SmolVLM-Instruct-med-vqav1

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

• Loss: 0.1712

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: 0.001
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_HF with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.3885	0.4454	100	0.2168
0.1862	0.8909	200	0.1728
0.1258	1.3341	300	0.1678
0.1131	1.7795	400	0.1615
0.0885	2.2227	500	0.1682
0.0656	2.6682	600	0.1712

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.0
• Tokenizers 0.21.0