[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
@@ -463,8 +463,7 @@ class VocabParallelEmbedding(PluggableLayer):
output_dim = getattr(param, "output_dim", None)
packed_dim = getattr(param, "packed_dim", None)
# If parameter does not have output dim, then it should
# be copied onto all gpus (e.g. g_idx for act_order gptq).
# Parameters without an output dimension are copied onto all GPUs.
if output_dim is None:
if (
loaded_weight.ndim == 0