🐛 Describe the bug
Hi,
in my use case, I have to wrap LLaMAv2 with an additional class.
There are two issues when I try to apply HybridParallelPlugin on my customized model:
HybridParallelPlugin can not assign the customized policy for ShardFormer, which makes HybridParallePlugin incompatible with the customized model.
- Policy
LlamaPolicy lacks the attribute self_attn.num_key_value_heads, which is a new property for LLaMAv2.
Environment
No response
🐛 Describe the bug
Hi,
in my use case, I have to wrap LLaMAv2 with an additional class.
There are two issues when I try to apply
HybridParallelPluginon my customized model:HybridParallelPlugincan not assign the customized policy for ShardFormer, which makesHybridParallePluginincompatible with the customized model.LlamaPolicylacks the attributeself_attn.num_key_value_heads, which is a new property for LLaMAv2.Environment
No response