Code
Hub
Workspaces
Connect
Indexed graphs
Engine
MCP
copy
hub
/
github.com/NVIDIA/TensorRT-LLM
/ abs
Method
abs
tensorrt_llm/functional.py:444–448 ·
view source on GitHub ↗
See functional.abs.
(self)
Source
from the content-addressed store, hash-verified
442
return
max(self, dim, keepdim)
443
444
def
abs(self):
445
''
&
#x27;
446
See functional.abs.
447
''
&
#x27;
448
return
abs(self)
449
450
def
sqrt(self):
451
''
&
#x27;
Callers
15
assert_close
Function · 0.80
compute_actual_scale
Function · 0.80
downcast_to_mxfp_torch
Function · 0.80
_compute_quant_and_scale
Function · 0.80
thread_local_absmax
Function · 0.80
is_close
Function · 0.80
ref.py
File · 0.80
perr
Function · 0.80
mae
Function · 0.80
reshape_softmax
Function · 0.80
test_gpt_attention_IFB
Method · 0.80
test_smooth_quant_rms_norm
Method · 0.80
Calls
no outgoing calls
Tested by
15
assert_close
Function · 0.64
compute_actual_scale
Function · 0.64
test_gpt_attention_IFB
Method · 0.64
test_smooth_quant_rms_norm
Method · 0.64
test_smooth_quant_layer_norm_layer
Method · 0.64
test_quantization_activation_scales
Method · 0.64
test_quantization_weight_scales
Method · 0.64
test_quantization_per_tensor_scales
Method · 0.64
test_quantization_dequantization_activation
Method · 0.64
test_quantization_dequantization_weight
Method · 0.64
test_quantization_dequantization_per_tensor
Method · 0.64
test_smooth_quant_layer_norm
Method · 0.64