Public/automatic
1
0
Fork 0
mirror of https://github.com/vladmandic/automatic synced 2026-09-18 16:54:33 +02:00
Code Issues Packages Projects Releases Wiki Activity

Move post load quant functions to a single function in model_quant

Browse Source
This commit is contained in:
Disty0
2025-04-20 17:04:25 +03:00
parent 707d7483a4
commit 434bb660ce
2 changed files with 14 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/model_quant.py
+13
View File
@@ -508,3 +508,16 @@ def get_dit_args(load_config:dict={}, module:str=None, device_map:bool=False, al
else:
quant_args = {}
return config, quant_args
def do_post_load_quant(sd_model):
from modules import shared
if shared.opts.nncf_compress_weights and not (shared.opts.cuda_compile and shared.opts.cuda_compile_backend == "openvino_fx"):
sd_model = nncf_compress_weights(sd_model)
if shared.opts.optimum_quanto_weights:
sd_model = optimum_quanto_weights(sd_model)
if shared.opts.torchao_quantization and shared.opts.torchao_quantization_mode == 'post':
sd_model = torchao_quantization(sd_model)
if shared.opts.layerwise_quantization:
apply_layerwise(sd_model)
return sd_model