Use PEFT or Full-parameter to finetune 350+ LLMs or 100+ MLLMs. (LLM: Qwen2.5, Llama3.2, GLM4, Internlm2.5, Yi1.5, Mistral, Baichuan2, DeepSeek, Gemma2, ...; MLLM: Qwen2-VL, Qwen2-Audio, Llama3.2-Vision, Llava, InternVL2, MiniCPM-V-2.6, GLM4v, Xcomposer2.5, Yi-VL, DeepSeek-VL, Phi3.5-Vision, ...)
4.01k
stars
355
forks
source link
rope_scaling["type"] error in config.json of qwen2-vl merge lora exported file #2340
Open
wellhowtosay opened 3 days ago
Jintao-Huang
commented
2 days ago
Describe the bug SFT qwen2-vl model, then merge and export lora weight, value changed in exported config.json ["rope_scaling"]["type"] should be "mrope", not default { "_name_or_path": "/mnt/hwnas/baolujia/llms/Qwen2-VL-7B-Instruct", "architectures": [ "Qwen2VLForConditionalGeneration" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 3584, "image_token_id": 151655, "initializer_range": 0.02, "intermediate_size": 18944, "max_position_embeddings": 32768, "max_window_layers": 28, "model_type": "qwen2_vl", "num_attention_heads": 28, "num_hidden_layers": 28, "num_key_value_heads": 4, "rms_norm_eps": 1e-06, "rope_scaling": { "mrope_section": [ 16, 24, 24 ], "rope_type": "default", "type": "default" }, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": false, "torch_dtype": "float16", "transformers_version": "4.46.0.dev0", "use_cache": true, "use_sliding_window": false, "video_token_id": 151656, "vision_config": { "in_chans": 3, "model_type": "qwen2_vl", "spatial_patch_size": 14 }, "vision_end_token_id": 151653, "vision_start_token_id": 151652, "vision_token_id": 151654, "vocab_size": 152064 }