查找技能

Add or verify model support in SGLang's Rust processor so chat requests produce the same prompt text and token IDs as Python. Use reference-generated test cases to find differences in tools, reasoning settings, conversation history, and errors.

中文简介正在准备中。当前显示英语简介。

• Adding model support to sglang-processor
• Porting model rendering from sgl-router
• Rechecking prompt behavior after a Dynamo update

Guide test design for SGLang's scripted runtime by deciding when a test harness API adds real value. Favor direct reads, meaningful outcome checks, and the engine's actual execution loop over unnecessary wrappers.

中文简介正在准备中。当前显示英语简介。

• Decide whether a test harness needs a new API
• Review wrappers around directly accessible test state
• Test timeout behavior through the real engine loop

Review existing tests and CI, the automated checks run during development, for redundant coverage, misplaced tests, and opportunities to reduce runtime or cost. Receive file-specific recommendations based on documented improvement patterns, with uncertain findings separated.

中文简介正在准备中。当前显示英语简介。

• Investigating a slow test suite
• Finding redundant or misplaced tests
• Reviewing CI runtime and cost

Launch an SGLang server, run a quick accuracy check, and capture an execution trace for performance inspection. Return the trace location and generated files, ready to open in Perfetto or Chrome tracing.

中文简介正在准备中。当前显示英语简介。

• Capture an execution trace of SGLang inference
• Check basic accuracy before profiling
• Record input processing and token generation separately

Add a CUDA/C++ GPU operation to sgl-kernel using AOT, meaning compilation before use. Follow the implementation through PyTorch registration, Python API exposure, SGLang integration, correctness tests, and benchmarks that account for cache reuse.

中文简介正在准备中。当前显示英语简介。

• Adding a CUDA/C++ operation to sgl-kernel
• Registering an operation with PyTorch and exposing a Python API
• Testing numerical results and invalid inputs

Investigate SGLang serving failures by collecting server state, preserving failing requests, and replaying them on a clean target. Summarize evidence and likely causes, then choose the next debugging tool for crashes, stalls, or performance regressions.

中文简介正在准备中。当前显示英语简介。

• Investigate failed health checks and growing queues
• Reproduce request-specific timeouts or crashes
• Investigate output changes after deployment

Add a diffusion model or Diffusers generation pipeline to SGLang. Examine reference code, choose a suitable pipeline structure, reuse existing components, and implement model settings and processing stages, with guidance for later multi-GPU support.

中文简介正在准备中。当前显示英语简介。

• Adding a new diffusion model to SGLang
• Porting a Diffusers generation pipeline
• Adding a variant by reusing existing model components

Collect and review SGLang startup logs to identify unnecessary output and trace its source. Get saved logs and specific cleanup proposals while preserving useful diagnostics, with runtime changes left for user approval.

中文简介正在准备中。当前显示英语简介。

• Investigating repeated messages in startup logs
• Checking which startup warnings need attention
• Comparing startup logs across models and GPU counts

Investigate recurring SGLang automated test failures by comparing logs, commit history, and GPU environments. Narrow down code changes or machine-specific problems, optionally reproduce failures remotely, and summarize the evidence in a structured report.

中文简介正在准备中。当前显示英语简介。

• Finding the change behind a recurring test failure
• Checking whether failures depend on a GPU or runner
• Reproducing a test failure on a remote GPU server

Calculate how to divide serving memory between recurrent state and KV cache, two memory pools used by hybrid models. Use startup logs and workload settings to recommend a ratio, estimate concurrency, and identify the limiting pool.

中文简介正在准备中。当前显示英语简介。

• Choosing a memory ratio for a hybrid model
• Investigating limits on concurrent requests
• Recalculating memory allocation for context length

Review SGLang Cookbook pull requests for configuration errors, documentation mismatches, benchmark issues, and broken links. Get findings for each file and an overall recommendation on whether to approve, request changes, or block the contribution.

中文简介正在准备中。当前显示英语简介。

• Reviewing a new model's Cookbook contribution
• Checking documentation against generated launch commands
• Checking benchmark definitions and measurement consistency

Analyze LLM serving traces to identify GPU time usage, opportunities to run operations concurrently, and known patterns for combining operations. Produce three consistent tables from saved traces or live captures on supported servers.

中文简介正在准备中。当前显示英语简介。

• Identify GPU operations that dominate inference time
• Compare input processing and token generation
• Trace GPU operations back to Python code

Investigate CUDA crashes in SGLang using logs and inputs saved before execution. Inspect tensor shapes, data types, and numerical anomalies, with guidance for separate process logs and additional CUDA debugging tools.

中文简介正在准备中。当前显示英语简介。

• Investigating illegal memory access and device-side assertions
• Tracing NaN or Inf values through kernel calls
• Saving inputs before a CUDA crash

Apply consistent names to SGLang speculative decoding code, which verifies draft tokens to generate text. Distinguish tokens with or without the bonus token, counts, counters, and arrays when adding identifiers, renaming code, or reviewing changes.

中文简介正在准备中。当前显示英语简介。

• Naming variables in speculative decoding code
• Renaming identifiers to match their meaning
• Reviewing metric and counter names

Prepare batches of merged SGLang pull requests for a release branch. It checks conflicts and previous attempts, starts the cherry-pick workflow, monitors completion, and reports created pull requests or failure reasons.

中文简介正在准备中。当前显示英语简介。

• Prepare multiple merged PRs for a release branch
• Check conflicts before starting workflows
• Detect previous cherry-pick PRs

Develop lightweight CUDA operations for SGLang using JIT, which compiles code on first use. A tensor-scaling example explains implementation conventions, input validation, memory access helpers, and when to choose this approach.

中文简介正在准备中。当前显示英语简介。

• Add a lightweight CUDA operation to SGLang
• Choose between JIT and ahead-of-time compilation
• Validate tensor inputs in a kernel launcher

Write and register SGLang unit and end-to-end tests, from choosing models and file locations to reusing server fixtures. Follow project rules for meaningful assertions, CI registration, dependency mocking, and safe server cleanup after tests.

中文简介正在准备中。当前显示英语简介。

• Adding CPU-only unit tests for component logic
• Testing features through a running SGLang server
• Registering tests in an appropriate CI suite

Integrate or upgrade cache-dit in SGLang diffusion, the image generation system. Follow guidance for resolving code conflicts, adding cache settings, testing feature combinations, and comparing generated images using quality measurements and logs.

中文简介正在准备中。当前显示英语简介。

• Upgrading cache-dit in SGLang diffusion
• Resolving conflicts when porting upstream changes
• Adding per-request cache settings

Verify file splits, function moves, module extractions, and renames by reproducing changes from a base commit and comparing files byte for byte. Generate repeatable proof scripts and reports that expose changes beyond relocation.

中文简介正在准备中。当前显示英语简介。

• Splitting a refactor into preparation, movement, and follow-up commits
• Generating repeatable proof scripts for code moves
• Verifying every provable commit in a branch

Quantize diffusion transformers with NVIDIA ModelOpt, using FP8 or NVFP4 lower-precision numbers, and prepare the saved weights for SGLang Diffusion. Follow a workflow for loading, checking output quality, and comparing speed under matching conditions.

中文简介正在准备中。当前显示英语简介。

• Preparing ModelOpt FP8 weights for SGLang Diffusion
• Checking quality after diffusion model quantization
• Comparing BF16 and quantized generation speed

Monitor selected GitHub Actions checks on an SGLang pull request until they pass on its latest commit. Diagnose failures, push small validated fixes, and retry only failed jobs within defined limits.

中文简介正在准备中。当前显示英语简介。

• Monitor selected PR checks until they pass
• Separate code failures from infrastructure problems
• Apply small fixes and track the resulting checks

Guide changes and reviews for SGLang’s Scheduler, TokenizerManager, and ModelRunner classes. Keep initialization in small, replaceable helpers and separate computation from coordination so large classes retain clear responsibilities and explicit dependencies.

中文简介正在准备中。当前显示英语简介。

• Reviewing changes to SGLang’s three large classes
• Splitting initialization into replaceable helpers
• Moving computation out of ModelRunner

Understand and modify SGLang's CI pipelines, including test stage order, start conditions, early stopping after failures, and workload splitting. Use the guide to add jobs and investigate failed, skipped, or cancelled checks.

中文简介正在准备中。当前显示英语简介。

• Add a test job to an SGLang CI stage
• Investigate failed or skipped CI checks
• Update test partitions and waiting job counts

Investigate SGLang runs that freeze across multiple GPUs. It uses stack traces, communication logs, GPU memory dumps, and comparisons between processes to locate state differences, trace their causes, and guide fixes.

中文简介正在准备中。当前显示英语简介。

• Locate where a multi-GPU inference run freezes
• Compare logs from distributed processes
• Trace the first difference in process state