error with sentence_transformers load model
#18
by kosung - opened
ValueError: Unrecognized configuration class <class 'transformers_modules.gme-Qwen2-VL-2B-Instruct.modeling_gme_qwen2vl.GmeQwen2VLConfig'> for this kind of AutoModel: AutoModelForVision2Seq.
Model type should be one of BlipConfig, Blip2Config, ChameleonConfig, GitConfig, Idefics2Config, Idefics3Config, InstructBlipConfig, InstructBlipVideoConfig, Kosmos2Config, LlavaConfig, LlavaNextConfig, LlavaNextVideoConfig, LlavaOnevisionConfig, Mistral3Config, MllamaConfig, PaliGemmaConfig, Pix2StructConfig, Qwen2_5_VLConfig, Qwen2VLConfig, VideoLlavaConfig, VipLlavaConfig, VisionEncoderDecoderConfig.
AttributeError: 'Qwen2VLForConditionalGeneration' object has no attribute 'embed_tokens'
solved
transformers 4.52.4 : self.auto_model.base_model.language_model.embed_tokens(features["input_ids"])
transformers 4.51.3: self.auto_model.base_model.embed_tokens(features["input_ids"])
izhx changed discussion status to closed