langleu/qmd-query-expansion-lfm2.5-1.2b-instruct-v2

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.2BQuant:BF16Context Size:32kPublished:Jul 25, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

langleu/qmd-query-expansion-lfm2.5-1.2b-instruct-v2 is a 1.2 billion parameter instruction-tuned model, fine-tuned from LiquidAI/LFM2.5-1.2B-Instruct using LoRA for Query-Metadata (QMD) query expansion. It is specifically designed to expand search queries for QMD systems, supporting directives like /only:lex, /only:vec, and /only:hyde. The model provides both a BF16 Transformers checkpoint and QMD-ready GGUF quantizations, demonstrating high format compliance and entity preservation in QMD evaluations.

Loading preview...

Overview

This model, langleu/qmd-query-expansion-lfm2.5-1.2b-instruct-v2, is a 1.2 billion parameter instruction-tuned variant of LiquidAI/LFM2.5-1.2B-Instruct. It has been fine-tuned using LoRA specifically for Query-Metadata (QMD) query expansion with a v2 data recipe. The repository offers both a merged BF16 Transformers checkpoint and GGUF quantizations optimized for QMD/llama.cpp.

Key Capabilities

  • Specialized Query Expansion: Designed to expand search queries for QMD systems, supporting specific directives.
  • QMD-Specific Directives: Understands and processes /only:lex, /only:vec, and /only:hyde directives, with each output line prefixed accordingly.
  • High Performance: Achieves an average QMD reward of 94.42% and 99.68% format compliance in validation.
  • GGUF Quantizations: Provides q5_k_m.gguf files for efficient deployment with QMD and llama.cpp.

When to Use This Model

  • QMD Query Expansion: Ideal for applications requiring precise and structured query expansion within a QMD framework.
  • Integration with QMD: Directly compatible with QMD systems, especially when using the provided GGUF files.
  • Resource-Constrained Environments: The 1.2B parameter size and GGUF quantizations make it suitable for efficient inference.