MCPcopy Index your code
hub / github.com/NVIDIA/TensorRT-LLM / forward

Method forward

tensorrt_llm/quantization/layers.py:82–83  ·  view source on GitHub ↗
(self, x)

Source from the content-addressed store, hash-verified

80 """
81
82 def forward(self, x):
83 return quantize_per_token(x)
84
85
86class Dequantize(Module):

Callers 1

test_quantize_per_tokenMethod · 0.95

Calls 1

quantize_per_tokenFunction · 0.85

Tested by 1

test_quantize_per_tokenMethod · 0.76