A issue about DDPGAgent

[Paradisepass]

When training my model, I found that after two or three iterations, the model would encounter a problem where two matrices could not be multiplied because of their different dimensions. The log information is as follows:
2026-03-26 14:56:11,410-PID(544763) env.py[179]: Game play with 1 seconds later.
2026-03-26 14:56:11,454-PID(545274) monitor.py[170]: Small moving 1, num++, num now - 1.
79.6454471726882
2026-03-26 14:56:12,421-PID(544763) mavlink.py[155]: name: MC_ROLL_P value: 0.600000
2026-03-26 14:56:12,421-PID(544763) mavlink.py[155]: name: MC_PITCH_P value: 11.600000
2026-03-26 14:56:12,421-PID(544763) mavlink.py[155]: name: MC_YAW_P value: 0.200000
2026-03-26 14:56:12,421-PID(544763) mavlink.py[155]: name: MC_YAW_WEIGHT value: 0.900000
2026-03-26 14:56:12,421-PID(544763) mavlink.py[155]: name: MPC_XY_P value: 1.900000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MPC_Z_P value: 0.400000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MC_PITCHRATE_P value: 0.150000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MC_ROLLRATE_P value: 0.480000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MC_YAWRATE_P value: 0.580000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MIS_YAW_ERR value: 1.000000
2026-03-26 14:56:12,422-PID(544763) mavlink.py[155]: name: MPC_TKO_SPEED value: 4.900000
5
2026-03-26 14:56:12,422-PID(544763) env.py[186]: Set parameters successfully.
2026-03-26 14:56:12,443-PID(545319) boardMavlink.py[615]: Start Bin monitor.
79.6454471726882
2026-03-26 14:56:14,946-PID(544763) boardMavlink.py[53]: Return a captured segment.
2026-03-26 14:56:14,948-PID(544763) learning_agent.py[396]: Observation deviation 4.78584833626519
2026-03-26 14:56:14,948-PID(544763) learning_agent.py[411]: Deviation over threshold, start repair.
2026-03-26 14:56:14,950-PID(544763) mavlink.py[155]: name: MC_PITCHRATE_P value: 0.150000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_PITCH_P value: 11.600000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_ROLLRATE_P value: 0.480000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_ROLL_P value: 0.600000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_YAWRATE_P value: 0.580000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_YAW_P value: 0.200000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MC_YAW_WEIGHT value: 0.900000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MIS_YAW_ERR value: 1.000000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MPC_TKO_SPEED value: 4.900000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MPC_XY_P value: 1.900000
2026-03-26 14:56:14,951-PID(544763) mavlink.py[155]: name: MPC_Z_P value: 0.400000
79.6454471726882
79.6454471726882
2026-03-26 14:56:15,918-PID(545274) monitor.py[170]: Small moving 1, num++, num now - 1.
79.6454471726882
79.6454471726882
2026-03-26 14:56:16,865-PID(545274) monitor.py[170]: Small moving 1, num++, num now - 1.
79.6454471726882
2026-03-26 14:56:17,054-PID(544763) boardMavlink.py[53]: Return a captured segment.
2026-03-26 14:56:17,055-PID(544763) env.py[259]: Change Deviation: 4.7858 -> 4.887: reward: -0.1012, done: False
2026-03-26 14:56:17,055-PID(544763) learning_agent.py[177]: Item add to buffer: 0 - len:4154
2026-03-26 14:56:17,055-PID(544763) learning_agent.py[244]: Current buffer size is 1, skip.
79.6454471726882
2026-03-26 14:56:19,158-PID(544763) boardMavlink.py[53]: Return a captured segment.
2026-03-26 14:56:19,159-PID(544763) learning_agent.py[396]: Observation deviation 4.9239496405862155
2026-03-26 14:56:19,159-PID(544763) learning_agent.py[411]: Deviation over threshold, start repair.
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_ROLL_P value: 5.800000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_PITCH_P value: 7.200000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_YAW_P value: 1.400000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_YAW_WEIGHT value: 0.600000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MPC_XY_P value: 1.200000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MPC_Z_P value: 0.800000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_PITCHRATE_P value: 0.330000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_ROLLRATE_P value: 0.310000
2026-03-26 14:56:19,180-PID(544763) mavlink.py[155]: name: MC_YAWRATE_P value: 0.270000
2026-03-26 14:56:19,181-PID(544763) mavlink.py[155]: name: MIS_YAW_ERR value: 36.000000
2026-03-26 14:56:19,181-PID(544763) mavlink.py[155]: name: MPC_TKO_SPEED value: 1.400000
79.6454471726882
79.6454471726882
2026-03-26 14:56:21,283-PID(544763) boardMavlink.py[53]: Return a captured segment.
2026-03-26 14:56:21,284-PID(544763) env.py[259]: Change Deviation: 4.9239 -> 3.1951: reward: 0.3091, done: False
2026-03-26 14:56:21,284-PID(544763) learning_agent.py[177]: Item add to buffer: 1 - len:3482
2026-03-26 14:56:21,284-PID(544763) learning_agent.py[244]: Current buffer size is 2, skip.
79.6454471726882
79.6454471726882
79.6454471726882
2026-03-26 14:56:23,487-PID(544763) boardMavlink.py[53]: Return a captured segment.
2026-03-26 14:56:23,488-PID(544763) learning_agent.py[396]: Observation deviation 3.4235570763485157
2026-03-26 14:56:23,488-PID(544763) learning_agent.py[411]: Deviation over threshold, start repair.
2026-03-26 14:56:23,489-PID(544763) learning_agent.py[456]: Exception: <class 'RuntimeError'>, learning_agent.py, 413, mat1 and mat2 shapes cannot be multiplied (1x168 and 240x256).