QCRI/AZERG-T3-Mistral

TEXT GENERATIONConcurrent Unit Cost:1Model Size:7BQuant:FP8Context Size:4kTool Calling:SupportedPublished:Jul 23, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

QCRI/AZERG-T3-Mistral is a fine-tuned Mistral-7B-Instruct-v0.3 model developed by QCRI, specialized for Task 3: Related Pair Detection within the AZERG framework. This specialist model is trained on the QCRI/AZERG-Dataset to identify semantic relationships between two STIX entities in a given text context. It is designed for high performance in extracting potential STIX entities from security text passages.

Loading preview...

Overview

QCRI/AZERG-T3-Mistral is a specialized language model developed by QCRI, fine-tuned from mistralai/Mistral-7B-Instruct-v0.3. Its primary function is Related Pair Detection within the AZERG framework. This model has been specifically trained on the QCRI/AZERG-Dataset to accurately determine if a semantic relationship exists between two STIX (Structured Threat Information eXpression) entities within a provided text passage.

Key Capabilities

  • Specialized Entity Relationship Detection: Focuses on identifying semantic relationships between STIX entities.
  • Contextual Analysis: Evaluates relationships based on the surrounding text passage.
  • High Performance: Designed for high accuracy in its specific task within the AZERG framework.

Good For

  • Threat Intelligence Analysis: Automating the extraction of actionable intelligence from security texts.
  • STIX Entity Extraction: Identifying potential STIX entities and their relationships from unstructured security data.
  • Security Text Processing: Analyzing security-related text to uncover connections between threat indicators.