yanguangqing

四卡3090失败

``` import os import platform import signal from transformers import AutoTokenizer, AutoModel import torch tokenizer = AutoTokenizer.from_pretrained("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000", trust_remote_code=True) model = AutoModel.from_pretrained("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000", trust_remote_code=True) prefix_state_dict = torch.load(os.path.join("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000", "pytorch_model.bin")) new_prefix_state_dict = {} for...

``` import os import platform import signal from transformers import AutoTokenizer, AutoModel, AutoConfig import torch tokenizer = AutoTokenizer.from_pretrained("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000", trust_remote_code=True) config = AutoConfig.from_pretrained("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000", trust_remote_code=True, pre_seq_len=128) model = AutoModel.from_pretrained("ptuning/output/adgen-chatglm-6b-pt-128-2e-2/checkpoint-1000",config=config, trust_remote_code=True) prefix_state_dict =...


``` import os import platform import signal from transformers import AutoTokenizer, AutoModel, AutoConfig import torch tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True) config = AutoConfig.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True, pre_seq_len=128) model = AutoModel.from_pretrained("THUDM/chatglm-6b",config=config, trust_remote_code=True) prefix_state_dict =...

> > > > > ``` > > > > import os > > import platform > > import signal > > from transformers import AutoTokenizer, AutoModel, AutoConfig > >...

`relation = [] for relation_one in relation_data: relation.append( {"source": int(node_id_mapping[relation_one[0]]), "destination": int(node_id_mapping[relation_one[2]])}) json.dump({ "label": "cause", "fields": "uid", "edge": relation }, open("relationship.json", "w", encoding="utf-8"), ensure_ascii=False, indent=4)`

```relation = [] for relation_one in relation_data: relation.append( {"source": int(node_id_mapping[relation_one[0]]), "destination": int(node_id_mapping[relation_one[2]])}) json.dump({ "label": "cause", "fields": "uid", "edge": relation }, open("relationship.json", "w", encoding="utf-8"), ensure_ascii=False, indent=4)```

数据大概长这个样子

同样遇到了这个问题，用的deepspeed的finetune的脚本。生成时候无法返回任何信息。