Views
No views yet
| File | Quantization | Size |
|---|---|---|
| gte-small-f32.gguf | F32 | 0 MB |
| gte-small-q4_k.gguf | Q4_K | 0 MB |
| gte-small-q8_0.gguf | Q8_0 | 0 MB |
| gte-small.gguf | F32 | 0 MB |
1./crispembed -m gte-small "Hello world"
2./crispembed-server -m gte-small --port 80801# Create model
2echo "FROM gte-small-q8_0.gguf" > Modelfile
3ollama create gte-small -f Modelfile
4
5# Embed
6curl http://localhost:11434/api/embed -d '{"model":"gte-small","input":["Hello world"]}'1from crispembed import CrispEmbed
2model = CrispEmbed("gte-small-q8_0.gguf")
3vectors = model.encode(["Hello world", "Goodbye world"])| Property | Value |
|---|---|
| Architecture | BERT |
| Parameters | 33M |
| Embedding Dimension | 384 |
| Layers | 12 |
| Pooling | mean |
| Tokenizer | WordPiece |
| Language | en |
| Q8_0 vs HuggingFace | 0.9999 |
| Q4_K vs HuggingFace | 0.991 |
POST /embed — nativePOST /v1/embeddings — OpenAI-compatiblePOST /api/embed — Ollama-compatiblePOST /api/embeddings — Ollama legacythenlper.mit. This repository redistributes under the same terms; it grants no rights the upstream licence does not.