split some options to config folder, translation fixes, markdown fixes, ui improvements to settings and ai, cleanup, add additional models
This commit is contained in:
parent
1857767213
commit
4409acd79d
8 changed files with 273 additions and 182 deletions
|
|
@ -34,7 +34,6 @@ import { Telegraf, Context } from "telegraf"
|
|||
import type { Message } from "telegraf/types"
|
||||
import { replyToMessageId } from "../utils/reply-to-message-id"
|
||||
import { getStrings } from "../plugins/checklang"
|
||||
import { languageCode } from "../utils/language-code"
|
||||
import axios from "axios"
|
||||
import { rateLimiter } from "../utils/rate-limiter"
|
||||
import { logger } from "../utils/log"
|
||||
|
|
@ -42,6 +41,7 @@ import { ensureUserInDb } from "../utils/ensure-user"
|
|||
import * as schema from '../db/schema'
|
||||
import type { NodePgDatabase } from "drizzle-orm/node-postgres"
|
||||
import { eq, sql } from 'drizzle-orm'
|
||||
import { models, unloadModelAfterB } from "../../config/ai"
|
||||
|
||||
const spamwatchMiddleware = spamwatchMiddlewareModule(isOnSpamWatch)
|
||||
export const flash_model = process.env.flashModel || "gemma3:4b"
|
||||
|
|
@ -51,7 +51,7 @@ type TextContext = Context & { message: Message.TextMessage }
|
|||
|
||||
type User = typeof schema.usersTable.$inferSelect
|
||||
|
||||
interface ModelInfo {
|
||||
export interface ModelInfo {
|
||||
name: string;
|
||||
label: string;
|
||||
descriptionEn: string;
|
||||
|
|
@ -67,59 +67,6 @@ interface OllamaResponse {
|
|||
response: string;
|
||||
}
|
||||
|
||||
export const models: ModelInfo[] = [
|
||||
{
|
||||
name: 'gemma3n',
|
||||
label: 'gemma3n',
|
||||
descriptionEn: 'Gemma3n is a family of open, light on-device models for general tasks.',
|
||||
descriptionPt: 'Gemma3n é uma família de modelos abertos, leves e para dispositivos locais, para tarefas gerais.',
|
||||
models: [
|
||||
{ name: 'gemma3n:e2b', label: 'Gemma3n e2b', parameterSize: '2B' },
|
||||
{ name: 'gemma3n:e4b', label: 'Gemma3n e4b', parameterSize: '4B' },
|
||||
]
|
||||
},
|
||||
{
|
||||
name: 'gemma3-abliterated',
|
||||
label: 'gemma3 Uncensored',
|
||||
descriptionEn: 'Gemma3-abliterated is a family of open, uncensored models for general tasks.',
|
||||
descriptionPt: 'Gemma3-abliterated é uma família de modelos abertos, não censurados, para tarefas gerais.',
|
||||
models: [
|
||||
{ name: 'huihui_ai/gemma3-abliterated:1b', label: 'Gemma3-abliterated 1B', parameterSize: '1b' },
|
||||
{ name: 'huihui_ai/gemma3-abliterated:4b', label: 'Gemma3-abliterated 4B', parameterSize: '4b' },
|
||||
]
|
||||
},
|
||||
{
|
||||
name: 'qwen3',
|
||||
label: 'Qwen3',
|
||||
descriptionEn: 'Qwen3 is a multilingual reasoning model series.',
|
||||
descriptionPt: 'Qwen3 é uma série de modelos multilingues.',
|
||||
models: [
|
||||
{ name: 'qwen3:4b', label: 'Qwen3 4B', parameterSize: '4B' },
|
||||
]
|
||||
},
|
||||
{
|
||||
name: 'deepseek',
|
||||
label: 'DeepSeek',
|
||||
descriptionEn: 'DeepSeek is a research model for reasoning tasks.',
|
||||
descriptionPt: 'DeepSeek é um modelo de pesquisa para tarefas de raciocínio.',
|
||||
models: [
|
||||
{ name: 'deepseek-r1:1.5b', label: 'DeepSeek 1.5B', parameterSize: '1.5B' },
|
||||
{ name: 'deepseek-r1:7b', label: 'DeepSeek 7B', parameterSize: '7B' },
|
||||
{ name: 'huihui_ai/deepseek-r1-abliterated:1.5b', label: 'DeepSeek Uncensored 1.5B', parameterSize: '1.5B' },
|
||||
{ name: 'huihui_ai/deepseek-r1-abliterated:7b', label: 'DeepSeek Uncensored 7B', parameterSize: '7B' },
|
||||
]
|
||||
},
|
||||
{
|
||||
name: 'phi3',
|
||||
label: 'Phi3',
|
||||
descriptionEn: 'Phi-3 is a family of lightweight 3B (Mini) and 14B (Medium) state-of-the-art open models by Microsoft.',
|
||||
descriptionPt: 'Phi-3 é uma família de modelos leves de 3B (Mini) e 14B (Médio) de última geração, abertos pela Microsoft.',
|
||||
models: [
|
||||
{ name: 'phi3:3.8b', label: 'Phi3 3.8B', parameterSize: '3.8B' },
|
||||
]
|
||||
}
|
||||
];
|
||||
|
||||
async function usingSystemPrompt(ctx: TextContext, db: NodePgDatabase<typeof schema>, botName: string, message: string): Promise<string> {
|
||||
const user = await db.query.usersTable.findMany({ where: (fields, { eq }) => eq(fields.telegramId, String(ctx.from!.id)), limit: 1 });
|
||||
if (user.length === 0) await ensureUserInDb(ctx, db);
|
||||
|
|
@ -263,16 +210,11 @@ function extractAxiosErrorMessage(error: unknown): string {
|
|||
return 'An unexpected error occurred.';
|
||||
}
|
||||
|
||||
function escapeMarkdown(text: string): string {
|
||||
return text.replace(/([_*\[\]()`>#\+\-=|{}.!~])/g, '\\$1');
|
||||
}
|
||||
|
||||
function containsUrls(text: string): boolean {
|
||||
return text.includes('http://') || text.includes('https://');
|
||||
return text.includes('http://') || text.includes('https://') || text.includes('.com') || text.includes('.net') || text.includes('.org') || text.includes('.io') || text.includes('.ai') || text.includes('.dev')
|
||||
}
|
||||
|
||||
async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Message, model: string, aiTemperature: number, originalMessage: string, db: NodePgDatabase<typeof schema>, userId: string): Promise<{ success: boolean; response?: string; error?: string }> {
|
||||
const Strings = getStrings(languageCode(ctx));
|
||||
async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Message, model: string, aiTemperature: number, originalMessage: string, db: NodePgDatabase<typeof schema>, userId: string, Strings: ReturnType<typeof getStrings>, showThinking: boolean): Promise<{ success: boolean; response?: string; error?: string }> {
|
||||
if (!ctx.chat) {
|
||||
return {
|
||||
success: false,
|
||||
|
|
@ -289,6 +231,8 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
await db.update(schema.usersTable)
|
||||
.set({ aiCharacters: sql`${schema.usersTable.aiCharacters} + ${promptCharCount}` })
|
||||
.where(eq(schema.usersTable.telegramId, userId));
|
||||
const paramSizeStr = models.find(m => m.name === model)?.models.find(m => m.name === model)?.parameterSize?.replace('B', '');
|
||||
const shouldKeepAlive = paramSizeStr ? Number(paramSizeStr) > unloadModelAfterB : false;
|
||||
|
||||
try {
|
||||
const aiResponse = await axios.post<unknown>(
|
||||
|
|
@ -297,6 +241,7 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
model,
|
||||
prompt,
|
||||
stream: true,
|
||||
keep_alive: shouldKeepAlive ? '1' : '0',
|
||||
options: {
|
||||
temperature: aiTemperature
|
||||
}
|
||||
|
|
@ -311,6 +256,16 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
let sentHeader = false;
|
||||
let firstChunk = true;
|
||||
const stream: NodeJS.ReadableStream = aiResponse.data as any;
|
||||
let thinkingMessageSent = false;
|
||||
let finalResponseText = '';
|
||||
|
||||
const formatThinkingMessage = (text: string) => {
|
||||
const withPlaceholders = text
|
||||
.replace(/___THINK_START___/g, `${Strings.ai.thinking}`)
|
||||
.replace(/___THINK_END___/g, `${Strings.ai.finishedThinking}`);
|
||||
return sanitizeMarkdownForTelegram(withPlaceholders);
|
||||
};
|
||||
|
||||
for await (const chunk of stream) {
|
||||
const lines = chunk.toString().split('\n');
|
||||
for (const line of lines) {
|
||||
|
|
@ -322,6 +277,22 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
console.error("[✨ AI | !] Error parsing chunk");
|
||||
continue;
|
||||
}
|
||||
if (model === thinking_model && !showThinking) {
|
||||
if (ln.response) {
|
||||
finalResponseText += ln.response;
|
||||
if (finalResponseText.includes('<think>') && !thinkingMessageSent) {
|
||||
await rateLimiter.editMessageWithRetry(
|
||||
ctx,
|
||||
ctx.chat.id,
|
||||
replyGenerating.message_id,
|
||||
modelHeader + Strings.ai.thinking,
|
||||
{ parse_mode: 'Markdown' }
|
||||
);
|
||||
thinkingMessageSent = true;
|
||||
}
|
||||
}
|
||||
continue;
|
||||
}
|
||||
if (model === thinking_model && ln.response) {
|
||||
if (ln.response.includes('<think>')) {
|
||||
const thinkMatch = ln.response.match(/<think>([\s\S]*?)<\/think>/);
|
||||
|
|
@ -338,9 +309,9 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
if (model === thinking_model) {
|
||||
let patchedThoughts = ln.response;
|
||||
const thinkTagRx = /<think>([\s\S]*?)<\/think>/g;
|
||||
patchedThoughts = patchedThoughts.replace(thinkTagRx, (p1) => p1.trim().length > 0 ? '`' + Strings.ai.thinking + '`' + p1 + '`' + Strings.ai.finishedThinking + '`' : '');
|
||||
patchedThoughts = patchedThoughts.replace(/<think>/g, '`' + Strings.ai.thinking + '`');
|
||||
patchedThoughts = patchedThoughts.replace(/<\/think>/g, '`' + Strings.ai.finishedThinking + '`');
|
||||
patchedThoughts = patchedThoughts.replace(thinkTagRx, (p1) => p1.trim().length > 0 ? '___THINK_START___' + p1.trim() + '___THINK_END___' : '');
|
||||
patchedThoughts = patchedThoughts.replace(/<think>/g, '___THINK_START___');
|
||||
patchedThoughts = patchedThoughts.replace(/<\/think>/g, '___THINK_END___');
|
||||
thoughts += patchedThoughts;
|
||||
fullResponse += patchedThoughts;
|
||||
} else {
|
||||
|
|
@ -356,7 +327,7 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
ctx,
|
||||
ctx.chat.id,
|
||||
replyGenerating.message_id,
|
||||
modelHeader + escapeMarkdown(fullResponse),
|
||||
modelHeader + formatThinkingMessage(fullResponse),
|
||||
{ parse_mode: 'Markdown' }
|
||||
);
|
||||
lastUpdateCharCount = fullResponse.length;
|
||||
|
|
@ -370,7 +341,7 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
ctx,
|
||||
ctx.chat.id,
|
||||
replyGenerating.message_id,
|
||||
modelHeader + escapeMarkdown(fullResponse),
|
||||
modelHeader + formatThinkingMessage(fullResponse),
|
||||
{ parse_mode: 'Markdown' }
|
||||
);
|
||||
lastUpdateCharCount = fullResponse.length;
|
||||
|
|
@ -379,6 +350,10 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
}
|
||||
}
|
||||
}
|
||||
if (model === thinking_model && !showThinking) {
|
||||
const cleanedResponse = finalResponseText.replace(/<think>[\s\S]*?<\/think>/g, '').trim();
|
||||
return { success: true, response: cleanedResponse };
|
||||
}
|
||||
status = Strings.ai.statusRendering;
|
||||
modelHeader = Strings.ai.modelHeader
|
||||
.replace("{model}", model)
|
||||
|
|
@ -388,7 +363,7 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
ctx,
|
||||
ctx.chat.id,
|
||||
replyGenerating.message_id,
|
||||
modelHeader + escapeMarkdown(fullResponse),
|
||||
modelHeader + formatThinkingMessage(fullResponse),
|
||||
{ parse_mode: 'Markdown' }
|
||||
);
|
||||
const responseCharCount = fullResponse.length;
|
||||
|
|
@ -432,13 +407,13 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
console.error("[✨ AI | !] Pull error:", pullMsg);
|
||||
return {
|
||||
success: false,
|
||||
error: `❌ Something went wrong while pulling ${escapeMarkdown(model)}: ${escapeMarkdown(pullMsg)}`,
|
||||
error: `❌ Something went wrong while pulling ${model}: ${pullMsg}`,
|
||||
};
|
||||
}
|
||||
console.log(`[✨ AI] ${model} pulled successfully`);
|
||||
return {
|
||||
success: true,
|
||||
response: Strings.ai.pulled.replace("{model}", escapeMarkdown(model)),
|
||||
response: Strings.ai.pulled.replace("{model}", model),
|
||||
};
|
||||
}
|
||||
}
|
||||
|
|
@ -449,9 +424,8 @@ async function getResponse(prompt: string, ctx: TextContext, replyGenerating: Me
|
|||
}
|
||||
}
|
||||
|
||||
async function handleAiReply(ctx: TextContext, model: string, prompt: string, replyGenerating: Message, aiTemperature: number, originalMessage: string, db: NodePgDatabase<typeof schema>, userId: string) {
|
||||
const Strings = getStrings(languageCode(ctx));
|
||||
const aiResponse = await getResponse(prompt, ctx, replyGenerating, model, aiTemperature, originalMessage, db, userId);
|
||||
async function handleAiReply(ctx: TextContext, model: string, prompt: string, replyGenerating: Message, aiTemperature: number, originalMessage: string, db: NodePgDatabase<typeof schema>, userId: string, Strings: ReturnType<typeof getStrings>, showThinking: boolean) {
|
||||
const aiResponse = await getResponse(prompt, ctx, replyGenerating, model, aiTemperature, originalMessage, db, userId, Strings, showThinking);
|
||||
if (!aiResponse) return;
|
||||
if (!ctx.chat) return;
|
||||
if (aiResponse.success && aiResponse.response) {
|
||||
|
|
@ -461,11 +435,17 @@ async function handleAiReply(ctx: TextContext, model: string, prompt: string, re
|
|||
.replace("{temperature}", aiTemperature)
|
||||
.replace("{status}", status) + "\n\n";
|
||||
const urlWarning = containsUrls(originalMessage) ? Strings.ai.urlWarning : '';
|
||||
let finalResponse = aiResponse.response;
|
||||
if (model === thinking_model) {
|
||||
finalResponse = finalResponse.replace(/___THINK_START___/g, `${Strings.ai.thinking}`)
|
||||
.replace(/___THINK_END___/g, `${Strings.ai.finishedThinking}`);
|
||||
}
|
||||
|
||||
await rateLimiter.editMessageWithRetry(
|
||||
ctx,
|
||||
ctx.chat.id,
|
||||
replyGenerating.message_id,
|
||||
modelHeader + sanitizeMarkdownForTelegram(aiResponse.response) + urlWarning,
|
||||
modelHeader + sanitizeMarkdownForTelegram(finalResponse) + urlWarning,
|
||||
{ parse_mode: 'Markdown' }
|
||||
);
|
||||
return;
|
||||
|
|
@ -480,7 +460,7 @@ async function handleAiReply(ctx: TextContext, model: string, prompt: string, re
|
|||
);
|
||||
}
|
||||
|
||||
async function getUserWithStringsAndModel(ctx: Context, db: NodePgDatabase<typeof schema>): Promise<{ user: User; Strings: ReturnType<typeof getStrings>; languageCode: string; customAiModel: string; aiTemperature: number }> {
|
||||
async function getUserWithStringsAndModel(ctx: Context, db: NodePgDatabase<typeof schema>): Promise<{ user: User; Strings: ReturnType<typeof getStrings>; languageCode: string; customAiModel: string; aiTemperature: number, showThinking: boolean }> {
|
||||
const userArr = await db.query.usersTable.findMany({ where: (fields, { eq }) => eq(fields.telegramId, String(ctx.from!.id)), limit: 1 });
|
||||
let user = userArr[0];
|
||||
if (!user) {
|
||||
|
|
@ -488,10 +468,10 @@ async function getUserWithStringsAndModel(ctx: Context, db: NodePgDatabase<typeo
|
|||
const newUserArr = await db.query.usersTable.findMany({ where: (fields, { eq }) => eq(fields.telegramId, String(ctx.from!.id)), limit: 1 });
|
||||
user = newUserArr[0];
|
||||
const Strings = getStrings(user.languageCode);
|
||||
return { user, Strings, languageCode: user.languageCode, customAiModel: user.customAiModel, aiTemperature: user.aiTemperature };
|
||||
return { user, Strings, languageCode: user.languageCode, customAiModel: user.customAiModel, aiTemperature: user.aiTemperature, showThinking: user.showThinking };
|
||||
}
|
||||
const Strings = getStrings(user.languageCode);
|
||||
return { user, Strings, languageCode: user.languageCode, customAiModel: user.customAiModel, aiTemperature: user.aiTemperature };
|
||||
return { user, Strings, languageCode: user.languageCode, customAiModel: user.customAiModel, aiTemperature: user.aiTemperature, showThinking: user.showThinking };
|
||||
}
|
||||
|
||||
export function getModelLabelByName(name: string): string {
|
||||
|
|
@ -547,7 +527,7 @@ export default (bot: Telegraf<Context>, db: NodePgDatabase<typeof schema>) => {
|
|||
|
||||
async function aiCommandHandler(ctx: TextContext, command: 'ask' | 'think' | 'ai') {
|
||||
const reply_to_message_id = replyToMessageId(ctx);
|
||||
const { user, Strings, customAiModel, aiTemperature } = await getUserWithStringsAndModel(ctx, db);
|
||||
const { user, Strings, customAiModel, aiTemperature, showThinking } = await getUserWithStringsAndModel(ctx, db);
|
||||
const message = ctx.message.text;
|
||||
const author = ("@" + ctx.from?.username) || ctx.from?.first_name || "Unknown";
|
||||
|
||||
|
|
@ -586,7 +566,7 @@ export default (bot: Telegraf<Context>, db: NodePgDatabase<typeof schema>) => {
|
|||
...(reply_to_message_id && { reply_parameters: { message_id: reply_to_message_id } })
|
||||
});
|
||||
const prompt = sanitizeForJson(await usingSystemPrompt(ctx, db, botName, fixedMsg));
|
||||
await handleAiReply(ctx, model, prompt, replyGenerating, aiTemperature, fixedMsg, db, user.telegramId);
|
||||
await handleAiReply(ctx, model, prompt, replyGenerating, aiTemperature, fixedMsg, db, user.telegramId, Strings, showThinking);
|
||||
};
|
||||
|
||||
if (isProcessing) {
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue