Llama-3.2-1B-Instruct-SchemaLinking-v1

This model is a fine-tuned version of meta-llama/Llama-3.2-1B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0956

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 14
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.3641	0.4668	500	0.1527
0.1066	0.9336	1000	0.1083
0.0794	1.4004	1500	0.0963
0.0739	1.8672	2000	0.0872
0.058	2.3341	2500	0.0910
0.057	2.8009	3000	0.0859
0.0432	3.2678	3500	0.0895
0.0452	3.7346	4000	0.0890
0.0406	4.2014	4500	0.0944
0.0363	4.6682	5000	0.0956

Framework versions

• PEFT 0.13.0
• Transformers 4.45.1
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.0