[NVIDIA] Support nvfp4 quantization (#12784)

This commit is contained in:
Kaixi Hou
2025-02-12 19:51:51 -08:00
committed by GitHub
parent 9f9704dca6
commit 4fc5c23bb6
12 changed files with 688 additions and 13 deletions
+4
View File
@@ -195,6 +195,10 @@ torch::Tensor gptq_gemm(torch::Tensor a, torch::Tensor b_q_weight,
void gptq_shuffle(torch::Tensor q_weight, torch::Tensor q_perm, int64_t bit);
void scaled_fp4_quant(torch::Tensor& output, torch::Tensor const& input,
torch::Tensor& output_scale,
torch::Tensor const& input_scale);
void static_scaled_fp8_quant(torch::Tensor& out, torch::Tensor const& input,
torch::Tensor const& scale);