TwT-6/cr-model

TEXT GENERATIONConcurrency Cost:1Model Size:14.2BQuant:FP8Ctx Length:32kPublished:Apr 8, 2024License:cc-by-nc-4.0Architecture:Transformer0.0K Open Weights Cold

The TwT-6/cr-model is a 14.2 billion parameter language processing AI developed by TwT-6, designed for understanding and generating human-like text. This model leverages deep learning to engage in a wide range of language tasks, providing information, recommendations, and conversational capabilities. It excels in complex language-based tasks and general text generation, demonstrating strong performance across various reasoning and common sense benchmarks.

Loading preview...

TwT-6/cr-model Overview

The TwT-6/cr-model is a 14.2 billion parameter language model developed by TwT-6, engineered for advanced natural language understanding and generation. It is designed to process and produce human-like text across diverse applications, from providing information and recommendations to engaging in nuanced conversations.

Key Capabilities & Performance

This model demonstrates robust performance on standard benchmarks, indicating strong reasoning and common sense abilities. Its evaluation on the Open LLM Leaderboard shows an average score of 68.09, with notable results:

  • AI2 Reasoning Challenge (25-Shot): 57.85
  • HellaSwag (10-Shot): 81.66
  • MMLU (5-Shot): 68.73
  • TruthfulQA (0-shot): 58.20
  • Winogrande (5-shot): 76.24
  • GSM8k (5-shot): 65.88

These metrics highlight its proficiency in various reasoning, common sense, and knowledge-based tasks. The model supports a context length of 32768 tokens, enabling it to handle extensive inputs and generate coherent, contextually relevant outputs.

When to Use This Model

  • General-purpose text generation: Creating human-like text for various applications.
  • Complex language-based tasks: Assisting with inquiries, summarization, and content creation.
  • Conversational AI: Developing chatbots or interactive agents that require nuanced understanding and response generation.
  • Reasoning and problem-solving: Applications benefiting from its strong performance on reasoning benchmarks like MMLU and GSM8k.

For detailed benchmark results, refer to the Open LLM Leaderboard and its specific details page.

Finetunes
4 models