learning_rate: 5e-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 4
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
4.8675	0.2	50	4.5785
2.796	0.4	100	2.7773
2.6219	0.6	150	2.6119
2.5757	0.8	200	2.5760
2.5706	1.0	250	2.5712
2.5678	1.2	300	2.5699
2.5673	1.4	350	2.5690
2.5681	1.6	400	2.5687
2.5659	1.8	450	2.5682
2.5662	2.0	500	2.5679
2.5663	2.2	550	2.5675
2.5641	2.4	600	2.5676
2.5666	2.6	650	2.5670
2.5659	2.8	700	2.5668
2.5671	3.0	750	2.5666

Samyam28/llava-1.5-7b-hf-1000-galaxy_class_v2q1

llava-1.5-7b-hf-1000-galaxy_class_v2q1