错误 RKNN_ERR_MODEL_INVALID

用 mac 转的模型（rknn-toolkit=1.2.1, tensorflow=1.11.0），在 mac 上接 rk1808 可以 inference：
D Save log info to: ./verbose.log
P14443 2019-12-02 10:26:10,520 I [enhancement.py:56] loaded rknn model
I npu_transfer_proxy pid: 4906, status: running
D NPUTransfer: Transfer spec = local:transfer_proxy

D NPUTransfer: Transfer interface successfully opened, fd = 5
D RKNNAPI: ==============================================
D RKNNAPI: RKNN VERSION:
D RKNNAPI:   API: 1.2.0 (1190a71 build: 2019-09-25 12:43:54)
D RKNNAPI:   DRV: 0.9.8 (943f8e8 build: 2019-07-03 17:51:17)
D RKNNAPI: ==============================================
P14443 2019-12-02 10:26:22,432 I [enhancement.py:62] initialized runtime
D NPUTransfer: Transfer client closed, fd = 5
P14443 2019-12-02 10:26:31,030 I [enhancement.py:74] released
2019-12-02 10:26:31.031177: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA
2019-12-02 10:26:31.031628: I tensorflow/core/common_runtime/process_util.cc:69] Creating new thread pool with default inter op setting: 8. Tune using inter_op_parallelism_threads for best performance.
P14443 2019-12-02 10:26:36,006 I [enhancement.py:92] rknn time: 8.389446
P14443 2019-12-02 10:26:36,007 I [enhancement.py:93] tf time: 4.899358
P14443 2019-12-02 10:26:36,007 I [enhancement.py:94] rknn shape: 409600, (1, 409600)
P14443 2019-12-02 10:26:36,008 I [enhancement.py:95] tf shape: 409600, (1, 640, 640, 1)
P14443 2019-12-02 10:26:36,008 I [enhancement.py:98] rknn output: [[0.49780273 0.49731445 0.4951172  ... 0.03842163 0.03582764 0.03793335]]
P14443 2019-12-02 10:26:36,010 I [enhancement.py:99] tf output: [[0.4979127  0.49742317 0.49518007 ... 0.4945533  0.4964797  0.4983984 ]]
P14443 2019-12-02 10:26:36,020 I [enhancement.py:101] diff: 0.000001
P14443 2019-12-02 10:26:36,021 I [enhancement.py:102] diff: [[-1.0997057e-04 -1.0871887e-04 -6.2882900e-05 ... -4.5613167e-01
  -4.6065205e-01 -4.6046504e-01]]


但是在 rk3399pro 上出现错误 RKNN_ERR_MODEL_INVALID（rknn-toolkit=1.2.1）：

D Save log info to: ./verbose.log
P6880 2019-12-01 21:26:15,797 I [enhancement.py:56] loaded rknn model
I npu_transfer_proxy pid: 682, status: sleeping
D NPUTransfer: efuseid = 3399:6c1234ab1cf918b3
D NPUTransfer: Transfer spec = local:transfer_proxy
D NPUTransfer: Transfer interface successfully opened, fd = 4
E RKNNAPI: rknn_init,  msg_load_ack fail, ack = 1(ACK_FAIL), expect 0(ACK_SUCC)!
E RKNNAPI: ==============================================
E RKNNAPI: RKNN VERSION:
E RKNNAPI:   API: 1.2.0 (1190a71 build: 2019-09-25 12:39:26)
E RKNNAPI:   DRV: 0.9.9 (8509ec7 build: 2019-08-05 10:54:34)
E RKNNAPI: ==============================================
D NPUTransfer: Transfer client closed, fd = 4
E Catch exception when init runtime!
E Traceback (most recent call last):
E   File "rknn/api/rknn_base.py", line 788, in rknn.api.rknn_base.RKNNBase.init_runtime
E   File "rknn/api/rknn_runtime.py", line 303, in rknn.api.rknn_runtime.RKNNRuntime.build_graph
E Exception: RKNN init failed. error code: RKNN_ERR_MODEL_INVALID
E Current device id is: None
E Devices connected:
E ['0123456789ABCDEF']
P6880 2019-12-01 21:26:24,391 E [enhancement.py:60] failed to initialize runtime