2186 commits

Author	SHA1	Message	Date
oobabooga	05e4842033	Fix image generation: default to SDPA attention backend	2026-04-05 20:03:06 -07:00
oobabooga	b1d06dcf96	UI: Add MCP server support	2026-04-05 19:46:01 -07:00
oobabooga	abc3487f4d	UI: Move cpu-moe checkbox to extra flags (no longer useful now that --fit exists)	2026-04-05 18:24:26 -07:00
oobabooga	d78fc46114	Fix "address already in use" on server restart (Linux/macOS)	2026-04-05 16:42:27 -07:00
oobabooga	422f42ca7f	Pre-compile LaTeX regex in html_generator.py	2026-04-04 23:51:15 -07:00
oobabooga	544fcb0b7f	Simplify modules/image_models.py	2026-04-04 23:29:57 -07:00
oobabooga	c63a79ee48	Image generation: Embed generation metadata in API image responses	2026-04-04 23:15:14 -07:00
oobabooga	dfd8ec9c49	UI: Make accordion outline styling global	2026-04-04 20:13:20 -07:00
oobabooga	1b403a4ffa	UI: Fix inline LaTeX rendering by protecting $...$ from markdown (closes #7423)	2026-04-04 19:33:05 -07:00
oobabooga	ffea8f282e	UI: Improve message text contrast	2026-04-04 18:53:13 -07:00
oobabooga	7fed60f90a	UI: Improve the hover menu looks	2026-04-04 18:29:36 -07:00
oobabooga	2eef90a323	API: Remove deprecated "settings" parameter from model load endpoint	2026-04-04 11:00:14 -07:00
oobabooga	9183dc444e	API: Fix loader args leaking between sequential model loads	2026-04-04 10:48:53 -07:00
oobabooga	e0ad4e60df	UI: Fix tool buffer check truncating visible text at end of generation	2026-04-04 09:57:07 -07:00
oobabooga	54b2f39c78	Cleanup modules/chat.py	2026-04-03 22:07:21 -07:00
oobabooga	fc35acab9b	API: Fix tool call parser crash on non-dict JSON output	2026-04-03 16:56:15 -07:00
oobabooga	8ecdb41078	fix(security): sanitize filenames in all prompt file operations (CWE-22) (#7462) ... --------- Co-authored-by: Alex Chen <ffulbtech@gmail.com>	2026-04-03 19:36:50 -03:00
oobabooga	95d6c53e13	Revert "API: Add warning about vanilla llama-server not supporting prompt logprobs + instructions" ... This reverts commit 42dfcdfc5b.	2026-04-03 07:30:48 -07:00
oobabooga	66d1a22c73	Fix crash when no model is selected (None passed to resolve_model_path)	2026-04-03 05:56:36 -07:00
oobabooga	000d776967	Revert "llama.cpp: Disable jinja by default (we use Python jinja, not cpp jinja)" ... This reverts commit a1cb5b5dc0.	2026-04-03 05:49:03 -07:00
oobabooga	a1cb5b5dc0	llama.cpp: Disable jinja by default (we use Python jinja, not cpp jinja) ... This was causing template compilation issues with qwen models.	2026-04-02 21:56:40 -07:00
oobabooga	42dfcdfc5b	API: Add warning about vanilla llama-server not supporting prompt logprobs + instructions	2026-04-02 20:46:27 -07:00
oobabooga	6e2b70bde6	Add Gemma 4 tool-calling support	2026-04-02 20:26:27 -07:00
oobabooga	b108c55353	Fix portable builds not starting due to missing ik element	2026-04-02 19:14:50 -07:00
oobabooga	7aab2fdf9a	API: Improve cache clearing in logprobs	2026-04-02 17:50:42 -07:00
oobabooga	091037ec20	Fix top_logprobs_ids missing for llama.cpp loader	2026-04-02 16:13:45 -03:00
oobabooga	ea1f8c71f2	API: Optimize prompt logprobs and refactor ExLlamav3 forward pass	2026-04-02 14:31:11 -03:00
oobabooga	c10c6e87ae	API: Add token ids to logprobs output	2026-04-02 07:17:27 -07:00
oobabooga	a32ce254f2	Don't pass torch_dtype to transformers, autodetect from model config	2026-04-02 00:44:14 -03:00
oobabooga	4073164be0	Fix ExLlamav3 OOM on prompt logprobs and qwen3_5_moe HF compat	2026-04-01 19:44:55 -07:00
oobabooga	71c1a52afe	API: Implement echo + logprobs for /v1/completions endpoint	2026-03-31 07:43:11 -07:00
oobabooga	6382fbef83	Several small code simplifications	2026-03-30 19:36:03 -07:00
oobabooga	0466b6e271	ik_llama.cpp: Auto-enable Hadamard KV cache rotation with quantized cache	2026-03-29 15:52:36 -07:00
oobabooga	4979e87e48	Add ik_llama.cpp support via ik_llama_cpp_binaries package	2026-03-28 12:09:00 -03:00
oobabooga	9dd04b86ce	Suppress EOS token at logit level for ExLlamav3 when ban_eos_token is set	2026-03-28 06:17:57 -07:00
oobabooga	bda95172bd	Fix stopping string detection for chromadb/context-1	2026-03-28 06:09:53 -07:00
oobabooga	4cbea02ed4	Add ik_llama.cpp support via --ik flag	2026-03-26 06:54:47 -07:00
oobabooga	e154140021	Rename "truncation length" to "context length" in logs	2026-03-25 07:21:02 -07:00
oobabooga	368f37335f	Fix --idle-timeout issues with encode/decode and parallel generation	2026-03-25 06:37:45 -07:00
oobabooga	d6f1485dd1	UI: Update the enable_thinking info message	2026-03-24 21:45:11 -07:00
oobabooga	807be11832	Remove obsolete models/config.yaml and related code	2026-03-24 18:48:50 -07:00
oobabooga	750502695c	Fix GPT-OSS tool-calling after 9ec20d97	2026-03-24 11:39:24 -07:00
oobabooga	a7ef430b38	Revert "llama.cpp: Don't suppress llama-server logs" ... This reverts commit 9488df3e48.	2026-03-23 20:22:51 -07:00
oobabooga	286bbb685d	Revert "Follow-up to previous commit" ... This reverts commit 1dda5e4711.	2026-03-23 20:22:46 -07:00
oobabooga	02f18a1d65	API: Add thinking block signature field, fix error codes, clean up logging	2026-03-23 07:06:38 -07:00
oobabooga	307d0c92be	UI polish	2026-03-23 06:35:14 -07:00
oobabooga	9ec20d9730	Strip thinking blocks before tool-call parsing	2026-03-22 19:19:14 -07:00
Phrosty1	bde496ea5d	Fix prompt corruption when continuing with context truncation (#7439)	2026-03-22 21:48:56 -03:00
oobabooga	1dda5e4711	Follow-up to previous commit	2026-03-21 20:58:45 -07:00
oobabooga	9488df3e48	llama.cpp: Don't suppress llama-server logs	2026-03-21 20:47:26 -07:00