v8.0.0

Author: BBC-Esq

src/chart_models_chat.py

@@ -15,8 +15,6 @@ def create_chat_models_comparison_plot():
     model_categories = {
         "coding": {
             "models": [
-                # "Qwen Coder - 1.5b",
-                # "Qwen Coder - 3b",
                 "Seed Coder - 8b"
             ],
             "color": "#DAA520",
@@ -40,8 +38,6 @@ def create_chat_models_comparison_plot():
         },
         "coding_and_thinking": {
             "models": [
-                # "Olympic Coder - 7b",
-                # "Olympic Coder - 32b"
             ],
             "color": "#8B0000",
 

src/check_packages.py

@@ -38,7 +38,7 @@ def get_latest_version(self, package_name):
                 if response.status != 200:
                     raise Exception(f"PyPI returned status code {response.status}")
                 return json.load(response)['info']['version']
-        except urllib.error.URLError as e:
+        except urllib.error.URLError as e: 
             raise Exception(f"Network error: {str(e)}")
         except TimeoutError:
             raise Exception("Connection timed out")
@@ -523,7 +523,7 @@ def open_context_menu(self, position: QPoint):
 
     def show_reverse_dependencies(self, package_name):
         try:
-            command = ["pipdeptree", "--reverse", "--packages", package_name]
+            command = ["pipdeptree", "--reverse", "--packages", package_name, "--depth", "1"]
             process = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
             stdout, stderr = process.communicate()
 

src/initialize.py

@@ -45,7 +45,7 @@ def get_supported_quantizations(device_type):
 
 def update_config_file(**system_info):
     full_config_path = Path('config.yaml').resolve()
-    
+
     with open(full_config_path, 'r', encoding='utf-8') as stream:
         config_data = yaml.safe_load(stream)
 

src/module_ask_jeeves.py

@@ -22,7 +22,7 @@
 from PySide6.QtWidgets import (
     QMainWindow, QWidget, QVBoxLayout, QTextEdit, 
     QLineEdit, QMessageBox, QPushButton, QLabel,
-    QHBoxLayout, QSizePolicy, QComboBox, QApplication
+    QHBoxLayout, QSizePolicy, QComboBox, QApplication, QSpinBox
 )
 from PySide6.QtCore import QThread, Signal, Qt, QTimer, QObject
 from PySide6.QtGui import QTextCursor, QPixmap
@@ -38,6 +38,7 @@
 from module_kokoro import KokoroTTS
 from utilities import normalize_chat_text
 
+
 class GenerationWorker(QThread):
     token_signal = Signal(str)
     finished_signal = Signal()
@@ -130,6 +131,17 @@ def __init__(self, parent=None):
         self.eject_button.setEnabled(False)
         model_layout.addWidget(self.eject_button)
 
+        self.context_label = QLabel("Contexts:")
+        self.context_label.setFixedHeight(30)
+
+        self.context_spin = QSpinBox()
+        self.context_spin.setRange(1, 10)      # allow 1-20
+        self.context_spin.setValue(5)          # default 5
+        self.context_spin.setFixedHeight(30)
+
+        model_layout.addWidget(self.context_label)
+        model_layout.addWidget(self.context_spin)
+
         self.layout.addLayout(model_layout)
 
         self.chat_display = QTextEdit()
@@ -226,6 +238,16 @@ def __init__(self, parent=None):
         self.tts_worker = None
         self.is_speaking = False
 
+    def _ensure_model(self) -> None:
+        """
+        Download or resume-download the model if *model.bin* is missing.
+        (Keeps everything else that is already in the cache.)
+        """
+        model_dir = Path(self.model_dir)
+        if not (model_dir / "model.bin").exists():
+            print("model.bin missing – redownloading just that file …")
+            self._download_model()
+
     def eject_model(self):
         if self.generator:
             del self.generator
@@ -294,6 +316,7 @@ def on_model_downloaded(self, model_name, model_type):
         self._load_model()
 
     def _load_model(self):
+        self._ensure_model()
         physical_cores = max(1, psutil.cpu_count(logical=False) - 1)
         device = "cuda" if torch.cuda.is_available() else "cpu"
 
@@ -361,6 +384,7 @@ def send_message(self):
         self.chat_display.clear()
 
         try:
+            k_value = self.context_spin.value()
             contexts, metadata = self.vector_db.search(user_message, k=5, score_threshold=0.9)
             if not contexts:
                 QMessageBox.warning(self, "No Contexts Found", "No relevant contexts were found for your query.")

src/utilities.py

@@ -24,29 +24,45 @@ def set_cuda_paths():
     import sys
     import os
     from pathlib import Path
+    # virtual environment path
     venv_base = Path(sys.executable).parent.parent
-    nvidia_base_path = venv_base / 'Lib' / 'site-packages' / 'nvidia'
-    cuda_path_runtime = nvidia_base_path / 'cuda_runtime' / 'bin'
-    cuda_path_runtime_lib = nvidia_base_path / 'cuda_runtime' / 'lib' / 'x64'
-    cuda_path_runtime_include = nvidia_base_path / 'cuda_runtime' / 'include'
-    cublas_path = nvidia_base_path / 'cublas' / 'bin'
-    cudnn_path = nvidia_base_path / 'cudnn' / 'bin'
-    nvrtc_path = nvidia_base_path / 'cuda_nvrtc' / 'bin'
-    nvcc_path = nvidia_base_path / 'cuda_nvcc' / 'bin'
+
+    # dependencies path
+    dependency_base = venv_base / 'Lib' / 'site-packages'
+
+    # nvidia base path
+    nvidia_base = dependency_base / 'nvidia'
+
+    # nvidia specific component paths
+    cuda_runtime = nvidia_base / 'cuda_runtime' / 'bin'
+    cuda_runtime_lib = nvidia_base / 'cuda_runtime' / 'lib' / 'x64'
+    cuda_runtime_include = nvidia_base / 'cuda_runtime' / 'include'
+    cublas = nvidia_base / 'cublas' / 'bin'
+    cudnn = nvidia_base / 'cudnn' / 'bin'
+    nvrtc = nvidia_base / 'cuda_nvrtc' / 'bin'
+    nvcc = nvidia_base / 'cuda_nvcc' / 'bin'
+    cusparse = nvidia_base / 'cusparse' / 'bin'
+
+    # unique cusparselt path
+    # cusparselt = dependency_base / 'cusparselt' / 'bin'
+
     paths_to_add = [
-        str(cuda_path_runtime),
-        str(cuda_path_runtime_lib),
-        str(cuda_path_runtime_include),
-        str(cublas_path),
-        str(cudnn_path),
-        str(nvrtc_path),
-        str(nvcc_path),
+        str(cuda_runtime),
+        str(cuda_runtime_lib),
+        str(cuda_runtime_include),
+        str(cublas),
+        str(cudnn),
+        str(nvrtc),
+        str(nvcc),
+        str(cusparse),
+        # str(cusparselt),
     ]
+
     current_value = os.environ.get('PATH', '')
     new_value = os.pathsep.join(paths_to_add + ([current_value] if current_value else []))
     os.environ['PATH'] = new_value
 
-    triton_cuda_path = nvidia_base_path / 'cuda_runtime'
+    triton_cuda_path = nvidia_base / 'cuda_runtime'
     current_cuda_path = os.environ.get('CUDA_PATH', '')
     new_cuda_path = os.pathsep.join([str(triton_cuda_path)] + ([current_cuda_path] if current_cuda_path else []))
     os.environ['CUDA_PATH'] = new_cuda_path