kuotient/Meta-Llama-3-8B

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kLicense:llama3Architecture:Transformer0.0K Warm

Meta-Llama-3-8B is an 8 billion parameter, instruction-tuned generative text model developed by Meta, utilizing an optimized transformer architecture. It is part of the Llama 3 family, trained on over 15 trillion tokens with a context length of 8192 tokens, and is specifically optimized for dialogue use cases, outperforming many open-source chat models on common industry benchmarks. This model is intended for commercial and research use in English, excelling in assistant-like chat applications.

Loading preview...

Overview

Meta-Llama-3-8B is an 8 billion parameter instruction-tuned large language model developed by Meta, released as part of the Llama 3 family. It features an optimized transformer architecture and is designed for generative text tasks, particularly excelling in dialogue-based applications. The model was trained on over 15 trillion tokens of publicly available data, with a knowledge cutoff of March 2023, and supports a context length of 8192 tokens. Both pre-trained and instruction-tuned variants are available, with the latter optimized using supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) for enhanced helpfulness and safety.

Key Capabilities

  • High Performance: Outperforms many open-source chat models on common industry benchmarks, demonstrating significant improvements over Llama 2 models across various categories like MMLU, AGIEval, and HumanEval.
  • Dialogue Optimization: Instruction-tuned specifically for assistant-like chat and dialogue use cases.
  • Robust Training: Benefits from a massive pretraining dataset (15T+ tokens) and fine-tuning with over 10 million human-annotated examples.
  • Safety & Responsibility: Incorporates extensive red teaming, adversarial evaluations, and safety mitigations, with a focus on reducing false refusals compared to previous versions.
  • Commercial & Research Use: Intended for a broad range of commercial and research applications in English.

Good For

  • Assistant-like Chatbots: Ideal for developing conversational AI agents and virtual assistants.
  • Natural Language Generation: Suitable for various text generation tasks where high-quality, coherent output is required.
  • Research & Development: Provides a strong foundation for further fine-tuning and exploration in LLM capabilities.
  • Benchmarking & Evaluation: Offers competitive performance against other models in its class, making it a good candidate for comparative studies.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p