Fix Transformers backend FP8 MoE and remove some boilerplate (#46820)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2026-06-27 00:16:05 +01:00
committed by GitHub
parent 2ff76a5e85
commit d8eb734d94
29 changed files with 80 additions and 156 deletions
@@ -195,8 +195,6 @@ class Glm4MoeMTP(nn.Module, Glm4MixtureOfExperts):
vllm_config=vllm_config, prefix=maybe_prefix(prefix, "model")
)
self.expert_weights = []
# Set MoE hyperparameters
self.num_moe_layers = self.config.num_nextn_predict_layers
self.num_expert_groups = self.config.n_group