CoolFace
Modelpublic

VisualArchiveSystem/vas-models

sourceHugging Faceapache-2.0updated 6mo agoView on Hugging Face
0likes25downloads
Model Card

VAS Pro — AI Models

Pre-quantized GGUF models for VAS Pro (Visual Archive System) local AI assistant.

Models

ModelFileSizePurposeOriginal
Phi-4 Miniphi-4-mini-Q4_K_M.gguf~2.5 GBFast responses, greetingsmicrosoft/Phi-4-mini-instruct
Gemma 3 4Bgemma-4-4b-it-Q4_K_M.gguf~2.5 GBStandard tasks, tool callinggoogle/gemma-3-4b-it
Qwen2.5-VL 7Bqwen3.5-9b-vision-Q4_K_M.gguf~4.7 GBVision, OCR, image analysisQwen/Qwen2.5-VL-7B-Instruct
MxBAI Embed Largemxbai-embed-large-v1-f16.gguf~670 MBSemantic search embeddingsmixedbread-ai/mxbai-embed-large-v1

Usage

These models are automatically downloaded by VAS Pro on first run. No manual setup required.

Quantization

  • —Text models use Q4_K_M quantization (best quality/size ratio for 4-bit)
  • —Embedding model uses F16 (full precision for maximum retrieval accuracy)

License

Models retain their original licenses:

  • —Phi-4 Mini: MIT License
  • —Gemma 3: Gemma Terms of Use
  • —Qwen2.5-VL: Apache 2.0
  • —MxBAI Embed: Apache 2.0