langroid 0.1.235__tar.gz → 0.1.237__tar.gz

{langroid-0.1.235 → langroid-0.1.237}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langroid
-Version: 0.1.235
+Version: 0.1.237
 Summary: Harness LLMs with Multi-Agent Programming
 License: MIT
 Author: Prasad Chalasani
@@ -29,12 +29,13 @@ Requires-Dist: chainlit (>=1.0.400,<2.0.0) ; extra == "chainlit"
 Requires-Dist: chromadb (>=0.4.21,<=0.4.23) ; extra == "chromadb"
 Requires-Dist: colorlog (>=6.7.0,<7.0.0)
 Requires-Dist: docstring-parser (>=0.15,<0.16)
-Requires-Dist: duckduckgo-search (>=4.4,<5.0)
+Requires-Dist: duckduckgo-search (>=5.3.0,<6.0.0)
 Requires-Dist: faker (>=18.9.0,<19.0.0)
 Requires-Dist: fakeredis (>=2.12.1,<3.0.0)
 Requires-Dist: fire (>=0.5.0,<0.6.0)
 Requires-Dist: flake8 (>=6.0.0,<7.0.0)
 Requires-Dist: google-api-python-client (>=2.95.0,<3.0.0)
+Requires-Dist: groq (>=0.5.0,<0.6.0)
 Requires-Dist: grpcio (>=1.62.1,<2.0.0)
 Requires-Dist: halo (>=0.0.31,<0.0.32)
 Requires-Dist: huggingface-hub (>=0.21.2,<0.22.0) ; extra == "transformers"
@@ -57,6 +58,7 @@ Requires-Dist: mkdocstrings[python] (>=0.21.2,<0.22.0) ; extra == "mkdocs"
 Requires-Dist: momento (>=1.10.2,<2.0.0)
 Requires-Dist: mypy (>=1.7.0,<2.0.0)
 Requires-Dist: neo4j (>=5.14.1,<6.0.0) ; extra == "neo4j"
+Requires-Dist: nest-asyncio (>=1.6.0,<2.0.0)
 Requires-Dist: nltk (>=3.8.1,<4.0.0)
 Requires-Dist: onnxruntime (==1.16.1)
 Requires-Dist: openai (>=1.14.0,<2.0.0)
@@ -81,6 +83,7 @@ Requires-Dist: pytest-postgresql (>=5.0.0,<6.0.0) ; extra == "postgres"
 Requires-Dist: pytest-redis (>=3.0.2,<4.0.0)
 Requires-Dist: python-docx (>=1.1.0,<2.0.0)
 Requires-Dist: python-dotenv (>=1.0.0,<2.0.0)
+Requires-Dist: python-magic (>=0.4.27,<0.5.0)
 Requires-Dist: python-socketio (>=5.11.0,<6.0.0) ; extra == "chainlit"
 Requires-Dist: qdrant-client (>=1.8.0,<2.0.0)
 Requires-Dist: rank-bm25 (>=0.2.2,<0.3.0)
@@ -230,6 +233,19 @@ teacher_task.run()
 <details>
 <summary> <b>Click to expand</b></summary>
 
+- **Apr 2024:**
+  - **0.1.236**: Support for open LLMs hosted on Groq, e.g. specify 
+    `chat_model="groq/llama3-8b-8192"`.
+      See [tutorial](https://langroid.github.io/langroid/tutorials/local-llm-setup/).
+  - **0.1.235**: `Task.run(), Task.run_async(), run_batch_tasks` have `max_cost` 
+    and `max_tokens` params to exit when tokens or cost exceed a limit. The result 
+    `ChatDocument.metadata` now includes a `status` field which is a code indicating a 
+     task completion reason code. Also `task.run()` etc can be invoked with an explicit
+     `session_id` field which is used as a key to look up various settings in Redis cache.
+    Currently only used to look up "kill status" - this allows killing a running task, either by `task.kill()`
+    or by the classmethod `Task.kill_session(session_id)`.
+    For example usage, see the `test_task_kill` in [tests/main/test_task.py](https://github.com/langroid/langroid/blob/main/tests/main/test_task.py)
+  
 - **Mar 2024:**
   - **0.1.216:** Improvements to allow concurrent runs of `DocChatAgent`, see the
     [`test_doc_chat_agent.py`](https://github.com/langroid/langroid/blob/main/tests/main/test_doc_chat_agent.py)

{langroid-0.1.235 → langroid-0.1.237}/README.md

@@ -122,6 +122,19 @@ teacher_task.run()
 <details>
 <summary> <b>Click to expand</b></summary>
 
+- **Apr 2024:**
+  - **0.1.236**: Support for open LLMs hosted on Groq, e.g. specify 
+    `chat_model="groq/llama3-8b-8192"`.
+      See [tutorial](https://langroid.github.io/langroid/tutorials/local-llm-setup/).
+  - **0.1.235**: `Task.run(), Task.run_async(), run_batch_tasks` have `max_cost` 
+    and `max_tokens` params to exit when tokens or cost exceed a limit. The result 
+    `ChatDocument.metadata` now includes a `status` field which is a code indicating a 
+     task completion reason code. Also `task.run()` etc can be invoked with an explicit
+     `session_id` field which is used as a key to look up various settings in Redis cache.
+    Currently only used to look up "kill status" - this allows killing a running task, either by `task.kill()`
+    or by the classmethod `Task.kill_session(session_id)`.
+    For example usage, see the `test_task_kill` in [tests/main/test_task.py](https://github.com/langroid/langroid/blob/main/tests/main/test_task.py)
+  
 - **Mar 2024:**
   - **0.1.216:** Improvements to allow concurrent runs of `DocChatAgent`, see the
     [`test_doc_chat_agent.py`](https://github.com/langroid/langroid/blob/main/tests/main/test_doc_chat_agent.py)

{langroid-0.1.235 → langroid-0.1.237}/langroid/agent/openai_assistant.py

@@ -7,6 +7,7 @@ import time
 from enum import Enum
 from typing import Any, Dict, List, Optional, Tuple, Type, cast, no_type_check
 
+import openai
 from openai.types.beta import Assistant, Thread
 from openai.types.beta.assistant_update_params import (
     ToolResources,
@@ -99,12 +100,14 @@ class OpenAIAssistant(ChatAgent):
         super().__init__(config)
         self.config: OpenAIAssistantConfig = config
         self.llm: OpenAIGPT = OpenAIGPT(self.config.llm)
+        if not isinstance(self.llm.client, openai.OpenAI):
+            raise ValueError("Client must be OpenAI")
         # handles for various entities and methods
-        self.client = self.llm.client
-        self.runs = self.llm.client.beta.threads.runs
-        self.threads = self.llm.client.beta.threads
-        self.thread_messages = self.llm.client.beta.threads.messages
-        self.assistants = self.llm.client.beta.assistants
+        self.client: openai.OpenAI = self.llm.client
+        self.runs = self.client.beta.threads.runs
+        self.threads = self.client.beta.threads
+        self.thread_messages = self.client.beta.threads.messages
+        self.assistants = self.client.beta.assistants
         # which tool_ids are awaiting output submissions
         self.pending_tool_ids: List[str] = []
         self.cached_tool_ids: List[str] = []
@@ -208,14 +211,14 @@ class OpenAIAssistant(ChatAgent):
 
     def _cache_thread_key(self) -> str:
         """Key to use for caching or retrieving thread id"""
-        org = self.llm.client.organization or ""
+        org = self.client.organization or ""
         uid = generate_user_id(org)
         name = self.config.name
         return "Thread:" + name + ":" + uid
 
     def _cache_assistant_key(self) -> str:
         """Key to use for caching or retrieving assistant id"""
-        org = self.llm.client.organization or ""
+        org = self.client.organization or ""
         uid = generate_user_id(org)
         name = self.config.name
         return "Assistant:" + name + ":" + uid
@@ -317,7 +320,7 @@ class OpenAIAssistant(ChatAgent):
         cached = self._cache_thread_lookup()
         if cached is not None:
             if self.config.use_cached_thread:
-                self.thread = self.llm.client.beta.threads.retrieve(thread_id=cached)
+                self.thread = self.client.beta.threads.retrieve(thread_id=cached)
             else:
                 logger.warning(
                     f"""
@@ -326,7 +329,7 @@ class OpenAIAssistant(ChatAgent):
                     """
                 )
                 try:
-                    self.llm.client.beta.threads.delete(thread_id=cached)
+                    self.client.beta.threads.delete(thread_id=cached)
                 except Exception:
                     logger.warning(
                         f"""
@@ -337,7 +340,7 @@ class OpenAIAssistant(ChatAgent):
         if self.thread is None:
             if self.assistant is None:
                 raise ValueError("Assistant is None")
-            self.thread = self.llm.client.beta.threads.create()
+            self.thread = self.client.beta.threads.create()
             hash_key_str = (
                 (self.assistant.instructions or "")
                 + str(self.config.use_tools)
@@ -371,7 +374,7 @@ class OpenAIAssistant(ChatAgent):
         cached = self._cache_assistant_lookup()
         if cached is not None:
             if self.config.use_cached_assistant:
-                self.assistant = self.llm.client.beta.assistants.retrieve(
+                self.assistant = self.client.beta.assistants.retrieve(
                     assistant_id=cached
                 )
             else:
@@ -382,7 +385,7 @@ class OpenAIAssistant(ChatAgent):
                     """
                 )
                 try:
-                    self.llm.client.beta.assistants.delete(assistant_id=cached)
+                    self.client.beta.assistants.delete(assistant_id=cached)
                 except Exception:
                     logger.warning(
                         f"""
@@ -391,7 +394,7 @@ class OpenAIAssistant(ChatAgent):
                     )
                 self.llm.cache.delete_keys([self._cache_assistant_key()])
         if self.assistant is None:
-            self.assistant = self.llm.client.beta.assistants.create(
+            self.assistant = self.client.beta.assistants.create(
                 name=self.config.name,
                 instructions=self.config.system_message,
                 tools=[],

{langroid-0.1.235 → langroid-0.1.237}/langroid/agent/task.py

@@ -84,6 +84,7 @@ class Task:
 
     # class variable called `cache` that is a RedisCache object
     cache: RedisCache = RedisCache(RedisCacheConfig(fake=False))
+
     def __init__(
         self,
         agent: Optional[Agent] = None,
@@ -331,9 +332,9 @@ class Task:
         self._cache_session_store("kill", "0")
 
     @classmethod
-    def kill_session(cls, session_id=""):
+    def kill_session(cls, session_id: str = "") -> None:
         """
-        Kill the current session.
+        Kill the session with the given session_id.
         """
         session_id_kill_key = f"{session_id}:kill"
         cls.cache.store(session_id_kill_key, "1")

{langroid-0.1.235 → langroid-0.1.237}/langroid/language_models/openai_gpt.py

@@ -21,6 +21,7 @@ from typing import (
 )
 
 import openai
+from groq import AsyncGroq, Groq
 from httpx import Timeout
 from openai import AsyncOpenAI, OpenAI
 from pydantic import BaseModel
@@ -347,6 +348,9 @@ class OpenAIGPT(LanguageModel):
     Class for OpenAI LLMs
     """
 
+    client: OpenAI | Groq
+    async_client: AsyncOpenAI | AsyncGroq
+
     def __init__(self, config: OpenAIGPTConfig = OpenAIGPTConfig()):
         """
         Args:
@@ -448,18 +452,31 @@ class OpenAIGPT(LanguageModel):
                 self.api_key = os.getenv("OPENAI_API_KEY", DUMMY_API_KEY)
         else:
             self.api_key = DUMMY_API_KEY
-        self.client = OpenAI(
-            api_key=self.api_key,
-            base_url=self.api_base,
-            organization=self.config.organization,
-            timeout=Timeout(self.config.timeout),
-        )
-        self.async_client = AsyncOpenAI(
-            api_key=self.api_key,
-            organization=self.config.organization,
-            base_url=self.api_base,
-            timeout=Timeout(self.config.timeout),
-        )
+
+        self.is_groq = self.config.chat_model.startswith("groq/")
+
+        if self.is_groq:
+            self.config.chat_model = self.config.chat_model.replace("groq/", "")
+            self.api_key = os.getenv("GROQ_API_KEY", DUMMY_API_KEY)
+            self.client = Groq(
+                api_key=self.api_key,
+            )
+            self.async_client = AsyncGroq(
+                api_key=self.api_key,
+            )
+        else:
+            self.client = OpenAI(
+                api_key=self.api_key,
+                base_url=self.api_base,
+                organization=self.config.organization,
+                timeout=Timeout(self.config.timeout),
+            )
+            self.async_client = AsyncOpenAI(
+                api_key=self.api_key,
+                organization=self.config.organization,
+                base_url=self.api_base,
+                timeout=Timeout(self.config.timeout),
+            )
 
         self.cache: MomentoCache | RedisCache
         if settings.cache_type == "momento":
@@ -855,6 +872,9 @@ class OpenAIGPT(LanguageModel):
         if self.config.use_chat_for_completion:
             return self.chat(messages=prompt, max_tokens=max_tokens)
 
+        if self.is_groq:
+            raise ValueError("Groq does not support pure completions")
+
         if settings.debug:
             print(f"[grey37]PROMPT: {escape(prompt)}[/grey37]")
 
@@ -869,6 +889,7 @@ class OpenAIGPT(LanguageModel):
             else:
                 if self.config.litellm:
                     from litellm import completion as litellm_completion
+                assert isinstance(self.client, OpenAI)
                 completion_call = (
                     litellm_completion
                     if self.config.litellm
@@ -929,6 +950,9 @@ class OpenAIGPT(LanguageModel):
         if self.config.use_chat_for_completion:
             return await self.achat(messages=prompt, max_tokens=max_tokens)
 
+        if self.is_groq:
+            raise ValueError("Groq does not support pure completions")
+
         if settings.debug:
             print(f"[grey37]PROMPT: {escape(prompt)}[/grey37]")
 
@@ -948,6 +972,7 @@ class OpenAIGPT(LanguageModel):
                     from litellm import acompletion as litellm_acompletion
                 # TODO this may not work: text_completion is not async,
                 # and we didn't find an async version in litellm
+                assert isinstance(self.async_client, AsyncOpenAI)
                 acompletion_call = (
                     litellm_acompletion
                     if self.config.litellm

{langroid-0.1.235 → langroid-0.1.237}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "langroid"
-version = "0.1.235"
+version = "0.1.237"
 description = "Harness LLMs with Multi-Agent Programming"
 authors = ["Prasad Chalasani <pchalasani@gmail.com>"]
 readme = "README.md"
@@ -90,12 +90,15 @@ grpcio = "^1.62.1"
 metaphor-python = {version = "^0.1.23", optional = true}
 chainlit = {version = "^1.0.400", optional = true}
 python-socketio = {version="^5.11.0", optional=true}
-duckduckgo-search = "^4.4"
+duckduckgo-search = "^5.3.0"
 huggingface-hub = {version="^0.21.2", optional=true}
 pdf2image = "^1.17.0"
 pytesseract = "^0.3.10"
 types-pyyaml = "^6.0.12.20240311"
 types-pillow = "^10.2.0.20240406"
+groq = "^0.5.0"
+nest-asyncio = "^1.6.0"
+python-magic = "^0.4.27"
 
 [tool.poetry.extras]
 # install these using `poetry install -E [...]` where [...] is one of the extras below