Skip to content

update Trie APIs #2417

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

Open
wants to merge 6 commits into
base: main
Choose a base branch
from
Open

update Trie APIs #2417

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
7a70202
update Trie APIs
lisaliu1 Oct 3, 2025
bbc3d39
address reviews
lisaliu1 Oct 3, 2025
155bce0
add trie test to accuracy test
lisaliu1 Oct 3, 2025
27f759b
Merge remote-tracking branch 'origin/main' into lisal.trie_api
lisaliu1 Oct 3, 2025
dffc15f
fix accuracy test
lisaliu1 Oct 3, 2025
f299377
Merge remote-tracking branch 'origin/main' into lisal.trie_api
lisaliu1 Oct 4, 2025
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
28 changes: 0 additions & 28 deletions app_tests/integration_tests/llm/shortfin/tinystories_llama2_25m_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -102,19 +102,12 @@ class TestLLMServer:
def test_basic_generation(
self,
server: tuple[Any, int, ServerConfig],
request: pytest.FixtureRequest,
) -> None:
"""Tests basic text generation capabilities.

Args:
server: Tuple of (process, port) from server fixture
"""
test_id = request.node.callspec.id
if "trie" in test_id:
pytest.skip(
reason="TrieAttentionCache APIs are under development, skip it for now."
)

process, port, config = server
assert process.poll() is None, "Server process terminated unexpectedly"
dataset = (
Expand Down Expand Up @@ -145,19 +138,12 @@ def test_basic_generation(
def test_multi_page_generation(
self,
server: tuple[Any, int, ServerConfig],
request: pytest.FixtureRequest,
) -> None:
"""Tests multi-page text generation capabilities.

Args:
server: Tuple of (process, port) from server fixture
"""
test_id = request.node.callspec.id
if "trie" in test_id:
pytest.skip(
reason="TrieAttentionCache APIs are under development, skip it for now."
)

process, port, config = server
assert process.poll() is None, "Server process terminated unexpectedly"
dataset = (
Expand Down Expand Up @@ -209,10 +195,6 @@ def test_concurrent_generation(
pytest.skip(
reason="Known issue with chunked prefill in batch case: https://github.com/nod-ai/shark-ai/issues/2235"
)
if "trie" in test_id:
pytest.skip(
reason="TrieAttentionCache APIs are under development, skip it for now."
)

process, port, config = server
assert process.poll() is None, "Server process terminated unexpectedly"
Expand Down Expand Up @@ -257,18 +239,12 @@ def _generate_task(prompt: str, port: int):
def test_single_greedy_switch(
self,
server: tuple[Any, int, ServerConfig],
request: pytest.FixtureRequest,
):
"""Tests switching to single-beam greedy generation.

Args:
server: Tuple of (process, port, config) from server fixture
"""
test_id = request.node.callspec.id
if "trie" in test_id:
pytest.skip(
reason="TrieAttentionCache APIs are under development, skip it for now."
)
process, port, _ = server
assert process.poll() is None, "Server process terminated unexpectedly"

Expand Down Expand Up @@ -316,10 +292,6 @@ def test_beam_search_switch(
pytest.skip(
"Beam search with 2 beams isn't compatible with logits returned by GPU argmax model."
)
if "trie" in test_id:
pytest.skip(
reason="TrieAttentionCache APIs are under development, skip it for now."
)

process, port, _ = server
assert process.poll() is None, "Server process terminated unexpectedly"
Expand Down
Loading
Loading