lemonade-sdk 8.0.3__py3-none-any.whl → 8.0.4__py3-none-any.whl

lemonade/tools/server/static/webapp.html

@@ -33,7 +33,47 @@
                         <input type="text" id="chat-input" placeholder="Type your message..." /> 
                         <button id="send-btn">Send</button> 
                     </div> 
-                </div> 
+                </div>
+                <!-- App Suggestions Section -->
+                <div class="app-suggestions-section">
+                    <div class="suggestion-text">
+                        Use Lemonade with your favorite app
+                    </div>
+                    <div class="app-logos-grid">
+                        <a href="https://lemonade-server.ai/docs/server/apps/open-webui/" target="_blank" class="app-logo-item" title="Open WebUI">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/openwebui.jpg" alt="Open WebUI" class="app-logo-img">
+                            <span class="app-name">Open WebUI</span>
+                        </a>
+                        <a href="https://lemonade-server.ai/docs/server/apps/continue/" target="_blank" class="app-logo-item" title="Continue">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/continue_dev.png" alt="Continue" class="app-logo-img">
+                            <span class="app-name">Continue</span>
+                        </a>
+                        <a href="https://github.com/amd/gaia" target="_blank" class="app-logo-item" title="Gaia">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/gaia.ico" alt="Gaia" class="app-logo-img">
+                            <span class="app-name">Gaia</span>
+                        </a>
+                        <a href="https://lemonade-server.ai/docs/server/apps/anythingLLM/" target="_blank" class="app-logo-item" title="AnythingLLM">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/anything_llm.png" alt="AnythingLLM" class="app-logo-img">
+                            <span class="app-name">AnythingLLM</span>
+                        </a>
+                        <a href="https://lemonade-server.ai/docs/server/apps/ai-dev-gallery/" target="_blank" class="app-logo-item" title="AI Dev Gallery">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/ai_dev_gallery.webp" alt="AI Dev Gallery" class="app-logo-img">
+                            <span class="app-name">AI Dev Gallery</span>
+                        </a>
+                        <a href="https://lemonade-server.ai/docs/server/apps/lm-eval/" target="_blank" class="app-logo-item" title="LM-Eval">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/lm_eval.png" alt="LM-Eval" class="app-logo-img">
+                            <span class="app-name">LM-Eval</span>
+                        </a>
+                        <a href="https://lemonade-server.ai/docs/server/apps/codeGPT/" target="_blank" class="app-logo-item" title="CodeGPT">
+                            <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/codegpt.jpg" alt="CodeGPT" class="app-logo-img">
+                            <span class="app-name">CodeGPT</span>
+                        </a>
+                    <a href="https://github.com/lemonade-sdk/lemonade/blob/main/docs/server/apps/ai-toolkit.md" target="_blank" class="app-logo-item" title="AI Toolkit">
+                        <img src="https://raw.githubusercontent.com/lemonade-sdk/assets/refs/heads/main/partner_logos/ai_toolkit.png" alt="AI Toolkit" class="app-logo-img">
+                        <span class="app-name">AI Toolkit</span>
+                    </a>
+                    </div>
+                </div>
             </div> 
             <div class="tab-content" id="content-models">                <div class="model-mgmt-register-form collapsed">                    <h3 class="model-mgmt-form-title" onclick="toggleAddModelForm()">
                         Add a Model
@@ -109,7 +149,66 @@
         <div class="copyright">Copyright 2025 AMD</div>
     </footer>
     <script src="https://cdn.jsdelivr.net/npm/openai@4.21.0/dist/openai.min.js"></script> 
-    <script>    // Tab switching logic 
+    <script src="https://cdn.jsdelivr.net/npm/marked@9.1.0/marked.min.js"></script>
+    <script src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script>
+    <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
+    <script>
+    // Configure MathJax
+    window.MathJax = {
+        tex: {
+            inlineMath: [['\\(', '\\)'], ['$', '$']],
+            displayMath: [['\\[', '\\]'], ['$$', '$$']],
+            processEscapes: true,
+            processEnvironments: true
+        },
+        options: {
+            skipHtmlTags: ['script', 'noscript', 'style', 'textarea', 'pre']
+        }
+    };
+    </script>
+    <script>
+    // Configure marked.js for safe HTML rendering
+    marked.setOptions({
+        breaks: true,
+        gfm: true,
+        sanitize: false,
+        smartLists: true,
+        smartypants: true
+    });
+
+    // Function to unescape JSON strings
+    function unescapeJsonString(str) {
+        try {
+            return str.replace(/\\n/g, '\n')
+                     .replace(/\\t/g, '\t')
+                     .replace(/\\r/g, '\r')
+                     .replace(/\\"/g, '"')
+                     .replace(/\\\\/g, '\\');
+        } catch (error) {
+            console.error('Error unescaping string:', error);
+            return str;
+        }
+    }
+
+    // Function to safely render markdown with MathJax support
+    function renderMarkdown(text) {
+        try {
+            const html = marked.parse(text);
+            // Trigger MathJax to process the new content
+            if (window.MathJax && window.MathJax.typesetPromise) {
+                // Use a timeout to ensure DOM is updated before typesetting
+                setTimeout(() => {
+                    window.MathJax.typesetPromise();
+                }, 0);
+            }
+            return html;
+        } catch (error) {
+            console.error('Error rendering markdown:', error);
+            return text; // fallback to plain text
+        }
+    }
+
+    // Tab switching logic 
     function showTab(tab, updateHash = true) { 
         document.getElementById('tab-chat').classList.remove('active'); 
         document.getElementById('tab-models').classList.remove('active'); 
@@ -163,6 +262,44 @@
         form.classList.toggle('collapsed');
     }
 
+    // Handle image load failures for app logos
+    function handleImageFailure(img) {
+        const logoItem = img.closest('.app-logo-item');
+        if (logoItem) {
+            logoItem.classList.add('image-failed');
+        }
+    }
+
+    // Set up image error handlers when DOM is loaded
+    document.addEventListener('DOMContentLoaded', function() {
+        const logoImages = document.querySelectorAll('.app-logo-img');
+        logoImages.forEach(function(img) {
+            let imageLoaded = false;
+            
+            img.addEventListener('load', function() {
+                imageLoaded = true;
+            });
+            
+            img.addEventListener('error', function() {
+                if (!imageLoaded) {
+                    handleImageFailure(this);
+                }
+            });
+            
+            // Also check if image is already broken (cached failure)
+            if (img.complete && img.naturalWidth === 0) {
+                handleImageFailure(img);
+            }
+            
+            // Timeout fallback for slow connections (5 seconds)
+            setTimeout(function() {
+                if (!imageLoaded && !img.complete) {
+                    handleImageFailure(img);
+                }
+            }, 5000);
+        });
+    });
+
     // Helper to get server base URL
     function getServerBaseUrl() {
         const port = window.SERVER_PORT || 8000;
@@ -184,17 +321,37 @@
                 select.innerHTML = '<option>No models available</option>';
                 return;
             }
+            
+            // Filter out embedding models from chat interface
+            const allModels = window.SERVER_MODELS || {};
+            let filteredModels = [];
             let defaultIndex = 0;
-            data.data.forEach(function(model, index) {
+            
+            data.data.forEach(function(model) {
                 const modelId = model.id || model.name || model;
+                const modelInfo = allModels[modelId] || {};
+                const labels = modelInfo.labels || [];
+                
+                // Skip models with "embeddings" or "reranking" label
+                if (labels.includes('embeddings') || labels.includes('reranking')) {
+                    return;
+                }
+                
+                filteredModels.push(modelId);
                 const opt = document.createElement('option');
                 opt.value = modelId;
                 opt.textContent = modelId;
                 if (modelId === 'Llama-3.2-1B-Instruct-Hybrid') {
-                    defaultIndex = index;
+                    defaultIndex = filteredModels.length - 1;
                 }
                 select.appendChild(opt);
             });
+            
+            if (filteredModels.length === 0) {
+                select.innerHTML = '<option>No chat models available</option>';
+                return;
+            }
+            
             select.selectedIndex = defaultIndex;
         } catch (e) {
             const select = document.getElementById('model-select');
@@ -217,26 +374,24 @@
         
         // Add labels if they exist
         const modelData = allModels[modelId];
-        if (modelData) {
-            // Add reasoning label if reasoning is true
-            if (modelData.reasoning === true) {
-                const reasoningLabel = document.createElement('span');
-                reasoningLabel.className = 'model-label reasoning';
-                reasoningLabel.textContent = 'reasoning';
-                container.appendChild(reasoningLabel);
-            }
-            
-            // Add other labels if they exist
-            if (modelData.labels && Array.isArray(modelData.labels)) {
-                modelData.labels.forEach(label => {
-                    const labelSpan = document.createElement('span');
-                    const labelLower = label.toLowerCase();
-                    const labelClass = (labelLower === 'vision') ? 'vision' : 'other';
-                    labelSpan.className = `model-label ${labelClass}`;
-                    labelSpan.textContent = label;
-                    container.appendChild(labelSpan);
-                });
-            }
+        if (modelData && modelData.labels && Array.isArray(modelData.labels)) {
+            modelData.labels.forEach(label => {
+                const labelSpan = document.createElement('span');
+                const labelLower = label.toLowerCase();
+                let labelClass = 'other';
+                if (labelLower === 'vision') {
+                    labelClass = 'vision';
+                } else if (labelLower === 'embeddings') {
+                    labelClass = 'embeddings';
+                } else if (labelLower === 'reasoning') {
+                    labelClass = 'reasoning';
+                } else if (labelLower === 'reranking') {
+                    labelClass = 'reranking';
+                }
+                labelSpan.className = `model-label ${labelClass}`;
+                labelSpan.textContent = label;
+                container.appendChild(labelSpan);
+            });
         }
         
         return container;
@@ -358,16 +513,110 @@
     const modelSelect = document.getElementById('model-select');
     let messages = [];
 
-    function appendMessage(role, text) {
+    function appendMessage(role, text, isMarkdown = false) {
         const div = document.createElement('div');
         div.className = 'chat-message ' + role;
         // Add a bubble for iMessage style
         const bubble = document.createElement('div');
         bubble.className = 'chat-bubble ' + role;
-        bubble.innerHTML = text;
+        
+        if (role === 'llm' && isMarkdown) {
+            bubble.innerHTML = renderMarkdownWithThinkTokens(text);
+        } else {
+            bubble.textContent = text;
+        }
+        
         div.appendChild(bubble);
         chatHistory.appendChild(div);
         chatHistory.scrollTop = chatHistory.scrollHeight;
+        return bubble; // Return the bubble element for streaming updates
+    }
+
+    function updateMessageContent(bubbleElement, text, isMarkdown = false) {
+        if (isMarkdown) {
+            bubbleElement.innerHTML = renderMarkdownWithThinkTokens(text);
+        } else {
+            bubbleElement.textContent = text;
+        }
+    }
+
+    function renderMarkdownWithThinkTokens(text) {
+        // Check if text contains opening think tag
+        if (text.includes('<think>')) {
+            if (text.includes('</think>')) {
+                // Complete think block - handle as before
+                const thinkMatch = text.match(/<think>(.*?)<\/think>/s);
+                if (thinkMatch) {
+                    const thinkContent = thinkMatch[1].trim();
+                    const mainResponse = text.replace(/<think>.*?<\/think>/s, '').trim();
+                    
+                    // Create collapsible structure
+                    let html = '';
+                    if (thinkContent) {
+                        html += `
+                            <div class="think-tokens-container">
+                                <div class="think-tokens-header" onclick="toggleThinkTokens(this)">
+                                    <span class="think-tokens-chevron">▼</span>
+                                    <span class="think-tokens-label">Thinking...</span>
+                                </div>
+                                <div class="think-tokens-content">
+                                    ${renderMarkdown(thinkContent)}
+                                </div>
+                            </div>
+                        `;
+                    }
+                    if (mainResponse) {
+                        html += `<div class="main-response">${renderMarkdown(mainResponse)}</div>`;
+                    }
+                    return html;
+                }
+            } else {
+                // Partial think block - only opening tag found, still being generated
+                const thinkMatch = text.match(/<think>(.*)/s);
+                if (thinkMatch) {
+                    const thinkContent = thinkMatch[1];
+                    const beforeThink = text.substring(0, text.indexOf('<think>'));
+                    
+                    let html = '';
+                    if (beforeThink.trim()) {
+                        html += `<div class="main-response">${renderMarkdown(beforeThink)}</div>`;
+                    }
+                    
+                    html += `
+                        <div class="think-tokens-container">
+                            <div class="think-tokens-header" onclick="toggleThinkTokens(this)">
+                                <span class="think-tokens-chevron">▼</span>
+                                <span class="think-tokens-label">Thinking...</span>
+                            </div>
+                            <div class="think-tokens-content">
+                                ${renderMarkdown(thinkContent)}
+                            </div>
+                        </div>
+                    `;
+                    
+                    return html;
+                }
+            }
+        }
+        
+        // Fallback to normal markdown rendering
+        return renderMarkdown(text);
+    }
+
+    function toggleThinkTokens(header) {
+        const container = header.parentElement;
+        const content = container.querySelector('.think-tokens-content');
+        const chevron = header.querySelector('.think-tokens-chevron');
+        
+        if (content.style.display === 'none') {
+            content.style.display = 'block';
+            chevron.textContent = '▼';
+            container.classList.remove('collapsed');
+        } else {
+            content.style.display = 'none';
+            chevron.textContent = '▶';
+            container.classList.add('collapsed');
+        }
     }
 
     async function sendMessage() {
@@ -379,8 +628,7 @@
         sendBtn.disabled = true;
         // Streaming OpenAI completions (placeholder, adapt as needed)
         let llmText = '';
-        appendMessage('llm', '...');
-        const llmDiv = chatHistory.lastChild.querySelector('.chat-bubble.llm');
+        const llmBubble = appendMessage('llm', '...');
         try {
             // Use the correct endpoint for chat completions
             const resp = await fetch(getServerBaseUrl() + '/api/v1/chat/completions', {
@@ -395,22 +643,40 @@
             if (!resp.body) throw new Error('No stream');
             const reader = resp.body.getReader();
             let decoder = new TextDecoder();
-            llmDiv.textContent = '';
+            llmBubble.textContent = '';
             while (true) {
                 const { done, value } = await reader.read();
                 if (done) break;
                 const chunk = decoder.decode(value);
                 if (chunk.trim() === 'data: [DONE]' || chunk.trim() === '[DONE]') continue;
-                // Try to extract the content from the OpenAI chunk
-                const match = chunk.match(/"content"\s*:\s*"([^"]*)"/);
-                if (match && match[1]) {
-                    llmText += match[1];
-                    llmDiv.textContent = llmText;
+                
+                // Handle Server-Sent Events format
+                const lines = chunk.split('\n');
+                for (const line of lines) {
+                    if (line.startsWith('data: ')) {
+                        const jsonStr = line.substring(6).trim();
+                        if (jsonStr === '[DONE]') continue;
+                        
+                        try {
+                            const parsed = JSON.parse(jsonStr);
+                            if (parsed.choices && parsed.choices[0] && parsed.choices[0].delta && parsed.choices[0].delta.content) {
+                                llmText += parsed.choices[0].delta.content;
+                                updateMessageContent(llmBubble, llmText, true);
+                            }
+                        } catch (e) {
+                            // Fallback to regex parsing if JSON parsing fails
+                            const match = jsonStr.match(/"content"\s*:\s*"((?:\\.|[^"\\])*)"/);
+                            if (match && match[1]) {
+                                llmText += unescapeJsonString(match[1]);
+                                updateMessageContent(llmBubble, llmText, true);
+                            }
+                        }
+                    }
                 }
             }
             messages.push({ role: 'assistant', content: llmText });
         } catch (e) {
-            llmDiv.textContent = '[Error: ' + e.message + ']';
+            llmBubble.textContent = '[Error: ' + e.message + ']';
         }
         sendBtn.disabled = false;
     }

lemonade/version.py

@@ -1 +1 @@
-__version__ = "8.0.3"
+__version__ = "8.0.4"

{lemonade_sdk-8.0.3.dist-info → lemonade_sdk-8.0.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lemonade-sdk
-Version: 8.0.3
+Version: 8.0.4
 Summary: Lemonade SDK: Your LLM Aide for Validation and Deployment
 Author-email: lemonade@amd.com
 Requires-Python: >=3.10, <3.12
@@ -82,7 +82,7 @@ Dynamic: summary
 
 [![Lemonade tests](https://github.com/lemonade-sdk/lemonade/actions/workflows/test_lemonade.yml/badge.svg)](https://github.com/lemonade-sdk/lemonade/tree/main/test "Check out our tests")
 [![OS - Windows | Linux](https://img.shields.io/badge/OS-windows%20%7C%20linux-blue)](docs/README.md#installation "Check out our instructions")
-[![Made with Python](https://img.shields.io/badge/Python-3.8,3.10-blue?logo=python&logoColor=white)](docs/README.md#installation "Check out our instructions")
+[![Made with Python](https://img.shields.io/badge/Python-3.10-blue?logo=python&logoColor=white)](docs/README.md#installation "Check out our instructions")
 
 ## 🍋 Lemonade SDK: Quickly serve, benchmark and deploy LLMs
 
@@ -97,8 +97,8 @@ The [Lemonade SDK](./docs/README.md) makes it easy to run Large Language Models
 The [Lemonade SDK](./docs/README.md) is comprised of the following:
 
 - 🌐 **[Lemonade Server](https://lemonade-server.ai/docs)**: A local LLM server for running ONNX and GGUF models using the OpenAI API standard. Install and enable your applications with NPU and GPU acceleration in minutes.
-- 🐍 **Lemonade API**: High-level Python API to directly integrate Lemonade LLMs into Python applications.
-- 🖥️ **Lemonade CLI**: The `lemonade` CLI lets you mix-and-match LLMs (ONNX, GGUF, SafeTensors) with measurement tools to characterize your models on your hardware. The available tools are:
+- 🐍 **[Lemonade API](./docs/lemonade_api.md)**: High-level Python API to directly integrate Lemonade LLMs into Python applications.
+- 🖥️ **[Lemonade CLI](./docs/dev_cli/README.md)**: The `lemonade` CLI lets you mix-and-match LLMs (ONNX, GGUF, SafeTensors) with measurement tools to characterize your models on your hardware. The available tools are:
   - Prompting with templates.
   - Measuring accuracy with a variety of tests.
   - Benchmarking to get the time-to-first-token and tokens per second.
@@ -153,14 +153,7 @@ Maximum LLM performance requires the right hardware accelerator with the right i
   </tbody>
 </table>
 
-
-
-#### Inference Engines Overview
-| Engine | Description |
-| :--- | :--- |
-| **OnnxRuntime GenAI (OGA)** | Microsoft engine that runs `.onnx` models and enables hardware vendors to provide their own execution providers (EPs) to support specialized hardware, such as neural processing units (NPUs). |
-| **llamacpp** | Community-driven engine with strong GPU acceleration, support for thousands of `.gguf` models, and advanced features such as vision-language models (VLMs) and mixture-of-experts (MoEs). |
-| **Hugging Face (HF)** | Hugging Face's `transformers` library can run the original `.safetensors` trained weights for models on Meta's PyTorch engine, which provides a source of truth for accuracy measurement. |
+To learn more about the supported hardware and software, visit the documentation [here](./docs/README.md#software-and-hardware-overview).
 
 ## Integrate Lemonade Server with Your Application
 

{lemonade_sdk-8.0.3.dist-info → lemonade_sdk-8.0.4.dist-info}/RECORD

@@ -4,7 +4,7 @@ lemonade/cache.py,sha256=djr2qgyUUAWlQv8FehU9qlNtCwK0IZqo82hcBDyZ3-A,2850
 lemonade/cli.py,sha256=9Pcs3PcrWC2F8_pcBaz09xHUICIJTvpemBdPGyXkjIk,4395
 lemonade/sequence.py,sha256=KSH7BPsiyDKsOsg_ziQKEGsDwMmuO_YbgPRBxkZd0pw,13267
 lemonade/state.py,sha256=sdSezla7Cd7KYL90xY3p9kcNV4ndSyN6UvNLOr3vBMA,5261
-lemonade/version.py,sha256=GImAlzwPDxsACkYFf5rTrX8QMH23tcqdm6vgjfFYD10,22
+lemonade/version.py,sha256=8H4GfArMIlRTCgSsTERRXsD3PA6Y67z17oTQOJnuUME,22
 lemonade/common/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 lemonade/common/build.py,sha256=zTb0m1-kuUx6zw5QHp2SNnVuN6jOTMQ2FCdj9iH374U,6140
 lemonade/common/cli_helpers.py,sha256=hjBfXrTtFl8gmCFlL-ksviXR0mOcdPtTWVNKoEp3PG4,4993
@@ -22,10 +22,10 @@ lemonade/tools/__init__.py,sha256=_6xRc-FHxmujoLjLjWtpYrWYEXtCSneSy-5ya01kyPk,53
 lemonade/tools/accuracy.py,sha256=9HCmczDngkBUuUrt49d2CkRo4J0qyWoFYs5cj20bGkg,11714
 lemonade/tools/adapter.py,sha256=HG54iMd6HDPZ4vnQIl7codq3HzffWbcHSIs_jVbNbhU,2958
 lemonade/tools/bench.py,sha256=aN5LMA_EH6-ZhAH3Gf26JYL7s0eKpUd3j-bReRhzvEY,10016
-lemonade/tools/humaneval.py,sha256=9lzsOaCSECf8LzqkQLFNwy1doAiZtK5gRN-RbZH7GLI,9532
+lemonade/tools/humaneval.py,sha256=JbxuoOzvR4iyxZv4R6MI7a3gUt5ef_Jj6Ie-9VP2wzY,9531
 lemonade/tools/management_tools.py,sha256=RO-lU-hjZhrP9KD9qcLI7MrLu-Rxnkrxzn45qqwKInE,8554
-lemonade/tools/mmlu.py,sha256=aEp9nMKTX5yaSaVZ15YmXbWE0YugjeAacnqjMZ13hHM,11072
-lemonade/tools/perplexity.py,sha256=xHl4cTBpJOCNcVxXhMv6eMp8fgUQmFM0G8DeRnx_rUk,5631
+lemonade/tools/mmlu.py,sha256=c2QaIMDzjqxCvgHlMXmy_dP1sAFkwkDxL7RO2nogI6s,11071
+lemonade/tools/perplexity.py,sha256=eiaTZ3yhqF2pfwOffVbKKJLwjSri7Im2pC-tBJr7LLU,5638
 lemonade/tools/prompt.py,sha256=cy6McZeLgk26xG1dJEY-cYnY2x8FUdyOOSG86WfBKCg,9348
 lemonade/tools/tool.py,sha256=UsxVYukfm_iM3BfeGYPZxQlTK5UfDfDOl3RIyLr8A1Y,13256
 lemonade/tools/huggingface/bench.py,sha256=-mTfldCtquL4mspq8ykVwDc9Mut5Ecv_jHJnSb0CYGE,6734
@@ -35,36 +35,36 @@ lemonade/tools/llamacpp/bench.py,sha256=A1X8ULQMxPVsff-AdiUsbWQUKpx7U7nFRNHFJRPd
 lemonade/tools/llamacpp/load.py,sha256=o3vVlefdxmdkHnuvFR3TOxiJkpNAuNFcs9Whfp24jpg,9236
 lemonade/tools/oga/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 lemonade/tools/oga/bench.py,sha256=T3c40NevM3NA7CT98B6vBj1nXfdITDqpfMHYSjhjwpA,5061
-lemonade/tools/oga/load.py,sha256=xSP0DWoGd5zBRozSafj1MMyIQyHJuIRj_vNlCTx8mfs,28309
+lemonade/tools/oga/load.py,sha256=XSznW8lOX_KafSq5J5mIBJzj8YJEBpK0RFGcTE1wnE8,28317
 lemonade/tools/oga/utils.py,sha256=p7faMNfT-rLURC9t_s1S_STQRzzLADqbngUliTOOXeQ,16144
 lemonade/tools/quark/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-lemonade/tools/quark/quark_load.py,sha256=tNy-G9yEJ5cTsxw9LmGUYmmdlEzMo_iy-KSIc2YVz6U,5581
-lemonade/tools/quark/quark_quantize.py,sha256=LZrcbLf9oIw7FW2ccP_qkCP32jxmz5YnNEaoY6rsAuY,16583
+lemonade/tools/quark/quark_load.py,sha256=FJ4LJKTToZbHHWVEOBLadae1a3jCnnY4KvXySHbkJMA,5589
+lemonade/tools/quark/quark_quantize.py,sha256=hwoaXhpBIORvJ16MvewphPkaDEQn3BAgXq5o82Gc-_s,16599
 lemonade/tools/report/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 lemonade/tools/report/llm_report.py,sha256=bVHhwCINA-Ok2EdSwAsLubsc83N3KWOVuwTguw7jDcE,6676
 lemonade/tools/report/table.py,sha256=wJFzKtlmGQH0RQ5O9nevtpMe_-zQ-8zNOndINQuzsjM,27793
 lemonade/tools/server/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-lemonade/tools/server/llamacpp.py,sha256=vjFNelm_VyKBBgWmltsAwLI7ncQ9AwVFQD7krZnF42w,16199
-lemonade/tools/server/serve.py,sha256=3_jBpi6THnnAmtKOxvPlOkIhSTTmrlZE3fr2Dpto-Q4,52794
+lemonade/tools/server/llamacpp.py,sha256=e1MYKSJBu-jlOE5GQSBsC9CUPAeqw5wXXxoxBKA5zb8,20038
+lemonade/tools/server/serve.py,sha256=ORffC4bcBJ-L5-JbmZX91X3yHt1JWxZcIjrZuu9x8TQ,56165
 lemonade/tools/server/tool_calls.py,sha256=xrAlQwKG-nv2xLlf8f9CDSaUbyMn8ZtHkds9iZLG9K8,5230
 lemonade/tools/server/tray.py,sha256=4Kf3x8YfRaItPW7lxlEwerD7c5Q2snzcNk3ZrEoae58,17259
 lemonade/tools/server/webapp.py,sha256=8Das5yXOaSBLZmSZ_eddJajQFxBhvl5D6GI_hHlGbE0,1040
 lemonade/tools/server/static/favicon.ico,sha256=hMmP9qGJNeZ0mFS86JIqPbZstXMZn0Z76_HfHQpREAU,126745
-lemonade/tools/server/static/styles.css,sha256=u-SzZ-vh5qEFMDSKLHJ7MsQwvwpJLB_DdJxocf06Sro,16880
-lemonade/tools/server/static/webapp.html,sha256=kPzORaogVRdFQewXyNI_JaH2ZZCTaq5zfMSyzuoFTuA,22414
+lemonade/tools/server/static/styles.css,sha256=x-pf7xts0te9JWAafcNFqzE7r1fl6n_H362Eiz49ixI,24722
+lemonade/tools/server/static/webapp.html,sha256=AS61ZBDnZkIUpT-iZFlTnWpkp6Yeozs4obzauX4crlU,35004
 lemonade/tools/server/utils/port.py,sha256=XnIg2qS73QRrsJn6LgHcrJPmku30Tv6vsYcBVMj82K4,2186
 lemonade/tools/server/utils/system_tray.py,sha256=b9lvNv9chJKQxvmH7qzAuUe6H9HsLu7pdHFqGlAJaL0,12654
 lemonade/tools/server/utils/thread.py,sha256=pK9K_6DNWoQ78NArkAX3Ym2WsxLnCs9sKTk6TitlYnI,2804
 lemonade_install/__init__.py,sha256=26zohKg2jgr_5y7tObduWMYQg8zCTWMZHL8lfi2zZVQ,40
 lemonade_install/install.py,sha256=DJWR36QSjZtvEwRjYPNSjhYgoxLjI_6OPrCMZjL0ChY,28263
-lemonade_sdk-8.0.3.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-lemonade_sdk-8.0.3.dist-info/licenses/NOTICE.md,sha256=B8lEqi4QE41J9ljz4Riv2JgHD1v8GCZE6nNBHO3KIA0,2135
+lemonade_sdk-8.0.4.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+lemonade_sdk-8.0.4.dist-info/licenses/NOTICE.md,sha256=B8lEqi4QE41J9ljz4Riv2JgHD1v8GCZE6nNBHO3KIA0,2135
 lemonade_server/cli.py,sha256=z6ojwFaOIz0hbUbVtZWMLP4YDpkcVOmqwmdm55dhKA4,11980
-lemonade_server/model_manager.py,sha256=Yvlsl0wipKfryKULH5ASQ9INhLQXPq9dTGQVBXf2_h0,16167
-lemonade_server/pydantic_models.py,sha256=nsbpHqAkd6nkz5QT16u9xMZbCXqccGiy5O0fWecOM88,2338
-lemonade_server/server_models.json,sha256=O5zk94gH_zRq6GSwbqvi2SNwx51eY9uqgAl_kxTi0iM,7271
-lemonade_sdk-8.0.3.dist-info/METADATA,sha256=WesWziLri9jQjZILRENliiJbggTVF8LmXKVIERInVbE,8285
-lemonade_sdk-8.0.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-lemonade_sdk-8.0.3.dist-info/entry_points.txt,sha256=gJppn0ETtXXR6ceKWEIRdk42kMC7ps59EmU3NCPyPUk,144
-lemonade_sdk-8.0.3.dist-info/top_level.txt,sha256=10ap5GNiPhalO4V50LRoxA1FqRT9g3Xkia6BITu880k,42
-lemonade_sdk-8.0.3.dist-info/RECORD,,
+lemonade_server/model_manager.py,sha256=0HqLR38uOu_hxRWVYQ_P6YmwaR-jkDuaAqGYo60X8C0,16702
+lemonade_server/pydantic_models.py,sha256=rp_FFhoTwg6jNmgol-kShwffnRDGbt7jTbIeELvgOIo,2876
+lemonade_server/server_models.json,sha256=Y-j9KAvHmfv77welC0rfRao4inLBce6AVySb-oy_uNE,7519
+lemonade_sdk-8.0.4.dist-info/METADATA,sha256=FqA9Jtgx1QE1EjLg_lxcfcAMI3j0cKpZxoe4GnaGLRA,7754
+lemonade_sdk-8.0.4.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+lemonade_sdk-8.0.4.dist-info/entry_points.txt,sha256=gJppn0ETtXXR6ceKWEIRdk42kMC7ps59EmU3NCPyPUk,144
+lemonade_sdk-8.0.4.dist-info/top_level.txt,sha256=10ap5GNiPhalO4V50LRoxA1FqRT9g3Xkia6BITu880k,42
+lemonade_sdk-8.0.4.dist-info/RECORD,,

lemonade_server/model_manager.py

@@ -54,6 +54,17 @@ class ModelManager:
                 for model_name, model_info in user_models.items()
             }
 
+            # Backwards compatibility for user models that were created before version 8.0.4
+            # "reasoning" was a boolean, but as of 8.0.4 it became a label
+            for _, model_info in user_models.items():
+                if "reasoning" in model_info:
+                    model_info["labels"] = (
+                        ["reasoning"]
+                        if not model_info["labels"]
+                        else model_info["labels"] + ["reasoning"]
+                    )
+                    del model_info["reasoning"]
+
             models.update(user_models)
 
         # Add the model name as a key in each entry, to make it easier
@@ -268,9 +279,8 @@ class ModelManager:
                 new_user_model = {
                     "checkpoint": checkpoint,
                     "recipe": recipe,
-                    "reasoning": reasoning,
                     "suggested": True,
-                    "labels": ["custom"],
+                    "labels": ["custom"] + (["reasoning"] if reasoning else []),
                 }
 
                 if mmproj:

lemonade_server/pydantic_models.py

@@ -1,4 +1,4 @@
-from typing import Optional
+from typing import Optional, Union, List, Any
 
 from pydantic import BaseModel
 
@@ -65,6 +65,30 @@ class ChatCompletionRequest(BaseModel):
     response_format: dict | None = None
 
 
+class EmbeddingsRequest(BaseModel):
+    """
+    Request model for embeddings API endpoint.
+
+    Generates embeddings for the provided input text or tokens.
+    """
+
+    input: Union[str, List]
+    model: Optional[str] = None
+    encoding_format: Optional[str] = "float"  # "float" or "base64"
+
+
+class RerankingRequest(BaseModel):
+    """
+    Request model for reranking API endpoint.
+
+    Reranks a list of documents based on their relevance to a query.
+    """
+
+    query: str
+    documents: List[str]
+    model: str
+
+
 class ResponsesRequest(BaseModel):
     """
     Request model for responses API endpoint.