828	0.4139	0.0066
0.0593	44.21	840	0.5238	0.5795	0.4967	0.0828	0.4139	0.0066
0.0445	45.26	860	0.5287	0.5795	0.4967	0.0828	0.4139	0.0066
0.0541	46.32	880	0.5298	0.5828	0.4967	0.0861	0.4106	0.0066
0.0578	47.37	900	0.5318	0.5795	0.4967	0.0828	0.4139	0.0066
0.0514	48.42	920	0.5330	0.5828	0.4967	0.0861	0.4106	0.0066
0.0451	49.47	940	0.5333	0.5828	0.4967	0.0861	0.4106	0.0066

Framework versions

Transformers 4.26.1
Pytorch 1.13.1
Datasets 2.10.1
Tokenizers 0.13.2