README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - sentiment140
metrics:
  - accuracy
base_model: bert-base-cased
model-index:
  - name: Sentiment140_BERT_5E
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: sentiment140
          type: sentiment140
          config: sentiment140
          split: train
          args: sentiment140
        metrics:
          - type: accuracy
            value: 0.82
            name: Accuracy

Sentiment140_BERT_5E

This model is a fine-tuned version of bert-base-cased on the sentiment140 dataset. It achieves the following results on the evaluation set:

• Loss: 0.7061
• Accuracy: 0.82

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6882	0.08	50	0.6047	0.7
0.6223	0.16	100	0.5137	0.8067
0.5463	0.24	150	0.4573	0.8067
0.4922	0.32	200	0.4790	0.8
0.4821	0.4	250	0.4207	0.8267
0.4985	0.48	300	0.4267	0.8067
0.4455	0.56	350	0.4301	0.8133
0.469	0.64	400	0.4294	0.82
0.4906	0.72	450	0.4059	0.8067
0.4006	0.8	500	0.4181	0.8133
0.445	0.88	550	0.3948	0.8267
0.4302	0.96	600	0.3976	0.84
0.4442	1.04	650	0.3887	0.8533
0.3424	1.12	700	0.4119	0.8267
0.3589	1.2	750	0.4083	0.8533
0.3737	1.28	800	0.4253	0.8333
0.334	1.36	850	0.4147	0.86
0.3637	1.44	900	0.3926	0.8533
0.3388	1.52	950	0.4084	0.8267
0.3375	1.6	1000	0.4132	0.8467
0.3725	1.68	1050	0.3965	0.8467
0.3649	1.76	1100	0.3956	0.8333
0.3799	1.84	1150	0.3923	0.8333
0.3695	1.92	1200	0.4266	0.84
0.3233	2.0	1250	0.4225	0.8333
0.2313	2.08	1300	0.4672	0.8333
0.231	2.16	1350	0.5212	0.8133
0.2526	2.24	1400	0.5392	0.8067
0.2721	2.32	1450	0.4895	0.82
0.2141	2.4	1500	0.5258	0.8133
0.2658	2.48	1550	0.5046	0.8267
0.2386	2.56	1600	0.4873	0.8267
0.2493	2.64	1650	0.4950	0.8333
0.2692	2.72	1700	0.5080	0.8267
0.2226	2.8	1750	0.5016	0.8467
0.2522	2.88	1800	0.5068	0.8267
0.2556	2.96	1850	0.4937	0.8267
0.2311	3.04	1900	0.5103	0.8267
0.1703	3.12	1950	0.5680	0.82
0.1744	3.2	2000	0.5501	0.82
0.1667	3.28	2050	0.6142	0.82
0.1863	3.36	2100	0.6355	0.82
0.2543	3.44	2150	0.6000	0.8133
0.1565	3.52	2200	0.6618	0.8267
0.1531	3.6	2250	0.6595	0.8133
0.1915	3.68	2300	0.6647	0.8267
0.1601	3.76	2350	0.6729	0.8267
0.176	3.84	2400	0.6699	0.82
0.1815	3.92	2450	0.6819	0.8067
0.1987	4.0	2500	0.6543	0.8333
0.1236	4.08	2550	0.6686	0.8333
0.1599	4.16	2600	0.6583	0.8267
0.1256	4.24	2650	0.6871	0.8267
0.1291	4.32	2700	0.6855	0.82
0.1198	4.4	2750	0.6901	0.82
0.1245	4.48	2800	0.7152	0.8267
0.1784	4.56	2850	0.7053	0.82
0.1705	4.64	2900	0.7016	0.82
0.1265	4.72	2950	0.7013	0.82
0.1192	4.8	3000	0.7084	0.82
0.174	4.88	3050	0.7062	0.82
0.1328	4.96	3100	0.7061	0.82

Framework versions

• Transformers 4.24.0
• Pytorch 1.12.1+cu113
• Datasets 2.6.1
• Tokenizers 0.13.1