Datasets:
GPUMODE
/
backendbench_tests

Dataset card Data Studio Files
xet
 Community
6
operator name
stringclasses
180 values
used in model
stringclasses
155 values
args
stringlengths
19
5.24k
aten.nll_loss_forward.default
TIMM/nasnetalarge
((T([16, 1000], f16), T([16], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/pnasnet5large
((T([16, 1000], f16), T([16], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/LayoutLMForSequenceClassification
((T([16, 2], f16), T([16], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/cait_m36_384
((T([2, 1000], f16), T([2], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/AlbertForQuestionAnswering
((T([2, 512], f16), T([2], i64), None, 1, 512), {})
aten.nll_loss_forward.default
HuggingFace/DistilBertForMaskedLM
((T([2048, 30522], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/MobileBertForMaskedLM
((T([2048, 30522], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/XLNetLMHeadModel
((T([2048, 32000], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/PLBartForCausalLM
((T([2048, 50005], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/BartForConditionalGeneration
((T([2048, 50265], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/DebertaForMaskedLM
((T([2048, 50265], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/MBartForCausalLM
((T([2048, 50265], f16), T([2048], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/MegatronBertForCausalLM
((T([254, 29056], f16), T([254], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/M2M100ForConditionalGeneration
((T([256, 128112], f16), T([256], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/XGLMForCausalLM
((T([256, 256008], f16), T([256], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/convmixer_768_32
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/convnext_base
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/dpn107
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/gluon_senet154
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/gluon_xception65
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/legacy_senet154
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/resnest101e
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/swsl_resnext101_32x16d
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/twins_pcpvt_base
((T([32, 1000], f16), T([32], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/DistilBertForQuestionAnswering
((T([32, 128], f16), T([32], i64), None, 1, 128), {})
aten.nll_loss_forward.default
HuggingFace/MobileBertForQuestionAnswering
((T([32, 128], f16), T([32], i64), None, 1, 128), {})
aten.nll_loss_forward.default
HuggingFace/GPT2ForSequenceClassification
((T([4, 2], f16), T([4], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/DebertaForQuestionAnswering
((T([4, 512], f16), T([4], i64), None, 1, 512), {})
aten.nll_loss_forward.default
HuggingFace/BartForCausalLM
((T([4096, 50265], f16), T([4096], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/RobertaForCausalLM
((T([508, 30522], f16), T([508], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/OPTForCausalLM
((T([508, 50272], f16), T([508], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/ElectraForCausalLM
((T([511, 30522], f16), T([511], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/DistillGPT2
((T([511, 50257], f16), T([511], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/DebertaV2ForMaskedLM
((T([512, 128100], f16), T([512], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/YituTechConvBert
((T([512, 30522], f16), T([512], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/GoogleFnet
((T([512, 32000], f16), T([512], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/CamemBert
((T([512, 32005], f16), T([512], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/PegasusForConditionalGeneration
((T([512, 50265], f16), T([512], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/beit_base_patch16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/crossvit_9_240
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/cspdarknet53
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/deit_base_distilled_patch16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/densenet121
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/dla102
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/ecaresnet101d
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/gmixer_24_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/gmlp_s16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/jx_nest_base
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/mixer_b16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/mixnet_l
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/mobilevit_s
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/pit_b_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/poolformer_m36
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/res2net101_26w_4s
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/sebotnet33ts_256
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/swin_base_patch4_window7_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/tf_mixnet_l
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/tnt_s_patch16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/vit_base_patch16_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
TIMM/volo_d1_224
((T([64, 1000], f16), T([64], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/BertForQuestionAnswering
((T([64, 128], f16), T([64], i64), None, 1, 128), {})
aten.nll_loss_forward.default
HuggingFace/RobertaForQuestionAnswering
((T([64, 128], f16), T([64], i64), None, 1, 128), {})
aten.nll_loss_forward.default
HuggingFace/ElectraForQuestionAnswering
((T([64, 512], f16), T([64], i64), None, 1, 512), {})
aten.nll_loss_forward.default
HuggingFace/MegatronBertForQuestionAnswering
((T([8, 128], f16), T([8], i64), None, 1, 128), {})
aten.nll_loss_forward.default
HuggingFace/Speech2Text2ForCausalLM
((T([8192, 10000], f16), T([8192], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/BertForMaskedLM
((T([8192, 30522], f16), T([8192], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/LayoutLMForMaskedLM
((T([8192, 30522], f16), T([8192], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/BlenderbotSmallForCausalLM
((T([8192, 50265], f16), T([8192], i64), None, 1, -100), {})
aten.nll_loss_forward.default
HuggingFace/BlenderbotSmallForConditionalGeneration
((T([8192, 50265], f16), T([8192], i64), None, 1, -100), {})
aten.nonzero.default
TorchBench/vision_maskrcnn
((T([0], b8),), {})
aten.nonzero.default
TorchBench/vision_maskrcnn
((T([5000], b8),), {})
aten.norm.ScalarOpt_dim
TorchBench/tts_angular
((T([64, 256], f16, stride=(12800, 1)), 2, [1], True), {})
aten.pow.Scalar
HuggingFace/XLNetLMHeadModel
((10000, T([512], f32)), {})
aten.pow.Tensor_Scalar
HuggingFace/BigBird
((T([1, 1024, 3072], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/BigBird
((T([1, 1024, 3072], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/BigBird
((T([1, 1024, 768], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/BigBird
((T([1, 1024, 768], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GPTNeoForCausalLM
((T([1, 128, 8192], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GPTNeoForSequenceClassification
((T([1, 128, 8192], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GPTNeoForCausalLM
((T([1, 128, 8192], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GPTNeoForSequenceClassification
((T([1, 128, 8192], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/DistillGPT2
((T([1, 512, 3072], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GoogleFnet
((T([1, 512, 3072], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/DistillGPT2
((T([1, 512, 3072], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GoogleFnet
((T([1, 512, 3072], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GoogleFnet
((T([1, 512, 768], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/GoogleFnet
((T([1, 512, 768], f16), 3.0), {})
aten.pow.Tensor_Scalar
TorchBench/hf_BigBird
((T([2, 1024, 3072], f16), 2.0), {})
aten.pow.Tensor_Scalar
TorchBench/hf_BigBird
((T([2, 1024, 3072], f16), 3.0), {})
aten.pow.Tensor_Scalar
TorchBench/hf_BigBird
((T([2, 1024, 768], f16), 2.0), {})
aten.pow.Tensor_Scalar
TorchBench/hf_BigBird
((T([2, 1024, 768], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForMaskedLM
((T([2, 512, 128], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForMaskedLM
((T([2, 512, 128], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForMaskedLM
((T([2, 512, 16384], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForQuestionAnswering
((T([2, 512, 16384], f16), 2.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForMaskedLM
((T([2, 512, 16384], f16), 3.0), {})
aten.pow.Tensor_Scalar
HuggingFace/AlbertForQuestionAnswering
((T([2, 512, 16384], f16), 3.0), {})
aten.pow.Tensor_Scalar
TIMM/convnext_base
((T([32, 1, 14, 14], f16), 3), {})
aten.pow.Tensor_Scalar
TIMM/convnext_base
((T([32, 1, 28, 28], f16), 3), {})
aten.pow.Tensor_Scalar
TIMM/convnext_base
((T([32, 1, 56, 56], f16), 3), {})