README.md

metadata

license: other
library_name: peft
tags:
  - trl
  - sft
  - generated_from_trainer
base_model: google/gemma-7b
model-index:
  - name: outputs
    results: []
datasets:
  - gnumanth/gita
language:
  - en
pipeline_tag: question-answering

outputs

This model is a fine-tuned version of google/gemma-7b on gnumanth/gita dataset.

Model description

Experimental fine tune gemma-7b model

Intended uses & limitations

Gita interpretation, no limitations.

Training procedure

SFTTrainer

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 1
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• training_steps: 10
• mixed_precision_training: Native AMP

Training results

TrainOutput(global_step=10, training_loss=0.5232708215713501, metrics={'train_runtime': 13.8846, 'train_samples_per_second': 2.881, 'train_steps_per_second': 0.72, 'total_flos': 55989006336000.0, 'train_loss': 0.5232708215713501, 'epoch': 10.0})

Framework versions

• PEFT 0.8.2
• Transformers 4.38.1
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2