MiniLM_classification_tools_fr

This model is a fine-tuned version of microsoft/Multilingual-MiniLM-L12-H384 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7694
• Accuracy: 0.75
• Learning Rate: 0.0000

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: 24
• eval_batch_size: 192
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 60

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Rate
No log	1.0	7	2.0620	0.35	0.0001
No log	2.0	14	1.9515	0.425	0.0001
No log	3.0	21	1.7736	0.45	0.0001
No log	4.0	28	1.6055	0.475	0.0001
No log	5.0	35	1.5108	0.5	0.0001
No log	6.0	42	1.4074	0.45	9e-05
No log	7.0	49	1.3848	0.475	0.0001
No log	8.0	56	1.2533	0.625	0.0001
No log	9.0	63	1.2463	0.525	0.0001
No log	10.0	70	1.1593	0.6	0.0001
No log	11.0	77	1.1637	0.6	0.0001
No log	12.0	84	1.0900	0.625	8e-05
No log	13.0	91	0.9577	0.7	0.0001
No log	14.0	98	0.9465	0.675	0.0001
No log	15.0	105	0.9255	0.675	0.0001
No log	16.0	112	0.8836	0.675	0.0001
No log	17.0	119	0.8307	0.675	0.0001
No log	18.0	126	0.8335	0.725	7e-05
No log	19.0	133	0.8469	0.625	0.0001
No log	20.0	140	0.7384	0.75	0.0001
No log	21.0	147	0.7330	0.775	0.0001
No log	22.0	154	0.7811	0.775	0.0001
No log	23.0	161	0.6857	0.8	0.0001
No log	24.0	168	0.6733	0.825	6e-05
No log	25.0	175	0.6510	0.85	0.0001
No log	26.0	182	0.6363	0.85	0.0001
No log	27.0	189	0.6101	0.875	0.0001
No log	28.0	196	0.6434	0.8	0.0001
No log	29.0	203	0.6644	0.775	0.0001
No log	30.0	210	0.7162	0.75	5e-05
No log	31.0	217	0.7422	0.775	0.0000
No log	32.0	224	0.7120	0.775	0.0000
No log	33.0	231	0.6296	0.8	0.0000
No log	34.0	238	0.6522	0.775	0.0000
No log	35.0	245	0.7636	0.75	0.0000
No log	36.0	252	0.7703	0.75	4e-05
No log	37.0	259	0.7694	0.75	0.0000

Framework versions

• Transformers 4.34.0
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.14.1