Mel-Iza0/Llama2-13B_ZeroShot-20K_classe_other_port

TEXT GENERATIONConcurrency Cost:1Model Size:13BQuant:FP8Ctx Length:4kArchitecture:Transformer Cold

Mel-Iza0/Llama2-13B_ZeroShot-20K_classe_other_port is a 13 billion parameter Llama 2-based language model, fine-tuned using PEFT with 4-bit quantization for efficient deployment. This model is designed for zero-shot classification tasks, leveraging its Llama 2 architecture and specific training to categorize inputs without explicit examples. Its primary strength lies in performing classification on diverse data with minimal setup.

Loading preview...

Model Overview

Mel-Iza0/Llama2-13B_ZeroShot-20K_classe_other_port is a 13 billion parameter language model built upon the Llama 2 architecture. It has been fine-tuned using the PEFT (Parameter-Efficient Fine-Tuning) framework, specifically employing 4-bit quantization for optimized performance and reduced memory footprint.

Key Characteristics

  • Architecture: Based on the Llama 2 family of models.
  • Parameter Count: 13 billion parameters, offering a balance between capability and computational requirements.
  • Quantization: Utilizes bitsandbytes 4-bit quantization (nf4 type with double quantization and bfloat16 compute dtype) for efficient inference and training.
  • Context Length: Supports a context window of 4096 tokens.

Good For

  • Zero-Shot Classification: This model is specifically adapted for zero-shot classification tasks, enabling it to categorize inputs effectively without requiring explicit in-context examples.
  • Resource-Efficient Deployment: The 4-bit quantization makes it suitable for environments where memory and computational resources are a concern, allowing for more accessible deployment compared to full-precision models of similar size.