feat(ui): update display of properties and add certifi

- updated project files
- added certifi to backend download and update checking
- add and fix type hints
- small file formatting changes
- update formatting of KV pairs to be cleaner
- update tensor data formatting and remove redundant KV pairs property
- add human readable mappings from KV pairs into model properties
- update CUDA backend check for latest llama.cpp format
- use urllib globally

SECURITY.md

@@ -4,10 +4,10 @@ ## Supported Versions
 
 | Version         | Supported          |
 |-----------------|--------------------|
-| stable (v1.9.x) | :white_check_mark: |
+| stable (v2.0.x) | :white_check_mark: |
 
 Beta versions are not officially supported and may contain unknown security vulnerabilities. Use them at your own risk.
 
 ## Reporting a Vulnerability
 
-Use the Issues tab, or for severe vulnerabilities please contact the maintainers via email.
+Use the Issues tab, or for severe vulnerabilities, please contact the maintainers via email.

requirements.txt

@@ -6,8 +6,9 @@ safetensors~=0.5.3
 numpy<2.0.0
 torch~=2.7.0
 sentencepiece~=0.2.0
-setuptools~=80.4.0
-huggingface-hub~=0.31.1
+setuptools~=80.7.1
+huggingface-hub~=0.31.2
 transformers~=4.51.3
 fastapi~=0.115.12
 uvicorn~=0.34.2
+certifi~=2025.4.26

src/AutoGGUF.py

@@ -2,6 +2,8 @@
 import shutil
 import urllib.error
 import urllib.request
+import certifi
+import ssl
 from datetime import datetime
 from functools import partial, wraps
 from typing import List
@@ -1148,7 +1150,10 @@ def check_for_updates(self) -> None:
             url = "https://api.github.com/repos/leafspark/AutoGGUF/releases/latest"
             req = urllib.request.Request(url)
 
-            with urllib.request.urlopen(req) as response:
+            # Create SSL context with certifi certificates
+            ssl_context = ssl.create_default_context(cafile=certifi.where())
+
+            with urllib.request.urlopen(req, context=ssl_context) as response:
                 if response.status != 200:
                     raise urllib.error.HTTPError(
                         url, response.status, "HTTP Error", response.headers, None

src/CustomTitleBar.py

@@ -98,7 +98,7 @@ def mouseMoveEvent(self, event) -> None:
     def mouseReleaseEvent(self, event) -> None:
         self.pressing = False
 
-    def toggle_maximize(self):
+    def toggle_maximize(self) -> None:
         if self.isMaximized:
             self.parent.showNormal()
             if self.normal_size:

src/DownloadThread.py

@@ -2,6 +2,8 @@
 import urllib.request
 import urllib.error
 import zipfile
+import ssl
+import certifi
 from PySide6.QtCore import QThread, Signal
 
 
@@ -19,7 +21,10 @@ def run(self) -> None:
         try:
             req = urllib.request.Request(self.url)
 
-            with urllib.request.urlopen(req) as response:
+            # Create SSL context with certifi certificates
+            ssl_context = ssl.create_default_context(cafile=certifi.where())
+
+            with urllib.request.urlopen(req, context=ssl_context) as response:
                 if response.status != 200:
                     raise urllib.error.HTTPError(
                         self.url, response.status, "HTTP Error", response.headers, None

src/KVOverrideEntry.py

@@ -22,6 +22,7 @@ def __init__(self, parent=None) -> None:
 
         self.key_input = QLineEdit()
         self.key_input.setPlaceholderText("Key")
+
         # Set validator for key input (letters and dots only)
         key_validator = QRegularExpressionValidator(QRegularExpression(r"[A-Za-z.]+"))
         self.key_input.setValidator(key_validator)

src/Localizations.py

@@ -365,7 +365,7 @@ def __init__(self):
         self.ADDING_LORA_ADAPTER = "Adding LoRA Adapter..."
         self.DELETING_LORA_ADAPTER = "Deleting LoRA Adapter..."
         self.SELECT_LORA_ADAPTER_FILE = "Select LoRA Adapter File"
-        self.STARTING_LORA_EXPORT = "Starting LoRA export..."
+        self.STARTING_LORA_EXPORT = "Starting LoRA export"
         self.SELECT_OUTPUT_TYPE = "Select Output Type (GGUF or GGML)"
         self.BASE_MODEL = "Base Model"
         self.SELECT_BASE_MODEL_FILE = "Select Base Model File (GGUF)"

src/ModelInfoDialog.py

@@ -24,8 +24,21 @@ def __init__(self, model_info, parent=None) -> None:
     def format_model_info(self, model_info) -> str:
         html = "<h2>Model Information</h2>"
         html += f"<p><b>Architecture:</b> {model_info.get('architecture', 'N/A')}</p>"
-        html += f"<p><b>Quantization Type:</b> {model_info.get('quantization_type', 'N/A')}</p>"
-        html += f"<p><b>KV Pairs:</b> {model_info.get('kv_pairs', 'N/A')}</p>"
+
+        # Format quantization types
+        quant_types = model_info.get("quantization_type", [])
+        if quant_types:
+            # Clean up the format: remove "- type " prefix and join with " | "
+            formatted_types = []
+            for qtype in quant_types:
+                # Remove "- type " prefix if present
+                clean_type = qtype.replace("- type ", "").strip()
+                formatted_types.append(clean_type)
+            quant_display = " | ".join(formatted_types)
+        else:
+            quant_display = "N/A"
+
+        html += f"<p><b>Quantization Type:</b> {quant_display}</p>"
         html += f"<p><b>Tensors:</b> {model_info.get('tensors', 'N/A')}</p>"
 
         html += "<h3>Key-Value Pairs:</h3>"

src/QuantizationThread.py

@@ -59,6 +59,34 @@ def run(self) -> None:
             self.error_signal.emit(str(e))
 
     def parse_model_info(self, line) -> None:
+        # Mapping of technical keys to human-readable names
+        key_mappings = {
+            "general.architecture": "Architecture",
+            "general.name": "Model Name",
+            "general.file_type": "File Type",
+            "general.quantization_version": "Quantization Version",
+            "llama.block_count": "Layers",
+            "llama.context_length": "Context Length",
+            "llama.embedding_length": "Embedding Size",
+            "llama.feed_forward_length": "Feed Forward Length",
+            "llama.attention.head_count": "Attention Heads",
+            "llama.attention.head_count_kv": "Key-Value Heads",
+            "llama.attention.layer_norm_rms_epsilon": "RMS Norm Epsilon",
+            "llama.rope.freq_base": "RoPE Frequency Base",
+            "llama.rope.dimension_count": "RoPE Dimensions",
+            "llama.vocab_size": "Vocabulary Size",
+            "tokenizer.ggml.model": "Tokenizer Model",
+            "tokenizer.ggml.pre": "Tokenizer Preprocessing",
+            "tokenizer.ggml.tokens": "Tokens",
+            "tokenizer.ggml.token_type": "Token Types",
+            "tokenizer.ggml.merges": "BPE Merges",
+            "tokenizer.ggml.bos_token_id": "Begin of Sequence Token ID",
+            "tokenizer.ggml.eos_token_id": "End of Sequence Token ID",
+            "tokenizer.chat_template": "Chat Template",
+            "tokenizer.ggml.padding_token_id": "Padding Token ID",
+            "tokenizer.ggml.unk_token_id": "Unknown Token ID",
+        }
+
         # Parse output for model information
         if "llama_model_loader: loaded meta data with" in line:
             parts = line.split()
@@ -66,10 +94,25 @@ def parse_model_info(self, line) -> None:
             self.model_info["tensors"] = parts[9]
         elif "general.architecture" in line:
             self.model_info["architecture"] = line.split("=")[-1].strip()
-        elif line.startswith("llama_model_loader: - kv"):
-            key = line.split(":")[2].strip()
-            value = line.split("=")[-1].strip()
-            self.model_info.setdefault("kv_data", {})[key] = value
+        elif line.startswith("llama_model_loader: - kv") and "=" in line:
+            # Split on '=' and take the parts
+            parts = line.split("=", 1)  # Split only on first '='
+            left_part = parts[0].strip()
+            value = parts[1].strip()
+
+            # Extract key and type from left part
+            # Format: "llama_model_loader: - kv N: key type"
+            kv_parts = left_part.split(":")
+            if len(kv_parts) >= 3:
+                key_type_part = kv_parts[2].strip()  # This is "key type"
+                key = key_type_part.rsplit(" ", 1)[
+                    0
+                ]  # Everything except last word (type)
+
+                # Use human-readable name if available, otherwise use original key
+                display_key = key_mappings.get(key, key)
+
+                self.model_info.setdefault("kv_data", {})[display_key] = value
         elif line.startswith("llama_model_loader: - type"):
             parts = line.split(":")
             if len(parts) > 1:

src/TaskListItem.py

@@ -95,11 +95,8 @@ def show_task_context_menu(self, position) -> None:
 
     def show_task_properties(self, item) -> None:
         self.logger.debug(SHOWING_PROPERTIES_FOR_TASK.format(item.text()))
-        task_item = self.task_list.itemWidget(item)
         for thread in self.quant_threads:
-            if thread.log_file == task_item.log_file:
             model_info_dialog = ModelInfoDialog(thread.model_info, self)
-
             model_info_dialog.exec()
             break
 

src/globals.py

@@ -1,7 +1,7 @@
 import os
 import re
 import sys
-from typing import Any, List, TextIO, Union
+from typing import Any, IO, List, TextIO, Union
 
 from PySide6.QtWidgets import (
     QMessageBox,
@@ -86,9 +86,9 @@ def show_about(self) -> None:
         
 A tool for managing and converting GGUF models.
 This application is licensed under the Apache License 2.0.
-Copyright (c) 2025 leafspark.
+Copyright (c) 2024-2025 leafspark.
 It also utilizes llama.cpp, licensed under the MIT License.
-Copyright (c) 2023-2024 The ggml authors."""
+Copyright (c) 2023-2025 The ggml authors."""
     QMessageBox.about(self, "About AutoGGUF", about_text)
 
 
@@ -97,7 +97,7 @@ def ensure_directory(path) -> None:
         os.makedirs(path)
 
 
-def open_file_safe(file_path, mode="r") -> TextIO:
+def open_file_safe(file_path, mode="r") -> IO[Any]:
     encodings = ["utf-8", "latin-1", "ascii", "utf-16"]
     for encoding in encodings:
         try:

src/ui_update.py

@@ -159,7 +159,9 @@ def update_cuda_backends(self) -> None:
         for item in os.listdir(llama_bin):
             item_path = os.path.join(llama_bin, item)
             if os.path.isdir(item_path) and "cudart-llama" not in item.lower():
-                if "cu1" in item.lower():  # Only include CUDA-capable backends
+                if (
+                    "cu1" in item.lower() or "cuda-1" in item.lower()
+                ):  # Only include CUDA-capable backends
                     self.backend_combo_cuda.addItem(item, userData=item_path)
 
     if self.backend_combo_cuda.count() == 0:

src/utils.py

@@ -1,6 +1,10 @@
 from typing import Any, Union
 
-import requests
+import urllib.request
+import urllib.error
+import json
+import ssl
+import certifi
 from PySide6.QtCore import Qt
 from PySide6.QtWidgets import QFileDialog, QInputDialog, QMenu
 
@@ -188,16 +192,28 @@ def refresh_releases(self) -> None:
         owner, repo = get_repo_from_env()
         url = f"https://api.github.com/repos/{owner}/{repo}/releases"
 
-        response = requests.get(url)
-        response.raise_for_status()
+        # Create SSL context with certifi certificates
+        ssl_context = ssl.create_default_context(cafile=certifi.where())
+
+        # Create request
+        req = urllib.request.Request(url)
+
+        # Make the request
+        with urllib.request.urlopen(req, context=ssl_context) as response:
+            if response.status != 200:
+                raise urllib.error.HTTPError(
+                    url, response.status, "HTTP Error", response.headers, None
+                )
+
+            releases = json.loads(response.read().decode("utf-8"))
 
-        releases = response.json()
         self.release_combo.clear()
         for release in releases:
             self.release_combo.addItem(release["tag_name"], userData=release)
         self.release_combo.currentIndexChanged.connect(self.update_assets)
         self.update_assets()
+
     except ValueError as e:
         show_error(self.logger, f"Invalid repository configuration: {str(e)}")
-    except requests.exceptions.RequestException as e:
+    except (urllib.error.URLError, urllib.error.HTTPError) as e:
         show_error(self.logger, ERROR_FETCHING_RELEASES.format(str(e)))