feat: Update Qwen3 to Qwen3.5

README.md

@@ -117,7 +117,7 @@ my_config = RAGLiteConfig(
 # Example 'local' config with a DuckDB database and a llama.cpp LLM:
 my_config = RAGLiteConfig(
     db_url="duckdb:///raglite.db",
-    llm="llama-cpp-python/unsloth/Qwen3-8B-GGUF/*Q4_K_M.gguf@8192",
+    llm="llama-cpp-python/unsloth/Qwen3.5-9B-GGUF/*Q4_K_M.gguf@8192",
     embedder="llama-cpp-python/lm-kit/bge-m3-gguf/*F16.gguf@512", # More than 512 tokens degrades bge-m3's performance
 )
 ```
@@ -386,7 +386,7 @@ RAGLite comes with an [MCP server](https://modelcontextprotocol.io) implemented
 ```
 raglite \
     --db-url duckdb:///raglite.db \
-    --llm llama-cpp-python/unsloth/Qwen3-4B-GGUF/*Q4_K_M.gguf@8192 \
+    --llm llama-cpp-python/unsloth/Qwen3.5-4B-GGUF/*Q4_K_M.gguf@8192 \
     --embedder llama-cpp-python/lm-kit/bge-m3-gguf/*F16.gguf@512 \
     mcp install
 ```
@@ -422,7 +422,7 @@ You can specify the database URL, LLM, and embedder directly in the Chainlit fro
 ```sh
 raglite \
     --db-url duckdb:///raglite.db \
-    --llm llama-cpp-python/unsloth/Qwen3-4B-GGUF/*Q4_K_M.gguf@8192 \
+    --llm llama-cpp-python/unsloth/Qwen3.5-4B-GGUF/*Q4_K_M.gguf@8192 \
     --embedder llama-cpp-python/lm-kit/bge-m3-gguf/*F16.gguf@512 \
     chainlit
 ```

pyproject.toml

@@ -79,7 +79,7 @@ dev = [
 # Frontend:
 chainlit = ["chainlit (>=2.0.0)"]
 # Large Language Models:
-llama-cpp-python = ["llama-cpp-python (>=0.3.9)"]
+llama-cpp-python = ["llama-cpp-python (>=0.3.16)"]  # Qwen 3.5 support
 # Markdown conversion:
 mistral-ocr = ["mistralai (>=1.10.1)"]
 pandoc = ["pypandoc-binary (>=1.13)"]

src/raglite/_config.py

@@ -68,9 +68,9 @@ class RAGLiteConfig:
     # LLM config used for generation.
     llm: str = field(
         default_factory=lambda: (
-            "llama-cpp-python/unsloth/Qwen3-8B-GGUF/*Q4_K_M.gguf@8192"
+            "llama-cpp-python/unsloth/Qwen3.5-9B-GGUF/*Q4_K_M.gguf@8192"
             if llama_supports_gpu_offload()
-            else "llama-cpp-python/unsloth/Qwen3-4B-GGUF/*Q4_K_M.gguf@8192"
+            else "llama-cpp-python/unsloth/Qwen3.5-4B-GGUF/*Q4_K_M.gguf@8192"
         )
     )
     llm_max_tries: int = 4

src/raglite/_litellm.py

@@ -56,7 +56,7 @@ class LlamaCppPythonLLM(CustomLLM):
     from litellm import completion
 
     response = completion(
-        model="llama-cpp-python/unsloth/Qwen3-8B-GGUF/*Q4_K_M.gguf@8192",
+        model="llama-cpp-python/unsloth/Qwen3.5-9B-GGUF/*Q4_K_M.gguf@8192",
         messages=[{"role": "user", "content": "Hello world!"}],
         # stream=True
     )

src/raglite/_rag.py

@@ -178,6 +178,14 @@ def _limit_chunkspans(
     # Early exit if we're already under the limit
     if total_tokens == 0 or total_tokens <= max_tokens:
         return tool_chunk_spans
+    # If the context window is completely exhausted, return empty spans.
+    if max_tokens <= 0 or total_tokens == 0:
+        logger.warning(
+            "RAG context was limited to 0 out of %d chunks due to context window size. "
+            "Consider using a model with a bigger context window or reducing the number of retrieved chunks.",
+            total_chunk_spans,
+        )
+        return {tool_id: [] for tool_id in tool_chunk_spans}
     # Allocate tokens proportionally and truncate
     new_total_chunk_spans = 0
     scale_ratio = max_tokens / total_tokens

tests/conftest.py

@@ -83,10 +83,10 @@ def database(request: pytest.FixtureRequest) -> str:
     params=[
         pytest.param(
             (
-                "llama-cpp-python/unsloth/Qwen3-4B-GGUF/*Q4_K_M.gguf@8192",  # mistralai/Ministral-3-3B-Instruct-2512
+                "llama-cpp-python/unsloth/Qwen3.5-4B-GGUF/*Q4_K_M.gguf@6144",  # 8192 exceeds Metal GPU memory when embedding model is co-loaded.
                 "llama-cpp-python/lm-kit/bge-m3-gguf/*Q4_K_M.gguf@512",  # More context degrades performance.
             ),
-            id="qwen3_4B-bge_m3",
+            id="qwen3.5_4B-bge_m3",
         ),
         pytest.param(
             ("gpt-4o-mini", "text-embedding-3-small"),

tests/test_chatml_function_calling.py

@@ -63,10 +63,10 @@ def is_accelerator_available() -> bool:
 @pytest.mark.parametrize(
     "llm_repo_id",
     [
-        pytest.param("unsloth/Qwen3-4B-GGUF", id="qwen3_4B"),
+        pytest.param("unsloth/Qwen3.5-4B-GGUF", id="qwen3.5_4B"),
         pytest.param(
-            "unsloth/Qwen3-8B-GGUF",
-            id="qwen3_8B",
+            "unsloth/Qwen3.5-9B-GGUF",
+            id="qwen3.5_9B",
             marks=pytest.mark.skipif(
                 not is_accelerator_available(), reason="Accelerator not available"
             ),