Option field llm is not necessary for loadEvaluator with type embedding_distance

[trend-ouki-wang]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain.js documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain.js rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

const {loadEvaluator} = require("langchain/evaluation");
const {AzureOpenAI, AzureChatOpenAI, AzureOpenAIEmbeddings} = require("@langchain/openai");
...
const chain = await loadEvaluator("embedding_distance", {
        embedding: azure_embedding,
        distanceMetric: "euclidean",
    });
...

Error Message and Stack Trace (if applicable)

Error: OpenAI or Azure OpenAI API key or Token Provider not found
at new ChatOpenAI (C:\Users\ouki_wang\Desktop\Team\sandbox-companion\node_modules@langchain\openai\dist\chat_models.cjs:1032:19)
at loadEvaluator (C:\Users\ouki_wang\Desktop\Team\sandbox-companion\node_modules\langchain\dist\evaluation\loader.cjs:21:9)
at main (C:\Users\ouki_wang\Desktop\Team\sandbox-companion\05-sandbox.js:78:29)
at Object. (C:\Users\ouki_wang\Desktop\Team\sandbox-companion\05-sandbox.js:131:1)
at Module._compile (node:internal/modules/cjs/loader:1364:14)
at Module._extensions..js (node:internal/modules/cjs/loader:1422:10)
at Module.load (node:internal/modules/cjs/loader:1203:32)
at Module._load (node:internal/modules/cjs/loader:1019:12)
at Function.executeUserEntryPoint [as runMain] (node:internal/modules/run_main:128:12)
at node:internal/main/run_main_module:28:49

Description

we have tried to call loadEvaluator to get euclidean distance with embedding model.
However get error message without llm missing in option fields of loadEvaluator.
I think llm field is not necessary when call with type embedding_distance.

System Info

"@langchain/openai": "^0.3.11",
"langchain": "^0.3.3"

node version: v18.20.4

[dosubot]

The error you're encountering is due to the OpenAIEmbeddings class requiring an API key or token provider, even if the llm field is not necessary for the embedding_distance evaluator. To resolve this, ensure that you provide either an OpenAI API key, an Azure OpenAI API key, or an Azure AD token provider. This requirement is part of the OpenAIEmbeddings class, which throws an error if none of these credentials are provided ^[1][2].

To continue talking to Dosu, mention @dosu.