Il-format JSONL u t-tlugħ permezz tal-API tal-Fajls
F'din il-paġna
Aġġornat fit-2 ta’ Novembru, 2026 abbażi tad-dokumentazzjoni ta’ xAI: l-iskont għall-batches huwa ta’ 20% fuq mudelli tat-test eliġibbli (mhux 50%), l-eżempji jsegwu l-API REST u l-
xai-sdkuffiċjali (client.batch.*; l-API tal-Batch ta’ xAI ma tużax il-format ta’ OpenAI, allura l-SDK ta’ OpenAI ma japplikax), u l-API tar-Risposti għandha preċedenza fuq Chat Completions, li huwa endpoint antikwat f’xAI. Sorsi: Batch API · Referenza REST · Prezzijiet tal-Batch.
Ipproċessar bil-kwantità b’JSONL
Għal volumi kbar imħejjija minn qabel, fajl JSONL huwa l-aktar rotta sempliċi. Tikteb it-talbiet kollha tiegħek f’fajl, ittella’ permezz tal-API tal-Fajls, imbagħad toħloq il-batch minn dak il-fajl. Fajl wieħed jista’ jżomm sa 50,000 talba.
Struttura tal-fajl JSONL
Fajl JSONL (JSON Lines) fih oġġett JSON wieħed kull linja. Kull linja tirrappreżenta talba sħiħa b’erba’ oqsma:
{"custom_id": "req-1", "method": "POST", "url": "/v1/responses", "body": {"model": "grok-4.3", "input": [{"role": "user", "content": "Summarize this text in one sentence: ..."}]}}
{"custom_id": "req-2", "method": "POST", "url": "/v1/responses", "body": {"model": "grok-4.3", "input": [{"role": "user", "content": "Classify this ticket: bug, feature, or question. ..."}]}}
{"custom_id": "req-3", "method": "POST", "url": "/v1/chat/completions", "body": {"model": "grok-4.3", "messages": [{"role": "user", "content": "Translate into English: ..."}]}}
custom_id: identifikatur uniku fil-fajl; isir il-batch_request_idfir-riżultatimethod: dejjem"POST"url: l-endpoint fil-mira, wieħed mis-segħa aċċettatibody: il-korp tat-talba, li jaqbel mar-referenza tal-API għal dak l-endpoint
Kull linja trid tkun JSON validu u awtosuffiċjenti: l-ebda virgola li tifred, l-ebda parentesi li tgħaqqad. L-endpoints jistgħu jitħalltu fl-istess fajl; kull talba tiġi rruttata indipendentement.
Iġġenera fajl JSONL f’Python
import json
documents = [
{"id": "doc-001", "text": "First article to summarize..."},
{"id": "doc-002", "text": "Second article to summarize..."},
# ... sa 50,000 dokument kull fajl
]
with open("batch_requests.jsonl", "w", encoding="utf-8") as f:
for doc in documents:
ligne = {
"custom_id": doc["id"],
"method": "POST",
"url": "/v1/responses",
"body": {
"model": "grok-4.3",
"input": [
{"role": "system", "content": "Summarize in three sentences."},
{"role": "user", "content": doc["text"]},
],
},
}
f.write(json.dumps(ligne, ensure_ascii=False) + "\n")
Tlugħ permezz tal-API tal-Fajls
Ttella’ l-fajl permezz tal-endpoint /v1/files, mingħajr ebda parametru ieħor:
curl -X POST https://api.x.ai/v1/files \
-H "Authorization: Bearer $XAI_API_KEY" \
-F file="@batch_requests.jsonl"
Bl-SDK ta’ Python:
from xai_sdk import Client
client = Client()
with open("batch_requests.jsonl", "rb") as f:
fichier = client.files.upload(f, filename="batch_requests.jsonl")
print(f"Uploaded file: {fichier.id}")
Oħloq il-batch mill-fajl
Il-fajl ma jiġix miżjud ma’ batch eżistenti: tispeċifikah fil-ħin tal-ħolqien b’input_file_id:
curl -X POST https://api.x.ai/v1/batches \
-H "Authorization: Bearer $XAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"name": "article_summaries", "input_file_id": "file-abc123"}'
batch = client.batch.create(batch_name="article_summaries", input_file_id=fichier.id)
print(f"Batch created: {batch.batch_id}")
Il-fajl imbagħad jinqara fl-isfond. Jekk sa linja waħda tkun invalida, il-batch jitħassar b’messaġġ ta’ żball. It-traċċar u l-irkupru tar-riżultati mbagħad jaħdmu eżattament bħal għal batch sottomess permezz tal-API.
Limiti tal-format JSONL
- Massimu ta’ 50,000 talba kull fajl
- Massimu ta’ 200 MB kull fajl
custom_iduniku kull linja- Batch maħluq minn fajl huwa issiġillat: ma tistax iżid miegħu aktar tard permezz tal-API
Jekk il-batch tiegħek taqbeż dawn il-limiti, aqsmu f’diversi fajls u oħloq batch wieħed kull fajl. M’hemm l-ebda limitu fuq in-numru ta’ batches kull tim.
Iġġenera fajl JSONL f’JavaScript
import fs from "fs";
const documents = [
{ id: "doc-001", text: "First article..." },
{ id: "doc-002", text: "Second article..." },
];
const lines = documents.map(doc => JSON.stringify({
custom_id: doc.id,
method: "POST",
url: "/v1/responses",
body: {
model: "grok-4.3",
input: [
{ role: "user", content: `Summarize: ${doc.text}` }
]
}
}));
fs.writeFileSync("batch_requests.jsonl", lines.join("\n") + "\n");
Validazzjoni u dijbukkar
Linja waħda invalida tħassar il-batch kollu: ivvalida l-fajl qabel ma ttellgħu. Skript sempliċi jkopri l-essenzjali:
import json
import os
ENDPOINTS = {
"/v1/responses", "/v1/chat/completions",
"/v1/images/generations", "/v1/images/edits",
"/v1/videos/generations", "/v1/videos", "/v1/videos/edits", "/v1/videos/extensions",
}
def valider_jsonl(chemin: str) -> list[str]:
erreurs, ids = [], set()
if os.path.getsize(chemin) > 200 * 1024 * 1024:
erreurs.append("file larger than 200 MB")
with open(chemin, encoding="utf-8") as f:
for i, ligne in enumerate(f, 1):
try:
obj = json.loads(ligne)
except json.JSONDecodeError as e:
erreurs.append(f"line {i}: invalid JSON ({e})")
continue
cid = obj.get("custom_id")
if not cid or cid in ids:
erreurs.append(f"line {i}: missing or duplicate custom_id")
ids.add(cid)
if obj.get("method") != "POST" or obj.get("url") not in ENDPOINTS:
erreurs.append(f"line {i}: invalid method or url")
if not isinstance(obj.get("body"), dict):
erreurs.append(f"line {i}: missing body")
if len(ids) > 50_000:
erreurs.append("more than 50,000 requests")
return erreurs
print(valider_jsonl("batch_requests.jsonl") or "File is valid")
Punti Ewlenin
- Il-format JSONL fih oġġett JSON wieħed kull linja:
custom_id,method,url,body - Ittella’ l-fajl permezz ta’
/v1/files, imbagħad ispeċifikah meta toħloq il-batch (input_file_id) - Massimu ta’ 50,000 talba u 200 MB kull fajl; lil hinn minn hekk, batch wieħed kull fajl
- Batch maħluq minn fajl huwa ssiġillat, u linja waħda invalida tħassru kompletament
- Ivvalida l-fajl qabel ma ttellgħu