10 actions in 9 pieces of software · ranked by words, meaning and measured trust · 651 ms
modelrecord6,630 downloads/month
- text-classification
Task: text-classification
- text-classification
modelrecord10,469 downloads/month
- text-ranking
Task: text-ranking
- text-ranking
modelrecord2,811 downloads/month
- sentence-similarity
Task: sentence-similarity
- sentence-similarity
modelrecord16,175 downloads/month
- feature-extraction
Task: feature-extraction
- feature-extraction
modelrecord1,568,999 downloads/month
- fill-mask
Task: fill-mask
- fill-mask
modelrecord582,820 downloads/month
- token-classification
Task: token-classification
- token-classification
modelrecord7,162 downloads/month
- summarization
Task: summarization
- summarization
MCP serverrecord
- laya_multilingual_demo__do_mod
- laya_multilingual_demo__do_guard
MCP serverrecord
- ettin_reranker_68m_v1_rerank
Rerank candidate passages by relevance to a query. Returns: A ranking table (rank, relevance score, passage) sorted by score descending,, plus a performance summary (inference latency and pairs-per-second throughput).
- ettin_reranker_68m_v1_rerank
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.