Edit model card

zephyr-7b-gemma-ipo

This model is a fine-tuned version of HuggingFaceH4/zephyr-7b-gemma-sft-v0.1 on the argilla/dpo-mix-7k dataset. It achieves the following results on the evaluation set:

  • Loss: 61.0152
  • Rewards/chosen: -0.4988
  • Rewards/rejected: -0.6909
  • Rewards/accuracies: 0.8021
  • Rewards/margins: 0.1921
  • Logps/rejected: -15.3755
  • Logps/chosen: -11.4268
  • Logits/rejected: 99.7522
  • Logits/chosen: 99.5411

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: 5e-07
  • train_batch_size: 2
  • eval_batch_size: 4
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 128
  • total_eval_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 2

Training results

Training Loss Epoch Step Validation Loss Rewards/chosen Rewards/rejected Rewards/accuracies Rewards/margins Logps/rejected Logps/chosen Logits/rejected Logits/chosen
54.5261 1.8957 100 60.8626 -0.5007 -0.6906 0.8021 0.1899 -15.3697 -11.4648 99.7591 99.5497

Framework versions

  • Transformers 4.40.1
  • Pytorch 2.1.2+cu121
  • Datasets 2.19.0
  • Tokenizers 0.19.1
Downloads last month
7
Safetensors
Model size
8.54B params
Tensor type
BF16
·
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Finetuned from

Dataset used to train chrlu/zephyr-7b-gemma-ipo