README.md

metadata

base_model: Fsoft-AIC/videberta-xsmall
tags:
  - generated_from_trainer
datasets:
  - vietnamese_students_feedback
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: videberta-sentiment-analysis
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: vietnamese_students_feedback
          type: vietnamese_students_feedback
          config: default
          split: validation
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9470198675496688
          - name: Precision
            type: precision
            value: 0.9480840543881335
          - name: Recall
            type: recall
            value: 0.9527950310559006
          - name: F1
            type: f1
            value: 0.9504337050805451

videberta-sentiment-analysis

This model is a fine-tuned version of Fsoft-AIC/videberta-xsmall on the vietnamese_students_feedback dataset. It achieves the following results on the evaluation set:

• Loss: 0.2787
• Accuracy: 0.9470
• Precision: 0.9481
• Recall: 0.9528
• F1: 0.9504

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: 2e-05
• 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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.6152	0.58	100	0.4777	0.8007	0.8580	0.7503	0.8005
0.408	1.16	200	0.3241	0.8669	0.8943	0.8509	0.8721
0.3268	1.74	300	0.2726	0.8954	0.8837	0.9255	0.9041
0.2654	2.33	400	0.2296	0.9199	0.9212	0.9292	0.9252
0.253	2.91	500	0.2088	0.9159	0.9206	0.9217	0.9212
0.2014	3.49	600	0.2318	0.9172	0.9028	0.9466	0.9242
0.1939	4.07	700	0.2131	0.9212	0.9224	0.9304	0.9264
0.1698	4.65	800	0.2005	0.9311	0.9499	0.9193	0.9343
0.1822	5.23	900	0.2249	0.9245	0.9089	0.9540	0.9309
0.1441	5.81	1000	0.2038	0.9311	0.9311	0.9404	0.9357
0.1403	6.4	1100	0.2044	0.9338	0.9315	0.9453	0.9383
0.1377	6.98	1200	0.1991	0.9417	0.9567	0.9329	0.9447
0.1191	7.56	1300	0.2955	0.9119	0.8792	0.9677	0.9213
0.1227	8.14	1400	0.2362	0.9318	0.9199	0.9553	0.9372
0.1023	8.72	1500	0.2221	0.9358	0.9286	0.9528	0.9405
0.1049	9.3	1600	0.1940	0.9424	0.9454	0.9466	0.9460
0.1002	9.88	1700	0.1949	0.9404	0.9649	0.9217	0.9428
0.0946	10.47	1800	0.2232	0.9404	0.9625	0.9242	0.9430
0.0911	11.05	1900	0.2016	0.9457	0.9641	0.9329	0.9482
0.0818	11.63	2000	0.2636	0.9311	0.9128	0.9627	0.9371
0.0889	12.21	2100	0.2279	0.9450	0.9524	0.9441	0.9482
0.0668	12.79	2200	0.2460	0.9411	0.9409	0.9491	0.9450
0.0635	13.37	2300	0.2764	0.9424	0.9465	0.9453	0.9459
0.072	13.95	2400	0.2519	0.9437	0.9390	0.9565	0.9477
0.0697	14.53	2500	0.2705	0.9404	0.9408	0.9478	0.9443
0.0602	15.12	2600	0.2686	0.9450	0.9513	0.9453	0.9483
0.065	15.7	2700	0.2629	0.9450	0.9501	0.9466	0.9484
0.0628	16.28	2800	0.2644	0.9450	0.9547	0.9416	0.9481
0.0505	16.86	2900	0.2704	0.9424	0.9400	0.9528	0.9463
0.0471	17.44	3000	0.2787	0.9470	0.9481	0.9528	0.9504
0.0568	18.02	3100	0.2766	0.9450	0.9424	0.9553	0.9488
0.0523	18.6	3200	0.2659	0.9424	0.9421	0.9503	0.9462
0.0487	19.19	3300	0.3091	0.9338	0.9222	0.9565	0.9390
0.0529	19.77	3400	0.3575	0.9272	0.9045	0.9652	0.9339
0.0484	20.35	3500	0.3228	0.9358	0.9214	0.9615	0.9410
0.0456	20.93	3600	0.2694	0.9437	0.9412	0.9540	0.9476
0.0424	21.51	3700	0.2793	0.9404	0.9376	0.9516	0.9445
0.045	22.09	3800	0.2953	0.9417	0.9356	0.9565	0.9459
0.0395	22.67	3900	0.2840	0.9417	0.9377	0.9540	0.9458
0.0418	23.26	4000	0.3527	0.9305	0.9108	0.9640	0.9366

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3