Safetensors
ParagonLight's picture
upload loras
b1775c3
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: logical_deduction_no_sys
    results: []

logical_deduction_no_sys

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

  • Loss: 0.1318

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
0.121 1.16 50 0.1113
0.0574 2.33 100 0.0727
0.0405 3.49 150 0.0608
0.0319 4.65 200 0.0899
0.0236 5.81 250 0.0640
0.0152 6.98 300 0.0996
0.0053 8.14 350 0.1150
0.0021 9.3 400 0.1312

Framework versions

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