lmingze

### Describe the issue this is my training file. #!/bin/bash deepspeed --include localhost:3 --master_port=29510 /data10/LLaVA-main/llava/train/train_mem.py \ --lora_enable True \ --lora_r 64 \ --lora_alpha 256 \ --mm_projector_lr 2e-5 \ --deepspeed /data10//LLaVA-main/scripts/zero2.json...

### Describe the issue #!/bin/bash this is my .sh file. deepspeed --include localhost:3 --master_port=29510 /data10/mingze/LLaVA-main/llava/train/train_mem.py \ --lora_enable True \ --lora_r 64 \ --lora_alpha 256 \ --mm_projector_lr 2e-5 \ --deepspeed /data10/mingze/LLaVA-main/scripts/zero2.json...

[Bug]: vllm部署qwen2.5_vl_72b之后，你们有出现，刚部署好之后调用一切正常3-5秒一条，然后使用一段时间，就越来越慢了的情况吗60s一条

5

### Your current environment 这是部署启动命令： CUDA_VISIBLE_DEVICES=2,34,5 vllm serve /mnt/cfs/ljc/ckpt/Qwen/Qwen2___5-VL-72B-Instruct --tensor-parallel-size 4 --gpu-memory-utilization 0.8 --port 20772 --limit-mm-per-prompt image=5 > qwen2.5_vl_72B_instruct_20772_new.log 2>&1 & ### 🐛 Describe the bug 这是部署启动命令： CUDA_VISIBLE_DEVICES=2,34,5 vllm serve...

[Bug]: 使用vllm部署qwen2.5-vl系列的模型，出现调用卡死的问题，怎么解决？刚刚部署调用正常，显卡占用50%，后来使用一段时间之后，调用一次请求时间会很长，显卡占用到了90%。日志报错如下，怎么解决？

1

### Your current environment 部署命令： CUDA_VISIBLE_DEVICES=2 vllm serve /home/li_mingze/models/Qwen/Qwen2___5-VL-7B-Instruct --tensor-parallel-size 1 --gpu-memory-utilization 0.8 --port 2003 --limit-mm-per-prompt image=5 > qwen2.5_vl_7B_instruct_2003.log 2>&1 & 日志报错： ERROR 03-03 02:48:16 client.py:300] RuntimeError('Engine process (pid 2218951)...

[Bug]: 为什么在部署qwen2.5-vl-32b-instruct的时候，部署过程被卡死不动了

5

### Your current environment 我的环境environment： Package Version --------------------------------- ------------- accelerate 1.5.2 addict 2.4.0 aiohappyeyeballs 2.6.1 aiohttp 3.11.14 aiosignal 1.3.2 airportsdata 20250224 annotated-types 0.7.0 anyio 4.9.0 astor 0.8.1 attrs 25.3.0 blake3...

[Bug] 使用lmdeploy部署internvl-38b-mpo-awq，在dify上使用接口的时候报错，是什么原因导致？

1

### Checklist - [ ] 1. I have searched related issues but cannot get the expected help. - [ ] 2. The bug has not been fixed in the latest...

使用verl进行多模态RL训练的时候（图片存储在本地服务器），生成训练和测试数据的时候，这个"images"字段中应该填入什么？求大神解答一下哦

1

data = { "data_source": data_source, "prompt": [ { "role": "user", "content": prompt, } ], "images": images, "ability": "math", "reward_model": {"style": "rule", "ground_truth": answer}, "extra_info": { "split": split, "index": idx, "answer":...

[Bug] 使用v0.10.1版本部署internvl2.5-38b-mpo-awq报错，如何解决？

1

### Checklist - [ ] 1. I have searched related issues but cannot get the expected help. - [ ] 2. The bug has not been fixed in the latest...

[Bug] 使用lmdeploy部署模型internvl2.5/3系列模型，部署在24小时之内会断掉怎么回事？

2

### Checklist - [ ] 1. I have searched related issues but cannot get the expected help. - [ ] 2. The bug has not been fixed in the latest...