- Cloudflare-hosted
- Deprecated
Meta Llama 3.1 多语言大语言模型(LLM)系列是一组经过预训练和指令微调的生成式模型。Llama 3.1 指令微调的纯文本模型针对多语言对话用例进行了优化,在常见行业基准测试中优于许多现有的开源和闭源聊天模型。
| 模型信息 | |
|---|---|
| 已弃用 | 2026/5/30 |
| 上下文窗口 ↗ | 24,000 个 token |
| 条款和许可证 | 链接 ↗ |
试用环境
使用 Workers AI LLM Playground 试用此模型。无需设置或身份验证,即可直接在浏览器中预览和测试模型。
启动 LLM Playground用法
export interface Env {
AI: Ai;
}
export default {
async fetch(request, env): Promise<Response> {
const messages = [
{ role: "system", content: "You are a friendly assistant" },
{
role: "user",
content: "What is the origin of the phrase Hello, World",
},
];
const stream = await env.AI.run("@cf/meta/llama-3.1-70b-instruct", {
messages,
stream: true,
});
return new Response(stream, {
headers: { "content-type": "text/event-stream" },
});
},
} satisfies ExportedHandler<Env>;
export interface Env {
AI: Ai;
}
export default {
async fetch(request, env): Promise<Response> {
const messages = [
{ role: "system", content: "You are a friendly assistant" },
{
role: "user",
content: "What is the origin of the phrase Hello, World",
},
];
const response = await env.AI.run("@cf/meta/llama-3.1-70b-instruct", { messages });
return Response.json(response);
},
} satisfies ExportedHandler<Env>;
import os
import requests
ACCOUNT_ID = "your-account-id"
AUTH_TOKEN = os.environ.get("CLOUDFLARE_AUTH_TOKEN")
prompt = "Tell me all about PEP-8"
response = requests.post(
f"https://api.cloudflare.com/client/v4/accounts/{ACCOUNT_ID}/ai/run/@cf/meta/llama-3.1-70b-instruct",
headers={"Authorization": f"Bearer {AUTH_TOKEN}"},
json={
"messages": [
{"role": "system", "content": "You are a friendly assistant"},
{"role": "user", "content": prompt}
]
}
)
result = response.json()
print(result)
curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run/@cf/meta/llama-3.1-70b-instruct \
-X POST \
-H "Authorization: Bearer $CLOUDFLARE_AUTH_TOKEN" \
-d '{ "messages": [{ "role": "system", "content": "You are a friendly assistant" }, { "role": "user", "content": "Why is pizza so good" }]}'参数
输入
frequency_penalty
numbermaximum: 2minimum: 0降低模型逐字重复相同内容的可能性。▶image
one oflora
string用于微调基础模型的 LoRA(低秩适配,Low-Rank Adaptation)模型名称。max_tokens
integerdefault: 256响应中生成的最大 token 数。presence_penalty
numbermaximum: 2minimum: 0提高模型引入新话题的可能性。prompt
stringrequiredmaxLength: 131072minLength: 1供模型生成响应的输入文本提示词。raw
booleandefault: false如果为 true,则不应用聊天模板,您必须遵循特定模型所要求的格式。repetition_penalty
numbermaximum: 2minimum: 0对重复 token 的惩罚;值越高,越能抑制重复。seed
integermaximum: 9999999999minimum: 1用于使生成结果可复现的随机种子。stream
booleandefault: false如果为 true,响应将通过 SSE(Server Sent Events,服务器发送事件)以增量方式流式返回。temperature
numberdefault: 0.6maximum: 5minimum: 0控制输出的随机性;值越高,结果越随机。top_k
integermaximum: 50minimum: 1限制 AI 仅从概率最高的前 k 个词中进行选择。较低的值使响应更聚焦;较高的值会带来更多变化和意外结果。top_p
numbermaximum: 2minimum: 0通过控制模型考虑的候选词数量来调整 AI 响应的创造性。较低的值使输出更可预测;较高的值使响应更加多样、更具创意。输出
Synchronous — Send a request and receive a complete response
response
string模型生成的文本响应▶tool_calls[]
array生成响应过程中产生的工具调用请求数组Streaming — Send a request with `stream: true` and receive server-sent events
contentType
text/event-streamformat
binarytype
string