Skip to content

fix multimodal data and max_token arguments when using vllm tokenizer (text mode) #4

fix multimodal data and max_token arguments when using vllm tokenizer (text mode)

fix multimodal data and max_token arguments when using vllm tokenizer (text mode) #4

Triggered via pull request February 2, 2026 03:15
Status Failure
Total duration 22s
Artifacts –

copyright-checks.yml

on: pull_request
copyright-checks
19s
copyright-checks
Fit to window
Zoom out
Zoom in

Annotations

1 error and 2 warnings
copyright-checks
Docker pull failed with exit code 1
copyright-checks
Docker pull failed with exit code 1, back off 7.816 seconds before retry.
copyright-checks
Docker pull failed with exit code 1, back off 8.149 seconds before retry.