sunhaohai

icon company Huazhong University of science and technology icon location Wuhan, China icon location Machine Learning; Deep Learning; NLP

Results 2 issues of sunhaohai

[ERROR] AttributeErrorAttributeErrorAttributeError: 'NoneType' object has no attribute 'get_tensor_model_parallel_group'

2 comment

https://github.com/microsoft/DeepSpeed/blob/d10b8ca011b18eba3a6ca56f4208a732d7fbb744/deepspeed/runtime/hybrid_engine.py#LL321C16-L321C37 Should the judgment condition be "self.mpu is None"?

[BUG] The text generated by the hybrid engine does not meet expectations, the model is GPTNeoX

2 comment

huggingface model config: { "activation_function": "gelu", "architectures": [ "GPTNeoXForCausalLM" ], "bos_token_id": 0, "eos_token_id": 2, "hidden_act": "gelu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 6144, "layer_norm_eps": 1e-05, "max_position_embeddings": 4096, "model_type": "gpt_neox", "num_attention_heads": 12,...

bug
deepspeed-chat