transformer_asr,facenet ixrt报错

### 该问题是怎么引起的？

### 重现步骤

### 报错信息

transformer_asr：

```
root@908560e41ac9:~/deepsparkinference/models/audio/speech_recognition/transformer_asr/ixrt# python3 inference.py hparams/train_ASR_transformer.yaml --data_folder=/home/data/speechbrain/aishell --engine_path transformer.engine
Loaded plugin from /usr/local/corex-4.2.0/lib64/python3/dist-packages/ixrt/lib/libixrt_plugin.so
/home/data/speechbrain/aishell/rirs_noises.zip exists. Skipping download
/usr/local/lib/python3.10/site-packages/speechbrain/lobes/augment.py:509: ResourceWarning: unclosed file <_io.TextIOWrapper name='/home/data/speechbrain/aishell/RIRS_NOISES/pointsource_noises/noise_list' mode='r' encoding='UTF-8'>
  for line in open(filelist):
ResourceWarning: Enable tracemalloc to get the object allocation traceback
speechbrain.core - Beginning experiment!
speechbrain.core - Experiment folder: results/transformer/8886
speechbrain.pretrained.fetching - Fetch tokenizer.ckpt: Using existing file/symlink in results/transformer/8886/save/tokenizer.ckpt.
speechbrain.utils.parameter_transfer - Loading pretrained files for: tokenizer
speechbrain.core - Info: ckpt_interval_minutes arg from hparam file is used
speechbrain.core - 31.0M trainable parameters in ASR
/usr/local/corex-4.2.0/lib64/python3/dist-packages/ixrt/__init__.py:28: DeprecationWarning: Context managers for IxRT types are deprecated. Memory will be freed automatically when the reference count reaches 0.
  warnings.warn(
[INFO] FCPluginDynamic deserialize
[INFO] FCPluginDynamic initialize
[INFO] FCPluginDynamic clone
[INFO] Half Weights(Host) => Half Array(Host)
[INFO] Half Weights(Host) => Half Array(Host)
[INFO] FCPluginDynamic initialize
*** start evaluation ***
speechbrain.utils.checkpoints - Loading a checkpoint from results/transformer/8886/save/CKPT+2021-04-20+23-20-18+00
/usr/local/lib/python3.10/site-packages/speechbrain/utils/checkpoints.py:93: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  obj.load_state_dict(torch.load(path, map_location=device), strict=True)
speechbrain.core - Exception:
Traceback (most recent call last):
  File "/root/deepsparkinference/models/audio/speech_recognition/transformer_asr/ixrt/inference.py", line 596, in <module>
    asr_brain.evaluate(
  File "/root/deepsparkinference/models/audio/speech_recognition/transformer_asr/ixrt/inference.py", line 380, in evaluate
    self.on_evaluate_start(max_key=max_key, min_key=min_key)
  File "/root/deepsparkinference/models/audio/speech_recognition/transformer_asr/ixrt/inference.py", line 348, in on_evaluate_start
    super().on_evaluate_start()
  File "/usr/local/lib/python3.10/site-packages/speechbrain/core.py", line 836, in on_evaluate_start
    self.checkpointer.recover_if_possible(
  File "/usr/local/lib/python3.10/site-packages/speechbrain/utils/checkpoints.py", line 840, in recover_if_possible
    self.load_checkpoint(chosen_ckpt, device)
  File "/usr/local/lib/python3.10/site-packages/speechbrain/utils/checkpoints.py", line 853, in load_checkpoint
    self._call_load_hooks(checkpoint, device)
  File "/usr/local/lib/python3.10/site-packages/speechbrain/utils/checkpoints.py", line 988, in _call_load_hooks
    default_hook(obj, loadpath, end_of_epoch, device)
  File "/usr/local/lib/python3.10/site-packages/speechbrain/utils/checkpoints.py", line 93, in torch_recovery
    obj.load_state_dict(torch.load(path, map_location=device), strict=True)
  File "/usr/local/corex-4.2.0/lib64/python3/dist-packages/torch/nn/modules/module.py", line 2215, in load_state_dict
    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for ModuleList:
        Unexpected key(s) in state_dict: "0.convblock_0.convs.norm_0.norm.running_mean", "0.convblock_0.convs.norm_0.norm.running_var", "0.convblock_0.convs.norm_0.norm.num_batches_tracked", "0.convblock_1.convs.norm_0.norm.running_mean", "0.convblock_1.convs.norm_0.norm.running_var", "0.convblock_1.convs.norm_0.norm.num_batches_tracked".
        size mismatch for 0.convblock_0.convs.norm_0.norm.weight: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([40, 256]).
        size mismatch for 0.convblock_0.convs.norm_0.norm.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([40, 256]).
        size mismatch for 0.convblock_1.convs.norm_0.norm.weight: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([20, 256]).
        size mismatch for 0.convblock_1.convs.norm_0.norm.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([20, 256]).
Segmentation fault (core dumped)
```

facenet: 需要tf1 导出onnx报错，待确认

DeepSpark/DeepSparkInference

内容风险标识

评论 (0)

DeepSpark/DeepSparkInference .gitee-modal { width: 500px !important; }

内容风险标识