Skip to content

feat(dataset api): (1.6/n) fix all iterrows callsites #1660

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 6 commits into from
Mar 16, 2025
Merged

feat(dataset api): (1.6/n) fix all iterrows callsites #1660

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f117407
fix all iterrows callsites
yanxi0830 Mar 16, 2025
b561cfd
fix all iterrows callsites
yanxi0830 Mar 16, 2025
659f5e8
eval
yanxi0830 Mar 16, 2025
ca5a6a5
fix notebook
yanxi0830 Mar 16, 2025
06f01f9
fix notebook
yanxi0830 Mar 16, 2025
1a07703
fix notebook
yanxi0830 Mar 16, 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
2,186 changes: 1,080 additions & 1,106 deletions docs/notebooks/Alpha_Llama_Stack_Post_Training.ipynb
View file
Open in desktop

Large diffs are not rendered by default.

352 changes: 268 additions & 84 deletions docs/notebooks/Llama_Stack_Benchmark_Evals.ipynb
View file
Open in desktop

Large diffs are not rendered by default.

24 changes: 9 additions & 15 deletions docs/source/references/evals_reference/index.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -114,23 +114,17 @@ pprint(response)
simpleqa_dataset_id = "huggingface::simpleqa"

_ = client.datasets.register(
dataset_id=simpleqa_dataset_id,
provider_id="huggingface",
url={"uri": "https://huggingface.co/datasets/llamastack/simpleqa"},
metadata={
"path": "llamastack/simpleqa",
"split": "train",
},
dataset_schema={
"input_query": {"type": "string"},
"expected_answer": {"type": "string"},
"chat_completion_input": {"type": "chat_completion_input"},
purpose="eval/messages-answer",
source={
"type": "uri",
"uri": "huggingface://datasets/llamastack/simpleqa?split=train",
},
dataset_id=simpleqa_dataset_id,
)

eval_rows = client.datasetio.get_rows_paginated(
eval_rows = client.datasets.iterrows(
dataset_id=simpleqa_dataset_id,
rows_in_page=5,
limit=5,
)
```

Expand All @@ -143,7 +137,7 @@ client.benchmarks.register(

response = client.eval.evaluate_rows(
benchmark_id="meta-reference::simpleqa",
input_rows=eval_rows.rows,
input_rows=eval_rows.data,
scoring_functions=["llm-as-judge::405b-simpleqa"],
benchmark_config={
"eval_candidate": {
Expand Down Expand Up @@ -191,7 +185,7 @@ agent_config = {

response = client.eval.evaluate_rows(
benchmark_id="meta-reference::simpleqa",
input_rows=eval_rows.rows,
input_rows=eval_rows.data,
scoring_functions=["llm-as-judge::405b-simpleqa"],
benchmark_config={
"eval_candidate": {
Expand Down
7 changes: 3 additions & 4 deletions llama_stack/distribution/ui/page/evaluations/native_eval.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -166,11 +166,10 @@ def run_evaluation_3():
eval_candidate = st.session_state["eval_candidate"]

dataset_id = benchmarks[selected_benchmark].dataset_id
rows = llama_stack_api.client.datasetio.iterrows(
rows = llama_stack_api.client.datasets.iterrows(
dataset_id=dataset_id,
rows_in_page=-1,
)
total_rows = len(rows.rows)
total_rows = len(rows.data)
# Add number of examples control
num_rows = st.number_input(
"Number of Examples to Evaluate",
Expand All @@ -195,7 +194,7 @@ def run_evaluation_3():
if st.button("Run Evaluation"):
progress_text = "Running evaluation..."
progress_bar = st.progress(0, text=progress_text)
rows = rows.rows
rows = rows.data
if num_rows < total_rows:
rows = rows[:num_rows]

Expand Down
4 changes: 2 additions & 2 deletions llama_stack/providers/inline/eval/meta_reference/eval.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,11 +92,11 @@ async def run_eval(
validate_dataset_schema(dataset_def.dataset_schema, get_valid_schemas(Api.eval.value))
all_rows = await self.datasetio_api.iterrows(
dataset_id=dataset_id,
rows_in_page=(-1 if benchmark_config.num_examples is None else benchmark_config.num_examples),
limit=(-1 if benchmark_config.num_examples is None else benchmark_config.num_examples),
)
res = await self.evaluate_rows(
benchmark_id=benchmark_id,
input_rows=all_rows.rows,
input_rows=all_rows.data,
scoring_functions=scoring_functions,
benchmark_config=benchmark_config,
)
Expand Down
4 changes: 2 additions & 2 deletions ...a_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -330,11 +330,11 @@ async def _setup_data(
async def fetch_rows(dataset_id: str):
return await self.datasetio_api.iterrows(
dataset_id=dataset_id,
rows_in_page=-1,
limit=-1,
)

all_rows = await fetch_rows(dataset_id)
rows = all_rows.rows
rows = all_rows.data

await validate_input_dataset_schema(
datasets_api=self.datasets_api,
Expand Down
4 changes: 2 additions & 2 deletions llama_stack/providers/inline/scoring/basic/scoring.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -86,10 +86,10 @@ async def score_batch(

all_rows = await self.datasetio_api.iterrows(
dataset_id=dataset_id,
rows_in_page=-1,
limit=-1,
)
res = await self.score(
input_rows=all_rows.rows,
input_rows=all_rows.data,
scoring_functions=scoring_functions,
)
if save_results_dataset:
Expand Down
4 changes: 2 additions & 2 deletions llama_stack/providers/inline/scoring/braintrust/braintrust.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -169,9 +169,9 @@ async def score_batch(

all_rows = await self.datasetio_api.iterrows(
dataset_id=dataset_id,
rows_in_page=-1,
limit=-1,
)
res = await self.score(input_rows=all_rows.rows, scoring_functions=scoring_functions)
res = await self.score(input_rows=all_rows.data, scoring_functions=scoring_functions)
if save_results_dataset:
# TODO: persist and register dataset on to server for reading
# self.datasets_api.register_dataset()
Expand Down
4 changes: 2 additions & 2 deletions llama_stack/providers/inline/scoring/llm_as_judge/scoring.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -74,10 +74,10 @@ async def score_batch(

all_rows = await self.datasetio_api.iterrows(
dataset_id=dataset_id,
rows_in_page=-1,
limit=-1,
)
res = await self.score(
input_rows=all_rows.rows,
input_rows=all_rows.data,
scoring_functions=scoring_functions,
)
if save_results_dataset:
Expand Down
8 changes: 4 additions & 4 deletions tests/integration/eval/test_eval.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,11 +20,11 @@ def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):
response = llama_stack_client.datasets.list()
assert any(x.identifier == "test_dataset_for_eval" for x in response)

rows = llama_stack_client.datasetio.get_rows_paginated(
rows = llama_stack_client.datasets.iterrows(
dataset_id="test_dataset_for_eval",
rows_in_page=3,
limit=3,
)
assert len(rows.rows) == 3
assert len(rows.data) == 3

scoring_functions = [
scoring_fn_id,
Expand All @@ -40,7 +40,7 @@ def test_evaluate_rows(llama_stack_client, text_model_id, scoring_fn_id):

response = llama_stack_client.eval.evaluate_rows(
benchmark_id=benchmark_id,
input_rows=rows.rows,
input_rows=rows.data,
scoring_functions=scoring_functions,
benchmark_config={
"eval_candidate": {
Expand Down