sam860/
qwen3-embedding:0.6b-Q8_0

765 3 months ago

tools

3 months ago

c3c880b0d5f9 · 639MB ·

model
qwen3
·
596M
·
Q8_0
params
{ "repeat_penalty": 1, "stop": [ "<|im_start|>", "<|im_end|>" ], "te
template
{{- if .Messages }} {{- if or .System .Tools }}<|im_start|>system {{- if .System }} {{ .System }} {{

Readme

Description:

The Qwen3-Embedding-0.6B-GGUF is a specialized text embedding model, part of the Qwen3 series, tailored for efficient and effective text embedding and reranking tasks.

Key Features & Use Cases:

  • Model Type: Primarily a Text Embedding model, also part of a series that includes Text Reranking models.
  • Size: A compact model with 0.6 billion parameters, making it suitable for scenarios where efficiency is prioritized. Other sizes (4B, 8B) are available in the series for different performance needs.
  • Multilingual Capability: Supports over 100 languages, including various programming languages, excelling in multilingual, cross-lingual, and code retrieval scenarios.
  • Context Length: Capable of processing texts up to 32,000 tokens.
  • Embedding Dimension: Offers flexible output dimensions, allowing users to define embeddings from 32 to 1024.
  • Instruction Awareness: Supports user-defined instructions to optimize performance for specific tasks, languages, or use cases. This can lead to a 1% to 5% improvement in performance for most downstream tasks, especially in retrieval scenarios.
  • Quantization: Available in q8_0 and f16 quantization formats.
  • Applications: Ideal for various text embedding and ranking tasks, including:
    • Text Retrieval
    • Code Retrieval
    • Text Classification
    • Text Clustering
    • Bitext Mining

Usage Notes:

  • When using the model, it’s recommended to append the <|endoftext|> token to your input context.
  • For optimal results, tailor your input instructions (instruct) to the specific task or language.

HuggingFace