[Core] Add MRV2 virtual-batch PCP for MLA (#46570)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> Signed-off-by: Matthew Bonanni <mbonanni@redhat.com> Co-authored-by: Codex <noreply@openai.com> Co-authored-by: OpenAI Codex <codex@openai.com> Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
co-authored by
Codex
OpenAI Codex
Matthew Bonanni
parent
9243e0124e
commit
b6ff8a2f50
@@ -385,7 +385,7 @@ def test_tiering_spec_aligns_row_size():
|
||||
assert spec.num_blocks == cpu_bytes_to_use // alignment
|
||||
|
||||
|
||||
def test_offloading_spec_resolves_prefill_context_parallel_block_sizes():
|
||||
def test_offloading_spec_kv_sharding_ignores_prefill_context_parallel():
|
||||
config = _make_layout_vllm_config(
|
||||
cpu_bytes_to_use=65536,
|
||||
extra_config={"block_size": 64},
|
||||
@@ -394,9 +394,9 @@ def test_offloading_spec_resolves_prefill_context_parallel_block_sizes():
|
||||
|
||||
spec = _create_spec(config, _make_kv_cache_config())
|
||||
|
||||
assert spec.tokens_per_block == (32,)
|
||||
assert spec.tokens_per_hash == 32
|
||||
assert spec.blocks_per_chunk == 2
|
||||
assert spec.tokens_per_block == (16,)
|
||||
assert spec.tokens_per_hash == 16
|
||||
assert spec.blocks_per_chunk == 4
|
||||
|
||||
|
||||
def test_offloading_config_preserves_data_parallel_index():
|
||||
|
||||
Reference in New Issue
Block a user