Sync: commaai/openpilot:master into sunnypilot/sunnypilot:master-new (#874)

common/model.h

@@ -1 +1 @@
-#define DEFAULT_MODEL "Tomb_Raider_6 (Default)"
+#define DEFAULT_MODEL "Filet o Fish (Default)"

selfdrive/car/tests/test_models.py

@@ -338,6 +338,7 @@ class TestCarModelBase(unittest.TestCase):
       prev_panda_gas = self.safety.get_gas_pressed_prev()
       prev_panda_brake = self.safety.get_brake_pressed_prev()
       prev_panda_regen_braking = self.safety.get_regen_braking_prev()
+      prev_panda_steering_disengage = self.safety.get_steering_disengage_prev()
       prev_panda_vehicle_moving = self.safety.get_vehicle_moving()
       prev_panda_vehicle_speed_min = self.safety.get_vehicle_speed_min()
       prev_panda_vehicle_speed_max = self.safety.get_vehicle_speed_max()
@@ -365,6 +366,9 @@ class TestCarModelBase(unittest.TestCase):
       if self.safety.get_regen_braking_prev() != prev_panda_regen_braking:
         self.assertEqual(CS.regenBraking, self.safety.get_regen_braking_prev())
 
+      if self.safety.get_steering_disengage_prev() != prev_panda_steering_disengage:
+        self.assertEqual(CS.steeringDisengage, self.safety.get_steering_disengage_prev())
+
       if self.safety.get_vehicle_moving() != prev_panda_vehicle_moving:
         self.assertEqual(not CS.standstill, self.safety.get_vehicle_moving())
 
@@ -440,6 +444,7 @@ class TestCarModelBase(unittest.TestCase):
           brake_pressed = False
       checks['brakePressed'] += brake_pressed != self.safety.get_brake_pressed_prev()
       checks['regenBraking'] += CS.regenBraking != self.safety.get_regen_braking_prev()
+      checks['steeringDisengage'] += CS.steeringDisengage != self.safety.get_steering_disengage_prev()
 
       if self.CP.pcmCruise:
         # On most pcmCruise cars, openpilot's state is always tied to the PCM's cruise state.

selfdrive/controls/lib/longitudinal_planner.py

@@ -94,9 +94,9 @@ class LongitudinalPlanner(LongitudinalPlannerSP):
 
   def update(self, sm):
     LongitudinalPlannerSP.update(self, sm)
-    self.mode = 'blended' if sm['selfdriveState'].experimentalMode else 'acc'
+    self.mpc.mode = 'blended' if sm['selfdriveState'].experimentalMode else 'acc'
     if dec_mpc_mode := self.get_mpc_mode():
-      self.mode = dec_mpc_mode
+      self.mpc.mode = dec_mpc_mode
 
     if len(sm['carControl'].orientationNED) == 3:
       accel_coast = get_coast_accel(sm['carControl'].orientationNED[1])
@@ -119,7 +119,7 @@ class LongitudinalPlanner(LongitudinalPlannerSP):
     # No change cost when user is controlling the speed, or when standstill
     prev_accel_constraint = not (reset_state or sm['carState'].standstill)
 
-    if self.mode == 'acc':
+    if self.mpc.mode == 'acc':
       accel_clip = [ACCEL_MIN, get_max_accel(v_ego)]
       steer_angle_without_offset = sm['carState'].steeringAngleDeg - sm['liveParameters'].angleOffsetDeg
       accel_clip = limit_accel_in_turns(v_ego, steer_angle_without_offset, accel_clip, self.CP)
@@ -166,17 +166,8 @@ class LongitudinalPlanner(LongitudinalPlannerSP):
     self.v_desired_filter.x = self.v_desired_filter.x + self.dt * (self.a_desired + a_prev) / 2.0
 
     action_t =  self.CP.longitudinalActuatorDelay + DT_MDL
-    output_a_target_mpc, output_should_stop_mpc = get_accel_from_plan(self.v_desired_trajectory, self.a_desired_trajectory, CONTROL_N_T_IDX,
+    output_a_target, self.output_should_stop = get_accel_from_plan(self.v_desired_trajectory, self.a_desired_trajectory, CONTROL_N_T_IDX,
                                                                         action_t=action_t, vEgoStopping=self.CP.vEgoStopping)
-    output_a_target_e2e = sm['modelV2'].action.desiredAcceleration
-    output_should_stop_e2e = sm['modelV2'].action.shouldStop
-
-    if self.mode == 'acc':
-      output_a_target = output_a_target_mpc
-      self.output_should_stop = output_should_stop_mpc
-    else:
-      output_a_target = min(output_a_target_mpc, output_a_target_e2e)
-      self.output_should_stop = output_should_stop_e2e or output_should_stop_mpc
 
     for idx in range(2):
       accel_clip[idx] = np.clip(accel_clip[idx], self.prev_accel_clip[idx] - 0.05, self.prev_accel_clip[idx] + 0.05)

selfdrive/modeld/fill_model_msg.py

@@ -102,7 +102,6 @@ def fill_model_msg(base_msg: capnp._DynamicStructBuilder, extended_msg: capnp._D
     temporal_pose.rot = net_output_data['plan'][0,0,Plan.ORIENTATION_RATE].tolist()
     temporal_pose.rotStd = net_output_data['plan_stds'][0,0,Plan.ORIENTATION_RATE].tolist()
 
-
   # poly path
   fill_xyz_poly(driving_model_data.path, ModelConstants.POLY_PATH_DEGREE, *net_output_data['plan'][0,:,Plan.POSITION].T)
 

selfdrive/modeld/modeld.py

@@ -26,7 +26,7 @@ from openpilot.common.transformations.camera import DEVICE_CAMERAS
 from openpilot.common.transformations.model import get_warp_matrix
 from openpilot.system import sentry
 from openpilot.selfdrive.controls.lib.desire_helper import DesireHelper
-from openpilot.selfdrive.controls.lib.drive_helpers import get_accel_from_plan, smooth_value, get_curvature_from_plan
+from openpilot.selfdrive.controls.lib.drive_helpers import get_accel_from_plan, smooth_value
 from openpilot.selfdrive.modeld.parse_model_outputs import Parser
 from openpilot.selfdrive.modeld.fill_model_msg import fill_model_msg, fill_pose_msg, PublishState
 from openpilot.selfdrive.modeld.constants import ModelConstants, Plan
@@ -41,8 +41,8 @@ POLICY_PKL_PATH = Path(__file__).parent / 'models/driving_policy_tinygrad.pkl'
 VISION_METADATA_PATH = Path(__file__).parent / 'models/driving_vision_metadata.pkl'
 POLICY_METADATA_PATH = Path(__file__).parent / 'models/driving_policy_metadata.pkl'
 
-LAT_SMOOTH_SECONDS = 0.3
-LONG_SMOOTH_SECONDS = 0.3
+LAT_SMOOTH_SECONDS = 0.0
+LONG_SMOOTH_SECONDS = 0.0
 MIN_LAT_CONTROL_SPEED = 0.3
 
 
@@ -54,9 +54,8 @@ def get_action_from_model(model_output: dict[str, np.ndarray], prev_action: log.
                                                      ModelConstants.T_IDXS,
                                                      action_t=long_action_t)
     desired_accel = smooth_value(desired_accel, prev_action.desiredAcceleration, LONG_SMOOTH_SECONDS)
-    desired_curvature = get_curvature_from_plan(plan[:, Plan.T_FROM_CURRENT_EULER][:, 2],
-                                                plan[:, Plan.ORIENTATION_RATE][:, 2],
-                                                ModelConstants.T_IDXS, v_ego, lat_action_t)
+
+    desired_curvature = model_output['desired_curvature'][0, 0]
     if v_ego > MIN_LAT_CONTROL_SPEED:
       desired_curvature = smooth_value(desired_curvature, prev_action.desiredCurvature, LAT_SMOOTH_SECONDS)
     else:
@@ -173,7 +172,7 @@ class ModelState:
     # TODO model only uses last value now
     self.full_prev_desired_curv[0,:-1] = self.full_prev_desired_curv[0,1:]
     self.full_prev_desired_curv[0,-1,:] = policy_outputs_dict['desired_curvature'][0, :]
-    self.numpy_inputs['prev_desired_curv'][:] = 0*self.full_prev_desired_curv[0, self.temporal_idxs]
+    self.numpy_inputs['prev_desired_curv'][:] = self.full_prev_desired_curv[0, self.temporal_idxs]
 
     combined_outputs_dict = {**vision_outputs_dict, **policy_outputs_dict}
     if SEND_RAW_PRED:

selfdrive/modeld/models/driving_policy.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a478376723ba79e2393ca95e2d3e497571ba6fed113e5f13a36f0e4b4d4a7c5
-size 15588463
+oid sha256:98f0121ccb6f850077b04cc91bd33d370fc6cbdc2bd35f1ab55628a15a813f36
+size 15966721

selfdrive/modeld/models/driving_vision.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dad289ae367cefcb862ef1d707fb4919d008f0eeaa1ebaf18df58d8de5a7d96e
-size 46265585
+oid sha256:897f80d0388250f99bba69b6a8434560cc0fd83157cbeb0bc134c67fe4e64624
+size 34882971

selfdrive/modeld/parse_model_outputs.py

@@ -88,12 +88,6 @@ class Parser:
     self.parse_mdn('pose', outs, in_N=0, out_N=0, out_shape=(ModelConstants.POSE_WIDTH,))
     self.parse_mdn('wide_from_device_euler', outs, in_N=0, out_N=0, out_shape=(ModelConstants.WIDE_FROM_DEVICE_WIDTH,))
     self.parse_mdn('road_transform', outs, in_N=0, out_N=0, out_shape=(ModelConstants.POSE_WIDTH,))
-    self.parse_mdn('lane_lines', outs, in_N=0, out_N=0, out_shape=(ModelConstants.NUM_LANE_LINES,ModelConstants.IDX_N,ModelConstants.LANE_LINES_WIDTH))
-    self.parse_mdn('road_edges', outs, in_N=0, out_N=0, out_shape=(ModelConstants.NUM_ROAD_EDGES,ModelConstants.IDX_N,ModelConstants.LANE_LINES_WIDTH))
-    self.parse_mdn('lead', outs, in_N=ModelConstants.LEAD_MHP_N, out_N=ModelConstants.LEAD_MHP_SELECTION,
-                   out_shape=(ModelConstants.LEAD_TRAJ_LEN,ModelConstants.LEAD_WIDTH))
-    for k in ['lead_prob', 'lane_lines_prob']:
-      self.parse_binary_crossentropy(k, outs)
     self.parse_categorical_crossentropy('desire_pred', outs, out_shape=(ModelConstants.DESIRE_PRED_LEN,ModelConstants.DESIRE_PRED_WIDTH))
     self.parse_binary_crossentropy('meta', outs)
     return outs
@@ -101,10 +95,17 @@ class Parser:
   def parse_policy_outputs(self, outs: dict[str, np.ndarray]) -> dict[str, np.ndarray]:
     self.parse_mdn('plan', outs, in_N=ModelConstants.PLAN_MHP_N, out_N=ModelConstants.PLAN_MHP_SELECTION,
                    out_shape=(ModelConstants.IDX_N,ModelConstants.PLAN_WIDTH))
+    self.parse_mdn('lane_lines', outs, in_N=0, out_N=0, out_shape=(ModelConstants.NUM_LANE_LINES,ModelConstants.IDX_N,ModelConstants.LANE_LINES_WIDTH))
+    self.parse_mdn('road_edges', outs, in_N=0, out_N=0, out_shape=(ModelConstants.NUM_ROAD_EDGES,ModelConstants.IDX_N,ModelConstants.LANE_LINES_WIDTH))
+    self.parse_mdn('sim_pose', outs, in_N=0, out_N=0, out_shape=(ModelConstants.POSE_WIDTH,))
+    self.parse_mdn('lead', outs, in_N=ModelConstants.LEAD_MHP_N, out_N=ModelConstants.LEAD_MHP_SELECTION,
+                   out_shape=(ModelConstants.LEAD_TRAJ_LEN,ModelConstants.LEAD_WIDTH))
     if 'lat_planner_solution' in outs:
       self.parse_mdn('lat_planner_solution', outs, in_N=0, out_N=0, out_shape=(ModelConstants.IDX_N,ModelConstants.LAT_PLANNER_SOLUTION_WIDTH))
     if 'desired_curvature' in outs:
       self.parse_mdn('desired_curvature', outs, in_N=0, out_N=0, out_shape=(ModelConstants.DESIRED_CURV_WIDTH,))
+    for k in ['lead_prob', 'lane_lines_prob']:
+      self.parse_binary_crossentropy(k, outs)
     self.parse_categorical_crossentropy('desire_state', outs, out_shape=(ModelConstants.DESIRE_PRED_WIDTH,))
     return outs
 

selfdrive/test/process_replay/ref_commit

@@ -1 +1 @@
-7bf4ae5b92a3ad1f073f675e24e28babad0f2aa0
+7bf4ae5b92a3ad1f073f675e24e28babad0f2aa0

sunnypilot/models/tests/model_hash

@@ -1 +1 @@
-82ead1a4ecffbde961382329a6cf71bc97a6b112fbfb1f175a51c28ee2472bec
+dec34c57c4131f6fca5d1035201d1afbf43e5250cede7bfdc798371af008afad