Inline code completions (#465)

* Draft inline completions implementation (server side)

* Implement inline completion provider (front)

* Add default debounce delay and error handling (front)

* Add `gpt-3.5-turbo-instruct` because text- models are deprecated.

OpenAI specifically recommends using `gpt-3.5-turbo-instruct` in
favour of text-davinci, text-ada, etc. See:
https://platform.openai.com/docs/deprecations/

* Improve/fix prompt template and add simple post-processing

* Handle missing `registerInlineProvider`, handle no model in name

* Remove IPython mention to avoid confusing languages

* Disable suggestions in markdown, move language logic

* Remove unused background and clip path from jupyternaut

* Implement toggling the AI completer via statusbar item

also adds the icon for provider re-using jupyternaut icon

* Implement streaming support

* Translate ipython to python for models, remove log

* Move `BaseLLMHandler` to `/completions` rename to `LLMHandlerMixin`

* Move frontend completions code to `/completions`

* Make `IStatusBar` required for now, lint

packages/jupyter-ai-magics/jupyter_ai_magics/providers.py

@@ -518,6 +518,7 @@ class OpenAIProvider(BaseProvider, OpenAI):
         "text-curie-001",
         "text-babbage-001",
         "text-ada-001",
+        "gpt-3.5-turbo-instruct",
         "davinci",
         "curie",
         "babbage",

packages/jupyter-ai/jupyter_ai/completions/handlers/__init__.py

@@ -0,0 +1,4 @@
+from .base import BaseInlineCompletionHandler
+from .default import DefaultInlineCompletionHandler
+
+__all__ = ["BaseInlineCompletionHandler", "DefaultInlineCompletionHandler"]

packages/jupyter-ai/jupyter_ai/completions/handlers/base.py

@@ -0,0 +1,76 @@
+import traceback
+
+# necessary to prevent circular import
+from typing import TYPE_CHECKING, AsyncIterator, Dict
+
+from jupyter_ai.completions.handlers.llm_mixin import LLMHandlerMixin
+from jupyter_ai.completions.models import (
+    CompletionError,
+    InlineCompletionList,
+    InlineCompletionReply,
+    InlineCompletionRequest,
+    InlineCompletionStreamChunk,
+    ModelChangedNotification,
+)
+from jupyter_ai.config_manager import ConfigManager, Logger
+
+if TYPE_CHECKING:
+    from jupyter_ai.handlers import InlineCompletionHandler
+
+
+class BaseInlineCompletionHandler(LLMHandlerMixin):
+    """Class implementing completion handling."""
+
+    handler_kind = "completion"
+
+    def __init__(
+        self,
+        log: Logger,
+        config_manager: ConfigManager,
+        model_parameters: Dict[str, Dict],
+        ws_sessions: Dict[str, "InlineCompletionHandler"],
+    ):
+        super().__init__(log, config_manager, model_parameters)
+        self.ws_sessions = ws_sessions
+
+    async def on_message(
+        self, message: InlineCompletionRequest
+    ) -> InlineCompletionReply:
+        try:
+            return await self.process_message(message)
+        except Exception as e:
+            return await self._handle_exc(e, message)
+
+    async def process_message(
+        self, message: InlineCompletionRequest
+    ) -> InlineCompletionReply:
+        """
+        Processes an inline completion request. Completion handlers
+        (subclasses) must implement this method.
+
+        The method definition does not need to be wrapped in a try/except block.
+        """
+        raise NotImplementedError("Should be implemented by subclasses.")
+
+    async def stream(
+        self, message: InlineCompletionRequest
+    ) -> AsyncIterator[InlineCompletionStreamChunk]:
+        """ "
+        Stream the inline completion as it is generated. Completion handlers
+        (subclasses) can implement this method.
+        """
+        raise NotImplementedError()
+
+    async def _handle_exc(self, e: Exception, message: InlineCompletionRequest):
+        error = CompletionError(
+            type=e.__class__.__name__,
+            title=e.args[0] if e.args else "Exception",
+            traceback=traceback.format_exc(),
+        )
+        return InlineCompletionReply(
+            list=InlineCompletionList(items=[]), error=error, reply_to=message.number
+        )
+
+    def broadcast(self, message: ModelChangedNotification):
+        for session in self.ws_sessions.values():
+            session.write_message(message.dict())

packages/jupyter-ai/jupyter_ai/completions/handlers/default.py

@@ -0,0 +1,184 @@
+from typing import Dict, Type
+
+from jupyter_ai_magics.providers import BaseProvider
+from langchain.prompts import (
+    ChatPromptTemplate,
+    HumanMessagePromptTemplate,
+    PromptTemplate,
+    SystemMessagePromptTemplate,
+)
+from langchain.schema.output_parser import StrOutputParser
+from langchain.schema.runnable import Runnable
+
+from ..models import (
+    InlineCompletionList,
+    InlineCompletionReply,
+    InlineCompletionRequest,
+    InlineCompletionStreamChunk,
+    ModelChangedNotification,
+)
+from .base import BaseInlineCompletionHandler
+
+SYSTEM_PROMPT = """
+You are an application built to provide helpful code completion suggestions.
+You should only produce code. Keep comments to minimum, use the
+programming language comment syntax. Produce clean code.
+The code is written in JupyterLab, a data analysis and code development
+environment which can execute code extended with additional syntax for
+interactive features, such as magics.
+""".strip()
+
+AFTER_TEMPLATE = """
+The code after the completion request is:
+
+```
+{suffix}
+```
+""".strip()
+
+DEFAULT_TEMPLATE = """
+The document is called `{filename}` and written in {language}.
+{after}
+
+Complete the following code:
+
+```
+{prefix}"""
+
+
+class DefaultInlineCompletionHandler(BaseInlineCompletionHandler):
+    llm_chain: Runnable
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+    def create_llm_chain(
+        self, provider: Type[BaseProvider], provider_params: Dict[str, str]
+    ):
+        lm_provider = self.config_manager.lm_provider
+        lm_provider_params = self.config_manager.lm_provider_params
+        next_lm_id = (
+            f'{lm_provider.id}:{lm_provider_params["model_id"]}'
+            if lm_provider
+            else None
+        )
+        self.broadcast(ModelChangedNotification(model=next_lm_id))
+
+        model_parameters = self.get_model_parameters(provider, provider_params)
+        llm = provider(**provider_params, **model_parameters)
+
+        if llm.is_chat_provider:
+            prompt_template = ChatPromptTemplate.from_messages(
+                [
+                    SystemMessagePromptTemplate.from_template(SYSTEM_PROMPT),
+                    HumanMessagePromptTemplate.from_template(DEFAULT_TEMPLATE),
+                ]
+            )
+        else:
+            prompt_template = PromptTemplate(
+                input_variables=["prefix", "suffix", "language", "filename"],
+                template=SYSTEM_PROMPT + "\n\n" + DEFAULT_TEMPLATE,
+            )
+
+        self.llm = llm
+        self.llm_chain = prompt_template | llm | StrOutputParser()
+
+    async def process_message(
+        self, request: InlineCompletionRequest
+    ) -> InlineCompletionReply:
+        if request.stream:
+            token = self._token_from_request(request, 0)
+            return InlineCompletionReply(
+                list=InlineCompletionList(
+                    items=[
+                        {
+                            # insert text starts empty as we do not pre-generate any part
+                            "insertText": "",
+                            "isIncomplete": True,
+                            "token": token,
+                        }
+                    ]
+                ),
+                reply_to=request.number,
+            )
+        else:
+            self.get_llm_chain()
+            model_arguments = self._template_inputs_from_request(request)
+            suggestion = await self.llm_chain.ainvoke(input=model_arguments)
+            suggestion = self._post_process_suggestion(suggestion, request)
+            return InlineCompletionReply(
+                list=InlineCompletionList(items=[{"insertText": suggestion}]),
+                reply_to=request.number,
+            )
+
+    async def stream(self, request: InlineCompletionRequest):
+        self.get_llm_chain()
+        token = self._token_from_request(request, 0)
+        model_arguments = self._template_inputs_from_request(request)
+        suggestion = ""
+        async for fragment in self.llm_chain.astream(input=model_arguments):
+            suggestion += fragment
+            if suggestion.startswith("```"):
+                if "\n" not in suggestion:
+                    # we are not ready to apply post-processing
+                    continue
+                else:
+                    suggestion = self._post_process_suggestion(suggestion, request)
+            yield InlineCompletionStreamChunk(
+                type="stream",
+                response={"insertText": suggestion, "token": token},
+                reply_to=request.number,
+                done=False,
+            )
+        # at the end send a message confirming that we are done
+        yield InlineCompletionStreamChunk(
+            type="stream",
+            response={"insertText": suggestion, "token": token},
+            reply_to=request.number,
+            done=True,
+        )
+
+    def _token_from_request(self, request: InlineCompletionRequest, suggestion: int):
+        """Generate a deterministic token (for matching streamed messages)
+        using request number and suggestion number"""
+        return f"t{request.number}s{suggestion}"
+
+    def _template_inputs_from_request(self, request: InlineCompletionRequest) -> Dict:
+        suffix = request.suffix.strip()
+        # only add the suffix template if the suffix is there to save input tokens/computation time
+        after = AFTER_TEMPLATE.format(suffix=suffix) if suffix else ""
+        filename = request.path.split("/")[-1] if request.path else "untitled"
+
+        return {
+            "prefix": request.prefix,
+            "after": after,
+            "language": request.language,
+            "filename": filename,
+            "stop": ["\n```"],
+        }
+
+    def _post_process_suggestion(
+        self, suggestion: str, request: InlineCompletionRequest
+    ) -> str:
+        """Remove spurious fragments from the suggestion.
+
+        While most models (especially instruct and infill models do not require
+        any pre-processing, some models such as gpt-4 which only have chat APIs
+        may require removing spurious fragments. This function uses heuristics
+        and request data to remove such fragments.
+        """
+        # gpt-4 tends to add "```python" or similar
+        language = request.language or "python"
+        markdown_identifiers = {"ipython": ["ipython", "python", "py"]}
+        bad_openings = [
+            f"```{identifier}"
+            for identifier in markdown_identifiers.get(language, [language])
+        ] + ["```"]
+        for opening in bad_openings:
+            if suggestion.startswith(opening):
+                suggestion = suggestion[len(opening) :].lstrip()
+                # check for the prefix inclusion (only if there was a bad opening)
+                if suggestion.startswith(request.prefix):
+                    suggestion = suggestion[len(request.prefix) :]
+                break
+        return suggestion

packages/jupyter-ai/jupyter_ai/completions/handlers/llm_mixin.py

@@ -0,0 +1,73 @@
+from typing import Dict, Type
+
+from jupyter_ai.config_manager import ConfigManager, Logger
+from jupyter_ai_magics.providers import BaseProvider
+
+
+class LLMHandlerMixin:
+    """Base class containing shared methods and attributes used by LLM handler classes."""
+
+    # This could be used to derive `BaseChatHandler` too (there is a lot of duplication!),
+    # but it was decided against it to avoid introducing conflicts for backports against 1.x
+
+    handler_kind: str
+
+    def __init__(
+        self,
+        log: Logger,
+        config_manager: ConfigManager,
+        model_parameters: Dict[str, Dict],
+    ):
+        self.log = log
+        self.config_manager = config_manager
+        self.model_parameters = model_parameters
+        self.llm = None
+        self.llm_params = None
+        self.llm_chain = None
+
+    def model_changed_callback(self):
+        """Method which can be overridden in sub-classes to listen to model change."""
+        pass
+
+    def get_llm_chain(self):
+        lm_provider = self.config_manager.lm_provider
+        lm_provider_params = self.config_manager.lm_provider_params
+
+        curr_lm_id = (
+            f'{self.llm.id}:{lm_provider_params["model_id"]}' if self.llm else None
+        )
+        next_lm_id = (
+            f'{lm_provider.id}:{lm_provider_params["model_id"]}'
+            if lm_provider
+            else None
+        )
+
+        if not lm_provider or not lm_provider_params:
+            return None
+
+        if curr_lm_id != next_lm_id:
+            self.log.info(
+                f"Switching {self.handler_kind} language model from {curr_lm_id} to {next_lm_id}."
+            )
+            self.create_llm_chain(lm_provider, lm_provider_params)
+            self.model_changed_callback()
+        elif self.llm_params != lm_provider_params:
+            self.log.info(
+                f"{self.handler_kind} model params changed, updating the llm chain."
+            )
+            self.create_llm_chain(lm_provider, lm_provider_params)
+
+        self.llm_params = lm_provider_params
+        return self.llm_chain
+
+    def get_model_parameters(
+        self, provider: Type[BaseProvider], provider_params: Dict[str, str]
+    ):
+        return self.model_parameters.get(
+            f"{provider.id}:{provider_params['model_id']}", {}
+        )
+
+    def create_llm_chain(
+        self, provider: Type[BaseProvider], provider_params: Dict[str, str]
+    ):
+        raise NotImplementedError("Should be implemented by subclasses")