# Issue with llm.bind\_tools() **URL:** \ **Category:** Discussion\ **Tags:** dev\ **Created:** [November 23, 2024, 8:19am UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665 "2024-11-23T08:19:39Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![maybinmichael91](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/maybinmichael91/32/1195_2.png) [@maybinmichael91](https://community.sambanova.ai/u/maybinmichael91)\ **Post date:** [November 23, 2024, 8:19am UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/1 "2024-11-23T08:19:39Z")
I am facing two errors with **llm.bind\_tools()** in two scenarios. Please guide what is the issue with my code. **My code:** from langchain\_openai import ChatOpenAI import os from langchain\_community.tools import ArxivQueryRun from langchain\_community.tools.tavily\_search import TavilySearchResults from langchain\_community.utilities import ArxivAPIWrapper from langchain\_community.tools import ArxivQueryRun from langchain import hub from langchain\_core.messages import HumanMessage, SystemMessage from langchain\_community.chat\_models.sambanova import ChatSambaNovaCloud os.environ[“SAMBANOVA\_API\_KEY”] = ‘66ea09\*\*\*\*’ os.environ[“TAVILY\_API\_KEY”] = ‘tvly-B\*\*\*\*’ os.environ[“LANGCHAIN\_API\_KEY”] = ‘lsv2\_pt\_22\*\*\*\*\*’ llm1 = ChatOpenAI( base\_url=“[https://api.sambanova.ai/v1/](https://api.sambanova.ai/v1/)”, api\_key=os.environ[“SAMBANOVA\_API\_KEY”], streaming=False, model=“Meta-Llama-3.1-70B-Instruct”, temperature=0.1, ) llm2 = ChatSambaNovaCloud( sambanova\_url = “[https://api.sambanova.ai/v1/chat/completions](https://api.sambanova.ai/v1/chat/completions)”, sambanova\_api\_key = os.environ[“SAMBANOVA\_API\_KEY”], model = “Meta-Llama-3.1-70B-Instruct”, max\_tokens = 1000, temperature = 0, ) arxiv\_wrapper = ArxivAPIWrapper(top\_k\_results=1, doc\_content\_chars\_max=200) arxiv\_tool = ArxivQueryRun(arxiv\_wrapper=arxiv\_wrapper) tavily\_tool =TavilySearchResults(max\_results=5, ) tools = [ tavily\_tool, arxiv\_tool ] llm1\_with\_tools = llm1.bind\_tools(tools) llm1\_with\_tools.invoke([HumanMessage(content=“what is the current weather in california?”)]) # Error 1 is with this line llm2\_with\_tools = llm2.bind\_tools(tools) # Error 2 is with this line **Error 1:** ## D:\OneDrive\mydirectory\venv\Lib\site-packages\pydantic\main.py:426: UserWarning: Pydantic serializer warnings: **Expected `str` but got `dict` with value `{'query': 'Current weather in California'}` - serialized value may not be as expected** return self. **pydantic\_serializer**.to\_python( ValidationError Traceback (most recent call last) Cell In[71], line 1 ----\> 1 llm1\_with\_tools.invoke([HumanMessage(content=“what is the current weather in california?”)]) File D:\OneDrive\mydirectoryvenv\Lib\site-packages\langchain\_core\runnables\base.py:5354, in RunnableBindingBase.invoke(self, input, config, \*\*kwargs) 5348 def invoke( 5349 self, 5350 input: Input, 5351 config: Optional[RunnableConfig] = None, 5352 \*\*kwargs: Optional[Any], 5353 ) → Output: → 5354 return self.bound.invoke( 5355 input, 5356 self.\_merge\_configs(config), 5357 \*\*{\*\*self.kwargs, \*\*kwargs}, 5358 ) File D:\OneDrive\Learning Room\GenAI Projects\Sambanova\iChat\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:286, in BaseChatModel.invoke(self, input, config, stop, \*\*kwargs) 275 def invoke( 276 self, 277 input: LanguageModelInput, (…) 281 \*\*kwargs: Any, 282 ) → BaseMessage: 283 config = ensure\_config(config) 284 return cast( 285 ChatGeneration, → 286 self.generate\_prompt( 287 [self.\_convert\_input(input)], 288 stop=stop, 289 callbacks=config.get(“callbacks”), 290 tags=config.get(“tags”), 291 metadata=config.get(“metadata”), 292 run\_name=config.get(“run\_name”), 293 run\_id=config.pop(“run\_id”, None), 294 \*\*kwargs, 295 ).generations[0][0], 296 ).message File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:786, in BaseChatModel.generate\_prompt(self, prompts, stop, callbacks, \*\*kwargs) 778 def generate\_prompt( 779 self, 780 prompts: list[PromptValue], (…) 783 \*\*kwargs: Any, 784 ) → LLMResult: 785 prompt\_messages = [p.to\_messages() for p in prompts] → 786 return self.generate(prompt\_messages, stop=stop, callbacks=callbacks, \*\*kwargs) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:643, in BaseChatModel.generate(self, messages, stop, callbacks, tags, metadata, run\_name, run\_id, \*\*kwargs) 641 if run\_managers: 642 run\_managers[i].on\_llm\_error(e, response=LLMResult(generations=)) → 643 raise e 644 flattened\_outputs = [ 645 LLMResult(generations=[res.generations], llm\_output=res.llm\_output) # type: ignore[list-item] 646 for res in results 647 ] 648 llm\_output = self.\_combine\_llm\_outputs([res.llm\_output for res in results]) File D:\OneDrive\mydirectory\iChat\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:633, in BaseChatModel.generate(self, messages, stop, callbacks, tags, metadata, run\_name, run\_id, \*\*kwargs) 630 for i, m in enumerate(messages): 631 try: 632 results.append( → 633 self.\_generate\_with\_cache( 634 m, 635 stop=stop, 636 run\_manager=run\_managers[i] if run\_managers else None, 637 \*\*kwargs, 638 ) 639 ) 640 except BaseException as e: 641 if run\_managers: File D:\OneDrive\Learning Room\GenAI Projects\Sambanova\iChat\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:851, in BaseChatModel.\_generate\_with\_cache(self, messages, stop, run\_manager, \*\*kwargs) 849 else: 850 if inspect.signature(self.\_generate).parameters.get(“run\_manager”): → 851 result = self.\_generate( 852 messages, stop=stop, run\_manager=run\_manager, \*\*kwargs 853 ) 854 else: 855 result = self.\_generate(messages, stop=stop, \*\*kwargs) File D:\OneDrive\Learning Room\GenAI Projects\Sambanova\iChat\venv\Lib\site-packages\langchain\_openai\chat\_models\base.py:690, in BaseChatOpenAI.\_generate(self, messages, stop, run\_manager, \*\*kwargs) 688 else: 689 response = self.client.create(\*\*payload) → 690 return self.\_create\_chat\_result(response, generation\_info) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_openai\chat\_models\base.py:727, in BaseChatOpenAI.\_create\_chat\_result(self, response, generation\_info) 725 token\_usage = response\_dict.get(“usage”) 726 for res in response\_dict[“choices”]: → 727 message = \_convert\_dict\_to\_message(res[“message”]) 728 if token\_usage and isinstance(message, AIMessage): 729 message.usage\_metadata = \_create\_usage\_metadata(token\_usage) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_openai\chat\_models\base.py:134, in \_convert\_dict\_to\_message(\_dict) 132 if audio := _dict.get(“audio”): 133 additional\_kwargs[“audio”] = audio → 134 return AIMessage( 135 content=content, 136 additional\_kwargs=additional\_kwargs, 137 name=name, 138 id=id_, 139 tool\_calls=tool\_calls, 140 invalid\_tool\_calls=invalid\_tool\_calls, 141 ) 142 elif role == “system”: 143 return SystemMessage(content=_dict.get(“content”, “”), name=name, id=id_) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_core\messages\ai.py:179, in AIMessage. **init** (self, content, \*\*kwargs) 170 def **init** ( 171 self, content: Union[str, list[Union[str, dict]]], \*\*kwargs: Any 172 ) → None: 173 “”“Pass in content as positional arg. 174 175 Args: 176 content: The content of the message. 177 kwargs: Additional arguments to pass to the parent class. 178 “”” → 179 super(). **init** (content=content, \*\*kwargs) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_core\messages\base.py:76, in BaseMessage. **init** (self, content, \*\*kwargs) 67 def **init** ( 68 self, content: Union[str, list[Union[str, dict]]], \*\*kwargs: Any 69 ) → None: 70 “”“Pass in content as positional arg. 71 72 Args: 73 content: The string contents of the message. 74 kwargs: Additional fields to pass to the 75 “”” —\> 76 super(). **init** (content=content, \*\*kwargs) File D:\OneDrive\mydirectory\venv\Lib\site-packages\langchain\_core\load\serializable.py:125, in Serializable. **init** (self, \*args, \*\*kwargs) 123 def **init** (self, \*args: Any, \*\*kwargs: Any) → None: 124 “”“”“” → 125 super(). **init** (\*args, \*\*kwargs) File D:\OneDrive\mydirectory\venv\Lib\site-packages\pydantic\main.py:214, in BaseModel. **init** (self, \*\*data) 212 # ` __tracebackhide__ ` tells pytest and some other tools to omit this function from tracebacks 213 **tracebackhide** = True → 214 validated\_self = self. **pydantic\_validator**.validate\_python(data, self\_instance=self) 215 if self is not validated\_self: 216 warnings.warn( 217 ‘A custom validator is returning a value other than `self`.\n’ 218 “Returning anything other than `self` from a top level model validator isn’t supported when validating via ` __init__ `.\n” 219 ‘See the `model_validator` docs ([Validators - Pydantic Validation](https://docs.pydantic.dev/latest/concepts/validators/#model-validators)) for more details.’, 220 stacklevel=2, 221 ) ValidationError: 1 validation error for AIMessage invalid\_tool\_calls.0.args Input should be a valid string [type=string\_type, input\_value={‘query’: ‘Current weather in California’}, input\_type=dict] For further information visit [Redirecting...](https://errors.pydantic.dev/2.10/v/string_type) **Error 2:** NotImplementedError Traceback (most recent call last) Cell In[74], line 1 ----\> 1 llm2\_with\_tools = llm2.bind\_tools(tools) 2 llm2\_with\_tools File D:\OneDrive\mydiretcory\venv\Lib\site-packages\langchain\_core\language\_models\chat\_models.py:1115, in BaseChatModel.bind\_tools(self, tools, \*\*kwargs) 1108 def bind\_tools( 1109 self, 1110 tools: Sequence[ (…) 1113 \*\*kwargs: Any, 1114 ) → Runnable[LanguageModelInput, BaseMessage]: → 1115 raise NotImplementedError NotImplementedError: ---
**Author:** ![hello1](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/hello1/32/2812_2.png) [@hello1](https://community.sambanova.ai/u/hello1)\ **Post date:** [November 24, 2024, 4:44pm UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/2 "2024-11-24T16:44:23Z")
In most cases, tools in wrappers like Langchain are not fully compatible. I recommend channeling these tools using the Sambanova tool call example. It might look more difficult to implement, but it is less framework-dependent. You can see the example on the following link: [https://community.sambanova.ai/t/function-calling-and-json-mode-in-sambanova-cloud/540](https://community.sambanova.ai/t/function-calling-and-json-mode-in-sambanova-cloud/540) ---
**Author:** ![Coby](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/coby/32/17_2.png) [@Coby](https://community.sambanova.ai/u/Coby)\ **Post date:** [November 24, 2024, 10:05pm UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/3 "2024-11-24T22:05:52Z")
@maybinmichael91 welcome to the community and I am sorry you are having issues. Did the thread suggested by @hello1 help? ---
**Author:** ![Coby](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/coby/32/17_2.png) [@Coby](https://community.sambanova.ai/u/Coby)\ **Post date:** [November 24, 2024, 10:06pm UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/4 "2024-11-24T22:06:23Z")
@hello1 Thank you for lending a hand. ---
**Author:** ![maybinmichael91](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/maybinmichael91/32/1195_2.png) [@maybinmichael91](https://community.sambanova.ai/u/maybinmichael91)\ **Post date:** [November 28, 2024, 1:34pm UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/5 "2024-11-28T13:34:46Z")
Thank you @hello1 . I willtry the recommended approach. ---
**Author:** ![maybinmichael91](https://sea1.discourse-cdn.com/sambanova/user_avatar/community.sambanova.ai/maybinmichael91/32/1195_2.png) [@maybinmichael91](https://community.sambanova.ai/u/maybinmichael91)\ **Post date:** [November 28, 2024, 1:36pm UTC](https://community.sambanova.ai/t/issue-with-llm-bind-tools/665/6 "2024-11-28T13:36:14Z")
Hi @Coby thank you, I am yet to try the approach.