Hugging Face's logo

MatanP
/
Qwen_fine_tuned

PEFT
Safetensors
trl
sft
Generated from Trainer
Model card Files Community
Qwen_fine_tuned / README.md
MatanP's picture
MatanP
End of training
f96ee9b verified
preview code
|
raw
history blame contribute delete
1.68 kB
metadata
base_model: Qwen/Qwen1.5-7B
library_name: peft
license: other
tags:
  - trl
  - sft
  - generated_from_trainer
model-index:
  - name: SFT_FineTuned_Qwen-7B
    results: []

SFT_FineTuned_Qwen-7B

This model is a fine-tuned version of Qwen/Qwen1.5-7B on the MatanP/stories-with_custom_prompts dataset. It achieves the following results on the evaluation set:

  • Loss: 1.7424

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: 1e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • 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: cosine
  • num_epochs: 6

Training results

Training Loss Epoch Step Validation Loss
1.904 1.0 90 1.8201
1.8606 2.0 180 1.7638
1.769 3.0 270 1.7466
1.7389 4.0 360 1.7426
1.7152 5.0 450 1.7422
1.326 6.0 540 1.7424

Framework versions

  • PEFT 0.12.0
  • Transformers 4.44.2
  • Pytorch 2.4.1+cu124
  • Datasets 3.0.0
  • Tokenizers 0.19.1