model_dnm

This model is a fine-tuned version of DinaSalama/symptom_to_disease_distb on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5334
• Accuracy: 0.6768

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	21	3.8845	0.0549
No log	2.0	42	3.6758	0.0732
No log	3.0	63	3.2875	0.1829
No log	4.0	84	2.3470	0.4024
No log	5.0	105	1.6458	0.5854
No log	6.0	126	1.3744	0.6585
No log	7.0	147	1.4220	0.6524
No log	8.0	168	1.4036	0.6768
No log	9.0	189	1.3755	0.6463
No log	10.0	210	1.4416	0.6768
No log	11.0	231	1.4216	0.6707
No log	12.0	252	1.5050	0.6951
No log	13.0	273	1.5019	0.6768
No log	14.0	294	1.4762	0.7073
No log	15.0	315	1.4628	0.6829
No log	16.0	336	1.4532	0.6829
No log	17.0	357	1.5105	0.6951
No log	18.0	378	1.4814	0.6829
No log	19.0	399	1.5126	0.6829
No log	20.0	420	1.5063	0.6890
No log	21.0	441	1.5246	0.6829
No log	22.0	462	1.5371	0.6768
No log	23.0	483	1.5238	0.6768
0.7037	24.0	504	1.5374	0.6829
0.7037	25.0	525	1.5310	0.6768
0.7037	26.0	546	1.5301	0.6768
0.7037	27.0	567	1.5339	0.6646
0.7037	28.0	588	1.5438	0.6829
0.7037	29.0	609	1.5338	0.6768
0.7037	30.0	630	1.5152	0.6768
0.7037	31.0	651	1.5189	0.6768
0.7037	32.0	672	1.5231	0.6829
0.7037	33.0	693	1.5269	0.6829
0.7037	34.0	714	1.5263	0.6768
0.7037	35.0	735	1.5288	0.6829
0.7037	36.0	756	1.5290	0.6768
0.7037	37.0	777	1.5302	0.6768
0.7037	38.0	798	1.5328	0.6768
0.7037	39.0	819	1.5330	0.6768
0.7037	40.0	840	1.5334	0.6768

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1