How to use estimator_config_dnn.yaml without GPU ?

[pan-cai]

If there is not GPU,

MODEL_CONFIG = {
                "loss": "mse",
                'input_dim': 158,
                'output_dim': 1,
                'lr': 0.002,
                'lr_decay': 0.96,
                'lr_decay_steps': 100,
                'optimizer': 'adam',
                'max_steps': 8000,
                'batch_size': 4096,
                # 'GPU': '0',
                }

     73 def _check_driver():
     74     if not hasattr(torch._C, '_cuda_isDriverSufficient'):
---> 75         raise AssertionError("Torch not compiled with CUDA enabled")
     76     if not torch._C._cuda_isDriverSufficient():
     77         if torch._C._cuda_getDriverVersion() == 0:

AssertionError: Torch not compiled with CUDA enabled

---

MODEL_CONFIG = {
                "loss": "mse",
                'input_dim': 158,
                'output_dim': 1,
                'lr': 0.002,
                'lr_decay': 0.96,
                'lr_decay_steps': 100,
                'optimizer': 'adam',
                'max_steps': 8000,
                'batch_size': 4096,
                'GPU': '0',
                }

     73 def _check_driver():
     74     if not hasattr(torch._C, '_cuda_isDriverSufficient'):
---> 75         raise AssertionError("Torch not compiled with CUDA enabled")
     76     if not torch._C._cuda_isDriverSufficient():
     77         if torch._C._cuda_getDriverVersion() == 0:

AssertionError: Torch not compiled with CUDA enabled

[bxdd]

Hi, thanks for your feedback. So far you can't run the baseline model without GPU. If your computer has multiple GPUs, the parameter "GPU" indicates which GPU to use instead of whether to use GPU. Please make sure that Cuda is installed on your computer before running the baseline model.

In the near future, we may make this example model compatible with CPU. Also, you can integrate your custom model into Qlib, and your PR will be welcome.

[bxdd]

If you want to integrate custom models into Qlib, please refer to the Integration document (https://qlib.readthedocs.io/en/latest/start/integration.html) for more details.

[bxdd]

Hi, We just now update the DNN model on PR #49. If Cuda is not installed on your computer, then Qlib will automatically use Cpu to train the model.

Looking forward to your further feedback.

[pan-cai]

Wow, thank you so much