feat(hf-transformer): use langchain's inbuilt transformer class

Dockerfile

@@ -17,22 +17,11 @@ RUN yarn build
 
 FROM node:24.5.0-slim
 
-RUN apt-get update && \
-    apt-get install -y \
-    python3 \
-    python3-pip \
-    python3-venv \
-    python3-dev \
-    sqlite3 \
-    git \
-    build-essential \
-    libxslt-dev \
-    zlib1g-dev \
-    libffi-dev \
-    libssl-dev \
-    uwsgi \
-    uwsgi-plugin-python3 \
-    curl \
+RUN apt-get update && apt-get install -y \
+    python3-dev python3-babel python3-venv python-is-python3 \
+    uwsgi uwsgi-plugin-python3 \
+    git build-essential libxslt-dev zlib1g-dev libffi-dev libssl-dev \
+    curl sudo \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /home/perplexica
@@ -45,28 +34,39 @@ COPY drizzle ./drizzle
 
 RUN mkdir /home/perplexica/uploads
 
-RUN useradd --system --home-dir /usr/local/searxng --shell /bin/sh searxng
-
-WORKDIR /usr/local/searxng
-RUN git clone https://github.com/searxng/searxng.git . && \
-    python3 -m venv venv && \
-    . venv/bin/activate && \
-    pip install --upgrade pip setuptools wheel pyyaml && \
-    pip install -r requirements.txt && \
-    pip install uwsgi
+RUN useradd --shell /bin/bash --system \
+    --home-dir "/usr/local/searxng" \
+    --comment 'Privacy-respecting metasearch engine' \
+    searxng
 
+RUN mkdir "/usr/local/searxng"
 RUN mkdir -p /etc/searxng
+RUN chown -R "searxng:searxng" "/usr/local/searxng"
+
 COPY searxng/settings.yml /etc/searxng/settings.yml
 COPY searxng/limiter.toml /etc/searxng/limiter.toml
 COPY searxng/uwsgi.ini /etc/searxng/uwsgi.ini
+RUN chown -R searxng:searxng /etc/searxng
 
-RUN chown -R searxng:searxng /usr/local/searxng /etc/searxng
+USER searxng
+
+RUN git clone "https://github.com/searxng/searxng" \
+                   "/usr/local/searxng/searxng-src"
+
+RUN python3 -m venv "/usr/local/searxng/searx-pyenv"
+RUN "/usr/local/searxng/searx-pyenv/bin/pip" install --upgrade pip setuptools wheel pyyaml msgspec
+RUN cd "/usr/local/searxng/searxng-src" && \
+    "/usr/local/searxng/searx-pyenv/bin/pip" install --use-pep517 --no-build-isolation -e .
+
+USER root
 
 WORKDIR /home/perplexica
 COPY entrypoint.sh ./entrypoint.sh
 RUN chmod +x ./entrypoint.sh
 RUN sed -i 's/\r$//' ./entrypoint.sh || true
 
+RUN echo "searxng ALL=(ALL) NOPASSWD: ALL" >> /etc/sudoers
+
 EXPOSE 3000 8080
 
 ENV SEARXNG_API_URL=http://localhost:8080

README.md

@@ -79,19 +79,19 @@ There are mainly 2 ways of installing Perplexica - With Docker, Without Docker.
 Perplexica can be easily run using Docker. Simply run the following command:
 
 ```bash
-docker run -p 3000:3000 --name perplexica itzcrazykns1337/perplexica:latest
+docker run -d -p 3000:3000 -v perplexica-data:/home/perplexica/data -v perplexica-uploads:/home/perplexica/uploads --name perplexica itzcrazykns1337/perplexica:latest
 ```
 
 This will pull and start the Perplexica container with the bundled SearxNG search engine. Once running, open your browser and navigate to http://localhost:3000. You can then configure your settings (API keys, models, etc.) directly in the setup screen.
 
-**Note**: The image includes both Perplexica and SearxNG, so no additional setup is required.
+**Note**: The image includes both Perplexica and SearxNG, so no additional setup is required. The `-v` flags create persistent volumes for your data and uploaded files.
 
 #### Using Perplexica with Your Own SearxNG Instance
 
 If you already have SearxNG running, you can use the slim version of Perplexica:
 
 ```bash
-docker run -p 3000:3000 -e SEARXNG_API_URL=http://your-searxng-url:8080 --name perplexica itzcrazykns1337/perplexica:slim-latest
+docker run -d -p 3000:3000 -e SEARXNG_API_URL=http://your-searxng-url:8080 -v perplexica-data:/home/perplexica/data -v perplexica-uploads:/home/perplexica/uploads --name perplexica itzcrazykns1337/perplexica:slim-latest
 ```
 
 **Important**: Make sure your SearxNG instance has:
@@ -118,7 +118,7 @@ If you prefer to build from source or need more control:
 
    ```bash
    docker build -t perplexica .
-   docker run -p 3000:3000 --name perplexica perplexica
+   docker run -d -p 3000:3000 -v perplexica-data:/home/perplexica/data -v perplexica-uploads:/home/perplexica/uploads --name perplexica perplexica
    ```
 
 5. Access Perplexica at http://localhost:3000 and configure your settings in the setup screen.

docker-compose.yaml

@@ -0,0 +1,15 @@
+services:
+  perplexica:
+    image: itzcrazykns1337/perplexica:latest
+    ports:
+      - '3000:3000'
+    volumes:
+      - data:/home/perplexica/data
+      - uploads:/home/perplexica/uploads
+    restart: unless-stopped
+
+volumes:
+  data:
+    name: 'perplexica-data'
+  uploads:
+    name: 'perplexica-uploads'

docs/API/SEARCH.md

@@ -17,6 +17,7 @@ Before making search requests, you'll need to get the available providers and th
 Returns a list of all active providers with their available chat and embedding models.
 
 **Response Example:**
+
 ```json
 {
   "providers": [

docs/installation/UPDATING.md

@@ -10,7 +10,7 @@ Simply pull the latest image and restart your container:
 docker pull itzcrazykns1337/perplexica:latest
 docker stop perplexica
 docker rm perplexica
-docker run -p 3000:3000 --name perplexica itzcrazykns1337/perplexica:latest
+docker run -d -p 3000:3000 -v perplexica-data:/home/perplexica/data -v perplexica-uploads:/home/perplexica/uploads --name perplexica itzcrazykns1337/perplexica:latest
 ```
 
 For slim version:
@@ -19,7 +19,7 @@ For slim version:
 docker pull itzcrazykns1337/perplexica:slim-latest
 docker stop perplexica
 docker rm perplexica
-docker run -p 3000:3000 -e SEARXNG_API_URL=http://your-searxng-url:8080 --name perplexica itzcrazykns1337/perplexica:slim-latest
+docker run -d -p 3000:3000 -e SEARXNG_API_URL=http://your-searxng-url:8080 -v perplexica-data:/home/perplexica/data -v perplexica-uploads:/home/perplexica/uploads --name perplexica itzcrazykns1337/perplexica:slim-latest
 ```
 
 Once updated, go to http://localhost:3000 and verify the latest changes. Your settings are preserved automatically.

entrypoint.sh

@@ -1,24 +1,32 @@
 #!/bin/sh
 set -e
 
-cd /usr/local/searxng
-export SEARXNG_SETTINGS_PATH=/etc/searxng/settings.yml
-
-# Start SearXNG in background with all output redirected to /dev/null
-/usr/local/searxng/venv/bin/uwsgi \
-  --http-socket 0.0.0.0:8080 \
-  --ini /etc/searxng/uwsgi.ini \
-  --virtualenv /usr/local/searxng/venv \
-  --disable-logging > /dev/null 2>&1 &
-
 echo "Starting SearXNG..."
+
+sudo -H -u searxng bash -c "cd /usr/local/searxng/searxng-src && export SEARXNG_SETTINGS_PATH='/etc/searxng/settings.yml' && export FLASK_APP=searx/webapp.py && /usr/local/searxng/searx-pyenv/bin/python -m flask run --host=0.0.0.0 --port=8080" &
+SEARXNG_PID=$!
+
+echo "Waiting for SearXNG to be ready..."
 sleep 5
 
+COUNTER=0
+MAX_TRIES=30
 until curl -s http://localhost:8080 > /dev/null 2>&1; do
+  COUNTER=$((COUNTER+1))
+  if [ $COUNTER -ge $MAX_TRIES ]; then
+    echo "Warning: SearXNG health check timeout, but continuing..."
+    break
+  fi
   sleep 1
 done
-echo "SearXNG started successfully"
+
+if curl -s http://localhost:8080 > /dev/null 2>&1; then
+  echo "SearXNG started successfully (PID: $SEARXNG_PID)"
+else
+  echo "SearXNG may not be fully ready, but continuing (PID: $SEARXNG_PID)"
+fi
 
 cd /home/perplexica
 echo "Starting Perplexica..."
+
 exec node server.js

package.json

@@ -1,6 +1,6 @@
 {
   "name": "perplexica-frontend",
-  "version": "1.11.0",
+  "version": "1.11.1",
   "license": "MIT",
   "author": "ItzCrazyKns",
   "scripts": {

src/components/Settings/Sections/Models/ModelProvider.tsx

@@ -112,100 +112,96 @@ const ModelProvider = ({
           >
             <div className="border-t border-light-200 dark:border-dark-200" />
             <div className="flex flex-col gap-y-4 px-5 py-4">
-              {modelProvider.chatModels.length > 0 && (
-                <div className="flex flex-col gap-y-2">
-                  <div className="flex flex-row w-full justify-between items-center">
-                    <p className="text-[11px] lg:text-xs text-black/70 dark:text-white/70">
-                      Chat models
-                    </p>
-                    <AddModel
-                      providerId={modelProvider.id}
-                      setProviders={setProviders}
-                      type="chat"
-                    />
-                  </div>
-                  <div className="flex flex-col gap-2">
-                    {modelProvider.chatModels.some((m) => m.key === 'error') ? (
-                      <div className="flex flex-row items-center gap-2 text-xs lg:text-sm text-red-500 dark:text-red-400 rounded-lg bg-red-50 dark:bg-red-950/20 px-3 py-2 border border-red-200 dark:border-red-900/30">
-                        <AlertCircle size={16} className="shrink-0" />
-                        <span className="break-words">
-                          {
-                            modelProvider.chatModels.find(
-                              (m) => m.key === 'error',
-                            )?.name
-                          }
-                        </span>
-                      </div>
-                    ) : (
-                      <div className="flex flex-row flex-wrap gap-2">
-                        {modelProvider.chatModels.map((model, index) => (
-                          <div
-                            key={`${modelProvider.id}-chat-${model.key}-${index}`}
-                            className="flex flex-row items-center space-x-1 text-xs lg:text-sm text-black/70 dark:text-white/70 rounded-lg bg-light-secondary dark:bg-dark-secondary px-3 py-1.5"
-                          >
-                            <span>{model.name}</span>
-                            <button
-                              onClick={() => {
-                                handleModelDelete('chat', model.key);
-                              }}
-                            >
-                              <X size={12} />
-                            </button>
-                          </div>
-                        ))}
-                      </div>
-                    )}
-                  </div>
+              <div className="flex flex-col gap-y-2">
+                <div className="flex flex-row w-full justify-between items-center">
+                  <p className="text-[11px] lg:text-xs text-black/70 dark:text-white/70">
+                    Chat models
+                  </p>
+                  <AddModel
+                    providerId={modelProvider.id}
+                    setProviders={setProviders}
+                    type="chat"
+                  />
                 </div>
-              )}
-              {modelProvider.embeddingModels.length > 0 && (
-                <div className="flex flex-col gap-y-2">
-                  <div className="flex flex-row w-full justify-between items-center">
-                    <p className="text-[11px] lg:text-xs text-black/70 dark:text-white/70">
-                      Embedding models
-                    </p>
-                    <AddModel
-                      providerId={modelProvider.id}
-                      setProviders={setProviders}
-                      type="embedding"
-                    />
-                  </div>
-                  <div className="flex flex-col gap-2">
-                    {modelProvider.embeddingModels.some(
-                      (m) => m.key === 'error',
-                    ) ? (
-                      <div className="flex flex-row items-center gap-2 text-xs lg:text-sm text-red-500 dark:text-red-400 rounded-lg bg-red-50 dark:bg-red-950/20 px-3 py-2 border border-red-200 dark:border-red-900/30">
-                        <AlertCircle size={16} className="shrink-0" />
-                        <span className="break-words">
-                          {
-                            modelProvider.embeddingModels.find(
-                              (m) => m.key === 'error',
-                            )?.name
-                          }
-                        </span>
-                      </div>
-                    ) : (
-                      <div className="flex flex-row flex-wrap gap-2">
-                        {modelProvider.embeddingModels.map((model, index) => (
-                          <div
-                            key={`${modelProvider.id}-embedding-${model.key}-${index}`}
-                            className="flex flex-row items-center space-x-1 text-xs lg:text-sm text-black/70 dark:text-white/70 rounded-lg bg-light-secondary dark:bg-dark-secondary px-3 py-1.5"
+                <div className="flex flex-col gap-2">
+                  {modelProvider.chatModels.some((m) => m.key === 'error') ? (
+                    <div className="flex flex-row items-center gap-2 text-xs lg:text-sm text-red-500 dark:text-red-400 rounded-lg bg-red-50 dark:bg-red-950/20 px-3 py-2 border border-red-200 dark:border-red-900/30">
+                      <AlertCircle size={16} className="shrink-0" />
+                      <span className="break-words">
+                        {
+                          modelProvider.chatModels.find(
+                            (m) => m.key === 'error',
+                          )?.name
+                        }
+                      </span>
+                    </div>
+                  ) : (
+                    <div className="flex flex-row flex-wrap gap-2">
+                      {modelProvider.chatModels.map((model, index) => (
+                        <div
+                          key={`${modelProvider.id}-chat-${model.key}-${index}`}
+                          className="flex flex-row items-center space-x-1 text-xs lg:text-sm text-black/70 dark:text-white/70 rounded-lg bg-light-secondary dark:bg-dark-secondary px-3 py-1.5"
+                        >
+                          <span>{model.name}</span>
+                          <button
+                            onClick={() => {
+                              handleModelDelete('chat', model.key);
+                            }}
                           >
-                            <span>{model.name}</span>
-                            <button
-                              onClick={() => {
-                                handleModelDelete('embedding', model.key);
-                              }}
-                            >
-                              <X size={12} />
-                            </button>
-                          </div>
-                        ))}
-                      </div>
-                    )}
-                  </div>
+                            <X size={12} />
+                          </button>
+                        </div>
+                      ))}
+                    </div>
+                  )}
                 </div>
-              )}
+              </div>
+              <div className="flex flex-col gap-y-2">
+                <div className="flex flex-row w-full justify-between items-center">
+                  <p className="text-[11px] lg:text-xs text-black/70 dark:text-white/70">
+                    Embedding models
+                  </p>
+                  <AddModel
+                    providerId={modelProvider.id}
+                    setProviders={setProviders}
+                    type="embedding"
+                  />
+                </div>
+                <div className="flex flex-col gap-2">
+                  {modelProvider.embeddingModels.some(
+                    (m) => m.key === 'error',
+                  ) ? (
+                    <div className="flex flex-row items-center gap-2 text-xs lg:text-sm text-red-500 dark:text-red-400 rounded-lg bg-red-50 dark:bg-red-950/20 px-3 py-2 border border-red-200 dark:border-red-900/30">
+                      <AlertCircle size={16} className="shrink-0" />
+                      <span className="break-words">
+                        {
+                          modelProvider.embeddingModels.find(
+                            (m) => m.key === 'error',
+                          )?.name
+                        }
+                      </span>
+                    </div>
+                  ) : (
+                    <div className="flex flex-row flex-wrap gap-2">
+                      {modelProvider.embeddingModels.map((model, index) => (
+                        <div
+                          key={`${modelProvider.id}-embedding-${model.key}-${index}`}
+                          className="flex flex-row items-center space-x-1 text-xs lg:text-sm text-black/70 dark:text-white/70 rounded-lg bg-light-secondary dark:bg-dark-secondary px-3 py-1.5"
+                        >
+                          <span>{model.name}</span>
+                          <button
+                            onClick={() => {
+                              handleModelDelete('embedding', model.key);
+                            }}
+                          >
+                            <X size={12} />
+                          </button>
+                        </div>
+                      ))}
+                    </div>
+                  )}
+                </div>
+              </div>
             </div>
           </motion.div>
         )}

src/components/Settings/SettingsButtonMobile.tsx

@@ -9,7 +9,7 @@ const SettingsButtonMobile = () => {
   return (
     <>
       <button className="lg:hidden" onClick={() => setIsOpen(true)}>
-        <Settings size={18}/>
+        <Settings size={18} />
       </button>
       <AnimatePresence>
         {isOpen && <SettingsDialogue isOpen={isOpen} setIsOpen={setIsOpen} />}

src/components/Setup/SetupConfig.tsx

@@ -63,8 +63,7 @@ const SetupConfig = ({
     }
   };
 
-  const hasProviders =
-    providers.filter((p) => p.chatModels.length > 0).length > 0;
+  const hasProviders = providers.length > 0;
 
   return (
     <div className="w-[95vw] md:w-[80vw] lg:w-[65vw] mx-auto px-2 sm:px-4 md:px-6 flex flex-col space-y-6">

src/lib/huggingfaceTransformer.ts

@@ -1,76 +0,0 @@
-import { Embeddings, type EmbeddingsParams } from '@langchain/core/embeddings';
-import { chunkArray } from '@langchain/core/utils/chunk_array';
-
-export interface HuggingFaceTransformersEmbeddingsParams
-  extends EmbeddingsParams {
-  modelName: string;
-
-  model: string;
-
-  timeout?: number;
-
-  batchSize?: number;
-
-  stripNewLines?: boolean;
-}
-
-export class HuggingFaceTransformersEmbeddings
-  extends Embeddings
-  implements HuggingFaceTransformersEmbeddingsParams
-{
-  modelName = 'Xenova/all-MiniLM-L6-v2';
-
-  model = 'Xenova/all-MiniLM-L6-v2';
-
-  batchSize = 512;
-
-  stripNewLines = true;
-
-  timeout?: number;
-
-  constructor(fields?: Partial<HuggingFaceTransformersEmbeddingsParams>) {
-    super(fields ?? {});
-
-    this.modelName = fields?.model ?? fields?.modelName ?? this.model;
-    this.model = this.modelName;
-    this.stripNewLines = fields?.stripNewLines ?? this.stripNewLines;
-    this.timeout = fields?.timeout;
-  }
-
-  async embedDocuments(texts: string[]): Promise<number[][]> {
-    const batches = chunkArray(
-      this.stripNewLines ? texts.map((t) => t.replace(/\n/g, ' ')) : texts,
-      this.batchSize,
-    );
-
-    const batchRequests = batches.map((batch) => this.runEmbedding(batch));
-    const batchResponses = await Promise.all(batchRequests);
-    const embeddings: number[][] = [];
-
-    for (let i = 0; i < batchResponses.length; i += 1) {
-      const batchResponse = batchResponses[i];
-      for (let j = 0; j < batchResponse.length; j += 1) {
-        embeddings.push(batchResponse[j]);
-      }
-    }
-
-    return embeddings;
-  }
-
-  async embedQuery(text: string): Promise<number[]> {
-    const data = await this.runEmbedding([
-      this.stripNewLines ? text.replace(/\n/g, ' ') : text,
-    ]);
-    return data[0];
-  }
-
-  private async runEmbedding(texts: string[]) {
-    const { pipeline } = await import('@huggingface/transformers');
-    const pipe = await pipeline('feature-extraction', this.model);
-
-    return this.caller.call(async () => {
-      const output = await pipe(texts, { pooling: 'mean', normalize: true });
-      return output.tolist();
-    });
-  }
-}

src/lib/models/providers/transformers.ts

@@ -4,8 +4,7 @@ import BaseModelProvider from './baseProvider';
 import { Embeddings } from '@langchain/core/embeddings';
 import { UIConfigField } from '@/lib/config/types';
 import { getConfiguredModelProviderById } from '@/lib/config/serverRegistry';
-import { HuggingFaceTransformersEmbeddings } from '@/lib/huggingfaceTransformer';
-
+import { HuggingFaceTransformersEmbeddings } from '@langchain/community/embeddings/huggingface_transformers';
 interface TransformersConfig {}
 
 const defaultEmbeddingModels: Model[] = [