Commits

Commit history

1. Commits on Aug 7, 2026

   • Upgrade llama.cpp from b10269 to b10270; rework TextToSpeech onto Qwen3-TTS

     Show description for d119cc2
     claude

     committed
     d119cc2View commit details

     Copy full SHA for d119cc2

     View code at this pointBrowse repository at this point
2. Commits on Jul 11, 2026

   • Fix review findings from PR #326 (JNI teardown, logger race, formatting)

     Show description for 9afba24
     claude

     committed
     9afba24View commit details

     Copy full SHA for 9afba24

     View code at this pointBrowse repository at this point

   • Merge pull request #327 from bernardladenthin/claude/update-b9959-s56dot

     Show description for 2797f96
     bernardladenthin

     authored
     2797f96View commit details

     Copy full SHA for 2797f96

     View code at this pointBrowse repository at this point

   • Expose pinned llama.cpp version to Java consumers

     Show description for 0a073f7
     claude

     committed
     0a073f7View commit details

     Copy full SHA for 0a073f7

     View code at this pointBrowse repository at this point
3. Commits on Jul 10, 2026

   • Fix concurrency, exception-safety, and correctness bugs in JNI layer and Java API

     Show description for 88fdd6d
     vaiju1981

     committed
     88fdd6dView commit details

     Copy full SHA for 88fdd6d

     View code at this pointBrowse repository at this point
4. Commits on Jul 8, 2026

   • Upgrade llama.cpp from b9912 to b9916; add QuantizationType.Q2_0

     Show description for b97f0d0
     claude

     committed
     b97f0d0View commit details

     Copy full SHA for b97f0d0

     View code at this pointBrowse repository at this point
5. Commits on Jul 6, 2026

   • Cover the loader backend chain with fixture-driven tests (Sonar gate)

     Show description for 42f9880
     claude

     committed
     42f9880View commit details

     Copy full SHA for 42f9880

     View code at this pointBrowse repository at this point

   • LlamaLoader: backend-manifest selection for multi-backend fat jars

     Show description for c90a061
     claude

     committed
     c90a061View commit details

     Copy full SHA for c90a061

     View code at this pointBrowse repository at this point

   • Correct stale Windows 3-DLL docs; clean up extracted ggml-metal.metal

     Show description for ad30f78
     claude

     committed
     ad30f78View commit details

     Copy full SHA for ad30f78

     View code at this pointBrowse repository at this point
6. Commits on Jul 5, 2026

   • Audio fixture wiring + REUSE coverage; promote emulator job to release gate

     Show description for a843ea8
     claude

     committed
     a843ea8View commit details

     Copy full SHA for a843ea8

     View code at this pointBrowse repository at this point

   • CI fixes: REUSE headers for new files + skip-safe router test teardown

     Show description for 0e7ab86
     claude

     committed
     0e7ab86View commit details

     Copy full SHA for 0e7ab86

     View code at this pointBrowse repository at this point

   • GGUF inspector, streaming tool calls (langchain4j), Session fork/rewind

     Show description for 6434b63
     claude

     committed
     6434b63View commit details

     Copy full SHA for 6434b63

     View code at this pointBrowse repository at this point

   • Typed Java API for router mode (RouterClient + RouterModel)

     Show description for 1ff8b5f
     claude

     committed
     1ff8b5fView commit details

     Copy full SHA for 1ff8b5f

     View code at this pointBrowse repository at this point

   • Add NativeServer attach mode, in-JVM router mode, and GGUF quantizer

     Show description for b605f8b
     claude

     committed
     b605f8bView commit details

     Copy full SHA for b605f8b

     View code at this pointBrowse repository at this point

   • feat: runtime LoRA control, batch embeddings, UTF-8-safe JNI strings

     Show description for 9c69b3b
     claude

     committed
     9c69b3bView commit details

     Copy full SHA for 9c69b3b

     View code at this pointBrowse repository at this point
7. Commits on Jul 3, 2026

   • Fix ArchUnit violations from the new server code (layering + no-sleep)

     Show description for d9a6a83
     claude

     committed
     d9a6a83View commit details

     Copy full SHA for d9a6a83

     View code at this pointBrowse repository at this point

   • Expose sse_ping_interval + audited completion params on InferenceParameters

     Show description for dca1b20
     claude

     committed
     dca1b20View commit details

     Copy full SHA for dca1b20

     View code at this pointBrowse repository at this point

   • Surface model ftype (quantization) through the Java layer and /v1/models

     Show description for e8abfc1
     claude

     committed
     e8abfc1View commit details

     Copy full SHA for e8abfc1

     View code at this pointBrowse repository at this point

   • server: simplify ServerLauncher dispatch to one primitive; rename flag to --openai-compat

     Show description for b274520
     claude

     committed
     b274520View commit details

     Copy full SHA for b274520

     View code at this pointBrowse repository at this point

   • server: add ServerLauncher — one fat-jar entry, pick mode via --open-ai-compat

     Show description for c288e43
     claude

     committed
     c288e43View commit details

     Copy full SHA for c288e43

     View code at this pointBrowse repository at this point

   • server: add NativeServer — run the full llama.cpp server (WebUI) in-DLL via JNI

     Show description for 3c8aeb8
     claude

     committed
     3c8aeb8View commit details

     Copy full SHA for 3c8aeb8

     View code at this pointBrowse repository at this point

   • server: add -b/-ub/-tb/-ctk/-ctv/--jinja/--chat-template-kwargs to OpenAiServerCli

     Show description for 926cd3c
     claude

     committed
     926cd3cView commit details

     Copy full SHA for 926cd3c

     View code at this pointBrowse repository at this point
8. Commits on Jul 2, 2026

   • Fix multi-turn tool-calling checkpoint starvation for recurrent models (Granite-4)

     Show description for cf0bf25
     claude

     committed
     cf0bf25View commit details

     Copy full SHA for cf0bf25

     View code at this pointBrowse repository at this point

   • Make merged #289 training code pass SpotBugs + PIT (keep pure @Builder)

     Show description for 18ed490
     claude

     committed
     18ed490View commit details

     Copy full SHA for 18ed490

     View code at this pointBrowse repository at this point

   • Fix reactor CI: keep class-level @Builder, re-root crash-dump paths, README

     Show description for d64a365
     claude

     committed
     d64a365View commit details

     Copy full SHA for d64a365

     View code at this pointBrowse repository at this point

   • Merge main (fine-tuning feature #289) into the reactor branch

     Show description for 91466f0
     claude

     committed
     91466f0View commit details

     Copy full SHA for 91466f0

     View code at this pointBrowse repository at this point
9. Commits on Jul 1, 2026

   • refactor: move core into llama/ submodule under an aggregator POM (Maven reactor)

     Show description for 776f117
     claude

     committed
     776f117View commit details

     Copy full SHA for 776f117

     View code at this pointBrowse repository at this point
10. Loading