Getting Error when calling openai/gpt-oss-20b from HuggingFaceEndpoint

#131
by DeepakParameswar - opened

from langchain_huggingface import HuggingFaceEndpoint
import os

llm = HuggingFaceEndpoint(
repo_id="openai/gpt-oss-20b",
provider="hf-inference", # use HF’s own inference API
max_new_tokens=256,
temperature=0.7,
huggingfacehub_api_token="",
)

response = llm.invoke("Explain quantum mechanics in simple terms.")
print(response)

When we are invoking this then we are getting error saying: Bad Request: The endpoint is paused, ask a maintainer to restart it

How we can use openai/gpt-oss-20b from HuggingFaceEndpoint, can anyone help on this !

Yes i am facing the same error, i think there is some error on the provider side

public async Task GenerateAsync(string prompt)
{
try
{
var textGen = _kernel.GetRequiredService();

     var response = await textGen.GetTextContentsAsync(prompt);

     return response.FirstOrDefault()?.Text ?? string.Empty;
 }
 catch (Exception ex)
 {
     _logger.LogError(ex, "Error generating text with HuggingFace model.");
     throw;
 }

}

If anyone knows, please tell me

Sign up or log in to comment