AITRADER/Amsi-fin-o1.5

VISIONConcurrency Cost:1Model Size:9BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Mar 15, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

AITRADER/Amsi-fin-o1.5 is a 9 billion parameter financial language model developed by AITRADER, based on the Qwen3.5-9B-Base architecture. It is specifically fine-tuned for advanced trading strategy reasoning, market structure interpretation, risk framing, and finance-related question answering. This model aims to provide stronger text-only financial reasoning capabilities compared to its predecessors, with a 32K context length.

Loading preview...

AITRADER/Amsi-fin-o1.5: Advanced Financial Reasoning Model

Amsi-fin-o1.5 is a specialized 9-billion parameter language model from AITRADER, built upon the Qwen/Qwen3.5-9B-Base architecture. This iteration represents a significant domain update, moving beyond previous AITRADER finance models by focusing on enhanced financial reasoning capabilities through full fine-tuning.

Key Capabilities

  • Trading Strategy Reasoning: Excels in analyzing and explaining complex trading strategies.
  • Market Structure Interpretation: Provides insights into market dynamics and structures.
  • Risk Framing: Offers improved explanations regarding risk assessment and management.
  • Finance Question Answering: Designed for accurate and nuanced responses to financial queries.
  • Improved Explanations: Delivers clearer explanations for market regimes, invalidations, and risk factors.

Training and Configuration

The model was fine-tuned using full_finetuning=true with an adafactor optimizer and a learning rate of 1e-05. It utilizes a sequence length of 1024 and gradient accumulation of 4 steps. This configuration aims for a cleaner supervised tuning footprint, facilitating deployment in OpenAI-compatible serving environments and downstream benchmarking.

Good For

  • Developers and researchers requiring a robust language model for financial analysis.
  • Applications focused on generating or evaluating trading strategies.
  • Systems needing to interpret complex financial market data and provide risk assessments.
  • Building advanced financial chatbots or question-answering systems.