[MoE] Add LFM2 MoE tuning support + tuned configs for H100/B200/MI325X (#22791)
Co-authored-by: Piotr Mazurek <piotr.mazurek@liquid.ai>
This commit is contained in:
co-authored by
Piotr Mazurek
parent
0e165ffbfc
commit
6cf0b004ca
@@ -138,6 +138,10 @@ def get_model_config(
|
||||
E = config.num_experts // ep_size
|
||||
topk = config.top_k_experts
|
||||
intermediate_size = config.moe_intermediate_size
|
||||
elif architecture == "Lfm2MoeForCausalLM":
|
||||
E = config.num_experts // ep_size
|
||||
topk = config.num_experts_per_tok
|
||||
intermediate_size = config.moe_intermediate_size
|
||||
else:
|
||||
# Default: Mixtral
|
||||
E = config.num_local_experts // ep_size
|
||||
|
||||
Reference in New Issue
Block a user