Safetensors
ParagonLight's picture
upload
6cb6113
metadata
license: other
library_name: peft
tags:
  - llama-factory
  - lora
  - generated_from_trainer
base_model: /data1/model/llama2/meta-llama/Llama2-13b
model-index:
  - name: strategyqa_no_sys
    results: []

strategyqa_no_sys

This model is a fine-tuned version of /data1/model/llama2/meta-llama/Llama2-13b on the strategyqa_no_sys dataset. It achieves the following results on the evaluation set:

  • Loss: 2.3844

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: 8
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 3
  • total_train_batch_size: 24
  • total_eval_batch_size: 24
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 20
  • num_epochs: 10.0

Training results

Training Loss Epoch Step Validation Loss
1.1108 0.77 50 1.1378
1.0652 1.54 100 1.1236
0.9679 2.31 150 1.1641
0.7321 3.08 200 1.3882
0.5562 3.85 250 1.3443
0.2699 4.62 300 1.6109
0.1135 5.38 350 1.9021
0.0511 6.15 400 2.2162
0.049 6.92 450 2.0983
0.0212 7.69 500 2.3844
0.0107 8.46 550 2.5324
0.006 9.23 600 2.5657
0.0131 10.0 650 2.5765

Framework versions

  • PEFT 0.9.0
  • Transformers 4.38.2
  • Pytorch 2.2.1
  • Datasets 2.18.0
  • Tokenizers 0.15.2