README.md

metadata

license: apache-2.0
base_model: t5-base
tags:
  - generated_from_trainer
datasets:
  - glue
metrics:
  - accuracy
model-index:
  - name: t5-base_cola_dense
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: glue
          type: glue
          config: cola
          split: validation
          args: cola
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8360498561840843

t5-base_cola_dense

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

• Loss: 0.5247
• Accuracy: 0.8360

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
• train_batch_size: 32
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6253	0.07	10	0.6218	0.6913
0.6283	0.15	20	0.6207	0.6913
0.6297	0.22	30	0.6193	0.6913
0.5936	0.3	40	0.6200	0.6913
0.609	0.37	50	0.6216	0.6913
0.5818	0.45	60	0.6182	0.6913
0.5984	0.52	70	0.6139	0.6913
0.5961	0.6	80	0.6089	0.6913
0.5594	0.67	90	0.6168	0.6913
0.597	0.75	100	0.6022	0.6913
0.6092	0.82	110	0.5537	0.6903
0.5454	0.9	120	0.5120	0.7296
0.5368	0.97	130	0.5120	0.7584
0.5185	1.04	140	0.4615	0.7987
0.4664	1.12	150	0.4893	0.7977
0.4938	1.19	160	0.4793	0.8044
0.3994	1.27	170	0.4912	0.8025
0.4989	1.34	180	0.5515	0.8092
0.4709	1.42	190	0.4909	0.8054
0.4099	1.49	200	0.5397	0.8121
0.4671	1.57	210	0.4736	0.8102
0.3893	1.64	220	0.4803	0.8178
0.4027	1.72	230	0.5195	0.8159
0.4208	1.79	240	0.4521	0.8188
0.4506	1.87	250	0.4943	0.8188
0.3647	1.94	260	0.4650	0.8255
0.4223	2.01	270	0.4865	0.8284
0.3584	2.09	280	0.4639	0.8284
0.3555	2.16	290	0.5321	0.8236
0.3433	2.24	300	0.5174	0.8303
0.3904	2.31	310	0.4811	0.8274
0.3418	2.39	320	0.5135	0.8265
0.3397	2.46	330	0.4854	0.8322
0.3336	2.54	340	0.5008	0.8332
0.3471	2.61	350	0.5065	0.8293
0.382	2.69	360	0.4708	0.8274
0.3533	2.76	370	0.4862	0.8265
0.3199	2.84	380	0.4904	0.8293
0.3757	2.91	390	0.4970	0.8332
0.3726	2.99	400	0.4965	0.8322
0.2957	3.06	410	0.4628	0.8303
0.3232	3.13	420	0.5174	0.8322
0.2836	3.21	430	0.5038	0.8351
0.2919	3.28	440	0.4987	0.8341
0.3578	3.36	450	0.5187	0.8313
0.398	3.43	460	0.5285	0.8380
0.3024	3.51	470	0.4971	0.8351
0.3153	3.58	480	0.5084	0.8351
0.307	3.66	490	0.5371	0.8332
0.2753	3.73	500	0.5247	0.8360
0.3515	3.81	510	0.4782	0.8360
0.2881	3.88	520	0.4784	0.8389
0.3203	3.96	530	0.5115	0.8351
0.2791	4.03	540	0.5294	0.8360
0.301	4.1	550	0.5218	0.8322
0.2652	4.18	560	0.4956	0.8360
0.2954	4.25	570	0.4878	0.8332
0.2345	4.33	580	0.5190	0.8313
0.3762	4.4	590	0.5315	0.8351
0.3614	4.48	600	0.5200	0.8341
0.3178	4.55	610	0.5237	0.8341
0.306	4.63	620	0.5232	0.8341
0.2828	4.7	630	0.5278	0.8360
0.3442	4.78	640	0.5270	0.8360
0.3268	4.85	650	0.5252	0.8351
0.2959	4.93	660	0.5284	0.8370
0.2861	5.0	670	0.5277	0.8351

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1