Skip to content

Bugs when use vllm with qwenvl2.5 #68

@zhangzef

Description

@zhangzef
using:  <class 'trainer.vllm_grpo_trainer.Qwen2VLGRPOVLLMTrainer'>                                                                                                                     
[rank0]: Traceback (most recent call last):                                                                                                                                            
[rank0]:   File "/root/paddlejob/workspace/output/zzf/codes/R1-V/src/open-r1-multimodal/./grpo.py", line 228, in <module>                                                              
[rank0]:     main(script_args, training_args, model_args)                                                                                                                              
[rank0]:   File "/root/paddlejob/workspace/output/zzf/codes/R1-V/src/open-r1-multimodal/./grpo.py", line 200, in main                                                                  
[rank0]:     trainer = trainer_cls(                                                                                                                                                    
[rank0]:               ^^^^^^^^^^^^                                                                                                                                                    
[rank0]:   File "/root/paddlejob/workspace/output/zzf/codes/R1-V/src/open-r1-multimodal/trainer/vllm_grpo_trainer.py", line 184, in __init__                                           
[rank0]:     model = AutoModelForCausalLM.from_pretrained(model, **model_init_kwargs)                                                                                                  
[rank0]:             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^                                                                                                  
[rank0]:   File "/root/miniconda3/envs/r1v/lib/python3.11/site-packages/transformers/models/auto/auto_factory.py", line 567, in from_pretrained                                        
[rank0]:     raise ValueError(                                                                                                                                                         
[rank0]: ValueError: Unrecognized configuration class <class 'transformers.models.qwen2_5_vl.configuration_qwen2_5_vl.Qwen2_5_VLConfig'> for this kind of AutoModel: AutoModelForCausal
LM.                                                                                                                                                                                    
[rank0]: Model type should be one of AriaTextConfig, BambaConfig, BartConfig, BertConfig, BertGenerationConfig, BigBirdConfig, BigBirdPegasusConfig, BioGptConfig, BlenderbotConfig, Bl
enderbotSmallConfig, BloomConfig, CamembertConfig, LlamaConfig, CodeGenConfig, CohereConfig, Cohere2Config, CpmAntConfig, CTRLConfig, Data2VecTextConfig, DbrxConfig, DiffLlamaConfig, 
ElectraConfig, Emu3Config, ErnieConfig, FalconConfig, FalconMambaConfig, FuyuConfig, GemmaConfig, Gemma2Config, GitConfig, GlmConfig, GotOcr2Config, GPT2Config, GPT2Config, GPTBigCode
Config, GPTNeoConfig, GPTNeoXConfig, GPTNeoXJapaneseConfig, GPTJConfig, GraniteConfig, GraniteMoeConfig, HeliumConfig, JambaConfig, JetMoeConfig, LlamaConfig, MambaConfig, Mamba2Confi
g, MarianConfig, MBartConfig, MegaConfig, MegatronBertConfig, MistralConfig, MixtralConfig, MllamaConfig, MoshiConfig, MptConfig, MusicgenConfig, MusicgenMelodyConfig, MvpConfig, Nemo
tronConfig, OlmoConfig, Olmo2Config, OlmoeConfig, OpenLlamaConfig, OpenAIGPTConfig, OPTConfig, PegasusConfig, PersimmonConfig, PhiConfig, Phi3Config, PhimoeConfig, PLBartConfig, Proph
etNetConfig, QDQBertConfig, Qwen2Config, Qwen2MoeConfig, RecurrentGemmaConfig, ReformerConfig, RemBertConfig, RobertaConfig, RobertaPreLayerNormConfig, RoCBertConfig, RoFormerConfig, 
RwkvConfig, Speech2Text2Config, StableLmConfig, Starcoder2Config, TransfoXLConfig, TrOCRConfig, WhisperConfig, XGLMConfig, XLMConfig, XLMProphetNetConfig, XLMRobertaConfig, XLMRoberta
XLConfig, XLNetConfig, XmodConfig, ZambaConfig, Zamba2Config.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions