search

CODA-Team / AnalogGym

BSD 3-Clause "New" or "Revised" License
19 stars 1 forks source link

Missing .whl file #1

Closed ShikaiWang-Sky closed 1 month ago

ShikaiWang-Sky commented 1 month ago

Hi guys, I tried to setup environment with docker; however, there seems missing a .whl file: torch-1.13.1+cpu-cp310-cp310-linux_x86_64.whl.

ShikaiWang-Sky commented 1 month ago

Also, I tried setup conda environment and run main_AMP.py, the error shows as following:

Register the environment cpu ERROR Simulation errors, no .OP simulation results. You need to run <_random_op_sims> to generate mean and std for transistor .OP parameters Traceback (most recent call last): File ".../main_AMP.py", line 82, in agent.train(num_steps) File ".../ddpg.py", line 220, in train state, info = self.env.reset() File ".../gymnasium/wrappers/time_limit.py", line 75, in reset return self.env.reset(kwargs) File ".../gymnasium/wrappers/order_enforcing.py", line 61, in reset return self.env.reset(kwargs) File ".../gymnasium/wrappers/env_checker.py", line 57, in reset return env_reset_passive_checker(self.env, kwargs) File ".../gymnasium/utils/passive_env_checker.py", line 186, in env_reset_passive_checker result = env.reset(kwargs) File ".../AMP_NMCF.py", line 123, in reset observation = self._get_obs() File ".../AMP_NMCF.py", line 211, in _get_obs self.OP_M0 = self.op_results['M0'] TypeError: 'NoneType' object is not subscriptable

krzhu commented 1 month ago

Please shows your running steps and give a reproducible case

Fican1 commented 1 month ago

Also, I tried setup conda environment and run main_AMP.py, the error shows as following:

Register the environment cpu ERROR Simulation errors, no .OP simulation results. You need to run <_random_op_sims> to generate mean and std for transistor .OP parameters Traceback (most recent call last): File ".../main_AMP.py", line 82, in agent.train(num_steps) File ".../ddpg.py", line 220, in train state, info = self.env.reset() File ".../gymnasium/wrappers/time_limit.py", line 75, in reset return self.env.reset(kwargs) File ".../gymnasium/wrappers/order_enforcing.py", line 61, in reset return self.env.reset(kwargs) File ".../gymnasium/wrappers/env_checker.py", line 57, in reset return env_reset_passive_checker(self.env, kwargs) File ".../gymnasium/utils/passive_env_checker.py", line 186, in env_reset_passive_checker result = env.reset(kwargs) File ".../AMP_NMCF.py", line 123, in reset observation = self._get_obs() File ".../AMP_NMCF.py", line 211, in _get_obs self.OP_M0 = self.op_results['M0'] TypeError: 'NoneType' object is not subscriptable

To normalize the transistor attributes used in the observation matrix, you'll need the mean and standard deviation. These values can be obtained by running initial experiments with the class method _init_random_sim.

For demonstration, we have already provided this data in the file AMP_NMCF_op_mean_std.json. When switching to a different circuit, you will need to generate this file yourself. To do so, simply set the parameter 'run_initial=True' at the indicated location in main_AMP.py.

Fican1 commented 1 month ago

Hi guys, I tried to setup environment with docker; however, there seems missing a .whl file: torch-1.13.1+cpu-cp310-cp310-linux_x86_64.whl.

Hi there! The torch-1.13.1+cpu file is quite large, so you'll need to download it manually from this link: PyTorch 1.13.1 Files.