Embeddings
MarkdownQidiruv, RAG va klasterlash uchun matnning vektorli ifodalari.
To‘liq qo‘llanma
POST
https://api.ailoot.ai/v1/embeddingscurl https://api.ailoot.ai/v1/embeddings \
-H "Authorization: Bearer $AILOOT_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "nvidia/nemotron-3-embed-1b", "input": ["Birinchi hujjat", "Ikkinchi hujjat"]}' emb = client.embeddings.create(
model="nvidia/nemotron-3-embed-1b",
input=["Birinchi hujjat", "Ikkinchi hujjat"],
)
vectors = [d.embedding for d in emb.data]{
"object": "list",
"data": [
{ "object": "embedding", "index": 0, "embedding": [0.0123, -0.0456, 0.0789, …] },
{ "object": "embedding", "index": 1, "embedding": [-0.0321, 0.0654, 0.0987, …] }
],
"model": "nvidia/nemotron-3-embed-1b",
"usage": { "prompt_tokens": 8, "total_tokens": 8 }
}Avtorizatsiya
- Authorizationheadermajburiy
Kalit quyidagi ko‘rinishda:
Bearer sk-…. Quyidagi sarlavhalar ham mos keladi:x-api-keyyokix-goog-api-key.
So‘rov tanasi
- modelstringmajburiy
Embedding modeli:
nvidia/nemotron-3-embed-1b,nvidia/llama-nemotron-embed-vl-1b-v2. - inputstring | arraymajburiy
Satr yoki satrlar massivi. Bitta kirish uzunligi model konteksti bilan cheklangan (NVIDIA modellarida — 4096 token).
- encoding_formatstringixtiyoriy
float(standart qiymat) yokibase64— agar model qo‘llab-quvvatlasa.
Javob
- data[].embeddingarray
Har bir kirish uchun sonlar vektori, xuddi shu tartibda.
- usageobject
Kirish tokenlari soni. Embedding chiqishi uchun to‘lov olinmaydi.
Savollaringiz qoldimi? Request ID va tavsifni yuboring — tushunishga yordam beramiz.Yordam xizmatiga yozish