Datasets:

---

optimum-benchmark

/

llm-perf-leaderboard

like 2

Follow

Optimum-Benchmark 11

Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 2 new columns ({'config.backend.half', 'config.backend.reshape'}) and 95 missing columns ({'report.decode.energy.total', 'report.decode.latency.unit', 'report.load.memory.max_process_vram', 'config.backend.auto_calibration', 'report.prefill.memory.max_reserved', 'report.per_token.latency.p95', 'report.load.energy', 'report.decode.efficiency.unit', 'report.per_token.energy', 'report.load.memory.max_global_vram', 'report.prefill.latency.p95', 'report.prefill.latency.stdev', 'report.prefill.throughput.value', 'report.decode.latency.values', 'report.decode.energy.ram', 'report.load.latency.count', 'report.decode.latency.p95', 'report.prefill.latency.stdev_', 'report.load.latency.values', 'report.load.latency.p99', 'report.per_token.latency.stdev_', 'report.decode.latency.count', 'report.load.latency.unit', 'report.decode.memory.unit', 'report.decode.latency.stdev', 'report.decode.throughput.value', 'report.per_token.latency.count', 'report.load.latency.p95', 'report.load.latency.p90', 'report.decode.energy.cpu', 'report.per_token.throughput.value', 'report.prefill.energy.unit', 'report.per_token.latency.p50', 'report.decode.energy.gpu', 'report.per_token.efficiency', 'report.load.efficiency', 'report.per_token.latency.p90', 'report.prefill.latency.unit', 'report.load.throughput', 'report.per_token.latency.p99', 'report.decode.memory.max_reserved', 'report.load.latency.stdev_', 'report.prefill.latency.p99', 'report.decode.latency.stdev_', 'report.load.latency.mean', 'report.load.memory.max_allocated', 'report.decode.energy.unit', 'report.prefill.energy.gpu', 'report.pr
...
'report.per_token.memory', 'report.load.latency.stdev', 'config.backend.optimization', 'report.load.memory.max_reserved', 'report.prefill.memory.max_ram', 'report.prefill.latency.values', 'report.per_token.throughput.unit', 'report.prefill.memory.unit', 'report.prefill.energy.total', 'report.prefill.latency.p50', 'report.prefill.efficiency.value', 'report.per_token.latency.total', 'report.prefill.memory.max_process_vram', 'report.prefill.latency.mean', 'report.prefill.memory.max_allocated', 'report.prefill.latency.total', 'report.per_token.latency.unit', 'report.decode.memory.max_global_vram', 'config.backend.use_io_binding', 'report.prefill.energy.ram', 'report.decode.latency.mean', 'report.load.latency.total', 'report.decode.throughput.unit', 'report.decode.latency.p50', 'report.decode.latency.p99', 'report.prefill.memory.max_global_vram', 'report.load.memory.max_ram', 'report.decode.latency.p90', 'report.prefill.latency.p90', 'report.decode.latency.total', 'config.backend.torch_dtype', 'config.backend.auto_quantization', 'report.prefill.throughput.unit', 'report.per_token.latency.stdev', 'report.load.latency.p50', 'report.load.memory.unit', 'report.decode.efficiency.value', 'report.prefill.efficiency.unit', 'report.decode.memory.max_ram', 'config.backend.auto_optimization', 'report.per_token.latency.mean', 'config.backend.provider', 'report.decode.memory.max_allocated', 'report.decode.memory.max_process_vram', 'report.prefill.energy.cpu', 'report.per_token.latency.values'}).

This happened while the csv dataset builder was generating data using

hf://datasets/optimum-benchmark/llm-perf-leaderboard/data/perf-df-openvino-cpu-unquantized-32vCPU-C7i.csv (at revision 94a9713e1842c87029a1dcc829e0003533a6d275)

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1870, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 622, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2292, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2240, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              config.name: string
              config.backend.name: string
              config.backend.version: string
              config.backend._target_: string
              config.backend.task: string
              config.backend.library: string
              config.backend.model_type: string
              config.backend.model: string
              config.backend.processor: string
              config.backend.device: string
              config.backend.device_ids: int64
              config.backend.seed: int64
              config.backend.inter_op_num_threads: double
              config.backend.intra_op_num_threads: double
              config.backend.model_kwargs.trust_remote_code: bool
              config.backend.no_weights: bool
              config.backend.export: bool
              config.backend.use_cache: bool
              config.backend.use_merged: bool
              config.backend.half: bool
              config.backend.reshape: bool
              config.backend.quantization: bool
              config.backend.calibration: bool
              config.scenario.name: string
              config.scenario._target_: string
              config.scenario.iterations: int64
              config.scenario.duration: int64
              config.scenario.warmup_runs: int64
              config.scenario.input_shapes.batch_size: int64
              config.scenario.input_shapes.num_choices: double
              config.scenario.input_shapes.sequence_length: int64
              config.scenario.new_tokens: double
              config.scenario.memory: bool
              config.scenario.latency: bool
              config.scenario.energy: bool
              config.scenario.generate_kwargs.max_new_tokens: int64
              config.scenario.generate_kwargs.min_new_tokens: int64
              config.launcher.name: string
              config.launcher._target_: string
              config.launcher.device_isolation: bool
              config.launcher.device_isolation_action: double
              config.launcher.numactl: bool
              config.launcher.start_method: string
              config.environment.cpu: string
              config.environment.cpu_count: int64
              config.environment.cpu_ram_mb: double
              config.environment.system: string
              config.environment.machine: string
              config.environment.platform: string
              config.environment.processor: string
              config.environment.python_version: string
              config.environment.optimum_benchmark_version: string
              config.environment.optimum_benchmark_commit: double
              config.environment.transformers_version: string
              config.environment.transformers_commit: double
              config.environment.accelerate_version: string
              config.environment.accelerate_commit: double
              config.environment.diffusers_version: double
              config.environment.diffusers_commit: double
              config.environment.optimum_version: string
              config.environment.optimum_commit: double
              config.environment.timm_version: double
              config.environment.timm_commit: double
              config.environment.peft_version: double
              config.environment.peft_commit: double
              config.print_report: bool
              config.log_report: bool
              report.traceback: string
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 10798
              to
              {'config.name': Value(dtype='string', id=None), 'config.backend.name': Value(dtype='string', id=None), 'config.backend.version': Value(dtype='string', id=None), 'config.backend._target_': Value(dtype='string', id=None), 'config.backend.task': Value(dtype='string', id=None), 'config.backend.library': Value(dtype='string', id=None), 'config.backend.model_type': Value(dtype='string', id=None), 'config.backend.model': Value(dtype='string', id=None), 'config.backend.processor': Value(dtype='string', id=None), 'config.backend.device': Value(dtype='string', id=None), 'config.backend.device_ids': Value(dtype='int64', id=None), 'config.backend.seed': Value(dtype='int64', id=None), 'config.backend.inter_op_num_threads': Value(dtype='float64', id=None), 'config.backend.intra_op_num_threads': Value(dtype='float64', id=None), 'config.backend.model_kwargs.trust_remote_code': Value(dtype='bool', id=None), 'config.backend.no_weights': Value(dtype='bool', id=None), 'config.backend.export': Value(dtype='bool', id=None), 'config.backend.use_cache': Value(dtype='bool', id=None), 'config.backend.use_merged': Value(dtype='bool', id=None), 'config.backend.torch_dtype': Value(dtype='string', id=None), 'config.backend.provider': Value(dtype='string', id=None), 'config.backend.use_io_binding': Value(dtype='bool', id=None), 'config.backend.auto_optimization': Value(dtype='float64', id=None), 'config.backend.auto_quantization': Value(dtype='float64', id=None), 'config.backend.auto_calibration': Value(dt
              ...
              ', id=None), 'report.decode.energy.unit': Value(dtype='string', id=None), 'report.decode.energy.cpu': Value(dtype='float64', id=None), 'report.decode.energy.ram': Value(dtype='float64', id=None), 'report.decode.energy.gpu': Value(dtype='float64', id=None), 'report.decode.energy.total': Value(dtype='float64', id=None), 'report.decode.efficiency.unit': Value(dtype='string', id=None), 'report.decode.efficiency.value': Value(dtype='float64', id=None), 'report.per_token.memory': Value(dtype='float64', id=None), 'report.per_token.latency.unit': Value(dtype='string', id=None), 'report.per_token.latency.values': Value(dtype='string', id=None), 'report.per_token.latency.count': Value(dtype='float64', id=None), 'report.per_token.latency.total': Value(dtype='float64', id=None), 'report.per_token.latency.mean': Value(dtype='float64', id=None), 'report.per_token.latency.p50': Value(dtype='float64', id=None), 'report.per_token.latency.p90': Value(dtype='float64', id=None), 'report.per_token.latency.p95': Value(dtype='float64', id=None), 'report.per_token.latency.p99': Value(dtype='float64', id=None), 'report.per_token.latency.stdev': Value(dtype='float64', id=None), 'report.per_token.latency.stdev_': Value(dtype='float64', id=None), 'report.per_token.throughput.unit': Value(dtype='string', id=None), 'report.per_token.throughput.value': Value(dtype='float64', id=None), 'report.per_token.energy': Value(dtype='float64', id=None), 'report.per_token.efficiency': Value(dtype='float64', id=None)}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1417, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1049, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 924, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1000, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1741, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 2 new columns ({'config.backend.half', 'config.backend.reshape'}) and 95 missing columns ({'report.decode.energy.total', 'report.decode.latency.unit', 'report.load.memory.max_process_vram', 'config.backend.auto_calibration', 'report.prefill.memory.max_reserved', 'report.per_token.latency.p95', 'report.load.energy', 'report.decode.efficiency.unit', 'report.per_token.energy', 'report.load.memory.max_global_vram', 'report.prefill.latency.p95', 'report.prefill.latency.stdev', 'report.prefill.throughput.value', 'report.decode.latency.values', 'report.decode.energy.ram', 'report.load.latency.count', 'report.decode.latency.p95', 'report.prefill.latency.stdev_', 'report.load.latency.values', 'report.load.latency.p99', 'report.per_token.latency.stdev_', 'report.decode.latency.count', 'report.load.latency.unit', 'report.decode.memory.unit', 'report.decode.latency.stdev', 'report.decode.throughput.value', 'report.per_token.latency.count', 'report.load.latency.p95', 'report.load.latency.p90', 'report.decode.energy.cpu', 'report.per_token.throughput.value', 'report.prefill.energy.unit', 'report.per_token.latency.p50', 'report.decode.energy.gpu', 'report.per_token.efficiency', 'report.load.efficiency', 'report.per_token.latency.p90', 'report.prefill.latency.unit', 'report.load.throughput', 'report.per_token.latency.p99', 'report.decode.memory.max_reserved', 'report.load.latency.stdev_', 'report.prefill.latency.p99', 'report.decode.latency.stdev_', 'report.load.latency.mean', 'report.load.memory.max_allocated', 'report.decode.energy.unit', 'report.prefill.energy.gpu', 'report.pr
              ...
              'report.per_token.memory', 'report.load.latency.stdev', 'config.backend.optimization', 'report.load.memory.max_reserved', 'report.prefill.memory.max_ram', 'report.prefill.latency.values', 'report.per_token.throughput.unit', 'report.prefill.memory.unit', 'report.prefill.energy.total', 'report.prefill.latency.p50', 'report.prefill.efficiency.value', 'report.per_token.latency.total', 'report.prefill.memory.max_process_vram', 'report.prefill.latency.mean', 'report.prefill.memory.max_allocated', 'report.prefill.latency.total', 'report.per_token.latency.unit', 'report.decode.memory.max_global_vram', 'config.backend.use_io_binding', 'report.prefill.energy.ram', 'report.decode.latency.mean', 'report.load.latency.total', 'report.decode.throughput.unit', 'report.decode.latency.p50', 'report.decode.latency.p99', 'report.prefill.memory.max_global_vram', 'report.load.memory.max_ram', 'report.decode.latency.p90', 'report.prefill.latency.p90', 'report.decode.latency.total', 'config.backend.torch_dtype', 'config.backend.auto_quantization', 'report.prefill.throughput.unit', 'report.per_token.latency.stdev', 'report.load.latency.p50', 'report.load.memory.unit', 'report.decode.efficiency.value', 'report.prefill.efficiency.unit', 'report.decode.memory.max_ram', 'config.backend.auto_optimization', 'report.per_token.latency.mean', 'config.backend.provider', 'report.decode.memory.max_allocated', 'report.decode.memory.max_process_vram', 'report.prefill.energy.cpu', 'report.per_token.latency.values'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/optimum-benchmark/llm-perf-leaderboard/data/perf-df-openvino-cpu-unquantized-32vCPU-C7i.csv (at revision 94a9713e1842c87029a1dcc829e0003533a6d275)
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)