Kosmos-2-Furnas-trl-2

This model is a fine-tuned version of microsoft/kosmos-2-patch14-224 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1544

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: 1e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 2
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.4556	100	0.2219
No log	0.9112	200	0.1850
0.2591	1.3667	300	0.1805
0.2591	1.8223	400	0.1691
0.1604	2.2779	500	0.1702
0.1604	2.7335	600	0.1626
0.1604	3.1891	700	0.1625
0.109	3.6446	800	0.1600
0.109	4.1002	900	0.1537
0.0932	4.5558	1000	0.1544

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.16.1
• Tokenizers 0.19.1