[Feature] The 45VL supports prompt_token_ids + messages input.

[kxz2002]

Motivation

When triggering multiple rounds with RLC, we aim to use the prompt_ids and completion_ids from the previous round as input for the current round without requiring concatenation. Additionally, we wish to pass image information via the message field.

Modifications

The process_request_dict function in ernie4_5_vl_processor now supports reading the prompt_token_ids field from the request as input. It also supports reading multimodal information from the messages field in this scenario.

Usage or Command

No change in manual command

Accuracy Tests

No need

Checklist

• Add at least a tag in the PR title.
  • Tag list: [[FDConfig],[APIServer],[Engine], [Scheduler], [PD Disaggregation], [Executor], [Graph Optimization], [Speculative Decoding], [RL], [Models], [Quantization], [Loader], [OP], [KVCache], [DataProcessor], [BugFix], [Docs], [CI], [Optimization], [Feature], [Benchmark], [Others], [XPU], [HPU], [GCU], [DCU], [Iluvatar], [Metax]]
  • You can add new tags based on the PR content, but the semantics must be clear.
• Format your code, run pre-commit before commit.
• Add unit tests. Please write the reason in this PR if no unit tests.
• Provide accuracy results.
• If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

[paddle-bot]

Thanks for your contribution!