[Quant][Kernel] Remove GPTQ Group/Dynamic Activation Ordering (#54809)

Signed-off-by: Roderick-Wu <roderickwu2003@gmail.com>
Signed-off-by: Roderick Wu <roderickwu2003@gmail.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: mgoin <mgoin64@gmail.com>
This commit is contained in:
Roderick Wu
2026-09-08 12:45:12 -07:00
committed by GitHub
co-authored by mgoin
parent 1a522b6949
commit 2c9d68f80b
99 changed files with 1182 additions and 3917 deletions
+2 -1
View File
@@ -432,8 +432,9 @@ class AutoWeightsLoader:
modules = (self.module, *self.module.children())
iterator = (m.quant_config for m in modules if hasattr(m, "quant_config"))
if quant_config := next(iterator, None):
# Get mappings and ignore prefixes for KV cache quantization scales
# Apply mappings for quantization-specific checkpoint tensors.
mapper |= quant_config.get_cache_scale_mapper()
mapper |= quant_config.get_checkpoint_weight_mapper()
ignore_unexpected_suffixes = quant_config._ignore_unexpected_suffixes
self.ignore_unexpected_suffixes.extend(ignore_unexpected_suffixes)
mapper |= self.REMOVE_UNUSED_ROTARY_EMBEDS_MAPPER