README.md

metadata

language:
  - en
license: apache-2.0
base_model: bert-base-multilingual-cased
tags:
  - generated_from_trainer
datasets:
  - tmnam20/VieGLUE
metrics:
  - accuracy
model-index:
  - name: bert-base-multilingual-cased-mnli-1
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: tmnam20/VieGLUE/MNLI
          type: tmnam20/VieGLUE
          config: mnli
          split: validation_matched
          args: mnli
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8031936533767291

bert-base-multilingual-cased-mnli-1

This model is a fine-tuned version of bert-base-multilingual-cased on the tmnam20/VieGLUE/MNLI dataset. It achieves the following results on the evaluation set:

• Loss: 0.5349
• Accuracy: 0.8032

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: 32
• eval_batch_size: 16
• seed: 1
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8082	0.04	500	0.7958	0.6485
0.7259	0.08	1000	0.7455	0.6895
0.7018	0.12	1500	0.6970	0.7118
0.7026	0.16	2000	0.6827	0.7127
0.6696	0.2	2500	0.6500	0.7323
0.6744	0.24	3000	0.6345	0.7380
0.6136	0.29	3500	0.6294	0.7402
0.632	0.33	4000	0.6269	0.7472
0.6735	0.37	4500	0.6195	0.7489
0.6202	0.41	5000	0.6336	0.7414
0.6495	0.45	5500	0.6125	0.7517
0.6235	0.49	6000	0.6097	0.7515
0.5852	0.53	6500	0.6068	0.7581
0.6395	0.57	7000	0.6039	0.7493
0.6009	0.61	7500	0.5878	0.7553
0.6059	0.65	8000	0.5876	0.7638
0.6019	0.69	8500	0.5829	0.7651
0.5989	0.73	9000	0.5922	0.7612
0.6195	0.77	9500	0.5868	0.7615
0.6028	0.81	10000	0.5724	0.7709
0.5741	0.86	10500	0.5670	0.7717
0.582	0.9	11000	0.5702	0.7732
0.5706	0.94	11500	0.5597	0.7755
0.5676	0.98	12000	0.5655	0.7735
0.5235	1.02	12500	0.5849	0.7662
0.521	1.06	13000	0.5646	0.7788
0.5122	1.1	13500	0.5717	0.7738
0.5102	1.14	14000	0.5667	0.7765
0.5152	1.18	14500	0.5598	0.7780
0.4904	1.22	15000	0.5693	0.7746
0.507	1.26	15500	0.5584	0.7804
0.5163	1.3	16000	0.5570	0.7787
0.4921	1.34	16500	0.5727	0.7798
0.5249	1.39	17000	0.5653	0.7789
0.4994	1.43	17500	0.5726	0.7783
0.5335	1.47	18000	0.5547	0.7848
0.543	1.51	18500	0.5541	0.7785
0.5138	1.55	19000	0.5569	0.7842
0.4626	1.59	19500	0.5625	0.7860
0.4828	1.63	20000	0.5434	0.7858
0.5121	1.67	20500	0.5495	0.7806
0.5012	1.71	21000	0.5318	0.7900
0.4609	1.75	21500	0.5485	0.7878
0.4928	1.79	22000	0.5462	0.7868
0.4922	1.83	22500	0.5305	0.7920
0.4913	1.87	23000	0.5396	0.7891
0.4992	1.91	23500	0.5341	0.7952
0.4732	1.96	24000	0.5277	0.7952
0.4925	2.0	24500	0.5339	0.7943
0.4098	2.04	25000	0.5643	0.7911
0.4168	2.08	25500	0.5534	0.7929
0.4099	2.12	26000	0.5674	0.7925
0.4142	2.16	26500	0.5652	0.7918
0.398	2.2	27000	0.5875	0.7899
0.3899	2.24	27500	0.5726	0.7975
0.403	2.28	28000	0.5596	0.7968
0.399	2.32	28500	0.5716	0.7885
0.4176	2.36	29000	0.5570	0.7941
0.3871	2.4	29500	0.5689	0.7926
0.4156	2.44	30000	0.5648	0.7918
0.386	2.49	30500	0.5650	0.7931
0.4131	2.53	31000	0.5525	0.7948
0.4202	2.57	31500	0.5585	0.7914
0.4129	2.61	32000	0.5495	0.7963
0.4215	2.65	32500	0.5524	0.7978
0.413	2.69	33000	0.5578	0.7954
0.4296	2.73	33500	0.5509	0.7966
0.3602	2.77	34000	0.5581	0.7974
0.3901	2.81	34500	0.5561	0.7985
0.4163	2.85	35000	0.5502	0.7955
0.3787	2.89	35500	0.5573	0.7951
0.4285	2.93	36000	0.5535	0.7958
0.3578	2.97	36500	0.5563	0.7964

Framework versions

• Transformers 4.35.2
• Pytorch 2.2.0.dev20231203+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0