Skip to content
Permalink

Comparing changes

Choose two branches to see what’s changed or to start a new pull request. If you need to, you can also or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: inference-sim/InferSim
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: main
Choose a base ref
...
head repository: alibaba/InferSim
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: main
Choose a head ref
Checking mergeability… Don’t worry, you can still create the pull request.
  • 6 commits
  • 11 files changed
  • 2 contributors

Commits on Mar 2, 2026

  1. Configuration menu
    Copy the full SHA
    fbc5098 View commit details
    Browse the repository at this point in the history
  2. update readme

    MtFitzRoy committed Mar 2, 2026
    Configuration menu
    Copy the full SHA
    76e74c3 View commit details
    Browse the repository at this point in the history
  3. Support qwen3.5

    MtFitzRoy committed Mar 2, 2026
    Configuration menu
    Copy the full SHA
    7ed3db7 View commit details
    Browse the repository at this point in the history
  4. Add mfu log

    MtFitzRoy committed Mar 2, 2026
    Configuration menu
    Copy the full SHA
    3e99354 View commit details
    Browse the repository at this point in the history
  5. Update readme

    MtFitzRoy committed Mar 2, 2026
    Configuration menu
    Copy the full SHA
    0d2b567 View commit details
    Browse the repository at this point in the history

Commits on Mar 18, 2026

  1. Adapt qwen3.5-35B-A3B-FP8 (alibaba#5)

    * Update benchmark data and kernel benchmarks
    
    * Update benchmark config for qwen3.5-35B: change world_size and batch sizes
    
    * fixed
    
    * feat: update sgl_chunk_gdn test config - seq_len 4096->8192, beta dtype bfloat16->float32, cu_seqlens size 2->4
    
    * feat(kernel_benchmark): add varlen mode test for sgl_causal_conv1d
    
    - Add generate_testcase_varlen() to simulate SGLang continuous batching scenario
    - Add run_test_varlen() for varlen mode performance testing
    - Update TestParam with seq_lens and num_cache_slots fields
    - Support x shape [dim, cu_seqlen] with query_start_loc and cache_indices
    
    * bug fix and add benchdata
    
    * fixed
    
    * fixed
    
    * fixed
    
    * fixed
    
    * fixed
    
    * fixed
    
    * Sync kernel benchmark files from upstream: sgl_causal_conv1d.py, sgl_causal_conv1d_update.py, sgl_chunk_gdn.py
    
    * fixed
    
    * fixed
    
    * fixed
    sheshuchen authored Mar 18, 2026
    Configuration menu
    Copy the full SHA
    58c8bc8 View commit details
    Browse the repository at this point in the history
Loading