Dynamic KV cache allocation #6025

Job	Run time
cpp-beam_search_causal_lm-Phi-2	1m 9s
cpp-beam_search_causal_lm-Qwen1_5-7B-Chat	1m 9s
cpp-beam_search_causal_lm-ubuntu (./build/samples/cpp/beam_search_causal_lm/beam_search_causal_lm)	1m 0s
cpp-greedy_causal_lm-Qwen-7B-Chat	1m 9s
cpp-beam_search_causal_lm-ubuntu (python ./samples/python/beam_search_causal_lm/beam_search_causa...	1m 12s
cpp-greedy_causal_lm-windows	48s
cpp-multinomial-greedy_causal_lm-ubuntu	1m 8s
cpp-greedy_causal_lm-redpajama-3b-chat	1m 8s
cpp-beam_search_causal_lm-notus-7b-v1	1m 5s
cpp-Phi-1_5	1m 4s
cpp-prompt_lookup_decoding_lm-ubuntu	56s
cpp-speculative_decoding_lm-ubuntu	59s
cpp-chat_sample-ubuntu	54s
cpp-continuous-batching-macos	1m 0s
cpp-continuous-batching-ubuntu	58s
cpp-continuous-batching-windows	47s
visual_language_chat_sample-ubuntu-internvl2	56s
visual_language_chat_sample-ubuntu-minicpm_v2_6	1m 8s
visual_language_chat_sample-ubuntu-llava_1_5 / visual_language_chat_sample-ubuntu-llava	1m 4s
visual_language_chat_sample-ubuntu-llava_next / visual_language_chat_sample-ubuntu-llava	1m 4s
ci/gha_overall_status_causal_lm	1s
	20m 39s

Provide feedback