add calc_server

[YYYYimo]

Calc server that orchestrates end-to-end runs for the Calc‑X example.
It starts an AgentLightning server, loads the dataset (prefers examples/calc_x/data/*.parquet, falls back to data.jsonl or a small demo), spawns a Trainer worker process running CalcAgent, queues all tasks, polls for completed rollouts, writes rollouts to rollouts.jsonl, and performs cleanup.

[microsoft-github-policy-service]

@YYYYimo please read the following Contributor License Agreement(CLA). If you agree with the CLA, please reply with the following information.

@microsoft-github-policy-service agree [company="{your company}"]

Options:

• (default - no company specified) I have sole ownership of intellectual property rights to my Submissions and I am not making Submissions in the course of work for my employer.

@microsoft-github-policy-service agree

• (when company given) I am making Submissions in the course of work for my employer (or my employer has intellectual property rights in my Submissions by contract or applicable law). I have permission from my employer to make Submissions and enter into this Agreement on behalf of my employer. By signing below, the defined term “You” includes me and my employer.

@microsoft-github-policy-service agree company="Microsoft"

Contributor License Agreement

Contribution License Agreement

This Contribution License Agreement (“Agreement”) is agreed to by the party signing below (“You”),
and conveys certain license rights to Microsoft Corporation and its affiliates (“Microsoft”) for Your
contributions to Microsoft open source projects. This Agreement is effective as of the latest signature
date below.

1. Definitions.
   “Code” means the computer software code, whether in human-readable or machine-executable form,
   that is delivered by You to Microsoft under this Agreement.
   “Project” means any of the projects owned or managed by Microsoft and offered under a license
   approved by the Open Source Initiative (www.opensource.org).
   “Submit” is the act of uploading, submitting, transmitting, or distributing code or other content to any
   Project, including but not limited to communication on electronic mailing lists, source code control
   systems, and issue tracking systems that are managed by, or on behalf of, the Project for the purpose of
   discussing and improving that Project, but excluding communication that is conspicuously marked or
   otherwise designated in writing by You as “Not a Submission.”
   “Submission” means the Code and any other copyrightable material Submitted by You, including any
   associated comments and documentation.
2. Your Submission. You must agree to the terms of this Agreement before making a Submission to any
   Project. This Agreement covers any and all Submissions that You, now or in the future (except as
   described in Section 4 below), Submit to any Project.
3. Originality of Work. You represent that each of Your Submissions is entirely Your original work.
   Should You wish to Submit materials that are not Your original work, You may Submit them separately
   to the Project if You (a) retain all copyright and license information that was in the materials as You
   received them, (b) in the description accompanying Your Submission, include the phrase “Submission
   containing materials of a third party:” followed by the names of the third party and any licenses or other
   restrictions of which You are aware, and (c) follow any other instructions in the Project’s written
   guidelines concerning Submissions.
4. Your Employer. References to “employer” in this Agreement include Your employer or anyone else
   for whom You are acting in making Your Submission, e.g. as a contractor, vendor, or agent. If Your
   Submission is made in the course of Your work for an employer or Your employer has intellectual
   property rights in Your Submission by contract or applicable law, You must secure permission from Your
   employer to make the Submission before signing this Agreement. In that case, the term “You” in this
   Agreement will refer to You and the employer collectively. If You change employers in the future and
   desire to Submit additional Submissions for the new employer, then You agree to sign a new Agreement
   and secure permission from the new employer before Submitting those Submissions.
5. Licenses.

• Copyright License. You grant Microsoft, and those who receive the Submission directly or
  indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license in the
  Submission to reproduce, prepare derivative works of, publicly display, publicly perform, and distribute
  the Submission and such derivative works, and to sublicense any or all of the foregoing rights to third
  parties.
• Patent License. You grant Microsoft, and those who receive the Submission directly or
  indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license under
  Your patent claims that are necessarily infringed by the Submission or the combination of the
  Submission with the Project to which it was Submitted to make, have made, use, offer to sell, sell and
  import or otherwise dispose of the Submission alone or with the Project.
• Other Rights Reserved. Each party reserves all rights not expressly granted in this Agreement.
  No additional licenses or rights whatsoever (including, without limitation, any implied licenses) are
  granted by implication, exhaustion, estoppel or otherwise.

6. Representations and Warranties. You represent that You are legally entitled to grant the above
   licenses. You represent that each of Your Submissions is entirely Your original work (except as You may
   have disclosed under Section 3). You represent that You have secured permission from Your employer to
   make the Submission in cases where Your Submission is made in the course of Your work for Your
   employer or Your employer has intellectual property rights in Your Submission by contract or applicable
   law. If You are signing this Agreement on behalf of Your employer, You represent and warrant that You
   have the necessary authority to bind the listed employer to the obligations contained in this Agreement.
   You are not expected to provide support for Your Submission, unless You choose to do so. UNLESS
   REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING, AND EXCEPT FOR THE WARRANTIES
   EXPRESSLY STATED IN SECTIONS 3, 4, AND 6, THE SUBMISSION PROVIDED UNDER THIS AGREEMENT IS
   PROVIDED WITHOUT WARRANTY OF ANY KIND, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTY OF
   NONINFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
7. Notice to Microsoft. You agree to notify Microsoft in writing of any facts or circumstances of which
   You later become aware that would make Your representations in this Agreement inaccurate in any
   respect.
8. Information about Submissions. You agree that contributions to Projects and information about
   contributions may be maintained indefinitely and disclosed publicly, including Your name and other
   information that You submit with Your Submission.
9. Governing Law/Jurisdiction. This Agreement is governed by the laws of the State of Washington, and
   the parties consent to exclusive jurisdiction and venue in the federal courts sitting in King County,
   Washington, unless no federal subject matter jurisdiction exists, in which case the parties consent to
   exclusive jurisdiction and venue in the Superior Court of King County, Washington. The parties waive all
   defenses of lack of personal jurisdiction and forum non-conveniens.
10. Entire Agreement/Assignment. This Agreement is the entire agreement between the parties, and
    supersedes any and all prior agreements, understandings or communications, written or oral, between
    the parties relating to the subject matter hereof. This Agreement may be assigned by Microsoft.

[Copilot]

Pull Request Overview

This PR adds a calc_server.py module that orchestrates end-to-end training runs for the Calc-X example. The server manages the complete workflow from data loading through result collection.

Key changes:

• Implements a comprehensive orchestration server for the Calc-X example
• Adds flexible dataset loading with multiple fallback options (parquet → jsonl → demo data)
• Provides task queuing, worker process management, and rollout collection functionality

Comments suppressed due to low confidence (1)

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The trainer process is not properly cleaned up if the main process exits unexpectedly. Consider using a context manager or try/finally block to ensure process cleanup.

[Copilot]

The fallback to rollout.__dict__ on line 116 may not work correctly if rollout is a Pydantic model, as Pydantic models store data differently. Consider using rollout.dict() as an alternative fallback for older Pydantic versions.

Suggested change

	rdict = rollout.__dict__
	# fallback: use dict() for Pydantic v1
	rdict = rollout.dict()

[Copilot]

After proc.join(timeout=5), there's no check if the process actually terminated. If the timeout expires, the process may still be running. Consider adding a force kill if the process doesn't terminate gracefully.

Suggested change

	if proc.is_alive():
	print("[server] WARNING: trainer process did not terminate after join; force killing")
	proc.terminate()
	proc.join(timeout=1)

[ultmaster]

You don't need to put server and client in the same script.

[YYYYimo]

Thanks, I’ll update it.

[YYYYimo]

Problem

After enabling instrument_vllm() in run_vllm_instrument_vllm.py, the vLLM server crashes when processing streaming chat requests. Although the client receives an HTTP/1.1 200 OK status, the server log immediately shows an ASGI application exception, which leads to a dropped connection.

Key Error Messages:

1. Server-Side:

   (APIServer pid=4177994) ERROR:    Exception in ASGI application
   ...
   (APIServer pid=4177994)     | AttributeError: 'tuple' object has no attribute 'encode'
   (APIServer pid=4177994)     +------------------------------------
   

   This AttributeError occurs in Starlette's response handling layer, indicating that our application passed it a tuple object when it expected a string that could be encoded into bytes via .encode().

2. Client-Side:

   🖇 AgentOps: [OPENAI WRAPPER] Error in async_chat_completion_stream_wrapper: Connection error.
   Failure: Connection error.
   answer: None ground_truth: 1_000 reward: 0.0
   

   Because the server crashed, the client's connection was unexpectedly closed. This caused AgentOps to report a connection error, ultimately resulting in a reward of 0.0 for the rollout.

Root Cause

The core flaw in the instrument_vllm patch was its incorrect assumption that chat_completion_full_generator would always return a single, complete ChatCompletionResponse object.

• This assumption holds true for non-streaming (stream=False) requests.
• However, for streaming (stream=True) requests, vLLM's behavior is entirely different: it returns an AsyncGenerator or StreamingResponse that yields data chunks. Our patch did not handle this case and incorrectly passed a tuple chunk from the generator directly to Starlette, triggering the fatal AttributeError.

[YYYYimo]

Run vLLM with:

python3 run_vllm_instrument_vllm.py serve Qwen/Qwen2-1.5B-Instruct \
  --tensor-parallel-size 1 \
  --enable-auto-tool-choice \
  --tool-call-parser hermes

[ultmaster]

You can ignore the failure when vLLM stream=True. Addressing that is a separate task that is being processed in parallel.

Could you explain how you were going to run the example and what's expected output of this example?

[YYYYimo]

You can ignore the failure when vLLM stream=True. Addressing that is a separate task that is being processed in parallel.

Could you explain how you were going to run the example and what's expected output of this example?

Yes, to run the calc_x example, you'll need three terminals:

1. Terminal 1: Start the vLLM server
   The command I put above is used to start instrumented vLLM, which will raise error.
   Run the uninstrumented vLLM server. Here I use Qwen/Qwen2-1.5B-Instruct model.

   # In agent-lightning/
   python3 -m vllm.entrypoints.openai.api_server \
        --model Qwen/Qwen2-1.5B-Instruct \
        --tensor-parallel-size 1 \
        --enable-auto-tool-choice \
        --tool-call-parser hermes

2. Terminal 2: Run the Calc server

   # In agent-lightning/
   python examples/calc_x/calc_server.py

3. Terminal 3: Run the Calc agent

   # In agent-lightning/
   python examples/calc_x/calc_agent.py

Expected Output:
The terminal running calc_agent.py will print the progress for each math problem, showing the agent's answer, the ground truth, and the calculated reward (1.0 for correct, 0.0 for incorrect). Simultaneously, detailed logs for each attempt are written to [rollouts.jsonl].