Adjusted GPT4All llm to streaming API and added support for GPT4All_J (#4131)

Fix for these issues:
https://github.com/hwchase17/langchain/issues/4126

https://github.com/hwchase17/langchain/issues/3839#issuecomment-1534258559

---------

Co-authored-by: Pawel Faron <ext-pawel.faron@vaisala.com>

docs/modules/models/llms/integrations/gpt4all.ipynb

@@ -125,7 +125,9 @@
                 "# Callbacks support token-wise streaming\n",
                 "callbacks = [StreamingStdOutCallbackHandler()]\n",
                 "# Verbose is required to pass to the callback manager\n",
-                "llm = GPT4All(model=local_path, callbacks=callbacks, verbose=True)"
+                "llm = GPT4All(model=local_path, callbacks=callbacks, verbose=True)\n",
+                "# If you want to use GPT4ALL_J model add the backend parameter\n",
+                "llm = GPT4All(model=local_path, backend='gptj', callbacks=callbacks, verbose=True)"
             ]
         },
         {

langchain/llms/gpt4all.py

@@ -28,6 +28,8 @@ class GPT4All(LLM):
     model: str
     """Path to the pre-trained GPT4All model file."""
 
+    backend: str = Field("llama", alias="backend")
+
     n_ctx: int = Field(512, alias="n_ctx")
     """Token context window."""
 
@@ -93,14 +95,11 @@ class GPT4All(LLM):
 
         extra = Extra.forbid
 
-    @property
-    def _default_params(self) -> Dict[str, Any]:
+    def _llama_default_params(self) -> Dict[str, Any]:
         """Get the identifying parameters."""
         return {
-            "seed": self.seed,
             "n_predict": self.n_predict,
             "n_threads": self.n_threads,
-            "n_batch": self.n_batch,
             "repeat_last_n": self.repeat_last_n,
             "repeat_penalty": self.repeat_penalty,
             "top_k": self.top_k,
@@ -108,6 +107,16 @@ class GPT4All(LLM):
             "temp": self.temp,
         }
 
+    def _gptj_default_params(self) -> Dict[str, Any]:
+        """Get the identifying parameters."""
+        return {
+            "n_predict": self.n_predict,
+            "n_threads": self.n_threads,
+            "top_k": self.top_k,
+            "top_p": self.top_p,
+            "temp": self.temp,
+        }
+
     @staticmethod
     def _llama_param_names() -> Set[str]:
         """Get the identifying parameters."""
@@ -122,14 +131,41 @@ class GPT4All(LLM):
             "embedding",
         }
 
+    @staticmethod
+    def _gptj_param_names() -> Set[str]:
+        """Get the identifying parameters."""
+        return set()
+
+    @staticmethod
+    def _model_param_names(backend: str) -> Set[str]:
+        if backend == "llama":
+            return GPT4All._llama_param_names()
+        else:
+            return GPT4All._gptj_param_names()
+
+    def _default_params(self) -> Dict[str, Any]:
+        if self.backend == "llama":
+            return self._llama_default_params()
+        else:
+            return self._gptj_default_params()
+
     @root_validator()
     def validate_environment(cls, values: Dict) -> Dict:
         """Validate that the python package exists in the environment."""
         try:
-            from pygpt4all.models.gpt4all import GPT4All as GPT4AllModel
-
-            llama_keys = cls._llama_param_names()
-            model_kwargs = {k: v for k, v in values.items() if k in llama_keys}
+            backend = values["backend"]
+            if backend == "llama":
+                from pygpt4all import GPT4All as GPT4AllModel
+            elif backend == "gptj":
+                from pygpt4all import GPT4All_J as GPT4AllModel
+            else:
+                raise ValueError(f"Incorrect gpt4all backend {cls.backend}")
+
+            model_kwargs = {
+                k: v
+                for k, v in values.items()
+                if k in GPT4All._model_param_names(backend)
+            }
             values["client"] = GPT4AllModel(
                 model_path=values["model"],
                 **model_kwargs,
@@ -147,11 +183,11 @@ class GPT4All(LLM):
         """Get the identifying parameters."""
         return {
             "model": self.model,
-            **self._default_params,
+            **self._default_params(),
             **{
                 k: v
                 for k, v in self.__dict__.items()
-                if k in GPT4All._llama_param_names()
+                if k in self._model_param_names(self.backend)
             },
         }
 
@@ -181,15 +217,14 @@ class GPT4All(LLM):
                 prompt = "Once upon a time, "
                 response = model(prompt, n_predict=55)
         """
+        text_callback = None
         if run_manager:
             text_callback = partial(run_manager.on_llm_new_token, verbose=self.verbose)
-            text = self.client.generate(
-                prompt,
-                new_text_callback=text_callback,
-                **self._default_params,
-            )
-        else:
-            text = self.client.generate(prompt, **self._default_params)
+        text = ""
+        for token in self.client.generate(prompt, **self._default_params()):
+            if text_callback:
+                text_callback(token)
+            text += token
         if stop is not None:
             text = enforce_stop_tokens(text, stop)
         return text