KeyError: 'data_files'

[Wjs-111]

尊敬的作者：
我按照README进行了配置，再运行bash ./evaluation/scripts/llava_video_7b/videomme_aks_llava_video_7b.sh时出现了错误，如下所示，试了很多方法还是无法解决，请求作者可以解答一下，万分感谢。

The following values were not passed to accelerate launch and had defaults used instead:
More than one GPU was found, enabling multi-GPU training.
If this was unintended please pass in --num_processes=1.
--num_machines was set to a value of 1
--mixed_precision was set to a value of 'no'
--dynamo_backend was set to a value of 'no'
To avoid this warning pass in values for each of the problematic parameters or run accelerate config.
2025-11-06 18:29:00.146 | INFO | main:cli_evaluate:294 - Verbosity set to DEBUG
2025-11-06 18:29:00.156 | INFO | main:cli_evaluate:294 - Verbosity set to DEBUG
2025-11-06 18:29:00.628 | DEBUG | lmms_eval.tasks:_get_task_and_group:452 - group and group_alias keys in tasks' configs will no longer be used in the next release of lmms-eval. tag will be used to allow to call a collection of tasks just like group. group will be removed in order to not cause confusion with the new ConfigurableGroup which will be the offical way to create groups with addition of group-wide configuations.
2025-11-06 18:29:00.781 | DEBUG | lmms_eval.tasks:_get_task_and_group:452 - group and group_alias keys in tasks' configs will no longer be used in the next release of lmms-eval. tag will be used to allow to call a collection of tasks just like group. group will be removed in order to not cause confusion with the new ConfigurableGroup which will be the offical way to create groups with addition of group-wide configuations.
2025-11-06 18:29:02.717 | INFO | main:cli_evaluate_single:377 - Evaluation tracker args: {'output_path': './results/selected_frames'}
2025-11-06 18:29:02.718 | INFO | main:cli_evaluate_single:466 - Selected Tasks: ['videomme']
2025-11-06 18:29:02.719 | INFO | lmms_eval.evaluator:simple_evaluate:155 - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234
2025-11-06 18:29:02.866 | INFO | main:cli_evaluate_single:377 - Evaluation tracker args: {'output_path': './results/selected_frames'}
2025-11-06 18:29:02.867 | INFO | main:cli_evaluate_single:466 - Selected Tasks: ['videomme']
2025-11-06 18:29:02.868 | INFO | lmms_eval.evaluator:simple_evaluate:155 - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234
force sample: False
force sample: False
Rank 0: Loaded LLaVA model: ./checkpoints/LLaVA-NeXT-Video-7B-Qwen2
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
Rank 0: Loading vision tower: google/siglip-so400m-patch14-384
Loading checkpoint shards: 100%|███████████████████| 4/4 [00:13<00:00, 3.40s/it]
Rank 0: Model Class: LlavaQwenForCausalLM
2025-11-06 18:29:22.500 | INFO | lmms_eval.models.llava_vid:init:218 - Using 2 devices with data parallelism
Fetching 24 files: 100%|█████████████████████| 24/24 [00:00<00:00, 139230.01it/s]
Loading checkpoint shards: 100%|███████████████████| 4/4 [00:13<00:00, 3.35s/it]
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 470, in call
result = fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 1050, in download
data_files=dataset_kwargs['data_files'],
KeyError: 'data_files'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 532, in
cli_evaluate()
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 345, in cli_evaluate
raise e
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 329, in cli_evaluate
results, samples = cli_evaluate_single(args)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 470, in cli_evaluate_single
results = evaluator.simple_evaluate(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/utils.py", line 533, in _wrapper
return fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/evaluator.py", line 180, in simple_evaluate
task_dict = get_task_dict(tasks, task_manager)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 558, in get_task_dict
task_name_from_string_dict = task_manager.load_task_or_group(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 372, in load_task_or_group
all_loaded_tasks = dict(collections.ChainMap(*map(self._load_individual_task_or_group, task_list)))
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 289, in _load_individual_task_or_group
return _load_task(task_config, task=name_or_config)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 259, in _load_task
task_object = ConfigurableTask(config=config, model_name=self.model_name)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 718, in init
self.download(self.config.dataset_kwargs)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 330, in wrapped_f
return self(f, *args, **kw)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 467, in call
do = self.iter(retry_state=retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 368, in iter
result = action(retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 411, in exc_check
raise retry_exc from fut.exception()
tenacity.RetryError: RetryError[<Future at 0x72320f198850 state=finished raised KeyError>]
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 470, in call
result = fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 1050, in download
data_files=dataset_kwargs['data_files'],
KeyError: 'data_files'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 532, in
cli_evaluate()
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 345, in cli_evaluate
raise e
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 329, in cli_evaluate
results, samples = cli_evaluate_single(args)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 470, in cli_evaluate_single
results = evaluator.simple_evaluate(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/utils.py", line 533, in _wrapper
return fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/evaluator.py", line 180, in simple_evaluate
task_dict = get_task_dict(tasks, task_manager)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 558, in get_task_dict
task_name_from_string_dict = task_manager.load_task_or_group(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 372, in load_task_or_group
all_loaded_tasks = dict(collections.ChainMap(*map(self._load_individual_task_or_group, task_list)))
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 289, in _load_individual_task_or_group
return _load_task(task_config, task=name_or_config)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 259, in _load_task
task_object = ConfigurableTask(config=config, model_name=self.model_name)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 718, in init
self.download(self.config.dataset_kwargs)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 330, in wrapped_f
return self(f, *args, **kw)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 467, in call
do = self.iter(retry_state=retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 368, in iter
result = action(retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 411, in exc_check
raise retry_exc from fut.exception()
tenacity.RetryError: RetryError[<Future at 0x71aa399aeb50 state=finished raised KeyError>]
[2025-11-06 18:29:32,717] torch.distributed.elastic.multiprocessing.api: [WARNING] Sending process 442626 closing signal SIGTERM
[2025-11-06 18:29:32,782] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 1) local_rank: 0 (pid: 442625) of binary: /home/jiusheng/miniconda3/envs/AKS/bin/python3.9
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/bin/accelerate", line 7, in
sys.exit(main())
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/accelerate_cli.py", line 50, in main
args.func(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/launch.py", line 1226, in launch_command
multi_gpu_launcher(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/launch.py", line 853, in multi_gpu_launcher
distrib_run.run(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/run.py", line 797, in run
elastic_launch(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 134, in call
return launch_agent(self._config, self._entrypoint, list(args))
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 264, in launch_agent
raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError:

lmms_eval FAILED

Failures:
<NO_OTHER_FAILURES>

Root Cause (first observed failure):
[0]:
time : 2025-11-06_18:29:32
host : aim5880-8
rank : 0 (local_rank: 0)
exitcode : 1 (pid: 442625)
error_file: <N/A>
traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html

[ncTimTang]

Hi @Wjs-111 ,

Thank you for your interest in our work!

From the error information you sent, I think it is possible that you did not follow the following step in README, because we defined 'data_files' here.

For VideoMME，please delete the README.md, and replace the raw videomme.yaml and utils.py in lmms_eval/tasks/videomme/ with videomme.yaml and utils.py.

Hope these can help you.

谢谢你对我们工作的关注！

从你发的报错信息来看，我认为有可能是你没有按照readme.md执行这一步，因为在这里我们定义了“data_files”.

For VideoMME，please delete the README.md, and replace the raw videomme.yaml and utils.py in lmms_eval/tasks/videomme/ with videomme.yaml and utils.py.

希望这些能帮到你！

Best,

Tim

[Wjs-111]

Sorry, due to my mistake, I modified the YAML file incorrectly and didn't modify the YAML file in the home directory, which caused the error. The issue has been resolved now. However, I'm still encountering the "dtype" error that I couldn't solve before. I had previously set up the environment and tested it, but because of this persistent issue, I decided to re-download everything and reconfigure the environment, which led to the YAML replacement mistake. Could you please help me figure out how to solve this "dtype" error? Thank you so much! (I'm just starting out in research and trying to reproduce results - I appreciate your patience with a beginner!)

error:
(AKS) jiusheng@aim5880-8:/data/jiusheng/AKS$ bash ./evaluation/scripts/llava_video_7b/videomme_aks_llava_video_7b.sh
The following values were not passed to accelerate launch and had defaults used instead:
More than one GPU was found, enabling multi-GPU training.
If this was unintended please pass in --num_processes=1.
--num_machines was set to a value of 1
--mixed_precision was set to a value of 'no'
--dynamo_backend was set to a value of 'no'
To avoid this warning pass in values for each of the problematic parameters or run accelerate config.
2025-11-07 11:04:54.542 | INFO | main:cli_evaluate:294 - Verbosity set to DEBUG
2025-11-07 11:04:54.542 | INFO | main:cli_evaluate:294 - Verbosity set to DEBUG
2025-11-07 11:04:55.170 | DEBUG | lmms_eval.tasks:_get_task_and_group:452 - group and group_alias keys in tasks' configs will no longer be used in the next release of lmms-eval. tag will be used to allow to call a collection of tasks just like group. group will be removed in order to not cause confusion with the new ConfigurableGroup which will be the offical way to create groups with addition of group-wide configuations.
2025-11-07 11:04:55.188 | DEBUG | lmms_eval.tasks:_get_task_and_group:452 - group and group_alias keys in tasks' configs will no longer be used in the next release of lmms-eval. tag will be used to allow to call a collection of tasks just like group. group will be removed in order to not cause confusion with the new ConfigurableGroup which will be the offical way to create groups with addition of group-wide configuations.
2025-11-07 11:05:01.380 | INFO | main:cli_evaluate_single:377 - Evaluation tracker args: {'output_path': './results/selected_frames'}
2025-11-07 11:05:01.380 | INFO | main:cli_evaluate_single:377 - Evaluation tracker args: {'output_path': './results/selected_frames'}
2025-11-07 11:05:01.381 | INFO | main:cli_evaluate_single:466 - Selected Tasks: ['videomme']
2025-11-07 11:05:01.381 | INFO | main:cli_evaluate_single:466 - Selected Tasks: ['videomme']
2025-11-07 11:05:01.382 | INFO | lmms_eval.evaluator:simple_evaluate:155 - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234
2025-11-07 11:05:01.382 | INFO | lmms_eval.evaluator:simple_evaluate:155 - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234
force sample: False
Rank 0: Loaded LLaVA model: ./checkpoints/LLaVA-NeXT-Video-7B-Qwen2
force sample: False
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
You are using a model of type llava to instantiate a model of type llava_qwen. This is not supported for all configurations of models and can yield errors.
Rank 0: Loading vision tower: google/siglip-so400m-patch14-384
Loading checkpoint shards: 50%|████████████████████████████████████████████████████████████████████████████████▌ | 2/4 [00:08<00:08, 4.14s/it]Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:13<00:00, 3.31s/it]
Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:13<00:00, 3.37s/it]
Rank 0: Model Class: LlavaQwenForCausalLM
2025-11-07 11:05:20.707 | INFO | lmms_eval.models.llava_vid:init:218 - Using 2 devices with data parallelism
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 470, in call
result = fn(*args, kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 1048, in download
self.dataset = datasets.load_dataset(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 2523, in load_dataset
builder_instance = load_dataset_builder(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 2195, in load_dataset_builder
dataset_module = dataset_module_factory(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1756, in dataset_module_factory
return LocalDatasetModuleFactoryWithoutScript(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1071, in get_module
{
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1072, in
config_name: DatasetInfo.from_dict(dataset_info_dict)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/info.py", line 356, in from_dict
return cls({k: v for k, v in dataset_info_dict.items() if k in field_names})
File "", line 21, in init
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/info.py", line 174, in post_init
self.features = Features.from_dict(self.features)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1704, in from_dict
obj = generate_from_dict(dic)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1359, in generate_from_dict
return {key: generate_from_dict(value) for key, value in obj.items()}
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1359, in
return {key: generate_from_dict(value) for key, value in obj.items()}
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1367, in generate_from_dict
return class_type(**{k: v for k, v in obj.items() if k in field_names})
TypeError: init() missing 1 required positional argument: 'dtype'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 532, in
cli_evaluate()
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 345, in cli_evaluate
raise e
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 329, in cli_evaluate
results, samples = cli_evaluate_single(args)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 470, in cli_evaluate_single
results = evaluator.simple_evaluate(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/utils.py", line 533, in _wrapper
return fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/evaluator.py", line 180, in simple_evaluate
task_dict = get_task_dict(tasks, task_manager)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 558, in get_task_dict
task_name_from_string_dict = task_manager.load_task_or_group(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 372, in load_task_or_group
all_loaded_tasks = dict(collections.ChainMap(*map(self._load_individual_task_or_group, task_list)))
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 289, in _load_individual_task_or_group
return _load_task(task_config, task=name_or_config)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 259, in _load_task
task_object = ConfigurableTask(config=config, model_name=self.model_name)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 718, in init
self.download(self.config.dataset_kwargs)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 330, in wrapped_f
return self(f, *args, kw)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 467, in call
do = self.iter(retry_state=retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 368, in iter
result = action(retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 411, in exc_check
raise retry_exc from fut.exception()
tenacity.RetryError: RetryError[<Future at 0x7fa76618b1c0 state=finished raised TypeError>]
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 470, in call
result = fn(*args, kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 1048, in download
self.dataset = datasets.load_dataset(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 2523, in load_dataset
builder_instance = load_dataset_builder(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 2195, in load_dataset_builder
dataset_module = dataset_module_factory(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1756, in dataset_module_factory
return LocalDatasetModuleFactoryWithoutScript(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1071, in get_module
{
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/load.py", line 1072, in
config_name: DatasetInfo.from_dict(dataset_info_dict)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/info.py", line 356, in from_dict
return cls({k: v for k, v in dataset_info_dict.items() if k in field_names})
File "", line 21, in init
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/info.py", line 174, in post_init
self.features = Features.from_dict(self.features)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1704, in from_dict
obj = generate_from_dict(dic)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1359, in generate_from_dict
return {key: generate_from_dict(value) for key, value in obj.items()}
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1359, in
return {key: generate_from_dict(value) for key, value in obj.items()}
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/datasets/features/features.py", line 1367, in generate_from_dict
return class_type({k: v for k, v in obj.items() if k in field_names})
TypeError: init() missing 1 required positional argument: 'dtype'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 532, in
cli_evaluate()
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 345, in cli_evaluate
raise e
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 329, in cli_evaluate
results, samples = cli_evaluate_single(args)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/main.py", line 470, in cli_evaluate_single
results = evaluator.simple_evaluate(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/utils.py", line 533, in _wrapper
return fn(*args, **kwargs)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/evaluator.py", line 180, in simple_evaluate
task_dict = get_task_dict(tasks, task_manager)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 558, in get_task_dict
task_name_from_string_dict = task_manager.load_task_or_group(
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 372, in load_task_or_group
all_loaded_tasks = dict(collections.ChainMap(*map(self._load_individual_task_or_group, task_list)))
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 289, in _load_individual_task_or_group
return _load_task(task_config, task=name_or_config)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/tasks/init.py", line 259, in _load_task
task_object = ConfigurableTask(config=config, model_name=self.model_name)
File "/home/jiusheng/AKS-main/lmms-eval/lmms_eval/api/task.py", line 718, in init
self.download(self.config.dataset_kwargs)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 330, in wrapped_f
return self(f, *args, **kw)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 467, in call
do = self.iter(retry_state=retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 368, in iter
result = action(retry_state)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/tenacity/init.py", line 411, in exc_check
raise retry_exc from fut.exception()
tenacity.RetryError: RetryError[<Future at 0x760223bf24c0 state=finished raised TypeError>]
[2025-11-07 11:05:28,948] torch.distributed.elastic.multiprocessing.api: [WARNING] Sending process 882541 closing signal SIGTERM
[2025-11-07 11:05:29,414] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 1) local_rank: 1 (pid: 882542) of binary: /home/jiusheng/miniconda3/envs/AKS/bin/python3.9
Traceback (most recent call last):
File "/home/jiusheng/miniconda3/envs/AKS/bin/accelerate", line 7, in
sys.exit(main())
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/accelerate_cli.py", line 50, in main
args.func(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/launch.py", line 1226, in launch_command
multi_gpu_launcher(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/accelerate/commands/launch.py", line 853, in multi_gpu_launcher
distrib_run.run(args)
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/run.py", line 797, in run
elastic_launch(
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 134, in call
return launch_agent(self._config, self._entrypoint, list(args))
File "/home/jiusheng/miniconda3/envs/AKS/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 264, in launch_agent
raise ChildFailedError(
torch.distributed.elastic.multiprocessing.errors.ChildFailedError:

lmms_eval FAILED

Failures:
<NO_OTHER_FAILURES>

Root Cause (first observed failure):
[0]:
time : 2025-11-07_11:05:28
host : aim5880-8
rank : 1 (local_rank: 1)
exitcode : 1 (pid: 882542)
error_file: <N/A>
traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html

[ncTimTang]

Hi @Wjs-111 ,

I'm sorry that just got back to you now because I was busy with some other matters.

The error message you provided is a bit unclear, and I haven't encountered a similar error before. You could try printing k and v to find the specific variable for 'dtype'; this might make it easier for us to identify the problem.

Hope these can help you.

不好意思，因为前段时间我在忙一些其他的事，所以现在才回复。

你提供的错误信息有点不清楚，我以前没有遇到过类似的错误。你可以尝试print k 和 v 来找到 'dtype' 的具体变量，这可能能更容易看到具体的问题。

希望这些能帮到你！祝好！

Best,

Tim

[Wjs-111]

Thank you very much for your reply. Currently, I have resolved these issues and successfully reproduced the results.