All pull requests

Pull requests

• Add mtplx serve --jit: load models on demand

  #659·jregeimbal opened Oct 23, 2026·

  ·

• fix: guard QSA cache frontiers during resize and extension

  #657·transcendence-x opened Oct 23, 2026·

  ·

• fix: partial-prefix SSD restore must skip tuple meta states

  #656·Phydra11 opened Oct 23, 2026·

  ·
  1

• A cancelled request keeps its cache outcome, and the dashboard shows it as cancelled

  #652·jvmenen opened Oct 20, 2026Contributor·

  ·

• SSD cold tier: the streaming spill uses the incremental-encode memo too

  #651·jvmenen opened Oct 20, 2026Contributor·

  ·

• fix: count pooled blocks from the restored keys, not the donor table

  #650·goodbadwolf opened Oct 20, 2026·

  ·
  1

• SSD cold tier: read restores in parallel into one buffer per tensor

  #648·jvmenen opened Oct 19, 2026Contributor·

  ·

• SSD session cache telemetry: where restores spend their time, where they land, and how much the cache writes

  #647·jvmenen opened Oct 19, 2026Contributor·

  ·

• SSD cold tier: evict and expire entries this server can never restore first

  #646·jvmenen opened Oct 19, 2026Contributor·

  ·

• Tests: drop the operator's MTPLX switches before collection

  #644·jvmenen opened Oct 19, 2026Contributor·

  ·

• GDN prefill on MLX's chunked gated_delta_update (MTPLX_GDN_FAST_PREFILL, off by default)

  #642·jvmenen opened Oct 19, 2026Contributor·

  ·

• Segmented KV: the segment kernel follows the quarter split and the wide verify route

  #639·jvmenen opened Oct 17, 2026Contributor·

  ·

• Session bank: the generation-final commit keeps the prompt's recurrent checkpoints

  #637·jvmenen opened Oct 17, 2026Contributor·

  ·

• Fix DeepSeek MTP logit projection controls

  #636·sypsyp97 opened Oct 15, 2026Contributor·

  ·
  1

• fix: apply output_config.effort on /v1/messages

  #634·goodbadwolf opened Oct 13, 2026·

  ·
  1

• Add MiniCPM-V 4.7 adapter and forward canvas attention arguments

  #633·shiftedx opened Oct 13, 2026Contributor·

  ·

• Leave Gemma 4 attention out of the generic SDPA hook

  #632·jvmenen opened Oct 13, 2026Contributor·

  ·

• Say why a request found nothing to reuse, and stop labelling normal misses as an SSD fault

  #631·jvmenen opened Oct 13, 2026Contributor·

  ·

• Segmented KV on Macs without tensor units (M1-M4): per-segment attention with a logsumexp merge

  #630·jvmenen opened Oct 13, 2026Contributor·

  ·

• Cut per-turn host cost of prefix lookups (cold-tier scan, prompt conversion, RAM compares)

  #629·jvmenen opened Oct 13, 2026Contributor·

  ·

• Share one common_prefix_len and one reader per session-prefix setting

  #628·jvmenen opened Oct 13, 2026Contributor·

  ·

• Dashboard: show time and tok/s at the cursor on the decode chart

  #626·jvmenen opened Oct 13, 2026Contributor·

  ·

• Run the fixed-M4 stage-3 lane for q2/g64 routed experts

  #625·trustos opened Oct 12, 2026·

  ·

• Dashboard: show what the server really runs with on the Settings tab

  #623·jvmenen opened Oct 11, 2026Contributor·

  ·

• Dashboard: show what is cached in RAM and on SSD instead of empty slot cards

  #622·jvmenen opened Oct 11, 2026Contributor·

  ·