pullrequests Search Results · bucket language:JavaScript language:JavaScript language:JavaScript language:Python language:Java
Filter by
1.1M results
... removes buckets that are full again, at most once per refill window. A full bucket gives the same
decision as a new one.
- New gauge overflow_redirect_rate_limiter_buckets (labels type, strategy ...
pl
... × prefill bucket pairs with asynchronous eager compilation. Warmup retains one destination
per prefill shape rather than every intermediate output.
Device-resident cache entries consume TPU memory ...
Multi-modal
tpu
zou3519
startr3k
dwiest... prompt del smoke de preview). En un commit aparte corrige dos
datos obsoletos de AGENTS.md: el bucket real es arte-chatbot-fichas-tecnicas y producción corre en Lambda + API Gateway
con estado en ...
creep1ng
Overview
When _get_kv_cache_groups_uniform_page_size groups KV cache specs by (is_draft, type_id, page_size_bytes), it previously
computed min_num_layers across all buckets—including draft-only ...
bug
kv-cache-manager
arashb
pattch-openclaw
Motivation
get_batch_sizes_to_capture drops the top decode bucket when --cuda-graph-max-bs is below the request pool and not a
multiple of the attention-TP alignment. Example: DP attention with attn ...
blackwell
lora
Denji-kk
... preserveAspectRatio= none . It had no axes, it distorted at every
width, and because the API skipped empty buckets, quiet periods disappeared and the time axis wasn t linear.
Changes
Server (timeline)
- Zero ...
jlaportebot
dependencies
python
dependabot[bot]