microsoft / microsoft/vattention

ModuleNotFoundError: No module named 'sarathi.model_executor.parallel_utils.pipeline_parallel

Open
#43 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C
Stars
523
Forks
46
Avg merge
9h 23m
Merged PRs (30d)
1

Description

I build sarathi-serve by pip install -e . --extra-index-url https://flashinfer.ai/whl/cu121/torch2.3/.

I run the code

from sarathi import AsyncLLMEngine, SamplingParams
from sarathi.engine.arg_utils import EngineArgs
def get_engine_config_from_args(args, arrival_rate):
    print(f"ratio: {args.ratio}, arrival_period: {args.arrival_period}, arrival_rate: {arrival_rate}")
    scheduling_policy = args.scheduling_policy 
    if args.scheduling_policy == "skip-join":
        scheduling_policy=  "skip-join"
    print(f"scheduling_policy: {args.scheduling_policy}")
    engine_args = {
        "model": args.model_type,
        "tensor_parallel_size" : args.tensor_parallel,
        "enable_chunked_prefill": True,  # Enable chunked prefill
        "enable_prefix_caching" : True,  # Enable prefix caching
        #Xiao: from sarathi-lean/sarathi/benchmark/config/default.yml
        "chunk_size" : 512,
        "low_chunk_size" : 128,
        "high_chunk_size" : 2048,
        "max_model_len" : 131072,  # Xiao Maximum model length
        # chunk_schedule_max_tokens: 4096
        "chunk_schedule_max_tokens" : 131072,
        "chunk_schedule_stages" : 16,
    }
    return engine_args

    eng_args = get_engine_config_from_args(args, arrival_rate)
    engine = AsyncLLMEngine.from_engine_args(
        eng_args
    )

engine.generate(xxx)

my error is

  File "/root/miniconda3/envs/pod/lib/python3.10/site-packages/sarathi/model_executor/models/internlm.py", line 18, in <module>
    from sarathi.model_executor.parallel_utils.pipeline_parallel.mappings import recv, send
ModuleNotFoundError: No module named 'sarathi.model_executor.parallel_utils.pipeline_parallel 

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing the reported pip installation and import failure, then inspect sarathi/model_executor/models/internlm.py and the sarathi/model_executor/parallel_utils/pipeline_parallel/mappings.py import path. Determine whether the installed package contains the referenced module and identify the expected installation or source layout. Done means the reported sarathi import and engine setup no longer raise ModuleNotFoundError.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
backend, machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.