refactor: rename the application package to docsgpt

The backend import package is now docsgpt, the name it will carry on PyPI;
application was far too generic to install into anyone's site-packages.
git mv plus a mechanical rewrite of every import, dotted string and path
reference: 734 Python files, the compose files, Dockerfile, workflows, docs,
setup scripts, devcontainer, k8s manifests, vscode config, pytest and coverage
config, .gitignore. Behaviour is unchanged.

Kept for one release:
- A top-level application package whose meta-path finder resolves
  application.x.y to the already-imported docsgpt.x.y object, so old imports
  and entry points (celery -A application.app.celery,
  uvicorn application.asgi:asgi_app) keep working with a FutureWarning.
- Celery registers every application.* task name as an alias of its
  docsgpt.* task on start-up, so messages queued by the previous release still
  run. The redbeat key prefix moves to redbeat:docsgpt:v2: so schedule entries
  the previous release wrote are left unread instead of firing twice.

The backend image builds from the repository root (docker build -f
docsgpt/Dockerfile .) so it can ship the alias package; a root .dockerignore
allow-lists docsgpt/ and application/ and keeps caches, local data, .env
files, the sample index files and the Dockerfile out. Compose and the image
workflows point at the new context.

docsgpt/llm/llama_cpp.py

@@ -0,0 +1,62 @@
+from docsgpt.llm.base import BaseLLM
+from docsgpt.core.settings import settings
+import threading
+
+
+class LlamaSingleton:
+    _instances = {}
+    _lock = threading.Lock()  # Add a lock for thread synchronization
+
+    @classmethod
+    def get_instance(cls, llm_name):
+        if llm_name not in cls._instances:
+            try:
+                from llama_cpp import Llama
+            except ImportError:
+                raise ImportError(
+                    "Please install llama_cpp using pip install llama-cpp-python"
+                )
+            cls._instances[llm_name] = Llama(model_path=llm_name, n_ctx=2048)
+        return cls._instances[llm_name]
+
+    @classmethod
+    def query_model(cls, llm, prompt, **kwargs):
+        with cls._lock:
+            return llm(prompt, **kwargs)
+
+
+class LlamaCpp(BaseLLM):
+    provider_name = "llama_cpp"
+
+    def __init__(
+        self,
+        api_key=None,
+        user_api_key=None,
+        llm_name=settings.LLM_PATH,
+        *args,
+        **kwargs,
+    ):
+        super().__init__(*args, **kwargs)
+        self.api_key = api_key
+        self.user_api_key = user_api_key
+        self.llama = LlamaSingleton.get_instance(llm_name)
+
+    def _raw_gen(self, baseself, model, messages, stream=False, **kwargs):
+        context = messages[0]["content"]
+        user_question = messages[-1]["content"]
+        prompt = f"### Instruction \n {user_question} \n ### Context \n {context} \n ### Answer \n"
+        result = LlamaSingleton.query_model(
+            self.llama, prompt, max_tokens=150, echo=False
+        )
+        return result["choices"][0]["text"].split("### Answer \n")[-1]
+
+    def _raw_gen_stream(self, baseself, model, messages, stream=True, **kwargs):
+        context = messages[0]["content"]
+        user_question = messages[-1]["content"]
+        prompt = f"### Instruction \n {user_question} \n ### Context \n {context} \n ### Answer \n"
+        result = LlamaSingleton.query_model(
+            self.llama, prompt, max_tokens=150, echo=False, stream=stream
+        )
+        for item in result:
+            for choice in item["choices"]:
+                yield choice["text"]