Modelpublic

AntoineD/MiniLM_uncased_classification_tools_fr

sourceHugging Facemitupdated 3y agoView on Hugging Face

0likes8downloads

Model Card

MiniLMuncasedclassificationtoolsfr

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

—Loss: 0.3165
—Accuracy: 0.925
—Learning Rate: 0.0

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
—trainbatchsize: 24
—evalbatchsize: 192
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 60

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Rate
No log	1.0	7	2.0607	0.35	0.0001
No log	2.0	14	1.9111	0.425	0.0001
No log	3.0	21	1.6543	0.45	0.0001
No log	4.0	28	1.4578	0.525	0.0001
No log	5.0	35	1.3136	0.65	0.0001
No log	6.0	42	1.2160	0.7	9e-05
No log	7.0	49	1.0786	0.725	0.0001
No log	8.0	56	1.0171	0.675	0.0001
No log	9.0	63	0.9491	0.7	0.0001
No log	10.0	70	0.8773	0.75	0.0001
No log	11.0	77	0.8019	0.75	0.0001
No log	12.0	84	0.7436	0.775	8e-05
No log	13.0	91	0.6747	0.825	0.0001
No log	14.0	98	0.7357	0.775	0.0001
No log	15.0	105	0.5386	0.85	0.0001
No log	16.0	112	0.6222	0.85	0.0001
No log	17.0	119	0.6284	0.85	0.0001
No log	18.0	126	0.4489	0.9	7e-05
No log	19.0	133	0.6431	0.85	0.0001
No log	20.0	140	0.6064	0.85	0.0001
No log	21.0	147	0.6948	0.825	0.0001
No log	22.0	154	0.5535	0.85	0.0001
No log	23.0	161	0.4672	0.875	0.0001
No log	24.0	168	0.4797	0.875	6e-05
No log	25.0	175	0.4908	0.9	0.0001
No log	26.0	182	0.5879	0.85	0.0001
No log	27.0	189	0.6601	0.85	0.0001
No log	28.0	196	0.6036	0.85	0.0001
No log	29.0	203	0.5495	0.85	0.0001
No log	30.0	210	0.5135	0.85	5e-05
No log	31.0	217	0.4767	0.875	0.0000
No log	32.0	224	0.4431	0.9	0.0000
No log	33.0	231	0.4681	0.875	0.0000
No log	34.0	238	0.5612	0.85	0.0000
No log	35.0	245	0.4495	0.9	0.0000
No log	36.0	252	0.4384	0.9	4e-05
No log	37.0	259	0.4378	0.875	0.0000
No log	38.0	266	0.4104	0.875	0.0000
No log	39.0	273	0.5060	0.875	0.0000
No log	40.0	280	0.4756	0.875	0.0000
No log	41.0	287	0.4558	0.875	0.0000
No log	42.0	294	0.4458	0.9	3e-05
No log	43.0	301	0.3969	0.875	0.0000
No log	44.0	308	0.4762	0.875	0.0000
No log	45.0	315	0.4891	0.875	0.0000
No log	46.0	322	0.4460	0.9	0.0000
No log	47.0	329	0.3892	0.925	0.0000
No log	48.0	336	0.4267	0.9	2e-05
No log	49.0	343	0.3327	0.9	0.0000
No log	50.0	350	0.3225	0.925	0.0000
No log	51.0	357	0.3223	0.925	0.0000
No log	52.0	364	0.3136	0.95	0.0000
No log	53.0	371	0.3109	0.925	0.0000
No log	54.0	378	0.3142	0.9	1e-05
No log	55.0	385	0.3168	0.925	0.0000
No log	56.0	392	0.3163	0.925	0.0000
No log	57.0	399	0.3174	0.925	5e-06
No log	58.0	406	0.3185	0.925	0.0000
No log	59.0	413	0.3168	0.925	0.0000
No log	60.0	420	0.3165	0.925	0.0

Framework versions

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