Model Overview

This model, llama_2_rlhf_safe_llama_3_70B_default_1000_full, is a 7 billion parameter language model derived from the meta-llama/Llama-2-7b-chat-hf base model. It has undergone a specific fine-tuning process using a generator dataset, resulting in an evaluation loss of 0.8687.

Key Training Details

• Base Model: meta-llama/Llama-2-7b-chat-hf
• Parameters: 7 Billion
• Fine-tuning Objective: Generator dataset
• Evaluation Loss: 0.8687
• Hyperparameters:
  • Learning Rate: 2e-05
  • Optimizer: Adam with betas=(0.9, 0.999) and epsilon=1e-08
  • Epochs: 1
  • Total Train Batch Size: 32 (across 4 GPUs)

Intended Use Cases

While specific intended uses and limitations are not detailed in the provided information, this model is suitable for developers looking for a Llama 2-7B variant that has been fine-tuned on a generator dataset. Its performance metrics suggest it has learned patterns from this specific dataset, which could be beneficial for tasks aligned with its training data.