agrobot-llama2-ft

This model is a fine-tuned version of TheBloke/Llama-2-7B-Chat-GPTQ on the None dataset. It achieves the following results on the evaluation set:

• Loss: nan

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
6139.9749	0.99	102	nan
906.7905	2.0	205	nan
7469.7591	3.0	308	nan
4521.9417	4.0	411	nan
10357.8309	4.99	513	nan
2193.7805	6.0	616	nan
415.3014	7.0	719	nan
5163.8981	8.0	822	nan
10537.2745	8.99	924	nan
1091.5675	9.93	1020	nan

Framework versions

• PEFT 0.11.1
• Transformers 4.39.3
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2