Spaces:

dmartincy
/

document-translation

Runtime error

App Files Files Community

dmartincy commited on Dec 25, 2024

Commit

e27f03b

1 Parent(s): c6d4e5d

Add document authoring

Browse files

Files changed (11) hide show

Dockerfile +48 -9
Modelfile +0 -2
Modelfile.embeddings +0 -1
app.py +0 -10
index.html +100 -3
nginx.conf +11 -1
public/javascript-app.js +90 -0
service-config.yml +3 -3
start-container.sh +0 -28
start-services.sh +23 -2
vite.config.ts +14 -3

Dockerfile CHANGED Viewed

@@ -1,5 +1,12 @@
 FROM pspdfkit/ai-assistant:nightly
 # Set environment variables
 ENV HOME=/home/node \
     PNPM_HOME=/root/.local/share/pnpm \
@@ -66,19 +73,15 @@ RUN mkdir -p /var/cache/nginx \
                       /run/nginx
 # Copy application files to /base BEFORE switching user
-COPY --chown=node:node app.py service-config.yml /base/app/
 COPY --chown=node:node nginx.conf /etc/nginx/nginx.conf
-COPY --chown=node:node healthcheck.sh start-services.sh /base/app/
 COPY --chown=node:node service-config.yml /service-config.yml
-# Create app directories and set up Python
 USER node
 RUN mkdir -p $HOME/models && \
-    mkdir -p $HOME/pgdata && \
-    python3 -m venv $HOME/venv && \
-    $HOME/venv/bin/pip install flask
-# pgvector installation (needs root)
 USER root
 RUN git clone https://github.com/pgvector/pgvector.git && \
     cd pgvector && \
@@ -91,9 +94,45 @@ RUN git clone https://github.com/pgvector/pgvector.git && \
 USER node
 RUN wget -q https://huggingface.co/bartowski/gemma-2-2b-it-GGUF/resolve/main/gemma-2-2b-it-Q4_K_M.gguf -O $HOME/models/gemma-2b.gguf && \
     wget -q https://huggingface.co/leliuga/all-MiniLM-L6-v2-GGUF/resolve/main/all-MiniLM-L6-v2.F16.gguf -O $HOME/models/embeddings.gguf && \
-    chmod +x /base/app/healthcheck.sh /base/app/start-services.sh
-EXPOSE 5432 7860 4000 8080 8081 8082
 USER root

 FROM pspdfkit/ai-assistant:nightly
+# The main ai assistant JS app is in /base/app.
+# Then I have a document authoring web app that is a Vite server that serves index.html and
+# communicates with the ai assistant.
+#
+# Where should we put this web app? In /base/app it's gonna be problematic.
+# New folder? /docauth?
 # Set environment variables
 ENV HOME=/home/node \
     PNPM_HOME=/root/.local/share/pnpm \
                       /run/nginx
 # Copy application files to /base BEFORE switching user
 COPY --chown=node:node nginx.conf /etc/nginx/nginx.conf
+COPY --chown=node:node start-services.sh /base/app/
 COPY --chown=node:node service-config.yml /service-config.yml
 USER node
 RUN mkdir -p $HOME/models && \
+    mkdir -p $HOME/pgdata
+# Install pgvector (needs root)
 USER root
 RUN git clone https://github.com/pgvector/pgvector.git && \
     cd pgvector && \
 USER node
 RUN wget -q https://huggingface.co/bartowski/gemma-2-2b-it-GGUF/resolve/main/gemma-2-2b-it-Q4_K_M.gguf -O $HOME/models/gemma-2b.gguf && \
     wget -q https://huggingface.co/leliuga/all-MiniLM-L6-v2-GGUF/resolve/main/all-MiniLM-L6-v2.F16.gguf -O $HOME/models/embeddings.gguf && \
+    chmod +x /base/app/start-services.sh
+USER root
+# Fix permissions for scripts and pnpm directory
+RUN chmod +x /base/app/start-services.sh && \
+    mkdir -p /root/.local/share/pnpm && \
+    chmod 755 /root && \
+    chmod 755 /root/.local && \
+    chmod 755 /root/.local/share && \
+    chmod -R 755 /root/.local/share/pnpm && \
+    chown -R node:node /root/.local/share/pnpm
+# Set up document authoring in a separate directory
+WORKDIR /docauth
+# Create directory and set permissions
+USER root
+RUN mkdir -p /docauth && \
+    chown -R node:node /docauth && \
+    chmod -R 755 /docauth
+USER node
+# Copy package files
+COPY --chown=node:node package*.json ./
+COPY --chown=node:node tsconfig*.json ./
+COPY --chown=node:node vite.config.ts ./
+COPY --chown=node:node .eslintrc.cjs ./
+# Install dependencies
+RUN npm install
+# Copy source files
+COPY --chown=node:node src ./src
+COPY --chown=node:node public ./public
+COPY --chown=node:node index.html ./
+# Expose Vite's port (using 7860 as configured in vite.config.ts)
+EXPOSE 7860
 USER root

Modelfile DELETED Viewed

	@@ -1,2 +0,0 @@
1	- FROM /models/gemma-2-2b-it-Q4_K_M.gguf
2	-

Modelfile.embeddings DELETED Viewed

	@@ -1 +0,0 @@
1	- FROM /models/mxbai-embed-large-v1-f16.gguf

app.py DELETED Viewed

@@ -1,10 +0,0 @@
-from flask import Flask
-app = Flask(__name__)
-@app.route('/')
-def home():
-    return "PostgreSQL is running with Flask!"
-if __name__ == '__main__':
-    app.run(host='0.0.0.0', port=7860)

index.html CHANGED Viewed

@@ -5,10 +5,107 @@
     <link rel="icon" type="image/svg+xml" href="/vite.svg" />
     <meta name="viewport" content="width=device-width, initial-scale=1.0" />
     <title>Document Authoring SDK</title>
   </head>
   <body>
-    <ul>
-      <li><a href='./javascript.html'>JavaScript example</a></li>
-    </ul>
   </body>
 </html>

     <link rel="icon" type="image/svg+xml" href="/vite.svg" />
     <meta name="viewport" content="width=device-width, initial-scale=1.0" />
     <title>Document Authoring SDK</title>
+    <style>
+      .spinner {
+        display: inline-block;
+        width: 16px;
+        height: 16px;
+        border: 2px solid #f3f3f3;
+        border-top: 2px solid #3498db;
+        border-radius: 50%;
+        animation: spin 1s linear infinite;
+        margin-right: 8px;
+        vertical-align: middle;
+      }
+      @keyframes spin {
+        0% { transform: rotate(0deg); }
+        100% { transform: rotate(360deg); }
+      }
+      #translateButton:disabled {
+        cursor: not-allowed;
+        opacity: 0.7;
+      }
+    </style>
   </head>
   <body>
+    <button id="translateButton">Translate to English</button>
+    <div id="editor" style="border: 1px solid black; width: 1024px; height: 600px; position: relative"></div>
+    <script src="https://document-authoring-cdn.pspdfkit.com/releases/document-authoring-1.0.26-umd.js" crossorigin></script>
+    <script>
+      (async () => {
+        const docAuthSystem = await DocAuth.createDocAuthSystem({
+          // assets: { base: '<CUSTOM_DOCUMENT_AUTHORING_ASSETS_LOCATION>' },
+          // licenseKey: '<YOUR_LICENSE_KEY>',
+        });
+        const editor = await docAuthSystem.createEditor(
+          document.getElementById("editor"),
+          {
+            document: await docAuthSystem.createDocumentFromPlaintext('Hi there!\n\nHow are you?'),
+          }
+        );
+        // Translation function.
+        async function translate(content, lang = "English") {
+          try {
+            const response = await fetch('/api/aia/v1/chat/completions', {
+              method: 'POST',
+              headers: {
+                'Content-Type': 'application/json',
+                'Authorization': 'Token token=eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9.eyJpYXQiOjE3MzMyNjQ2MDcsImV4cCI6MTczMzUyMzgwN30.WWyks6NOU43_2OcZX82z3P4QUwQiUA61SeTZsNobIGoVg3uDTkedXWqISIB7QqTnRyf_D92t_pGr5g_GzqBaOTqaXYJJPco9mtZjtAyUpdWr1UHu4jx-zoKUpCeguZQZbHG6vqYzu1BNV1iPdij2OIoUrr_DY--G-6m5-31E0LcDZb7ysUyZL1Ai_q7PPRwTS8mEtvuF4QO6Cw_a77X7Qh9x_V8a_WwNMFwQMRkIUAsRx3IOI10q0hdwP-k8xYtTbHQO-z66pRl6I71ATclR7FS5FraoLLmWlq8lqNKfzeOgER8Kcf8uB87zHQbfk0QR8AfJGmZDE8pG5F3QG29MeA'
+              },
+              body: JSON.stringify({
+                messages: [
+                  {
+                    role: "system",
+                    content: "You are a translator. Only provide the translation, no explanations or additional text."
+                  },
+                  {
+                    role: "user",
+                    content: `Translate the following text to ${lang}: ${content}`
+                  }
+                ],
+                model: "gemma-2b",
+                stream: false
+              })
+            });
+            if (!response.ok) {
+              throw new Error('Translation request failed');
+            }
+            const data = await response.json();
+            console.log('Translation API response:', data);
+            return data.choices[0].message.content;
+          } catch (error) {
+            console.error("Translation error:", error);
+            throw new Error("Failed to generate translation");
+          }
+        }
+        async function testDummy() {
+          try {
+            const response = await fetch('/dummy');
+            const text = await response.text();
+            console.log('Dummy response:', text);
+          } catch (error) {
+            console.error('Dummy error:', error);
+          }
+        }
+        // Dev exports
+        window.DocAuth = DocAuth;
+        window.docAuthSystem = docAuthSystem;
+        window.editor = editor;
+        const testButton = document.createElement('button');
+        testButton.textContent = 'Test Dummy';
+        document.body.insertBefore(testButton, document.getElementById('editor'));
+        testButton.addEventListener('click', testDummy);
+      })();
+    </script>
   </body>
 </html>

nginx.conf CHANGED Viewed

@@ -4,8 +4,18 @@ events {
 http {
     server {
-        listen 8080;
         location /v1/embeddings {
             proxy_pass http://127.0.0.1:8081;
         }

 http {
     server {
+        listen 7860;
+        location / {
+            return 200 'Hello World!<br><a href="https://dmartincy-document-translation.hf.space/dummy">Go to dummy page</a>';
+            add_header Content-Type text/html;
+        }
+        location /dummy {
+            return 200 'Hello dummy';
+            add_header Content-Type text/html;
+        }
         location /v1/embeddings {
             proxy_pass http://127.0.0.1:8081;
         }

public/javascript-app.js ADDED Viewed

	@@ -0,0 +1,90 @@

+import DocAuth from "https://document-authoring-cdn.pspdfkit.com/releases/document-authoring-1.0.26-esm.js";
+const docAuthSystem = await DocAuth.createDocAuthSystem({
+  // assets: { base: '<CUSTOM_DOCUMENT_AUTHORING_ASSETS_LOCATION>' },
+  // licenseKey: '<YOUR_LICENSE_KEY>',
+});
+const editor = await docAuthSystem.createEditor(
+  document.getElementById("editor"),
+  {
+    document: await docAuthSystem.importDOCX(fetch("./Spanish.docx")),
+  }
+);
+// Translation function.
+async function translate(content, lang = "English") {
+  try {
+    const response = await fetch('/api/aia/v1/chat/completions', {
+      method: 'POST',
+      headers: {
+        'Content-Type': 'application/json',
+        'Authorization': 'Token token=eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9.eyJpYXQiOjE3MzMyNjQ2MDcsImV4cCI6MTczMzUyMzgwN30.WWyks6NOU43_2OcZX82z3P4QUwQiUA61SeTZsNobIGoVg3uDTkedXWqISIB7QqTnRyf_D92t_pGr5g_GzqBaOTqaXYJJPco9mtZjtAyUpdWr1UHu4jx-zoKUpCeguZQZbHG6vqYzu1BNV1iPdij2OIoUrr_DY--G-6m5-31E0LcDZb7ysUyZL1Ai_q7PPRwTS8mEtvuF4QO6Cw_a77X7Qh9x_V8a_WwNMFwQMRkIUAsRx3IOI10q0hdwP-k8xYtTbHQO-z66pRl6I71ATclR7FS5FraoLLmWlq8lqNKfzeOgER8Kcf8uB87zHQbfk0QR8AfJGmZDE8pG5F3QG29MeA'
+      },
+      body: JSON.stringify({
+        messages: [
+          {
+            role: "system",
+            content: "You are a translator. Only provide the translation, no explanations or additional text."
+          },
+          {
+            role: "user",
+            content: `Translate the following text to ${lang}: ${content}`
+          }
+        ],
+        model: "gemma-2b",
+        stream: false
+      })
+    });
+    if (!response.ok) {
+      throw new Error('Translation request failed');
+    }
+    const data = await response.json();
+    console.log('Translation API response:', data);
+    return data.choices[0].message.content;
+  } catch (error) {
+    console.error("Translation error:", error);
+    throw new Error("Failed to generate translation");
+  }
+}
+// Function to recursively translate text in the JSON structure.
+async function translateRecursive(obj) {
+  for (let key in obj) {
+    if (typeof obj[key] === "string" && key === "text" && obj[key].length > 0) {
+      obj[key] = await translate(obj[key]);
+    } else if (typeof obj[key] === "object") {
+      await translateRecursive(obj[key]);
+    }
+  }
+}
+// Function to translate the document.
+async function translateDocument() {
+  try {
+    translateButton.disabled = true;
+    translateButton.innerHTML = '<span class="spinner"></span> Translating...';
+    const jsonDoc = await editor.currentDocument().saveDocument();
+    await translateRecursive(jsonDoc.container.document);
+    const translatedDoc = await docAuthSystem.loadDocument(jsonDoc);
+    editor.setCurrentDocument(translatedDoc);
+  } catch (error) {
+    console.error("Translation error:", error);
+  } finally {
+    translateButton.disabled = false;
+    translateButton.innerHTML = 'Translate to English';
+  }
+}
+// Add click event listener to the translate button.
+const translateButton = document.getElementById("translateButton");
+translateButton.addEventListener("click", translateDocument);
+// Dev exports.
+window.DocAuth = DocAuth;
+window.docAuthSystem = docAuthSystem;
+window.editor = editor;

service-config.yml CHANGED Viewed

@@ -4,17 +4,17 @@ aiServices:
   chat:
     provider:
       name: 'openai-compat'
-      baseUrl: http://127.0.0.1:8080/v1
     model: 'gemma-2b'
   textEmbeddings:
     provider:
       name: 'openai-compat'
-      baseUrl: http://127.0.0.1:8080/v1
     model: 'all-MiniLM-L6-v2'
   headless:
     - provider:
         name: 'openai-compat'
-        baseUrl: http://127.0.0.1:8080/v1
       model:
         name: 'gemma-2b'
         id: 'gemma-2b'

   chat:
     provider:
       name: 'openai-compat'
+      baseUrl: http://127.0.0.1:7861/v1
     model: 'gemma-2b'
   textEmbeddings:
     provider:
       name: 'openai-compat'
+      baseUrl: http://127.0.0.1:7861/v1
     model: 'all-MiniLM-L6-v2'
   headless:
     - provider:
         name: 'openai-compat'
+        baseUrl: http://127.0.0.1:7861/v1
       model:
         name: 'gemma-2b'
         id: 'gemma-2b'

start-container.sh DELETED Viewed

@@ -1,28 +0,0 @@
-#!/bin/sh
-# Start nginx
-nginx
-# Start llamafiler instances
-/usr/local/bin/llamafiler --model /base/gemma-2b.gguf --listen 0.0.0.0:8082 &
-GEMMA_PID=$!
-/usr/local/bin/llamafiler --model /base/embeddings.gguf --listen 0.0.0.0:8081 &
-EMBEDDINGS_PID=$!
-# Wait for services
-until curl -s --fail -X POST http://localhost:8080/v1/chat/completions \
-    -H "Content-Type: application/json" \
-    -d '{"model": "gemma-2b", "messages":[{"role":"user","content":"hi"}]}' >/dev/null 2>&1 && \
-    curl -s --fail -X POST http://localhost:8080/v1/embeddings \
-    -H "Content-Type: application/json" \
-    -d '{"input":"test"}' >/dev/null 2>&1
-do
-    sleep 1
-    echo "Waiting for services..."
-done
-echo 'Services are ready!'
-# Start the main application
-exec ./bin/entrypoint.sh node app/main.bundle.js

start-services.sh CHANGED Viewed

@@ -46,6 +46,27 @@ done
 echo "PostgreSQL is ready!"
 # 4. Start Llamafiler (reusing existing code)
 echo "Starting llamafiler services..."
 mkdir -p /tmp/llamafiler
@@ -63,10 +84,10 @@ echo "Waiting for initial model loading..."
 sleep 15  # Give more time for the models to load into memory
 echo 'Waiting for servers to start...'
-until curl -s --fail -X POST http://localhost:8080/v1/chat/completions \
     -H "Content-Type: application/json" \
     -d '{"model": "gemma-2b", "messages":[{"role":"user","content":"hi"}]}' >/dev/null 2>&1 && \
-      curl -s --fail -X POST http://localhost:8080/v1/embeddings \
     -H "Content-Type: application/json" \
     -d '{"input":"test"}' >/dev/null 2>&1
 do

 echo "PostgreSQL is ready!"
+# Start the web app
+echo "Starting web app..."
+cd /docauth
+export NODE_OPTIONS="--experimental-vm-modules --experimental-json-modules"
+# Start vite with the existing config.
+npx vite serve --host 0.0.0.0 --port 7860 &
+# Wait for the web app to start
+echo "Waiting for web app..."
+until curl -s http://127.0.0.1:7860 > /dev/null; do
+    echo "Waiting for web app..."
+    sleep 2
+    if [ $SECONDS -gt 30 ]; then
+        echo "Timeout waiting for web app"
+        exit 1
+    fi
+done
+echo "Web app is ready!"
 # 4. Start Llamafiler (reusing existing code)
 echo "Starting llamafiler services..."
 mkdir -p /tmp/llamafiler
 sleep 15  # Give more time for the models to load into memory
 echo 'Waiting for servers to start...'
+until curl -s --fail -X POST http://127.0.0.1:7861/v1/chat/completions \
     -H "Content-Type: application/json" \
     -d '{"model": "gemma-2b", "messages":[{"role":"user","content":"hi"}]}' >/dev/null 2>&1 && \
+      curl -s --fail -X POST http://127.0.0.1:7861/v1/embeddings \
     -H "Content-Type: application/json" \
     -d '{"input":"test"}' >/dev/null 2>&1
 do

vite.config.ts CHANGED Viewed

@@ -8,14 +8,25 @@ export default defineConfig({
   appType: 'mpa',
   server: {
     host: '0.0.0.0',  // Listen on all available network interfaces
-    port: 3000,
     headers: {
       'Cross-Origin-Embedder-Policy': 'require-corp',
-      'Cross-Origin-Opener-Policy': 'same-origin'
     },
     proxy: {
       '/api/aia': {
-        target: 'http://aia:4000',
         changeOrigin: true,
         rewrite: (path) => {
           console.log('Rewriting path:', path);

   appType: 'mpa',
   server: {
     host: '0.0.0.0',  // Listen on all available network interfaces
+    port: 7860,
     headers: {
       'Cross-Origin-Embedder-Policy': 'require-corp',
+      'Cross-Origin-Opener-Policy': 'same-origin',
     },
+    middleware: [
+      // Add custom middleware for /dummy route
+      (req, res, next) => {
+        if (req.url === '/dummy') {
+          res.writeHead(200, { 'Content-Type': 'text/plain' });
+          res.end('Hello World!');
+          return;
+        }
+        next();
+      }
+    ],
     proxy: {
       '/api/aia': {
+        target: 'http://127.0.0.1:4000',
         changeOrigin: true,
         rewrite: (path) => {
           console.log('Rewriting path:', path);