fix qwen3_vl by changing sequence_length to seq_length

[shifangx]

What does this PR do ?

This pr is used to fix the issue HFDatasetConversationProvider.init() got an unexpected keyword argument 'sequence_length' for qwen3vl finetune.

After sequence_length -> seq_length #1023 was merged, the input argment of HFDatasetConversationProvider changed from sequence_length to seq_length.
So we need to adaptive the param name in qwen3vl.py#L246 while init HFDatasetConversationProvider.

Changelog

• Add specific line by line info of high level changes in this PR.

GitHub Actions CI

See the CI sectionin the Contributing doc for how to trigger the CI. A Nvidia developer will need to approve and trigger the CI for external contributors.

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation?
• Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
  • Reviewer: Does the PR have correct import guards for all optional libraries?

If you haven't finished some of the above items you can still open "Draft" PR.

Additional Information

• Related to HFDatasetConversationProvider.__init__() got an unexpected keyword argument 'sequence_length' for qwen3vl finetune #1510

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[yaoyu-33]

/ok to test 602d7c8

[shifangx]

/ok to test 14989ca