Skip to content

fix(examples): update strands_sglang example to v0.3.x API #1684

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
yitianlian merged 1 commit into from
Mar 8, 2026
Merged

fix(examples): update strands_sglang example to v0.3.x API #1684

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
20 changes: 5 additions & 15 deletions examples/strands_sglang/generate_with_strands.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,8 +1,9 @@
# Updated with strands-sglang 0.3.2
import logging

from camel.interpreters import SubprocessInterpreter
from strands import Agent, tool
from strands_sglang import SGLangClient, SGLangModel, ToolLimiter
from strands_sglang import SGLangModel, ToolLimiter, get_client_from_slime_args
from strands_sglang.tool_parsers import HermesToolParser

from slime.rollout.rm_hub.math_dapo_utils import compute_score as math_dapo_compute_score
Expand All @@ -23,17 +24,6 @@
MAX_TOOL_ITERS = 5
MAX_TOOL_CALLS = None # No limit

_client_cache: dict[str, SGLangClient] = {}


def get_client(args) -> SGLangClient:
"""Get shared client for connection pooling."""
base_url = f"http://{args.sglang_router_ip}:{args.sglang_router_port}"
if base_url not in _client_cache:
_client_cache[base_url] = SGLangClient.from_slime_args(args, timeout=300.0)
return _client_cache[base_url]


@tool
def execute_python_code(code: str) -> str:
"""Execute Python code and return the output."""
Expand All @@ -55,7 +45,7 @@ async def generate(args, sample: Sample, sampling_params) -> Sample:
state = GenerateState(args)
model = SGLangModel(
tokenizer=state.tokenizer,
client=get_client(args),
client=get_client_from_slime_args(args, timeout=300.0),
tool_parser=HermesToolParser(), # tool parsing for wrapped JSON tool calls
sampling_params=sampling_params,
)
Expand All @@ -69,14 +59,14 @@ async def generate(args, sample: Sample, sampling_params) -> Sample:
system_prompt=SYSTEM_PROMPT,
)

# remember don't set --apply-chat-template in rollout args, it will make user prompt wrapped twice
prompt = sample.prompt if isinstance(sample.prompt, str) else sample.prompt[0]["content"]

try:
await agent.invoke_async(prompt)
sample.status = Sample.Status.COMPLETED
except Exception as e:
# Always use TRUNCATED instead of ABORTED because slime doesn't properly
# handle ABORTED samples in reward processing. See: https://github.com/THUDM/slime/issues/200
# Default all failed rollouts to TRUNCATED; cutomize your logic here if needed
sample.status = Sample.Status.TRUNCATED
logger.warning(f"TRUNCATED: {type(e).__name__}: {e}")

Expand Down
Loading