[Model] Deepseek GGUF support (#13167)

This commit is contained in:
Szymon Ożóg
2025-02-27 02:08:35 -08:00
committed by GitHub
parent edf309ebbe
commit 7f0be2aa24
8 changed files with 198 additions and 10 deletions
@@ -496,7 +496,6 @@ def gguf_quant_weights_iterator(
weight = tensor.data
weight_type = tensor.tensor_type
name = gguf_to_hf_name_map[tensor.name]
if weight_type.name != "F32":
name = name.replace("weight", "qweight")
param = torch.tensor(weight)