Added missing allocate_kv_cache() call in CausalLM class#1431

regisss merged 2 commits into

v1.14-releasehuggingface/optimum-habana:v1.14-releasefrom

gemma_fixhuggingface/optimum-habana:gemma_fixCopy head branch name to clipboard