Skip to content

[KVCache] Add tree attention with paged cache support #17326

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Sep 3, 2024
Merged

[KVCache] Add tree attention with paged cache support #17326

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion python/tvm/relax/frontend/nn/llm/kv_cache.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@
from tvm.target import Target

from .position_embedding import llama_rope_with_position_map, switch_rope_freq_func
from .tree_attn import tree_attn
from .tree_attn import tree_attn, tree_attn_with_paged_kv_cache


def get_max_num_threads_per_block(target: Target) -> int:
Expand Down Expand Up @@ -257,6 +257,7 @@ def __init__( # pylint: disable=too-many-locals
bb.add_func(_kv_cache_debug_get_kv(num_hidden_layers, num_key_value_heads, head_dim, dtype), "kv_cache_debug_get_kv"),
bb.add_func(_compact_kv_copy(num_key_value_heads, head_dim, dtype, target), "kv_cache_compact_kv_copy"),
bb.add_func(tree_attn(num_key_value_heads, num_attention_heads, head_dim, dtype, rope_scaling, target), "tir_attention_prefill_with_tree_mask"),
bb.add_func(tree_attn_with_paged_kv_cache(num_key_value_heads, num_attention_heads, head_dim, dtype, rope_scaling, target), "tir_attention_prefill_with_tree_mask_with_paged_kv_cache"),
rope_ext_factors,
# fmt: on
# pylint: enable=line-too-long
Expand Down
Loading