indischepartij/OpenMia-Indo-Engineering-7b

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Feb 4, 2024License:cc-by-nc-4.0Architecture:Transformer Open Weights Cold

OpenMia-Indo-Engineering-7b is a 7 billion parameter Mistral-based language model developed by indischepartij, fine-tuned for conversations in Bahasa Indonesia. This model specializes in engineering topics, offering domain-specific dialogue capabilities. It is an alpha-stage model with a 4096-token context length, designed for Indonesian-speaking users in technical fields.

Loading preview...

OpenMia-Indo-Engineering-7b Overview

OpenMia-Indo-Engineering-7b is a 7 billion parameter language model, a specialized branch of the OpenMia project. It is built upon the Mistral-7b architecture and has been fine-tuned specifically for Bahasa Indonesia conversations, with a particular focus on engineering topics.

Key Capabilities

  • Indonesian Language Proficiency: Designed for natural and effective communication in Bahasa Indonesia.
  • Engineering Domain Expertise: Optimized for discussions and queries related to engineering subjects.
  • Mistral-7b Foundation: Leverages the robust architecture of Mistral-7b for its underlying language understanding and generation.
  • Context Length: Supports a context window of 4096 tokens.

Performance Metrics

Evaluated on the Open LLM Leaderboard, OpenMia-Indo-Engineering-7b achieved an average score of 70.03. Notable scores include:

  • AI2 Reasoning Challenge (25-Shot): 67.15
  • HellaSwag (10-Shot): 85.01
  • MMLU (5-Shot): 62.86
  • GSM8k (5-Shot): 64.90

Good For

  • Developers and researchers requiring an LLM for technical discussions in Bahasa Indonesia.
  • Applications focused on engineering-related content generation or analysis for Indonesian-speaking audiences.
  • Experimentation with alpha-stage models for domain-specific fine-tuning.