Wagner Bruna
da1c06c33d
sd: cleanup and document sd device mappings
2026-05-19 14:32:27 -03:00
Wagner Bruna
cf3d374f14
sd: remove device offset for zero-based indexes
2026-05-19 14:31:55 -03:00
Concedo
c95d25abe2
gui visual breaking change - make GPU ID list 0-based index instead of 1-based index to match CLI
2026-05-19 22:00:43 +08:00
Wagner Bruna
592d12d0a3
sd: support for CLIP and VAE on different devices ( #2184 )
...
* sd: generalize internal interfaces to place generation on CPU
* sd: backend support for multi-device selection
* sd: frontend support for multi-device selection
* add deprecated flags to avoid breaking old cli args
---------
Co-authored-by: Concedo <39025047+LostRuins@users.noreply.github.com>
2026-05-19 21:51:23 +08:00
Concedo
7232096c11
Merge branch 'concedo' into concedo_experimental
...
# Conflicts:
# koboldcpp.py
2026-05-19 20:35:11 +08:00
Concedo
8bdef0bcfe
hotfix 1.113.2 bugs
2026-05-19 20:33:57 +08:00
Concedo
712ee6be64
try fix recent segfault on SIGINT https://github.com/LostRuins/koboldcpp/issues/2215
2026-05-18 22:37:14 +08:00
Concedo
7e08e8d8b4
add some rpc dependencies (+1 squashed commits)
...
Squashed commits:
[b092a94e5] add some rpc dependencies
2026-05-18 22:17:30 +08:00
Concedo
fecf2dc3fa
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# .github/workflows/server-self-hosted.yml
# CMakeLists.txt
# CODEOWNERS
# ci/run.sh
# cmake/llama-config.cmake.in
# common/chat.cpp
# examples/sycl/start-svr.sh
# examples/sycl/test.sh
# examples/sycl/win-start-svr.bat
# examples/sycl/win-test.bat
# ggml/src/ggml-sycl/ggml-sycl.cpp
# ggml/src/ggml-sycl/vecdotq.hpp
# ggml/src/ggml-vulkan/CMakeLists.txt
# scripts/wc2wt.sh
# tests/test-backend-ops.cpp
# tests/test-chat.cpp
2026-05-18 21:27:23 +08:00
Wagner Bruna
90326f8585
sd: sync to master-612-d7ecbe1 ( #2213 )
2026-05-18 21:19:12 +08:00
Aleksander Grygier
a135ec0baa
ui: Centralize monospace font styles in app.css ( #23272 )
Python Type-Check / python type-check (push) Has been cancelled
2026-05-18 15:10:14 +02:00
Martin Andersson
232f466583
webui: fix Tailwind v4 utility classes missing when built via cmake ( #23253 )
2026-05-18 14:08:02 +02:00
Andrei
49c21f97cd
llama: initialize pre-norm embedding mask flag ( #23256 )
2026-05-18 14:20:49 +03:00
Sigbjørn Skjæret
77e38d68f2
add myself to conversion ( #23261 )
2026-05-18 12:42:56 +02:00
Martin Klacer
053e01dff6
ci : added kleidiai-server to server-self-hosted workflow ( #22435 )
...
* kleidiai: added kleidiai-server to server-self-hosted workflow
* Added KleidiAI-enabled Arm64 Linux llama-server CI/integration test
workflow into the server-self-hosted.yml configuration file
Signed-off-by: Martin Klacer <martin.klacer@arm.com>
Change-Id: I032e33c525b7e26bc5d53719f638bee610cec1ee
* Added self-hosted executor for KleidiAI server workflow
Signed-off-by: Martin Klacer <martin.klacer@arm.com>
* Update .github/workflows/server-self-hosted.yml
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
---------
Signed-off-by: Martin Klacer <martin.klacer@arm.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
2026-05-18 11:14:57 +02:00
Georgi Gerganov
c3f95c1f06
scripts : allow wc2wt with an existing branch ( #23189 )
2026-05-18 08:57:28 +03:00
Intel AI Get-to Market Customer Success and Solutions
0caf2a1d48
sycl: scalar SWAR byte-subtract in Q6_K MMVQ dot product ( #22156 )
...
Signed-off-by: Chun Tao <chun.tao@intel.com>
Co-authored-by: Chun Tao <chun.tao@intel.com>
2026-05-18 08:12:21 +03:00
Intel AI Get-to Market Customer Success and Solutions
5511965b19
sycl: route small f32 matmuls to oneMKL, bypass oneDNN ( #22150 )
...
Signed-off-by: Chun Tao <chun.tao@intel.com>
Co-authored-by: Chun Tao <chun.tao@intel.com>
2026-05-18 08:11:51 +03:00
Neo Zhang
e98bcfec28
sycl : fix error when use -mg 1 error ( #23140 )
2026-05-18 08:11:19 +03:00
Incarnas
1867a0c692
update bid to match each layers MTP source ( #23237 )
...
* update bid to match each layers MTP source
* Update conversion/qwen.py
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
---------
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
2026-05-18 12:37:12 +08:00
Sigbjørn Skjæret
dd7cad7197
cmake : do not check for bin install dir ( #23234 )
2026-05-18 02:33:14 +02:00
Gabe Goodhart
726704a160
feat: Support d_conv=15 for ssm-conv.cu ( #23017 )
...
Branch: ModalityConditionalAdapters
AI-usage: none
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
2026-05-17 23:05:11 +02:00
Aldehir Rojas
87589042ca
cmake : fix LLAMA_BUILD_UI logic ( #23190 )
2026-05-17 14:42:26 -04:00
Sigbjørn Skjæret
e0de4c2419
cmake : do not install conversion script ( #23204 )
2026-05-17 18:07:21 +02:00
Oliver Simons
84c678242a
CUDA: Continue directly including cuda/iterator ( #23102 )
...
Cont of #22936 , forgot to update one site
2026-05-17 18:00:10 +02:00
Aman Gupta
3e12fbdea5
llama: avoid copying logits during prompt decode in MTP ( #23198 )
...
* llama: avoid copying logits during prompt decode in MTP
* review: update comment
* llama-graph: call set_output for t_h_pre_norm
2026-05-17 23:30:25 +08:00
Concedo
b91ee904d2
colab link to download directly, set a different fallback url
2026-05-17 23:11:32 +08:00
Aldehir Rojas
39cf5d6191
common : delegate assistant continuation to underlying template handlers ( #23089 )
...
* common : delegate assistant continuation to template handler
* server : implement echo parameter to exclude assistant prefill in the response
* server : fix tests for prefill
* server : use existing llama template
* cont : clean up
2026-05-17 13:36:05 +02:00
Jan Ekström
a6d6183dbc
ggml-vulkan/CMakeLists: add a check for SPIRV-Headers ( #22009 )
...
* ci/run: set explicit SPIR-V Headers search path for macOS vulkan CI
For whatever reason, the files are under additional sub-path
`vulkan/` under the cmake directory, which does not match either
current LunarG macOS Vulkan SDK structure (`lib/cmake/SPIRV-Headers`),
nor what gets installed when you run the cmake build+install for
SPIRV-Headers itself on at least Linux (`share/cmake/SPIRV-Headers`).
This allows for SPIRV-Headers to be found, as currently the CI
runner's setup does not seem to include the relevant path in
list of search locations.
* ggml-vulkan/CMakeLists: add a check for SPIRV-Headers
This is installed by the project if it is built and installed.
Receiving an error during the configuration step is generally
preferred to receiving an error in the middle of a build.
2026-05-17 13:12:11 +02:00
Pascal
fcae601e44
vulkan: add cpy bf16 -> f32 pipelines ( #22677 )
2026-05-17 11:31:20 +02:00
Jeff Bolz
7ba22c6a09
vulkan: Support unaligned tensors for ROPE ( #22637 )
2026-05-17 11:30:16 +02:00
Concedo
de2c5f1cef
detokenize add special token ids
2026-05-17 17:04:47 +08:00
Concedo
c1514e328b
updated lite
2026-05-17 16:47:19 +08:00
Aldehir Rojas
f4cc787b9f
common : enable streaming JSON argument values ( #23173 )
...
* common : remove atomic from json arguments
* common : remove parsing logic on JSON arguments
2026-05-17 03:44:34 -05:00
Jeff Bolz
3fbadb06dc
vulkan: fuse SSM_CONV + BIAS + SILU ( #22653 )
2026-05-17 10:25:50 +02:00
Rares Vernica
1a68ec9378
server : honor --embd-normalize CLI arg ( #23125 )
...
The --embd-normalize flag was registered only for the embedding and debug
examples, so llama-server rejected it and the /embedding handler used a
hard-coded default of 2 (L2). Add LLAMA_EXAMPLE_SERVER to the flag's
example set and read params.embd_normalize as the handler's default. The
per-request "embd_normalize" body field continues to override.
2026-05-17 09:39:04 +03:00
ddh0
a16cce81d3
ngram : reduce noisy logs ( #23185 )
...
* ngram : reduce noisy logs
* ngram : reduce noisy logs
2026-05-17 09:38:17 +03:00
Wagner Bruna
1ae9a79ecc
sd: sync to master-607-fd1a279 ( #2212 )
2026-05-17 11:37:16 +08:00
Concedo
1e828ccabf
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# common/common.cpp
# ggml/CMakeLists.txt
# scripts/sync-ggml.last
# scripts/sync_vendor.py
# src/llama-context.cpp
# tests/CMakeLists.txt
# tests/test-backend-ops.cpp
# tools/cli/README.md
# tools/completion/README.md
# tools/server/README.md
2026-05-17 11:26:18 +08:00
Judd
4f13cb7424
webui: support video files as input ( #22830 )
2026-05-17 02:13:44 +02:00
Xuan-Son Nguyen
b64739ea39
server: (router) alloc tmp buffer on heap ( #23159 )
2026-05-16 23:42:16 +02:00
Pascal
64b38b561b
server: skip device enumeration in router mode to avoid creating CUDA primary context ( #23137 )
2026-05-16 21:21:06 +02:00
Winston Ma
6049906133
vulkan: removed duplicate #include <memory> in headers ( #23144 )
2026-05-16 19:57:35 +02:00
Concedo
9d38a9edc0
quick fix for colab
2026-05-17 00:18:02 +08:00
Concedo
0d320f60a6
fix multiuser regression
2026-05-17 00:17:12 +08:00
Concedo
47d5772fbe
add batching failure spam logs
2026-05-16 23:21:01 +08:00
Concedo
9203b6a051
Merge branch 'upstream' into concedo_experimental
...
# Conflicts:
# .github/labeler.yml
# .github/workflows/build-self-hosted.yml
# .github/workflows/release.yml
# .github/workflows/server-sanitize.yml
# .github/workflows/server-self-hosted.yml
# .github/workflows/server.yml
# .github/workflows/ui-build.yml
# .github/workflows/ui-ci.yml
# .github/workflows/ui-publish.yml
# .gitignore
# CMakeLists.txt
# CODEOWNERS
# scripts/ui-download.cmake
# scripts/xxd.cmake
# tests/test-backend-ops.cpp
# tests/test-reasoning-budget.cpp
# tools/CMakeLists.txt
# tools/server/CMakeLists.txt
# tools/server/README.md
2026-05-16 22:56:33 +08:00
Concedo
3095da076a
only fetch new popped horde requests if model is not blocked queue
2026-05-16 22:27:12 +08:00
Aleksander Grygier
0253fb21f5
ui: Add request timeout for MCP tool calls ( #23138 )
...
Check Pre-Tokenizer Hashes / pre-tokenizer-hashes (push) Has been cancelled
Python check requirements.txt / check-requirements (push) Has been cancelled
Python Type-Check / python type-check (push) Has been cancelled
* feat: Add request timeout for MCP tool calls in llama-ui
* feat: MCP Settings tab with max timeout setting
2026-05-16 15:20:27 +02:00
Georgi Gerganov
3a92bc99db
sync : ggml
2026-05-16 16:11:29 +03:00
