XrayGLM

XrayGLM copied to clipboard

Traceback (most recent call last): File "web_demo.py", line 154, in main(args) File "web_demo.py", line 90, in main model, model_args = AutoModel.from_pretrained( File "/root/miniconda3/lib/python3.8/site-packages/sat/model/base_model.py", line 275, in from_pretrained load_checkpoint(model, args, load_path=model_path,...

silencewa

请问多轮对话数据格式如何设计？

1

请问是如何保证训练的垂直行业大模型仍然具有较强的广泛知识问答能力（多轮对话）的 如下图所示的效果：  数据格式如下图：   prompt问题如何设计？ label答案描述内容也要具备丰富性吗？ 如何设计多轮对话？ 谢谢

elesun2018

[ERROR] [launch.py:322:sigkill_handler]

1165787081

有没有尝试使用CogVLM进行训练

3

[CogVLM](https://github.com/THUDM/CogVLM)

yuyo15117

'Namespace' object has no attribute 'pad_token_id' 请问这个问题怎么解决呢！

2

Neuzly1

 参考https://github.com/THUDM/VisualGLM-6B/issues/218的建议，仍然如标题报错，请问大佬有什么解决方案吗？

StefanieWang123

测试报错，args中无 'pad_token_id' AttributeError: 'Namespace' object has no attribute 'pad_token_id'

4


shituo123456

大佬您好，我在用咱们提供的openi数据集对VisualGLM进行微调之后，检测模型的推理能力的时候，出现以下情况  是不是过拟合太严重了，以下是我的微调参数 ``` #! /bin/bash NUM_WORKERS=1 NUM_GPUS_PER_WORKER=4 MP_SIZE=1 script_path=$(realpath $0) script_dir=$(dirname $script_path) main_dir=$(dirname $script_dir) MODEL_TYPE="visualglm-6b" MODEL_ARGS="--max_source_length 64 \ --max_target_length 256 \ --lora_rank 10 \ --pre_seq_len 4" # OPTIONS_SAT="SAT_HOME=$1" #"SAT_HOME=/raid/dm/sat_models"...

20191864218

微调时出现TypeError: string indices must be integers错误

1

Traceback (most recent call last): File "/root/VisualGLM-6B/finetune_XrayGLM.py", line 194, in training_main(args, model_cls=model, forward_step_function=forward_step, create_dataset_function=create_dataset_function, collate_fn=data_collator) File "/root/miniconda3/lib/python3.10/site-packages/sat/training/deepspeed_training.py", line 67, in training_main train_data, val_data, test_data = make_loaders(args, hooks['create_dataset_function'], collate_fn=collate_fn) File "/root/miniconda3/lib/python3.10/site-packages/sat/data_utils/configure_data.py",...

20191864218

运行cli_demo.py "slow_conv2d_cpu" not implemented for 'Half'

2


shituo123456