MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT-LLM / dim

Method dim

triton_kernels/tensor.py:160–161  ·  view source on GitHub ↗
(self)

Source from the content-addressed store, hash-verified

158 return t.data if isinstance(t, Storage) else t
159
160 def dim(self):
161 return self.ndim
162
163 def size(self, i=None):
164 if i is None:

Callers 15

forwardMethod · 0.80
_test_additional_model_outputs_integrationFunction · 0.80
test_quantization_dequantization_per_tensorMethod · 0.80
per_token_cast_to_fp8Function · 0.80
per_block_cast_to_fp8Function · 0.80
per_token_cast_to_fp8_e8m0Function · 0.80
per_block_cast_to_fp8_e8m0Function · 0.80
per_rank_test_fused_moe_alltoall_fp8_blockwiseFunction · 0.80
per_custom_dims_cast_to_fp8Function · 0.80
verify_local_weight_sizesFunction · 0.80
test_fp4_batched_quantizeFunction · 0.80
mamba_conv1d_refFunction · 0.80

Calls

no outgoing calls

Tested by 9

_test_additional_model_outputs_integrationFunction · 0.64
test_quantization_dequantization_per_tensorMethod · 0.64
per_rank_test_fused_moe_alltoall_fp8_blockwiseFunction · 0.64
per_custom_dims_cast_to_fp8Function · 0.64
verify_local_weight_sizesFunction · 0.64
test_fp4_batched_quantizeFunction · 0.64
test_llava_onevision_utils_videoFunction · 0.64
test_llava_onevision_utils_imageFunction · 0.64
test_qwen2_vl_utils_compute_mropeFunction · 0.64