From cbfc5a9afb27f2bc77ebe045c2e34a73bce0faf2 Mon Sep 17 00:00:00 2001 From: biondizzle Date: Thu, 7 May 2026 16:34:37 +0000 Subject: [PATCH] Update nvfp4_experts_only to use dequantized BF16 model --- scripts/model_opt_nvfp4_experts_only.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/model_opt_nvfp4_experts_only.py b/scripts/model_opt_nvfp4_experts_only.py index 70d080a..4558251 100644 --- a/scripts/model_opt_nvfp4_experts_only.py +++ b/scripts/model_opt_nvfp4_experts_only.py @@ -30,7 +30,7 @@ import sys import os # ── Config ────────────────────────────────────────────────────────────────── -MODEL = "/root/nvidia-meeting/DeepSeek-V4-Pro" # Mixed-precision source +MODEL = "/root/nvidia-meeting/DeepSeek-V4-Pro-BF16" # Dequantized BF16 (from scripts/dequant_fp8_to_bf16.py) QUANT = "nvfp4_experts_only" TP = 8 CALIB = 256