deepseek-ai / DeepSeek-V2

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
MIT License
3.43k stars 138 forks source link

DeepSeek-V2-Lite-Chat模型启动依赖问题 #78

Open Malowking opened 1 month ago

Malowking commented 1 month ago

DeepSeek-V2-Lite-Chat模型必须使用flash_attn这个包么,如果不必要该怎么指定禁用flash_attn 这个包

itaowei commented 2 weeks ago

I have the same question. May I ask which version of flash_attn is required? I installed the latest version (2.6.3) but it will make the error below:


Traceback (most recent call last):
  File "test.py", line 335, in <module>
    model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True, device_map="sequential", 
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/workspace/anaconda3/envs/deepseek/lib/python3.12/site-packages/transformers/models/auto/auto_factory.py", line 551, in from_pretrained
    model_class = get_class_from_dynamic_module(
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/workspace/anaconda3/envs/deepseek/lib/python3.12/site-packages/transformers/dynamic_module_utils.py", line 502, in get_class_from_dynamic_module
    final_module = get_cached_module_file(
                   ^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/workspace/anaconda3/envs/deepseek/lib/python3.12/site-packages/transformers/dynamic_module_utils.py", line 327, in get_cached_module_file
    modules_needed = check_imports(resolved_module_file)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/data/workspace/anaconda3/envs/deepseek/lib/python3.12/site-packages/transformers/dynamic_module_utils.py", line 182, in check_imports
    raise ImportError(
ImportError: This modeling file requires the following packages that were not found in your environment: flash_attn. Run `pip install flash_attn`