Spaces:
Sleeping
Sleeping
Commit
•
9be5ab5
1
Parent(s):
c2e468f
⚡️ Load balance endpoints (#106)
Browse files
.env
CHANGED
@@ -3,15 +3,17 @@
|
|
3 |
|
4 |
MONGODB_URL=#your mongodb URL here
|
5 |
MONGODB_DB_NAME=chat-ui
|
6 |
-
HF_TOKEN=#your huggingface token here
|
7 |
COOKIE_NAME=hf-chat
|
8 |
|
9 |
PUBLIC_MAX_INPUT_TOKENS=1024
|
10 |
PUBLIC_ORIGIN=#https://hf.co
|
11 |
-
PUBLIC_MODEL_ENDPOINT=https://api-inference.huggingface.co/models/OpenAssistant/oasst-sft-6-llama-30b
|
12 |
PUBLIC_MODEL_NAME=OpenAssistant/oasst-sft-6-llama-30b # public facing link
|
13 |
PUBLIC_MODEL_ID=OpenAssistant/oasst-sft-6-llama-30b-xor # used to link to model page
|
14 |
PUBLIC_DISABLE_INTRO_TILES=false
|
15 |
PUBLIC_USER_MESSAGE_TOKEN=<|prompter|>
|
16 |
PUBLIC_ASSISTANT_MESSAGE_TOKEN=<|assistant|>
|
17 |
PUBLIC_SEP_TOKEN=<|endoftext|>
|
|
|
|
|
|
|
|
|
|
3 |
|
4 |
MONGODB_URL=#your mongodb URL here
|
5 |
MONGODB_DB_NAME=chat-ui
|
|
|
6 |
COOKIE_NAME=hf-chat
|
7 |
|
8 |
PUBLIC_MAX_INPUT_TOKENS=1024
|
9 |
PUBLIC_ORIGIN=#https://hf.co
|
|
|
10 |
PUBLIC_MODEL_NAME=OpenAssistant/oasst-sft-6-llama-30b # public facing link
|
11 |
PUBLIC_MODEL_ID=OpenAssistant/oasst-sft-6-llama-30b-xor # used to link to model page
|
12 |
PUBLIC_DISABLE_INTRO_TILES=false
|
13 |
PUBLIC_USER_MESSAGE_TOKEN=<|prompter|>
|
14 |
PUBLIC_ASSISTANT_MESSAGE_TOKEN=<|assistant|>
|
15 |
PUBLIC_SEP_TOKEN=<|endoftext|>
|
16 |
+
|
17 |
+
# Array<{endpoint: string, authorization: "Bearer XXX", weight: number}> to load balance
|
18 |
+
# Eg if one endpoint has weight 2 and the other has weight 1, the first endpoint will be called twice as often
|
19 |
+
MODEL_ENDPOINTS=`[]`
|
src/lib/server/modelEndpoint.ts
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import { MODEL_ENDPOINTS } from "$env/static/private";
|
2 |
+
import { sum } from "$lib/utils/sum";
|
3 |
+
|
4 |
+
const endpoints: Array<{ endpoint: string; authorization: string; weight: number }> =
|
5 |
+
JSON.parse(MODEL_ENDPOINTS);
|
6 |
+
const totalWeight = sum(endpoints.map((e) => e.weight));
|
7 |
+
|
8 |
+
/**
|
9 |
+
* Find a random load-balanced endpoint
|
10 |
+
*/
|
11 |
+
export function modelEndpoint(): { endpoint: string; authorization: string; weight: number } {
|
12 |
+
let random = Math.random() * totalWeight;
|
13 |
+
for (const endpoint of endpoints) {
|
14 |
+
if (random < endpoint.weight) {
|
15 |
+
return endpoint;
|
16 |
+
}
|
17 |
+
random -= endpoint.weight;
|
18 |
+
}
|
19 |
+
|
20 |
+
throw new Error("Invalid config, no endpoint found");
|
21 |
+
}
|
src/routes/conversation/[id]/+server.ts
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
-
import {
|
2 |
-
import { PUBLIC_MODEL_ENDPOINT, PUBLIC_SEP_TOKEN } from "$env/static/public";
|
3 |
import { buildPrompt } from "$lib/buildPrompt.js";
|
4 |
import { collections } from "$lib/server/database.js";
|
|
|
5 |
import type { Message } from "$lib/types/Message.js";
|
6 |
import { streamToAsyncIterable } from "$lib/utils/streamToAsyncIterable";
|
7 |
import { sum } from "$lib/utils/sum";
|
@@ -29,10 +29,12 @@ export async function POST({ request, fetch, locals, params }) {
|
|
29 |
const messages = [...conv.messages, { from: "user", content: json.inputs }] satisfies Message[];
|
30 |
const prompt = buildPrompt(messages);
|
31 |
|
32 |
-
const
|
|
|
|
|
33 |
headers: {
|
34 |
"Content-Type": request.headers.get("Content-Type") ?? "application/json",
|
35 |
-
Authorization:
|
36 |
},
|
37 |
method: "POST",
|
38 |
body: JSON.stringify({
|
|
|
1 |
+
import { PUBLIC_SEP_TOKEN } from "$env/static/public";
|
|
|
2 |
import { buildPrompt } from "$lib/buildPrompt.js";
|
3 |
import { collections } from "$lib/server/database.js";
|
4 |
+
import { modelEndpoint } from "$lib/server/modelEndpoint.js";
|
5 |
import type { Message } from "$lib/types/Message.js";
|
6 |
import { streamToAsyncIterable } from "$lib/utils/streamToAsyncIterable";
|
7 |
import { sum } from "$lib/utils/sum";
|
|
|
29 |
const messages = [...conv.messages, { from: "user", content: json.inputs }] satisfies Message[];
|
30 |
const prompt = buildPrompt(messages);
|
31 |
|
32 |
+
const randomEndpoint = modelEndpoint();
|
33 |
+
|
34 |
+
const resp = await fetch(randomEndpoint.endpoint, {
|
35 |
headers: {
|
36 |
"Content-Type": request.headers.get("Content-Type") ?? "application/json",
|
37 |
+
Authorization: randomEndpoint.authorization,
|
38 |
},
|
39 |
method: "POST",
|
40 |
body: JSON.stringify({
|
src/routes/conversation/[id]/summarize/+server.ts
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
-
import {
|
2 |
-
import { PUBLIC_MAX_INPUT_TOKENS, PUBLIC_MODEL_ENDPOINT } from "$env/static/public";
|
3 |
import { buildPrompt } from "$lib/buildPrompt";
|
4 |
import { collections } from "$lib/server/database.js";
|
|
|
5 |
import { textGeneration } from "@huggingface/inference";
|
6 |
import { error } from "@sveltejs/kit";
|
7 |
import { ObjectId } from "mongodb";
|
@@ -38,14 +38,20 @@ export async function POST({ params, locals, fetch }) {
|
|
38 |
return_full_text: false,
|
39 |
};
|
40 |
|
|
|
41 |
const { generated_text } = await textGeneration(
|
42 |
{
|
43 |
-
model:
|
44 |
inputs: prompt,
|
45 |
parameters,
|
46 |
-
accessToken: HF_TOKEN,
|
47 |
},
|
48 |
-
{
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
);
|
50 |
|
51 |
if (generated_text) {
|
|
|
1 |
+
import { PUBLIC_MAX_INPUT_TOKENS } from "$env/static/public";
|
|
|
2 |
import { buildPrompt } from "$lib/buildPrompt";
|
3 |
import { collections } from "$lib/server/database.js";
|
4 |
+
import { modelEndpoint } from "$lib/server/modelEndpoint.js";
|
5 |
import { textGeneration } from "@huggingface/inference";
|
6 |
import { error } from "@sveltejs/kit";
|
7 |
import { ObjectId } from "mongodb";
|
|
|
38 |
return_full_text: false,
|
39 |
};
|
40 |
|
41 |
+
const endpoint = modelEndpoint();
|
42 |
const { generated_text } = await textGeneration(
|
43 |
{
|
44 |
+
model: endpoint.endpoint,
|
45 |
inputs: prompt,
|
46 |
parameters,
|
|
|
47 |
},
|
48 |
+
{
|
49 |
+
fetch: (url, options) =>
|
50 |
+
fetch(url, {
|
51 |
+
...options,
|
52 |
+
headers: { ...options?.headers, Authorization: endpoint.authorization },
|
53 |
+
}),
|
54 |
+
}
|
55 |
);
|
56 |
|
57 |
if (generated_text) {
|