search

bark-simulator / bark-ml

Gym environments and agents for autonomous driving.
https://bark-simulator.github.io/
MIT License
92 stars 18 forks source link

Failed to run some projects #165

Open Netceor opened 2 years ago

Netceor commented 2 years ago

Sorry to bother you, for some reason many of my projects fail, such as fqf、iqn、qrdqn、tfa_counterfactual

I use code like bazel run //examples:Project to run these.The following are some error details, I do not know how to solve.

But among these items, I care most about item tfa_counterfactual. If you can solve my questions, I would be very grateful！

Traceback (most recent call last):
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/fqf.runfiles/bark_ml/examples/fqf.py", line 58, in <module>
    app.run(run_configuration)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/absl/app.py", line 308, in run
    _run_main(main, args)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/absl/app.py", line 254, in _run_main
    sys.exit(main(argv))
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/fqf.runfiles/bark_ml/examples/fqf.py", line 39, in run_configuration
    agent = FQFAgent(env=env, test_env=env, params = params)
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/fqf.runfiles/bark_ml/bark_ml/library_wrappers/lib_fqf_iqn_qrdqn/agent/fqf_agent.py", line 24, in __init__
    super(FQFAgent, self).__init__(*args, **kwargs)
TypeError: __init__() got an unexpected keyword argument 'test_env'

I try to modify the code in fqf.py. Change agent = FQFAgent(env=env, test_env=env, params = params) to agent = FQFAgent(env=env, params = params), but new error occurs.

AttributeError: attempted to get missing private attribute '_observer'

In base_agent.py, there has a function,but error will say that no attribute '_observer' and '_ml_behavior'

def reset_action_observer(self, env):
    self._observer = self._env._observer
    self._ml_behavior = self._env._ml_behavior
Traceback (most recent call last):
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/iqn.runfiles/bark_ml/examples/iqn.py", line 57, in <module>
    app.run(run_configuration)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/absl/app.py", line 308, in run
    _run_main(main, args)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/absl/app.py", line 254, in _run_main
    sys.exit(main(argv))
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/iqn.runfiles/bark_ml/examples/iqn.py", line 37, in run_configuration
    env = gym.make(FLAGS.env, params=params)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/gym/envs/registration.py", line 676, in make
    return registry.make(id, **kwargs)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/gym/envs/registration.py", line 490, in make
    versions = self.env_specs.versions(namespace, name)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/gym/envs/registration.py", line 220, in versions
    self._assert_name_exists(namespace, name)
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/gym/envs/registration.py", line 297, in _assert_name_exists
    raise error.NameNotFound(message)
gym.error.NameNotFound: Environment `highway` doesn't exist.

I add import bark_ml.environments.gym in iqn.py, but new error occurs.

RuntimeError: Unable to cast Python instance of type <class 'float'> to C++ type 'int'

This is the same mistake, but I don't know how to solove it.

RuntimeError: Unable to cast Python instance of type <class 'float'> to C++ type 'int'
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/tfa_counterfactual.runfiles/bark_ml/bark_ml/environments/counterfactual_runtime.py", line 254, in step
    replaced_agent=cf_key, num_virtual_world=i)
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/tfa_counterfactual.runfiles/bark_ml/bark_ml/environments/counterfactual_runtime.py", line 150, in SimulateWorld
    local_tracer.Trace(eval_state, **kwargs)
TypeError: Trace() got an unexpected keyword argument 'replaced_agent'

Similar to fqf, I changed the code in counterfactual_runtime.py

self.SimulateWorld(
        # cf_world[cf_key], local_tracer, N=self._cf_simulation_steps,replaced_agent=cf_key, num_virtual_world=i)
        cf_world[cf_key], local_tracer, N=self._cf_simulation_steps)

self.SimulateWorld(
      # gt_world, local_tracer, N=self._cf_simulation_steps,replaced_agent="None", num_virtual_world="None")
      gt_world, local_tracer, N=self._cf_simulation_steps)

And new error occurs.

  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/tfa_counterfactual.runfiles/bark_ml/bark_ml/environments/counterfactual_runtime.py", line 169, in TraceCounterfactualWorldStats
    collision_rate = local_tracer.collision_rate
AttributeError: 'Tracer' object has no attribute 'collision_rate'

I Find that Tracer import through from bark_ml.commons.tracer import Tracer,and in bark_ml/commons/tracer.py, the context as follows:

import pickle

class Tracer:
  """The tracer can be used to log certain values during episodes."""

  def __init__(self, states=None, trace_history=True):
    self._trace_history = trace_history
    self._states = []

  def Trace(self, eval_dict):
    """Traces and stores a state."""
    if self._trace_history:
      self._states.append(eval_dict)

  def Reset(self):
    self._trace_history = []
Netceor commented 2 years ago

Now I have solved part of the problem, currently there are only two types of error.

  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/qrdqn.runfiles/bark_ml/bark_ml/library_wrappers/lib_fqf_iqn_qrdqn/agent/base_agent.py", line 131, in reset_action_observer
    self._observer = self._env._observer
  File "/home/myx/anaconda3/envs/bark-ml/lib/python3.7/site-packages/gym/core.py", line 228, in __getattr__
    raise AttributeError(f"attempted to get missing private attribute '{name}'")
AttributeError: attempted to get missing private attribute '_observer'
  File "/home/myx/.cache/bazel/_bazel_myx/d2a33cf9a94393d7ef91227effa0186e/execroot/bark_ml/bazel-out/k8-fastbuild/bin/examples/tfa_counterfactual.runfiles/bark_ml/bark_ml/environments/counterfactual_runtime.py", line 169, in TraceCounterfactualWorldStats
    collision_rate = local_tracer.collision_rate
AttributeError: 'Tracer' object has no attribute 'collision_rate'

Tracer import through from bark_ml.commons.tracer import Tracer,and in bark_ml/commons/tracer.py, the context as follows:

import pickle
class Tracer:
  """The tracer can be used to log certain values during episodes."""

  def __init__(self, states=None, trace_history=True):
    self._trace_history = trace_history
    self._states = []

  def Trace(self, eval_dict):
    """Traces and stores a state."""
    if self._trace_history:
      self._states.append(eval_dict)

  def Reset(self):
    self._trace_history = []

There are no collision_rate、success_rate params which is running counterfactual_runtime needed

CesarLiu commented 2 years ago

Hi,

thanks a lot for your interest in our bark-ml. Sorry for these bugs, we noticed them in June, and just had time to fix them. These bugs should be fixed in the latest commit. If you still have problems with running them, pls. raise your question here again. Thanks and best