Hugging Face's logo

Datasets:
hongguoguo
/
Low-bit-Cutlass-GEMM

License:
Dataset card Data Studio Files
xet
 Community
Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code:   DatasetGenerationError
Exception:    ArrowInvalid
Message:      Float value 0.00304096 was truncated converting to int64
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1871, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 643, 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 2293, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2246, in cast_table_to_schema
                  arrays = [
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2247, in <listcomp>
                  cast_array_to_feature(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 1796, in wrapper
                  return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 1796, in <listcomp>
                  return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2103, in cast_array_to_feature
                  return array_cast(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 1798, in wrapper
                  return func(array, *args, **kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 1950, in array_cast
                  return array.cast(pa_type)
                File "pyarrow/array.pxi", line 996, in pyarrow.lib.Array.cast
                File "/src/services/worker/.venv/lib/python3.9/site-packages/pyarrow/compute.py", line 404, in cast
                  return call_function("cast", [arr], options, memory_pool)
                File "pyarrow/_compute.pyx", line 590, in pyarrow._compute.call_function
                File "pyarrow/_compute.pyx", line 385, in pyarrow._compute.Function.call
                File "pyarrow/error.pxi", line 154, in pyarrow.lib.pyarrow_internal_check_status
                File "pyarrow/error.pxi", line 91, in pyarrow.lib.check_status
              pyarrow.lib.ArrowInvalid: Float value 0.00304096 was truncated converting to int64
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1436, 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 1053, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 925, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1001, 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 1742, 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 1898, in _prepare_split_single
                  raise DatasetGenerationError("An error occurred while generating the dataset") from e
              datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

M
int64
N
int64
K
int64
number of SM
int64
first shape and stage
int64
second shape and stage
int64
third shape and stage
int64
fourth shape and stage
int64
fifth shape and stage
int64
sixth shape and stage
int64
seventh shape and stage
int64
eighth shape and stage
int64
ninth shape and stage
int64
64
64
256
48
9
8
7
5
6
4
2
3
1
224
224
256
48
8
9
7
5
6
4
2
3
1
384
384
256
48
9
7
8
4
5
6
2
3
1
544
544
256
48
7
8
9
4
5
6
1
2
3
704
704
256
48
7
8
9
4
5
6
1
2
3
864
864
256
48
8
9
5
4
7
2
3
1
6
1,024
1,024
256
48
9
7
8
5
3
2
1
4
6
1,184
1,184
256
48
7
8
5
9
4
3
2
6
1
1,344
1,344
256
48
8
9
4
5
7
6
1
3
2
1,504
1,504
256
48
7
9
5
4
6
1
2
8
3
1,664
1,664
256
48
7
9
4
5
6
1
2
3
8
1,824
1,824
256
48
7
8
9
4
5
6
1
3
2
1,984
1,984
256
48
7
9
8
5
4
6
2
3
1
2,144
2,144
256
48
5
4
6
9
8
7
3
2
1
2,304
2,304
256
48
8
7
9
5
4
6
1
3
2
2,464
2,464
256
48
7
9
8
5
4
6
2
3
1
2,624
2,624
256
48
7
9
8
5
4
6
1
3
2
2,784
2,784
256
48
9
8
7
5
4
6
2
3
1
2,944
2,944
256
48
5
4
6
8
7
9
1
3
2
3,104
3,104
256
48
8
9
5
7
4
6
2
3
1
3,264
3,264
256
48
9
7
8
5
6
1
4
2
3
3,424
3,424
256
48
4
5
9
8
6
7
1
2
3
3,584
3,584
256
48
7
4
5
9
8
6
2
3
1
3,744
3,744
256
48
9
4
5
8
7
6
1
3
2
3,904
3,904
256
48
8
4
5
9
7
6
2
3
1
4,064
4,064
256
48
7
4
5
9
8
6
2
3
1
4,224
4,224
256
48
4
5
8
9
6
7
2
3
1
4,384
4,384
256
48
4
5
7
9
8
6
3
2
1
4,544
4,544
256
48
5
9
7
4
6
8
2
3
1
4,704
4,704
256
48
5
9
6
7
4
8
2
3
1
4,864
4,864
256
48
7
6
4
5
9
8
2
3
1
5,024
5,024
256
48
7
4
5
6
9
8
2
3
1
5,184
5,184
256
48
7
9
4
5
6
8
2
3
1
5,344
5,344
256
48
7
8
6
5
9
4
2
3
1
5,504
5,504
256
48
9
6
5
4
8
7
2
3
1
5,664
5,664
256
48
8
7
9
4
5
6
2
3
1
5,824
5,824
256
48
7
8
9
5
4
6
3
2
1
5,984
5,984
256
48
9
8
7
5
6
4
2
3
1
6,144
6,144
256
48
8
9
7
4
6
5
3
2
1
6,304
6,304
256
48
9
8
7
5
6
4
3
2
1
6,464
6,464
256
48
7
8
9
4
5
6
3
2
1
6,624
6,624
256
48
9
7
5
8
6
2
3
4
1
6,784
6,784
256
48
7
8
9
5
4
6
2
3
1
6,944
6,944
256
48
9
8
7
4
5
6
2
3
1
7,104
7,104
256
48
9
8
7
5
4
6
2
3
1
7,264
7,264
256
48
9
7
8
4
5
6
2
3
1
7,424
7,424
256
48
9
7
3
8
4
5
2
6
1
7,584
7,584
256
48
7
9
8
5
4
2
3
6
1
7,744
7,744
256
48
8
9
7
4
5
6
3
2
1
7,904
7,904
256
48
9
8
7
5
4
3
2
6
1
8,064
8,064
256
48
9
8
7
5
4
2
3
6
1
8,224
8,224
256
48
9
7
8
4
5
3
2
6
1
8,384
8,384
256
48
9
8
7
2
3
4
5
6
1
8,544
8,544
256
48
7
9
8
3
2
4
5
6
1
8,704
8,704
256
48
8
7
9
2
3
4
5
6
1
8,864
8,864
256
48
9
8
7
2
3
4
5
6
1
9,024
9,024
256
48
7
8
9
2
3
5
4
6
1
9,184
9,184
256
48
8
9
7
3
2
4
5
1
6
9,344
9,344
256
48
7
2
3
8
5
4
6
9
1
9,504
9,504
256
48
9
8
7
2
3
4
5
6
1
9,664
9,664
256
48
9
8
7
2
3
5
4
6
1
9,824
9,824
256
48
9
7
2
3
8
4
1
5
6
9,984
9,984
256
48
9
8
7
3
2
5
4
6
1
10,144
10,144
256
48
9
3
2
8
7
1
5
4
6
10,304
10,304
256
48
9
7
8
2
3
5
1
4
6
10,464
10,464
256
48
9
2
3
8
7
5
1
4
6
10,624
10,624
256
48
9
2
3
4
5
8
6
1
7
10,784
10,784
256
48
9
3
2
8
7
1
4
5
6
10,944
10,944
256
48
9
7
2
3
8
4
5
1
6
11,104
11,104
256
48
7
9
2
3
8
1
5
4
6
11,264
11,264
256
48
9
7
3
2
8
1
4
5
6
11,424
11,424
256
48
7
2
3
9
1
4
5
8
6
11,584
11,584
256
48
7
9
3
2
8
1
4
5
6
11,744
11,744
256
48
1
3
2
7
9
4
8
5
6
11,904
11,904
256
48
9
1
2
3
8
7
5
6
4
12,064
12,064
256
48
1
2
3
9
4
5
8
6
7
12,224
12,224
256
48
9
1
2
3
7
8
5
4
6
12,384
12,384
256
48
1
9
2
3
7
8
4
5
6
12,544
12,544
256
48
1
2
3
9
8
7
4
5
6
12,704
12,704
256
48
1
3
2
9
4
5
6
7
8
12,864
12,864
256
48
1
2
3
9
5
7
8
6
4
64
64
512
48
9
8
7
5
6
4
2
3
1
224
224
512
48
8
9
7
5
6
4
2
3
1
384
384
512
48
7
8
9
4
5
6
1
3
2
544
544
512
48
8
9
7
4
5
6
2
3
1
704
704
512
48
8
9
7
5
4
6
1
2
3
864
864
512
48
8
7
9
4
5
1
3
6
2
1,024
1,024
512
48
9
7
8
4
5
1
3
2
6
1,184
1,184
512
48
7
8
5
9
6
4
3
1
2
1,344
1,344
512
48
7
9
8
4
5
6
3
1
2
1,504
1,504
512
48
9
7
8
5
4
6
3
1
2
1,664
1,664
512
48
9
5
4
6
3
2
1
7
8
1,824
1,824
512
48
9
7
5
4
6
8
2
3
1
1,984
1,984
512
48
8
7
6
5
9
4
3
2
1
2,144
2,144
512
48
7
8
5
6
9
4
2
3
1
2,304
2,304
512
48
9
7
5
4
6
8
3
2
1
2,464
2,464
512
48
9
7
8
5
4
6
3
2
1
2,624
2,624
512
48
7
9
6
5
8
4
3
2
1
2,784
2,784
512
48
9
8
7
5
6
4
3
2
1
2,944
2,944
512
48
8
7
5
6
4
9
3
2
1
End of preview.

This is a dataset for low-bit Cutlass template evaluation, which contains nine different CUTLASS GEMM templates for each low-bit data type on RTX 3080, RTX3090, and A100 NVIDIA Ampere GPUs.

All GEMM time data are in milliseconds.

Directory Original_data: 1 bit: contains matrix sizes MxN from 64x64 to a maximum of 26304x26304, step size 160x160, K from 256 to 12800, step size 256. 4 bit: contains matrix sizes MxN from 64x64 to a maximum of 26304x26304, step size 160x160, K from 64 to 3200, step size 64. 8 bit: contains matrix sizes MxN from 64x64 to a maximum of 26304x26304, step size 160x160, K from 32 to 1600, step size 32.

Directory Original_data_extend: 1 bit: contains matrix sizes MxN from 144x144 to a maximum of 26384x26384, step size 160x160, K from 256 to 12800, step size 256. 4 bit: contains matrix sizes MxN from 144x144 to a maximum of 26384x26384, step size 160x160, K from 64 to 3200, step size 64. 8 bit: contains matrix sizes MxN from 144x144 to a maximum of 26384x26384, step size 160x160, K from 32 to 1600, step size 32.
Downloads last month
56
Total file size:
26.3 MB