InternVL2 微调 报错 Message: 'Error occurs in lazy tokenize:'

[a815137111]

Message: 'Error occurs in lazy tokenize:' 这个报错怎么解决

[tastelikefeet]

发下完整的栈信息

[a815137111]

发下完整的栈信息 Traceback (most recent call last): File "/home/junhang/swift/swift/llm/utils/utils.py", line 266, in _try_fetch res = self.template.encode(data) File "/home/junhang/swift/swift/llm/utils/template.py", line 447, in encode return _encode(example) if not streaming else _encode(example)[0] File "/home/junhang/swift/swift/llm/utils/template.py", line 1714, in _encode pixel_values = [transform_image(image, max_num=1 if has_video else 12) for image in images] File "/home/junhang/swift/swift/llm/utils/template.py", line 1714, in pixel_values = [transform_image(image, max_num=1 if has_video else 12) for image in images] File "/home/junhang/swift/swift/llm/utils/vision_utils.py", line 167, in transform_image images = _dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num) File "/home/junhang/swift/swift/llm/utils/vision_utils.py", line 63, in _dynamic_preprocess resized_img = image.resize((target_width, target_height)) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/PIL/Image.py", line 2293, in resize self.load() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/PIL/ImageFile.py", line 312, in load raise _get_oserror(err_code, encoder=False) OSError: broken data stream when reading image file

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/home/junhang/anaconda3/envs/VL/lib/python3.10/logging/init.py", line 1100, in emit msg = self.format(record) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/logging/init.py", line 943, in format return fmt.format(record) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/logging/init.py", line 678, in format record.message = record.getMessage() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/logging/init.py", line 368, in getMessage msg = msg % self.args TypeError: not all arguments converted during string formatting Call stack: File "/home/junhang/swift/swift/cli/sft.py", line 5, in sft_main() File "/home/junhang/swift/swift/utils/run_utils.py", line 32, in x_main result = llm_x(args, kwargs) File "/home/junhang/swift/swift/llm/sft.py", line 405, in llm_sft trainer.train(training_args.resume_from_checkpoint) File "/home/junhang/swift/swift/trainers/mixin.py", line 538, in train res = super().train(resume_from_checkpoint, *args, *kwargs) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/transformers/trainer.py", line 1948, in train return inner_training_loop( File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/transformers/trainer.py", line 2246, in _inner_training_loop for step, inputs in enumerate(epoch_iterator): File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/accelerate/data_loader.py", line 451, in iter dataloader_iter = super().iter() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 440, in iter return self._get_iterator() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 388, in _get_iterator return _MultiProcessingDataLoaderIter(self) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 1038, in init w.start() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/process.py", line 121, in start self._popen = self._Popen(self) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/context.py", line 224, in _Popen return _default_context.get_context().Process._Popen(process_obj) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/context.py", line 281, in _Popen return Popen(process_obj) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/popen_fork.py", line 19, in init self._launch(process_obj) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/popen_fork.py", line 71, in _launch code = process_obj._bootstrap(parent_sentinel=child_r) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/process.py", line 314, in _bootstrap self.run() File "/home/junhang/anaconda3/envs/VL/lib/python3.10/multiprocessing/process.py", line 108, in run self._target(self._args, self._kwargs) File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/_utils/worker.py", line 309, in _worker_loop data = fetcher.fetch(index) # type: ignore[possibly-undefined] File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/junhang/anaconda3/envs/VL/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/junhang/swift/swift/llm/utils/utils.py", line 255, in getitem res = self._try_fetch(idx) File "/home/junhang/swift/swift/llm/utils/utils.py", line 268, in _try_fetch logger.error('Error occurs in lazy tokenize:', e) Message: 'Error occurs in lazy tokenize:' Arguments: (OSError('broken data stream when reading image file'),) 是有图片无法读取吗

[Jintao-Huang]

尝试拉取一下main分支最新的代码 如果不行的话，发一下sh看看