fix order, kompute whitelist, changelog entry,

Signed-off-by: Jared Van Bortel <jared@nomic.ai>

.circleci/continue_config.yml

@@ -1903,22 +1903,22 @@ workflows:
           context: gpt4all
       - build-offline-chat-installer-linux:
           context: gpt4all
-      - sign-online-chat-installer-macos:
+      - sign-offline-chat-installer-macos:
           context: gpt4all
           requires:
-            - build-online-chat-installer-macos
-      - notarize-online-chat-installer-macos:
+            - build-offline-chat-installer-macos
+      - notarize-offline-chat-installer-macos:
           context: gpt4all
           requires:
-            - sign-online-chat-installer-macos
-      - sign-online-chat-installer-windows:
+            - sign-offline-chat-installer-macos
+      - sign-offline-chat-installer-windows:
           context: gpt4all
           requires:
-            - build-online-chat-installer-windows
-      - sign-online-chat-installer-windows-arm:
+            - build-offline-chat-installer-windows
+      - sign-offline-chat-installer-windows-arm:
           context: gpt4all
           requires:
-            - build-online-chat-installer-windows-arm
+            - build-offline-chat-installer-windows-arm
   build-chat-installers-release:
     # only run on main branch tags that start with 'v' and a digit
     when:

MAINTAINERS.md

@@ -72,6 +72,6 @@ Discord: `@Tim453`
 - Flatpak
 
 Jack ([@wuodoo](https://github.com/wuodoo))<br/>
-E-mail: 2296103047@qq.com><br/>
+E-mail: 2296103047@qq.com<br/>
 Discord: `@mikage`
 - zh\_CN translation

gpt4all-backend/src/llamamodel.cpp

@@ -53,6 +53,7 @@ static const std::vector<const char *> KNOWN_ARCHES {
     "gpt2",
     // "gptj", -- no inference code
     "gptneox",
+    "granite",
     "granitemoe",
     "mpt",
     "baichuan",

gpt4all-chat/CHANGELOG.md

@@ -6,8 +6,12 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+### Added
+- Whitelist Granite (non-MoE) model architecture (by [@ThiloteE](https://github.com/ThiloteE) in [#3487](https://github.com/nomic-ai/gpt4all/pull/3487))
+
 ### Fixed
 - Fix several potential crashes ([#3465](https://github.com/nomic-ai/gpt4all/pull/3465))
+- Fix visual spacing issues with deepseek models ([#3470](https://github.com/nomic-ai/gpt4all/pull/3470))
 
 ## [3.9.0] - 2025-02-04
 

gpt4all-chat/qml/ChatItemView.qml

@@ -198,6 +198,7 @@ GridLayout {
                         isError: false
                         isThinking: true
                         thinkingTime: modelData.thinkingTime
+                        visible: modelData.content !== ""
                     }
                 }
             }

gpt4all-chat/src/chatllm.cpp

@@ -976,7 +976,8 @@ public:
     {
         Q_UNUSED(bufferIdx)
         try {
-            m_cllm->m_chatModel->setResponseValue(response);
+            QString r = response;
+            m_cllm->m_chatModel->setResponseValue(removeLeadingWhitespace(r));
         } catch (const std::exception &e) {
             // We have a try/catch here because the main thread might have removed the response from
             // the chatmodel by erasing the conversation during the response... the main thread sets
@@ -991,7 +992,7 @@ public:
     bool onRegularResponse() override
     {
         auto respStr = QString::fromUtf8(m_result->response);
-        return onBufferResponse(removeLeadingWhitespace(respStr), 0);
+        return onBufferResponse(respStr, 0);
     }
 
     bool getStopGenerating() const override
@@ -1078,7 +1079,7 @@ auto ChatLLM::promptInternal(
     auto respStr = QString::fromUtf8(result.response);
     if (!respStr.isEmpty() && (std::as_const(respStr).back().isSpace() || finalBuffers.size() > 1)) {
         if (finalBuffers.size() > 1)
-            m_chatModel->setResponseValue(finalBuffers.last());
+            m_chatModel->setResponseValue(finalBuffers.last().trimmed());
         else
             m_chatModel->setResponseValue(respStr.trimmed());
         emit responseChanged();

gpt4all-chat/src/embllm.cpp

@@ -281,9 +281,6 @@ void EmbeddingLLMWorker::docEmbeddingsRequested(const QVector<EmbeddingChunk> &c
         std::vector<float> result;
         result.resize(chunks.size() * m_model->embeddingSize());
         for (int j = 0; j < chunks.size(); j += BATCH_SIZE) {
-            if (m_stopGenerating)
-                return;
-
             QMutexLocker locker(&m_mutex);
             std::vector batchTexts(texts.begin() + j, texts.begin() + std::min(j + BATCH_SIZE, int(texts.size())));
             try {