phoenix-oss/llama-stack
1
0
Fork 0
forked from phoenix-oss/llama-stack-mirror
Code Pull requests Releases Packages 2 Activity Actions

Fix precommit check after moving to ruff (#927)

Browse source 
Lint check in main branch is failing. This fixes the lint check after we
moved to ruff in https://github.com/meta-llama/llama-stack/pull/921. We
need to move to a `ruff.toml` file as well as fixing and ignoring some
additional checks.

Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>
This commit is contained in:
Yuan Tang 2025-02-02 09:46:45 -05:00 • committed by GitHub
parent 4773092dd1
commit 34ab7a3b6c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
217 changed files with 981 additions and 2681 deletions
Download patch file Download diff file Expand all files Collapse all files

23
llama_stack/providers/inline/post_training/torchtune/datasets/format_adapter.py
View file

@ -15,18 +15,13 @@ from typing import Any, Mapping
from llama_stack.providers.utils.common.data_schema_validator import ColumnName
def llama_stack_instruct_to_torchtune_instruct(
sample: Mapping[str, Any]
) -> Mapping[str, Any]:
assert (
ColumnName.chat_completion_input.value in sample
and ColumnName.expected_answer.value in sample
), "Invalid input row"
def llama_stack_instruct_to_torchtune_instruct(sample: Mapping[str, Any]) -> Mapping[str, Any]:
assert ColumnName.chat_completion_input.value in sample and ColumnName.expected_answer.value in sample, (
"Invalid input row"
)
input_messages = eval(str(sample[ColumnName.chat_completion_input.value]))
assert (
len(input_messages) == 1
), "llama stack intruct dataset format only supports 1 user message"
assert len(input_messages) == 1, "llama stack intruct dataset format only supports 1 user message"
input_message = input_messages[0]
assert "content" in input_message, "content not found in input message"
@ -48,13 +43,9 @@ def llama_stack_chat_to_torchtune_chat(sample: Mapping[str, Any]) -> Mapping[str
roles = []
conversations = []
for message in dialog:
assert (
"role" in message and "content" in message
), "role and content must in message"
assert "role" in message and "content" in message, "role and content must in message"
roles.append(message["role"])
conversations.append(
{"from": role_map[message["role"]], "value": message["content"]}
)
conversations.append({"from": role_map[message["role"]], "value": message["content"]})
assert roles[0] == "user", "first message must be from user"
assert "assistant" in roles, "at least 1 message should be from assistant"

3
llama_stack/providers/inline/post_training/torchtune/datasets/sft.py
View file

@ -61,8 +61,7 @@ class SFTDataset(Dataset):
if not ("tokens" in tokenized_dict and "mask" in tokenized_dict):
keys_str = ", ".join(tokenized_dict.keys())
error_message = (
"model_transform returned the following keys: "
f"{keys_str}. Must return 'tokens' and 'mask' as keys."
f"model_transform returned the following keys: {keys_str}. Must return 'tokens' and 'mask' as keys."
)
raise ValueError(error_message)