solanaclawd/solana-nvidia-trading-factory-8b

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Sep 6, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The solanaclawd/solana-nvidia-trading-factory-8b is an 8 billion parameter merged Transformer model, built upon NousResearch/Hermes-3-Llama-3.1-8B with a LoRA adapter. It is specifically fine-tuned for Solana agent coding and reasoning tasks, leveraging a combination of 'claude-fable-5-claude-code' and 'Fable-5-traces' training data. This model is designed to assist developers in generating and understanding code related to Solana, particularly for trading applications.

Loading preview...

Overview

The solanaclawd/solana-nvidia-trading-factory-8b is an 8 billion parameter merged Transformer model. It is constructed by combining the NousResearch/Hermes-3-Llama-3.1-8B base model with a specialized LoRA adapter, solanaclawd/solana-nvidia-trading-factory-8b-lora. The model's training data includes armand0e/claude-fable-5-claude-code and Glint-Research/Fable-5-traces, indicating a focus on code generation and reasoning.

Key Capabilities

  • Solana Agent Coding: Optimized for generating and understanding code relevant to Solana blockchain agents.
  • Reasoning: Designed to perform reasoning tasks within the context of Solana development.
  • Trading Applications: Specifically tailored for use cases involving Solana-based trading systems.

Good For

  • Developers: Ideal for developers working on Solana-related projects, particularly those involving automated trading or agent-based systems.
  • Research: Suitable for research into AI applications for blockchain development and decentralized finance.

Important Safety Note

This model is intended for research and development purposes related to Solana agent coding and reasoning. It is crucial that any live trading, wallet operations, or deployment actions are handled by separate execution clients, require explicit operator approval, and incorporate robust risk checks. The model itself does not directly execute these sensitive operations.

Finetunes
3 models