Bracko/mistral-7B-Instruct-v0.3-finetuned-ds3

results

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8456

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_steps: 500
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
1.0544	0.3608	500	0.9443
0.9332	0.7215	1000	0.8999
0.8875	1.0823	1500	0.8851
0.8301	1.4430	2000	0.8664
0.8179	1.8038	2500	0.8555
0.762	2.1645	3000	0.8671
0.7358	2.5253	3500	0.8592
0.7384	2.8860	4000	0.8456

Framework versions

• PEFT 0.10.0
• Transformers 4.40.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1