Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on May 5, 2026

• feat(HybridCheckpointCache): add `on-device` hybrid checkpoint support

  Show description for 8eafd9e
  JamePeng

  committed
  8eafd9e

  Copy full SHA for 8eafd9e

Commits on Mar 15, 2026

• feat(core): add verbose debug logging to longest_token_prefix fast paths

  Show description for 90d46dc
  JamePeng

  committed
  90d46dc

  Copy full SHA for 90d46dc

Commits on Mar 10, 2026

• perf(cache): optimize LlamaDiskCache I/O and fix LRU behavior

  Show description for a36e8d2
  JamePeng

  committed
  a36e8d2

  Copy full SHA for a36e8d2

• perf(cache): upgrade `LlamaRAMCache` to O(1) eviction and set `LlamaTrieCache` as default

  Show description for 5c8e056
  JamePeng

  committed
  5c8e056

  Copy full SHA for 5c8e056

Commits on Mar 9, 2026

• fix(cache): fix namespace shadowing to prevent AttributeError

  Show description for 5e285fe
  JamePeng

  committed
  5e285fe

  Copy full SHA for 5e285fe

Commits on Mar 7, 2026

• perf(hybrid): eliminate PCIe I/O latency for single-turn workflows

  Show description for 7289b54
  JamePeng

  committed
  7289b54

  Copy full SHA for 7289b54

Commits on Feb 26, 2026

• fix(HybridCheckpointCache): ValueError: bytes must be in range(0, 256)

  JamePeng

  committed
  781790f

  Copy full SHA for 781790f

• fix(cache): add safety guards to checkpoint restore and optimize API calls

  Show description for 29b9522
  JamePeng

  committed
  29b9522

  Copy full SHA for 29b9522

• Remove redundant seq_id and add resource cleanup

  Show description for 5776f15
  JamePeng

  committed
  5776f15

  Copy full SHA for 5776f15

• feat(cache): implement HybridCheckpointCache for hybrid/recurrent models

  Show description for f6dda38
  JamePeng

  committed
  f6dda38

  Copy full SHA for f6dda38

Commits on Nov 2, 2025

• Implement LlamaTrieCache into llama_cache.py: Optimize LlamaCache lookup from O(N) to O(K) using a Trie

  JamePeng

  committed
  2419dc2

  Copy full SHA for 2419dc2

Commits on Sep 19, 2024

• misc: Format

  abetlen

  committed
  9b64bb5

  Copy full SHA for 9b64bb5

Commits on Jul 9, 2024

• fix(misc): Format

  abetlen

  committed
  c1ae815

  Copy full SHA for c1ae815

Commits on Jan 17, 2024

• Move cache classes to llama_cache submodule.

  abetlen

  committed
  3b92419

  Copy full SHA for 3b92419

Loading