langroid 0.1.31__py3-none-any.whl → 0.1.33__py3-none-any.whl

langroid/agent/special/retriever_agent.py

@@ -177,7 +177,7 @@ class RetrieverAgent(DocChatAgent, ABC):
             list of Document objects
         """
         doc_contents = "\n\n".join(
-            [f"DOC: ID={d.id()}, content={d.content}" for d in docs]
+            [f"DOC: ID={d.id()}, CONTENT: {d.content}" for d in docs]
         )
         prompt = f"""
         Given the following QUERY: 
@@ -186,8 +186,11 @@ class RetrieverAgent(DocChatAgent, ABC):
         {doc_contents}
         
         Find at most {self.config.n_matches} DOCs that are most relevant to the QUERY.
-        Return your as a sequence of DOC IDS ONLY, for example: 
+        Return your answer as a sequence of DOC IDS ONLY, for example: 
         "id1 id2 id3..."
+        If there are no relevant docs, simply say {NO_ANSWER}.
+        Even if there is only one relevant doc, return it as a single ID.
+        Do not give any explanations or justifications.
         """
         default_response = Document(
             content=NO_ANSWER,

langroid/agent/task.py

@@ -61,6 +61,7 @@ class Task:
         user_message: str = "",
         restart: bool = False,
         default_human_response: Optional[str] = None,
+        interactive: bool = True,
         only_user_quits_root: bool = True,
         erase_substeps: bool = False,
     ):
@@ -84,6 +85,9 @@ class Task:
             restart (bool): if true, resets the agent's message history
             default_human_response (str): default response from user; useful for
                 testing, to avoid interactive input from user.
+            interactive (bool): if true, wait for human input after each non-human
+                response (prevents infinite loop of non-human responses).
+                Default is true. If false, then `default_human_response` is set to ""
             only_user_quits_root (bool): if true, only user can quit the root task.
             erase_substeps (bool): if true, when task completes, erase intermediate
                 conversation with subtasks from this agent's `message_history`, and also
@@ -111,6 +115,9 @@ class Task:
         self.agent = agent
         self.name = name or agent.config.name
         self.default_human_response = default_human_response
+        self.interactive = interactive
+        if not interactive:
+            self.default_human_response = ""
         if default_human_response is not None:
             self.agent.default_human_response = default_human_response
         self.only_user_quits_root = only_user_quits_root

langroid/cachedb/momento_cachedb.py

@@ -15,7 +15,7 @@ logger = logging.getLogger(__name__)
 
 
 class MomentoCacheConfig(BaseModel):
-    """Configuration model for RedisCache."""
+    """Configuration model for Momento Cache."""
 
     ttl: int = 60 * 60 * 24 * 7  # 1 week
     cachename: str = "langroid_momento_cache"

langroid/language_models/openai_gpt.py

@@ -224,7 +224,7 @@ class OpenAIGPT(LanguageModel):
         hashed_key = hashlib.sha256(raw_key.encode()).hexdigest()
 
         if not settings.cache:
-            # when cacheing disabled, return the hashed_key and none result
+            # when caching disabled, return the hashed_key and none result
             return hashed_key, None
         # Try to get the result from the cache
         return hashed_key, self.cache.retrieve(hashed_key)

langroid/parsing/repo_loader.py

@@ -96,7 +96,7 @@ class RepoLoader:
                 json.dump({"junk": "ignore"}, f)
         with open(self.log_file, "r") as f:
             log = json.load(f)
-        if self.url in log:
+        if self.url in log and os.path.exists(log[self.url]):
             logger.info(f"Repo Already downloaded in {log[self.url]}")
             self.clone_path = log[self.url]
 

langroid/prompts/templates.py

@@ -55,7 +55,15 @@ SUMMARY_ANSWER_PROMPT_GPT4 = f"""
         and even if the answer contradicts other parts of the document. The only 
         important thing is that your answer is consistent with and supported by the 
         extracts. Compose your complete answer and cite all supporting sources on a 
-        separate separate line as "SOURCE:". 
+        separate separate line as follows. For each SOURCE, (a) show the source name,
+        and (b) the exact EXTRACT from the source that supports your answer.
+        DO NOT SHOW THE FULL EXTRACT; only show the FIRST 3 words and LAST 3 words.  
+        For example:
+        SOURCE:  https://en.wikipedia.org/wiki/Tree
+        EXTRACTS: "The tree species ... oak, maple, and birch."
+        
+        SOURCE: https://en.wikipedia.org/wiki/Oak
+        EXTRACTS: "The oak trees ... longevity and strength."
         
         {{extracts}}
         

langroid/utils/configuration.py

@@ -1,9 +1,12 @@
 from typing import List
 
+from dotenv import load_dotenv
 from pydantic import BaseSettings
 
 
 class Settings(BaseSettings):
+    # NOTE all of these can be overridden in your .env file with upper-case names,
+    # for example CACHE_TYPE=momento
     debug: bool = False  # show debug messages?
     progress: bool = False  # show progress spinners/bars?
     stream: bool = True  # stream output?
@@ -17,6 +20,7 @@ class Settings(BaseSettings):
         extra = "forbid"
 
 
+load_dotenv()  # get settings from .env file
 settings = Settings()
 
 

{langroid-0.1.31.dist-info → langroid-0.1.33.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: langroid
-Version: 0.1.31
+Version: 0.1.33
 Summary: Harness LLMs with Multi-Agent Programming
 License: MIT
 Author: Prasad Chalasani
@@ -154,7 +154,8 @@ Here is what it looks like in action:
   Agents with specific skills, wrap them in Tasks, and combine tasks in a flexible way.
 - **LLM Support**: Langroid supports OpenAI LLMs including GPT-3.5-Turbo,
   GPT-4-0613
-- **Caching of LLM prompts, responses:** Langroid uses [Redis](https://redis.com/try-free/) for caching.
+- **Caching of LLM responses:** Langroid supports [Redis](https://redis.com/try-free/) and 
+  [Momento](https://www.gomomento.com/) to cache LLM responses.
 - **Vector-stores**: [Qdrant](https://qdrant.tech/) and [Chroma](https://www.trychroma.com/) are currently supported.
   Vector stores allow for Retrieval-Augmented-Generation (RAG).
 - **Grounding and source-citation:** Access to external documents via vector-stores 
@@ -226,14 +227,16 @@ All of the below are optional and not strictly needed to run any of the examples
   If you skip setting up these, Langroid will use Qdrant in local-storage mode.
   Alternatively [Chroma](https://docs.trychroma.com/) is also currently supported. 
   We use the local-storage version of Chroma, so there is no need for an API key.
+  Langroid uses Qdrant by default.
 - **Redis** Password, host, port: This is optional, and only needed to cache LLM API responses
   using Redis Cloud. Redis [offers](https://redis.com/try-free/) a free 30MB Redis account
   which is more than sufficient to try out Langroid and even beyond.
   If you don't set up these, Langroid will use a pure-python 
   Redis in-memory cache via the [Fakeredis](https://fakeredis.readthedocs.io/en/latest/) library.
 - **Momento** Serverless Caching of LLM API responses (as an alternative to Redis). 
-   To use Momento instead of Redis, simply enter your Momento Token in the `.env` file,
-   as the value of `MOMENTO_AUTH_TOKEN` (see example file below).
+   To use Momento instead of Redis:
+  - enter your Momento Token in the `.env` file, as the value of `MOMENTO_AUTH_TOKEN` (see example file below),
+  - in the `.env` file set `CACHE_TYPE=momento` (instead of `CACHE_TYPE=redis` which is the default).
 - **GitHub** Personal Access Token (required for apps that need to analyze git
   repos; token-based API calls are less rate-limited). See this
   [GitHub page](https://docs.github.com/en/authentication/keeping-your-account-and-data-secure/managing-your-personal-access-tokens).
@@ -242,6 +245,7 @@ If you add all of these optional variables, your `.env` file should look like th
 ```bash
 OPENAI_API_KEY=your-key-here-without-quotes
 GITHUB_ACCESS_TOKEN=your-personal-access-token-no-quotes
+CACHE_TYPE=redis # or momento
 REDIS_PASSWORD=your-redis-password-no-quotes
 REDIS_HOST=your-redis-hostname-no-quotes
 REDIS_PORT=your-redis-port-no-quotes

{langroid-0.1.31.dist-info → langroid-0.1.33.dist-info}/RECORD

@@ -8,14 +8,14 @@ langroid/agent/junk,sha256=LxfuuW7Cijsg0szAzT81OjWWv1PMNI-6w_-DspVIO2s,339
 langroid/agent/special/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/agent/special/doc_chat_agent.py,sha256=NhoS02rlLr3nAfRvLB1YNcmfsWZJH24K8O-m-uYnb-I,15741
 langroid/agent/special/recipient_validator_agent.py,sha256=x2UprcGlh-fyxQCZbb_fkKrruU5Om0mgOnNzk_PYBNM,4527
-langroid/agent/special/retriever_agent.py,sha256=AiGKIJdndV4u5kr3woINw-vLKhwWnsINlvXtXxHcJgQ,6995
+langroid/agent/special/retriever_agent.py,sha256=dkG_-QdL5KvpeS4K4jnx1kV9h_KxVEhgHz2rzdilZss,7196
 langroid/agent/special/table_chat_agent.py,sha256=6xoFcb3-y3xb3NAADV-sXid77et1KRz5PcwN9GshP6g,4673
-langroid/agent/task.py,sha256=irYxLnJ6J7cJCEyJnq6KDIDc6T2VeaaixsWUm2P2x_s,25844
+langroid/agent/task.py,sha256=KJZt6lSfCAVctyFfBJWLFcfys73kjrQJV4Y3aMVwO1M,26233
 langroid/agent/tool_message.py,sha256=7OdVcV7UyOZD2ihYgV1C_1fIwiWM-2pR8FFxoA1IgOo,5379
 langroid/agent_config.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/cachedb/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/cachedb/base.py,sha256=F-QSDlRXrC0tBRbxL397MX8hulfBMAnZNs1e9zH71OQ,790
-langroid/cachedb/momento_cachedb.py,sha256=TG_6sYhKvgl2b7t1bAQHDs9pXvkS6iPCwo-9_pMHwhs,2339
+langroid/cachedb/momento_cachedb.py,sha256=mfLnAJO3-mPucHjwwgHeU7HQzE_jmUxca2qt3RUHVC0,2342
 langroid/cachedb/redis_cachedb.py,sha256=xuQ96FAqcHTfK8PEt1tjrh1BkMWUjojFHIgjDfF3SnU,2369
 langroid/embedding_models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/embedding_models/base.py,sha256=176jDrjEAAhNzdFCG8pfossd8SAhvHR8Q5Y8pOOm0LI,983
@@ -23,7 +23,7 @@ langroid/embedding_models/clustering.py,sha256=tZWElUqXl9Etqla0FAa7og96iDKgjqWju
 langroid/embedding_models/models.py,sha256=1xcv9hqmCTsbUbS8v7XeZRsf25Tu79JUoSipIYpvNoo,2765
 langroid/language_models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/language_models/base.py,sha256=7FMiUyJ3Kf9Uiqvjh-lNGfwnCibY0SbnK4jdORCv3SM,12657
-langroid/language_models/openai_gpt.py,sha256=NQYQqTj5OqsFtOxq9LBsJktEldbpeEKQTj_Jd-G7_XM,17974
+langroid/language_models/openai_gpt.py,sha256=TuertCPkjJzz35YYGHTwX2E5gevENEPdPoFVyGl8M6k,17973
 langroid/language_models/utils.py,sha256=rmnSn-sJ3aKl_wBdeLPkck0Li4Ed6zkCxZYYl7n1V34,4668
 langroid/mytypes.py,sha256=YA42IJcooJnTxAwk-B4FmZ1hqzIIF1ZZKcpUKzBTGGo,1537
 langroid/parsing/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -34,7 +34,7 @@ langroid/parsing/json.py,sha256=d56BDPjs7mC4Dc6aqm3pfNkTnA0Pq15tAJT7YkeQc0s,981
 langroid/parsing/para_sentence_split.py,sha256=2wa2_BWGhZx_Uh9TF8VsCOvn1TfLfQJcHmffJVtBtMg,1829
 langroid/parsing/parser.py,sha256=2qPjFfVjFI4jmt-21gYJTCyhYGv4HvssQj0aAGEz9ZY,7350
 langroid/parsing/pdf_parser.py,sha256=RFrck50VvqYl37xzUp-cj4uSC4FDIvJqTwv100Dilgg,1432
-langroid/parsing/repo_loader.py,sha256=_Wlv123kdvRMcVXll-VArRpciBHS2sGOUJRcBe8p6bo,26983
+langroid/parsing/repo_loader.py,sha256=ILlvBH-wSvfdLLQKHklAzuxU980_ajts_bz7_9IwtLY,27017
 langroid/parsing/table_loader.py,sha256=4nK5QzQkwNwwzKRpg40pVLGarQp8HA9JqBQoRsGqDHA,1080
 langroid/parsing/url_loader.py,sha256=tdYPMC2V9zI_B207LFdU4aroZzJW4sY9y794XePFAVg,1824
 langroid/parsing/url_loader_cookies.py,sha256=Lg4sNpRz9MByWq2mde6T0hKv68VZSV3mtMjNEHuFeSU,2327
@@ -43,11 +43,11 @@ langroid/parsing/utils.py,sha256=giJOG_idlSrJi3BJWpot3ggent2VYnrNQS5Dj2lGpPk,683
 langroid/prompts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/prompts/dialog.py,sha256=SpfiSyofSgy2pwD1YboHR_yHO3LEEMbv6j2sm874jKo,331
 langroid/prompts/prompts_config.py,sha256=EMK1Fm7EmS8y3CV4AkrVgn5K4NipiM4m7J8819W1KeM,98
-langroid/prompts/templates.py,sha256=8jmDiCN926mFMalMYRUUzYykFnVojVeSTazdDH9V1zA,5850
+langroid/prompts/templates.py,sha256=4OAujKJzKIhqt4SQL2GE1aPahlZuDotlT_dZAKx0bqE,6311
 langroid/prompts/transforms.py,sha256=GsQo1klGxUy0fACh6j0lTblk6XEl2erRnhRWlN2M4-c,2706
 langroid/scripts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 langroid/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-langroid/utils/configuration.py,sha256=ApAAL8w3TRtAolEl-nHy5fegLwA4WffsytbEbbWl0Tg,1564
+langroid/utils/configuration.py,sha256=ZTKzE9hRx7QgoSiaAXMUJreWB5E-vOsU9MtPOKTzphY,1760
 langroid/utils/constants.py,sha256=_keF-e-GIwEFvBxKJbC1Ry0u3SVWQBk4HRX0NpV2Qq8,485
 langroid/utils/docker.py,sha256=kJQOLTgM0x9j9pgIIqp0dZNZCTvoUDhp6i8tYBq1Jr0,1105
 langroid/utils/llms/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -64,7 +64,7 @@ langroid/vector_store/base.py,sha256=QZx3NUNwf2I0r3A7iuoUHIRGbqt_pFGD0hq1R-Yg8iM
 langroid/vector_store/chromadb.py,sha256=s5pQkKjaMP-Tt5A8M10EInFzttaALPbJAq7q4gf0TKg,5235
 langroid/vector_store/qdrant_cloud.py,sha256=3im4Mip0QXLkR6wiqVsjV1QvhSElfxdFSuDKddBDQ-4,188
 langroid/vector_store/qdrantdb.py,sha256=KRvIIj1IZG2zFqejofMnRs2hT86B-27LgBEnuczdqOU,9072
-langroid-0.1.31.dist-info/LICENSE,sha256=EgVbvA6VSYgUlvC3RvPKehSg7MFaxWDsFuzLOsPPfJg,1065
-langroid-0.1.31.dist-info/WHEEL,sha256=vVCvjcmxuUltf8cYhJ0sJMRDLr1XsPuxEId8YDzbyCY,88
-langroid-0.1.31.dist-info/METADATA,sha256=JcvdrnuK85jDSoXwGapFFUF5LnYLHMz3OgJBQTSlYFU,24978
-langroid-0.1.31.dist-info/RECORD,,
+langroid-0.1.33.dist-info/LICENSE,sha256=EgVbvA6VSYgUlvC3RvPKehSg7MFaxWDsFuzLOsPPfJg,1065
+langroid-0.1.33.dist-info/WHEEL,sha256=vVCvjcmxuUltf8cYhJ0sJMRDLr1XsPuxEId8YDzbyCY,88
+langroid-0.1.33.dist-info/METADATA,sha256=gGIWKbaKr2cNtJbwrH6rZvsx0GUKQJKyWLeUbADs3u0,25190
+langroid-0.1.33.dist-info/RECORD,,