niuzaisheng / ScreenAgent

ScreenAgent: A Computer Control Agent Driven by Visual Language Large Model (IJCAI-24)
https://arxiv.org/abs/2402.07945
Other
311 stars 29 forks source link

Windows Run error #30

Open 276585877 opened 2 months ago

276585877 commented 2 months ago

todo

276585877 commented 2 months ago

set RANK=0 && set LOCAL_WORLD_SIZE=1 && set LOCAL_RANK=0 && python ./cogagent_model_worker.py --host 0.0.0.0 --port 40000 --from_pretrained "./saved_models/ScreenAgent-2312" --bf16 --max_length 2048 [2024-09-04 10:04:16,616] [INFO] DeepSpeed/CUDA is not installed, fallback to Pytorch checkpointing. [2024-09-04 10:04:19,245] [WARNING] DeepSpeed Not Installed, you cannot import training_main from sat now. WARNING[XFORMERS]: xFormers can't load C++/CUDA extensions. xFormers was built for: PyTorch 2.4.0+cu121 with CUDA 1201 (you have 2.3.0+cu118) Python 3.10.11 (you have 3.10.11) Please reinstall xformers (see https://github.com/facebookresearch/xformers#installing-xformers) Memory-efficient attention, SwiGLU, sparse and more won't be available. Set XFORMERS_MORE_DETAILS=1 for more details Please 'pip install apex' The installed version of bitsandbytes was compiled without GPU support. 8-bit optimizers, 8-bit multiplication, and GPU quantization are unavailable. 2024-09-04 10:04:22 | INFO | model_worker | args: Namespace(host='0.0.0.0', port=40000, max_length=2048, top_p=0.4, top_k=1, temperature=0.8, chinese=False, version='chat', quant=None, stream_chat=False, from_pretrained='./saved_models/ScreenAgent-2312', local_tokenizer='lmsys/vicuna-7b-v1.5', fp16=False, bf16=True) [2024-09-04 10:04:22,783] [INFO] building FineTuneTrainCogAgentModel model ... 2024-09-04 10:04:22 | INFO | sat | building FineTuneTrainCogAgentModel model ... [2024-09-04 10:04:22,812] [INFO] [RANK 0] > initializing model parallel with size 1 2024-09-04 10:04:22 | INFO | sat | [RANK 0] > initializing model parallel with size 1 [2024-09-04 10:04:22,820] [INFO] [RANK 0] You are using model-only mode. For torch.distributed users or loading model parallel models, set environment variables RANK, WORLD_SIZE and LOCAL_RANK. 2024-09-04 10:04:22 | INFO | sat | [RANK 0] You are using model-only mode. For torch.distributed users or loading model parallel models, set environment variables RANK, WORLD_SIZE and LOCAL_RANK.

运行一会就退出了,符合预期么