Joseph Yan

Add support for the gemma2 language model in multimodal fine-tuning.

1

Add support for the gemma2 language model during the process of multimodal model fusion. The fine-tuning process has been successfully completed under 8*H100, hoping to make a contribution.

The model needs to be deeply copied.

Llama 3.2-11B-vision fully fine-tuned model file question

2

During the use of LoRA fine-tuning, everything was normal, but the following issue arose during full-scale fine-tuning. I use the following script for full fine-tuning : ```shell #!/bin/bash NNODES=1 NPROC_PER_NODE=4...

Issues Converting lit_model.pth to Huggingface Format Using convert_from_litgpt

7

Hello, I have been using **litgpt** to pretrain a model, which produces a `lit_model.pth` file. This model functions correctly when loaded with `LLM.load()` for inference. However, when I attempt to...

[rollout, vllm] fix: add compatibility shim for vllm process_weights_after_loading

1

### What does this PR do? This PR fixes a compatibility issue between VERL’s vLLM rollout and certain vLLM versions (e.g. `vllm==0.8.5.post1`) where importing: ```python from vllm.model_executor.model_loader.utils import process_weights_after_loading ````...