-
11300e3034
feat: more o n migration to v3
feat/migration-mlx-swift-lm-v3
hugo
2026-04-30 11:58:53 +02:00
-
3502266ff9
feat: migration to mlx-swift-lm v3
hugo
2026-04-30 09:18:37 +02:00
-
4ad46ec1ea
feat: added gemma 4 (not supported yet in mlx-swift-lm, though)
main
hugo
2026-04-28 21:52:32 +02:00
-
d5b9ae15cc
fix: weird model selector fixed
hugo
2026-04-17 14:13:38 +02:00
-
6e20bd538b
feat: generation defaults moved to model management
hugo
2026-03-21 20:13:44 +01:00
-
6b14d7d46c
feat: added model management
hugo
2026-03-21 19:15:13 +01:00
-
84a6b2229f
added new story model with 12B size
hugo
2026-03-21 16:08:40 +01:00
-
8ebf19408b
Merge pull request #1 from rfc1437/session-cache-upgrade-phase1
Georg Bauer
2026-03-21 13:35:53 +01:00
-
-
491a1e6ffe
fix: corrected loading of app with doubleclick on chat document
hugo
2026-03-21 10:56:17 +01:00
-
365f4e0531
feat: better visualisation in the fill bar for context
hugo
2026-03-21 09:50:17 +01:00
-
32bbf3f204
feat: context fill grade in chat UI
hugo
2026-03-21 09:39:25 +01:00
-
675183152a
fix: repaired scene prefs
hugo
2026-03-21 09:32:33 +01:00
-
941ec92dbf
feat: first take at scene-local generation settings
hugo
2026-03-21 09:24:14 +01:00
-
24b940d526
chore: close out phase 7 as out of scope
hugo
2026-03-21 08:52:51 +01:00
-
107ac0524b
feat: finished all open things up to and including phase 6
hugo
2026-03-21 08:41:13 +01:00
-
0325fa8964
feat: phase 6 implemented and tested
hugo
2026-03-21 07:59:48 +01:00
-
2c7195490e
chore: moved some tests to qwen3.5 0.8B for speed
hugo
2026-03-20 17:01:23 +01:00
-
7d25955042
chore: added more qwen 3.5 models
hugo
2026-03-20 16:47:10 +01:00
-
455eba7caf
chore: removed unslopNemo
hugo
2026-03-20 13:10:12 +01:00
-
0761254d17
feat: vision properly coverd with tests and completed
hugo
2026-03-20 12:57:27 +01:00
-
e59be9df1a
fix: app-close-crash
hugo
2026-03-20 12:20:26 +01:00
-
1f12fac5e2
feat: finally nailed down phases 1-4
hugo
2026-03-20 12:05:24 +01:00
-
5aed0107c6
fix: more hardening of cache behaviour and some fixes
hugo
2026-03-20 11:43:58 +01:00
-
ee34fd5e84
feat: implemented more of phase 4
hugo
2026-03-20 10:40:51 +01:00
-
aadcc308a5
feat: better hardening with unit tests and end-to-end tests
hugo
2026-03-20 10:27:39 +01:00
-
e40a2f3c45
feat: implement phase 2 of session-cache-upgrade.md
hugo
2026-03-20 08:57:54 +01:00
-
e98e5fd88b
Implement session cache upgrade phase 1 foundation
hugo
2026-03-20 08:35:37 +01:00
-
-
41199cb9bc
chore: added an implementation plan to rework the OpenAI API server to be more performant
hugo
2026-03-19 19:07:31 +01:00
-
577fdf8950
feat: more visibility of prefilling
hugo
2026-03-19 11:36:46 +01:00
-
49bd165ce7
fix: more telemetry and tighter implementation of cache
hugo
2026-03-19 11:30:18 +01:00
-
c2e80e4066
chore: small update on qwen 3.5 9b actually having vision
hugo
2026-03-19 11:20:26 +01:00
-
25df02daa1
feat: auto-save on close and auto-load on reopen
hugo
2026-03-19 09:44:35 +01:00
-
b52633a301
image support for qwen3.5 9b activated again
hugo
2026-03-19 08:56:03 +01:00
-
f0db0c0938
feat: document saving/loading
hugo
2026-03-18 15:29:02 +01:00
-
09b94b32d0
feat: scene management for RP settings
hugo
2026-03-18 14:57:29 +01:00
-
ed1c91cd2b
chore: added a plan for proper document support
hugo
2026-03-18 13:22:12 +01:00
-
27849ccbd7
feat: added stheno (llambda based) text-only model, too
hugo
2026-03-18 13:08:21 +01:00
-
6a87fe6f08
fix: export finally works
hugo
2026-03-18 11:59:51 +01:00
-
82a77fdb0a
feat: first tries at save dialog, so far failing
hugo
2026-03-18 11:40:43 +01:00
-
af8b8c9532
feat: copy-paste image files from finder
hugo
2026-03-18 09:25:26 +01:00
-
07b71f90ec
feat: start of support for thinking mode, qwen 3.5 9b addition and better idle time handling
hugo
2026-03-18 09:15:08 +01:00
-
ed6cc5f5d1
fix: better handling of API stuff, still not where internal chat is
hugo
2026-03-17 21:24:04 +01:00
-
20f9c0bcc4
feat: settings for default model
hugo
2026-03-17 20:20:54 +01:00
-
aa2712555a
feat: idle-unload of models
hugo
2026-03-17 20:01:44 +01:00
-
033443589c
fea: added a proper icon
hugo
2026-03-17 19:57:39 +01:00
-
1a67311874
feat: inference visualisation
hugo
2026-03-17 19:30:09 +01:00
-
5313b7175e
feat: complete rewrite to swift
hugo
2026-03-17 19:12:54 +01:00
-
c80fe97f41
feat: added gemma 3n E4B as another model for fast response
hugo
2026-03-17 13:24:43 +01:00
-
cc4f937d9a
feat: proper support for context size
hugo
2026-03-17 12:34:11 +01:00
-
540b187593
chore: added README
hugo
2026-03-17 12:07:45 +01:00
-
ef83c24b0b
feat: hot swapping of models
hugo
2026-03-17 11:58:24 +01:00
-
cc6e761ed4
feat: qwen now works, too
hugo
2026-03-17 11:44:24 +01:00
-
bdfbd14577
fix: trying to do kv prefix caching
hugo
2026-03-17 10:04:14 +01:00
-
5bf170cedb
removed kv quantization due to incompatibility with gemma3
hugo
2026-03-17 09:20:35 +01:00
-
df81afe8d7
initial commit
hugo
2026-03-17 09:14:27 +01:00