zsxkib/embedding-gemma-300m ❓📝✓ → ❓

▶️ 104.8K runs 📅 Sep 2025 ⚙️ Cog 0.16.2 🔗 GitHub 📄 Paper ⚖️ License
code-search semantic-search text-embedding

About

Turn any text into 768-dimensional vectors for search, classification, and AI apps 🧠✨

Example Output

Output

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

Performance Metrics

0.36s Prediction Time
95.98s Total Time
All Input Parameters
{
  "task": "retrieval_document",
  "text": "Hello World",
  "normalize": true,
  "embedding_dim": 768,
  "output_format": "base64"
}
Input Parameters
task Default: retrieval_document
Task type - affects the internal prompt used for embedding
text (required) Type: string
Text to embed. Automatically truncated to 2048 tokens if longer.
normalize Type: booleanDefault: true
Whether to L2-normalize the embedding to unit length
embedding_dim Default: 768
Output embedding dimensions using Matryoshka truncation
output_format Default: base64
Output format: base64 (efficient) or array (list of floats)
Output Schema

Output

Version Details
Version ID
d753bd5a898a96666f233f9a33ab1c3fe6527a7be308e1cc9fdcda46abf3e233
Version Created
September 5, 2025
Run on Replicate →