Skip to content

Actions: pythongiant/KVBoost

Actions

Deploy site to GitHub Pages

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
site.yml
60 workflow runs
60 workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
implement a CUDA graph self check Deploy site to GitHub Pages #60: Commit b6f253f pushed by pythongiant
5s main
main
5s
recompile thrash Deploy site to GitHub Pages #59: Commit 57cae9b pushed by pythongiant
6s main
main
6s
multi turn setup Deploy site to GitHub Pages #58: Commit 44d1b6f pushed by pythongiant
5s main
main
5s
replace torch.cuda.CUDAGraph Deploy site to GitHub Pages #57: Commit af59562 pushed by pythongiant
7s main
main
7s
samlpe (1,vocab) probs Deploy site to GitHub Pages #56: Commit e2a6837 pushed by pythongiant
5s main
main
5s
capture single token decode step into a cudagraph Deploy site to GitHub Pages #55: Commit b3b69ab pushed by pythongiant
5s main
main
5s
three kernels Deploy site to GitHub Pages #54: Commit f3e9e50 pushed by pythongiant
6s main
main
6s
2x faster decode with CUDA graph capture Deploy site to GitHub Pages #53: Commit 97b765b pushed by pythongiant
4s main
main
4s
Faster TTFT Deploy site to GitHub Pages #52: Commit 58c7eb9 pushed by pythongiant
5s main
main
5s
cache blend sparse Deploy site to GitHub Pages #51: Commit 393eb53 pushed by pythongiant
5s main
main
5s
raw detok fix Deploy site to GitHub Pages #50: Commit 6ca941e pushed by pythongiant
5s main
main
5s
pypi stats Deploy site to GitHub Pages #49: Commit 89b506a pushed by pythongiant
5s main
main
5s
chunked the cacheblend step-1 forward the same way _decode_with_kv ch… Deploy site to GitHub Pages #48: Commit 49ad7d2 pushed by pythongiant
4s main
main
4s
throughput was getting killed in launch_oom_planner Deploy site to GitHub Pages #47: Commit 0ad743a pushed by pythongiant
4s main
main
4s
production loads Deploy site to GitHub Pages #46: Commit 2897e99 pushed by pythongiant
3s main
main
3s
OOM Recovery tokens Deploy site to GitHub Pages #45: Commit f16bced pushed by pythongiant
4s main
main
4s
engine_worker to use planner.log_pre_request + explicit kwargs + log_… Deploy site to GitHub Pages #44: Commit 60cf88f pushed by pythongiant
4s main
main
4s
create a scoring metric Deploy site to GitHub Pages #43: Commit 3f723b4 pushed by pythongiant
4s main
main
4s
add a defensive pre hook Deploy site to GitHub Pages #42: Commit d64d554 pushed by pythongiant
6s main
main
6s
max tokens for server Deploy site to GitHub Pages #41: Commit a0f9fef pushed by pythongiant
5s main
main
5s
enforce new OOM policy Deploy site to GitHub Pages #40: Commit 5d508ea pushed by pythongiant
4s main
main
4s
OOM Recovery: reduce prefill chunk size Deploy site to GitHub Pages #39: Commit 09186f8 pushed by pythongiant
4s main
main
4s
increase default max tokens Deploy site to GitHub Pages #38: Commit 31be9a7 pushed by pythongiant
4s main
main
4s
sttreaming kernels Deploy site to GitHub Pages #37: Commit 2761b51 pushed by pythongiant
4s main
main
4s
tool-call auto Deploy site to GitHub Pages #36: Commit 47c7027 pushed by pythongiant
4s main
main
4s