oobabooga
dcbb323cb6
UI: Improve web search security by rejecting non-HTTP links
2026-05-16 17:39:10 -07:00
oobabooga
4c16b9458f
UI: Fix token count not being set in non-streaming mode
2026-05-16 17:35:30 -07:00
oobabooga
aa7d6bf048
Electron: Validate model_dir path before applying it
2026-05-16 17:34:58 -07:00
oobabooga
d803e29d7f
Fix continue-mode regressions across template families
2026-05-16 13:40:24 -07:00
MeemeeLab
9a396e1749
Fix incorrect prompts generated with continue mode
...
When using the continue function, add_generation_prompt was set to False,
bypassing any template logic that depends on it. This caused missing tokens
or headers that some models inject at the start of an assistant turn.
Gemma 4 was particularly sensitive to this, producing garbled output due
to the missing thought channel header (<|channel>thought\n<channel|>).
Fix by popping the last incomplete assistant message and re-rendering the
prompt with add_generation_prompt=True, then appending the partial content
afterward. This ensures the prompt is structurally identical to normal
generation regardless of the model's template.
GPT-OSS and Seed-OSS thinking block handling is preserved via the existing
fake-message approach, as those models manage thinking content differently.
2026-05-16 13:40:15 -07:00
oobabooga
be7f3a20f2
Electron: Add spellcheck toggle in the Session tab ( closes #7550 )
2026-05-16 10:48:02 -07:00
oobabooga
f327da7831
Improve the looks of the Session tab
2026-05-16 09:59:42 -07:00
oobabooga
6378c5cd57
Electron: Add "Check for updates" button in the Session tab
2026-05-16 08:53:40 -07:00
oobabooga
dbf922006c
Revert "Soften slider and checkbox label colors in light theme"
...
This reverts commit 3c4a1402a8
2026-05-15 12:47:50 -07:00
oobabooga
8e34c7b1b0
Show live context size while generating
2026-05-15 12:45:05 -07:00
oobabooga
3c4a1402a8
Soften slider and checkbox label colors in light theme
2026-05-15 12:10:27 -07:00
oobabooga
2a27ceae93
Hide reasoning and tools controls in chat mode
2026-05-15 11:44:09 -07:00
oobabooga
6033be53ac
Reorder right sidebar: Mode/Character/Chat style on top
2026-05-15 09:43:22 -07:00
oobabooga
bed909d46c
Auto-select sibling mmproj when loading a model ( closes #7564 )
2026-05-15 06:39:30 -07:00
oobabooga
0f88365b0e
Rename "Send dummy message/reply" to "Insert user/assistant message"
2026-05-14 19:59:56 -07:00
oobabooga
1e8136dc2a
Small simplifications
2026-05-14 11:21:18 -07:00
oobabooga
58774c8cb1
Also fix streaming leak across all other chat actions
2026-05-14 11:07:59 -07:00
oobabooga
630141912b
Fix streaming output leaking across chats ( closes #7555 )
2026-05-14 08:36:49 -07:00
oobabooga
a711849fa7
Detect mmproj files in the models folder (simplifies LM Studio migration)
2026-05-14 08:02:45 -07:00
oobabooga
66c3c49c54
Treat negative ctx-size as auto
2026-05-13 21:15:43 -07:00
oobabooga
47fdee9cb1
Electron: Add a folder picker for the models directory
2026-05-13 20:34:07 -07:00
oobabooga
13f5d378e5
Make web_search tool call results pretty
2026-05-12 17:50:36 -07:00
oobabooga
79630e9bb1
Remove backend="duckduckgo" from ddgs ( https://github.com/oobabooga/textgen/issues/7548#issuecomment-4433225429 )
2026-05-12 16:47:17 -07:00
oobabooga
f4f556a504
Add snippet support to the web_search tool ( closes #7548 )
2026-05-11 22:56:27 -07:00
oobabooga
88d47dc412
Don't clear chat input when no model is loaded ( closes #7542 )
2026-05-07 06:34:10 -07:00
oobabooga
4d3f355f04
Add a --no-electron flag
2026-05-06 22:02:56 -07:00
oobabooga
3e38f44a2e
Fix speculative decoding broken by llama.cpp arg rename ( closes #7541 )
2026-05-06 21:06:36 -07:00
oobabooga
71525d51a6
Fix truncation length reverting after model load on UI reload ( closes #7540 )
2026-05-06 07:00:39 -07:00
oobabooga
0f4d8aa9ca
API: Fix crash on list-format content in tool and assistant messages
2026-05-03 17:51:31 -07:00
oobabooga
31c2a82925
Electron: fix missing log colors on Windows
2026-05-03 08:49:55 -07:00
oobabooga
2ca61f21fb
Cleanup modules/web_search.py
2026-05-02 19:19:19 -07:00
oobabooga
8376284d7a
Fix extension settings not saving for user_data/extensions ( closes #7525 )
2026-05-02 18:54:49 -07:00
oobabooga
7facf669f8
Replace DuckDuckGo scraping with ddgs library
2026-05-02 18:50:22 -07:00
oobabooga
03b964a3e5
Minor color improvement
2026-05-02 09:25:11 -07:00
oobabooga
8e0b02a169
Fix Stop button being ignored during tool call approval
2026-05-02 08:26:38 -07:00
oobabooga
ddec64af71
UI: refinements after c6a57a4572 & electron user_data path
2026-05-02 08:18:39 -07:00
oobabooga
c6a57a4572
Major UI overhaul
2026-05-01 22:06:18 -07:00
oobabooga
54c4feeb4f
Support .jinja/.jinja2 instruction template files in the UI ( closes #7517 )
2026-04-26 18:57:07 -07:00
oobabooga
4b575a2fae
llama.cpp: Add --split-mode flag with tensor parallelism support
2026-04-26 17:44:32 -07:00
oobabooga
187b07a924
Fix race condition and clean up exllamav3 module
2026-04-25 15:14:44 -07:00
oobabooga
1d1a319597
llama.cpp: Fix multimodal by using server's random media marker
2026-04-23 06:21:19 -07:00
oobabooga
7338120bff
Fix GPT-OSS channel tokens leaking in UI after tool calls
2026-04-22 18:08:59 -07:00
oobabooga
1d871c0896
Fix Gemma 4 thinking tags not hidden after tool calls ( closes #7509 )
2026-04-22 15:17:17 -07:00
oobabooga
4737744f4a
Security: Fix SSRF vulnerabilities in URL fetching
2026-04-22 14:43:16 -07:00
oobabooga
83347ed7e4
Add preserve_thinking chat template parameter (UI and CLI)
2026-04-22 13:43:56 -07:00
oobabooga
9a9f3ae9e8
llama.cpp: Pass --draft-min 48 by default for draftless speculative decoding
2026-04-22 13:00:48 -07:00
oobabooga
a253d38953
Add tool call confirmation with inline approve/reject buttons
2026-04-22 11:49:39 -07:00
oobabooga
1e1d2d3728
Cache MCP tool discovery to avoid re-querying on each generation
2026-04-22 09:45:51 -07:00
oobabooga
707b5df4ea
Add stdio MCP server support via user_data/mcp.json
2026-04-19 20:47:59 -07:00
oobabooga
fe0b385dae
Merge pull request #7506 from Th-Underscore/fix/download-model
...
Add model download branch handling in download_model_wrapper
2026-04-18 21:09:44 -03:00
