"examples/offline_inference/vision_language.py" did not exist on "e24113a8fe5de5b96459d1f8509d1b48fd7ceebe"
Unverified Commit ccc00515 authored by Zhonghua Deng's avatar Zhonghua Deng Committed by GitHub
Browse files

[BugFix] Illegal memory access for MoE On H20 (#13693)

parent 781096e3
......@@ -1271,7 +1271,7 @@ def fused_experts_impl(hidden_states: torch.Tensor,
# so the cache size and config are already set correctly and
# do not need to be adjusted.
intermediate_cache1 = intermediate_cache1[:tokens_in_chunk]
intermediate_cache2 = intermediate_cache2[:tokens_in_chunk]
intermediate_cache2 = intermediate_cache2[:tokens_in_chunk * topk_ids.shape[1]]
intermediate_cache3 = intermediate_cache3[:tokens_in_chunk]
config = get_config_func(tokens_in_chunk)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment