[3/N] HiSparse: host-resident sparse-MLA decode hot-buffering (#53781)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com> Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com> Signed-off-by: Matthew Bonanni <matthew@neuralmagic.com> Signed-off-by: Nicolò Lucchesi <nicolo.lucchesi@mistral.ai> Co-authored-by: OpenAI Codex <codex@openai.com> Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com> Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com> Co-authored-by: Nicolò Lucchesi <nicolo.lucchesi@mistral.ai> Co-authored-by: OpenAI Codex <noreply@openai.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
This commit is contained in:
co-authored by
OpenAI Codex
Lucas Wilkinson
Lucas Wilkinson
Nicolò Lucchesi
OpenAI Codex
Claude Opus 5
Lucas Wilkinson
parent
120ec4ebd2
commit
d43bb2f37f
@@ -400,6 +400,8 @@ def test_register_mixed_page_sizes_in_one_cache_group(monkeypatch):
|
||||
vllm_config.cache_config = CacheConfig()
|
||||
vllm_config.cache_config.kv_cache_layout = layout.name
|
||||
vllm_config.cache_config.num_gpu_blocks_override = None
|
||||
vllm_config.attention_config.hisparse_config = None
|
||||
vllm_config.kv_transfer_config = None
|
||||
|
||||
pages = [spec.page_size_bytes for spec in specs.values()]
|
||||
num_blocks = 4
|
||||
@@ -458,6 +460,8 @@ def test_register_mixed_page_sizes_odd_block_counts(monkeypatch, rank_blocks):
|
||||
vllm_config.cache_config = CacheConfig()
|
||||
vllm_config.cache_config.kv_cache_layout = layout.name
|
||||
vllm_config.cache_config.num_gpu_blocks_override = None
|
||||
vllm_config.attention_config.hisparse_config = None
|
||||
vllm_config.kv_transfer_config = None
|
||||
|
||||
pages = [spec.page_size_bytes for spec in specs.values()]
|
||||
kv_cache_config = get_kv_cache_config_from_groups(
|
||||
@@ -507,6 +511,8 @@ def test_mixed_page_byte_placement_is_dcp_invariant():
|
||||
vllm_config.cache_config.kv_cache_layout = KVCacheLayout.LBNHC.name
|
||||
vllm_config.cache_config.num_gpu_blocks_override = None
|
||||
vllm_config.parallel_config.decode_context_parallel_size = dcp_size
|
||||
vllm_config.attention_config.hisparse_config = None
|
||||
vllm_config.kv_transfer_config = None
|
||||
|
||||
page_bytes = sum(spec.page_size_bytes for spec in specs.values())
|
||||
config = get_kv_cache_config_from_groups(vllm_config, [group], page_bytes * 4)
|
||||
|
||||
Reference in New Issue
Block a user