smiles_llava_ft

This model is a fine-tuned version of weathon/smiles_llava on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.0768
• Accuracy: 0.7191

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-06
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 20
• mixed_precision_training: Native AMP
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.3041	0.9569	100	3.5557	0.0
2.3241	1.9091	200	2.5052	0.1835
2.029	2.8612	300	2.2936	0.5056
1.9409	3.8134	400	2.2173	0.5693
1.9861	4.7656	500	2.1782	0.6030
1.9564	5.7177	600	2.1461	0.6217
1.9314	6.6699	700	2.1301	0.6704
1.8838	7.6220	800	2.1084	0.6854
1.9538	8.5742	900	2.1052	0.7154
1.8382	9.5263	1000	2.0955	0.7191
1.9399	10.4785	1100	2.1008	0.6554
1.8231	11.4306	1200	2.0939	0.6891
1.8172	12.3828	1300	2.0899	0.6929
1.8708	13.3349	1400	2.0800	0.7491
1.915	14.2871	1500	2.0776	0.7116
1.8387	15.2392	1600	2.0819	0.7041
1.8646	16.1914	1700	2.0771	0.7228
1.7943	17.1435	1800	2.0770	0.7041
1.8878	18.0957	1900	2.0768	0.7154
1.841	19.0478	2000	2.0768	0.7191

Framework versions

• Transformers 4.48.2
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0