fix yield + version number

cookbook/liteLLM_Ollama.ipynb

@@ -0,0 +1,135 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Defaulting to user installation because normal site-packages is not writeable\n",
+      "Collecting litellm==0.1.384\n",
+      "  Downloading litellm-0.1.384-py3-none-any.whl (43 kB)\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m43.1/43.1 kB\u001b[0m \u001b[31m2.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[?25hRequirement already satisfied: openai<0.28.0,>=0.27.8 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from litellm==0.1.384) (0.27.8)\n",
+      "Requirement already satisfied: python-dotenv<2.0.0,>=1.0.0 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from litellm==0.1.384) (1.0.0)\n",
+      "Requirement already satisfied: tiktoken<0.5.0,>=0.4.0 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from litellm==0.1.384) (0.4.0)\n",
+      "Requirement already satisfied: requests>=2.20 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from openai<0.28.0,>=0.27.8->litellm==0.1.384) (2.28.2)\n",
+      "Requirement already satisfied: tqdm in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from openai<0.28.0,>=0.27.8->litellm==0.1.384) (4.65.0)\n",
+      "Requirement already satisfied: aiohttp in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from openai<0.28.0,>=0.27.8->litellm==0.1.384) (3.8.4)\n",
+      "Requirement already satisfied: regex>=2022.1.18 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from tiktoken<0.5.0,>=0.4.0->litellm==0.1.384) (2023.6.3)\n",
+      "Requirement already satisfied: charset-normalizer<4,>=2 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from requests>=2.20->openai<0.28.0,>=0.27.8->litellm==0.1.384) (3.1.0)\n",
+      "Requirement already satisfied: idna<4,>=2.5 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from requests>=2.20->openai<0.28.0,>=0.27.8->litellm==0.1.384) (3.4)\n",
+      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from requests>=2.20->openai<0.28.0,>=0.27.8->litellm==0.1.384) (1.26.6)\n",
+      "Requirement already satisfied: certifi>=2017.4.17 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from requests>=2.20->openai<0.28.0,>=0.27.8->litellm==0.1.384) (2023.5.7)\n",
+      "Requirement already satisfied: attrs>=17.3.0 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (23.1.0)\n",
+      "Requirement already satisfied: multidict<7.0,>=4.5 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (6.0.4)\n",
+      "Requirement already satisfied: async-timeout<5.0,>=4.0.0a3 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (4.0.2)\n",
+      "Requirement already satisfied: yarl<2.0,>=1.0 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (1.9.2)\n",
+      "Requirement already satisfied: frozenlist>=1.1.1 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (1.3.3)\n",
+      "Requirement already satisfied: aiosignal>=1.1.2 in /Users/ishaanjaffer/Library/Python/3.9/lib/python/site-packages (from aiohttp->openai<0.28.0,>=0.27.8->litellm==0.1.384) (1.3.1)\n",
+      "Installing collected packages: litellm\n",
+      "  Attempting uninstall: litellm\n",
+      "    Found existing installation: litellm 0.1.379\n",
+      "    Uninstalling litellm-0.1.379:\n",
+      "      Successfully uninstalled litellm-0.1.379\n",
+      "Successfully installed litellm-0.1.384\n",
+      "\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m A new release of pip is available: \u001b[0m\u001b[31;49m23.1.2\u001b[0m\u001b[39;49m -> \u001b[0m\u001b[32;49m23.2.1\u001b[0m\n",
+      "\u001b[1m[\u001b[0m\u001b[34;49mnotice\u001b[0m\u001b[1;39;49m]\u001b[0m\u001b[39;49m To update, run: \u001b[0m\u001b[32;49m/Library/Developer/CommandLineTools/usr/bin/python3 -m pip install --upgrade pip\u001b[0m\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install litellm==0.1.384"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from litellm import completion\n",
+    "import asyncio"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "user_message = \"respond in 20 words. who are you?\"\n",
+    "messages = [{ \"content\": user_message,\"role\": \"user\"}]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "async def get_response(generator):\n",
+    "    response = \"\"\n",
+    "    async for elem in generator:\n",
+    "        print(elem)\n",
+    "        response += elem[\"content\"]\n",
+    "    return response"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "<async_generator object get_ollama_response_stream at 0x103d51940>\n"
+     ]
+    }
+   ],
+   "source": [
+    "response = completion(model=\"llama2\", messages=messages, custom_api_base=\"http://localhost:11434\", custom_llm_provider=\"ollama\", stream=True)\n",
+    "print(response)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "string_response = asyncio.run(get_response(response))"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.6"
+  },
+  "orig_nbformat": 4
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

litellm/utils.py

@@ -766,11 +766,9 @@ async def get_ollama_response_stream(api_base="http://localhost:11434", model="l
                             if chunk.strip() != "":
                                 j = json.loads(chunk)
                                 if "response" in j:
-                                    print(j["response"])
+                                    completion_obj ={ "role": "assistant", "content": ""}
-                                    yield {
+                                    completion_obj["content"] = j["response"]
-                                        "role": "assistant",
+                                    yield {"choices": [{"delta": completion_obj}]}
-                                        "content": j["response"]
-                                    }
                                     # self.responses.append(j["response"])
                                     # yield "blank"
                     except Exception as e:

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "litellm"
-version = "0.1.384"
+version = "0.1.385"
 description = "Library to easily interface with LLM API providers"
 authors = ["BerriAI"]
 license = "MIT License"