Closed houliangxue closed 6 months ago
self.model = AutoModelForCausalLM.from_pretrained(model_path,torch_dtype="auto",device_map="auto",trust_remote_code=True, bf16=True).eval() self.tokenizer = AutoTokenizer.from_pretrained(model_path) self.generation_Config = GenerationConfig.from_pretrained(model_path, trust_remote_code=True)
无
No response
4090显卡,transformer 4.37 ,使用bf16=True
TypeError: init() got an unexpected keyword argument 'bf16'
Don't mix the use of Qwen and Qwen1.5 code examples. For information on Qwen1.5, please refer to https://github.com/Qwen/Qwen1.5 or https://qwen.readthedocs.io/en/latest/.
是否已有关于该错误的issue或讨论? | Is there an existing issue / discussion for this?
该问题是否在FAQ中有解答? | Is there an existing answer for this in FAQ?
当前行为 | Current Behavior
self.model = AutoModelForCausalLM.from_pretrained(model_path,torch_dtype="auto",device_map="auto",trust_remote_code=True, bf16=True).eval() self.tokenizer = AutoTokenizer.from_pretrained(model_path) self.generation_Config = GenerationConfig.from_pretrained(model_path, trust_remote_code=True)
期望行为 | Expected Behavior
无
复现方法 | Steps To Reproduce
No response
运行环境 | Environment
备注 | Anything else?
TypeError: init() got an unexpected keyword argument 'bf16'