zilo-instruct-v2-sft-qlora

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on the arcee-ai/Zilo-Filtered-SQL-Instruct-v2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6246

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 2
• 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: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.7588	0.9899	49	0.7386
0.6201	2.0	99	0.6307
0.6087	2.9697	147	0.6246

Framework versions

• PEFT 0.11.1
• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1