Adrien Gallouët [Wed, 1 Oct 2025 17:22:18 +0000 (19:22 +0200)]
common: introduce http.h for httplib-based client (#16373)
* common: introduce http.h for httplib-based client
This change moves cpp-httplib based URL parsing and client setup into
a new header `common/http.h`, and integrates it in `arg.cpp` and `run.cpp`.
It is an iteration towards removing libcurl, while intentionally
minimizing changes to existing code to guarantee the same behavior when
`LLAMA_CURL` is used.
model : support GLM 4.6 (make a few NextN/MTP tensors not required) (#16359)
* Make a few GLM tensors not required
layer.nextn.shared_head_head and layer.nextn.embed_tokens are both excluded from GLM 4.6 resulting in the model not loading after conversion/quantization, this marks those tensors as not required which makes it work
* Update llama-model.cpp
layer.nextn.shared_head_norm also not required in case of future models
`test-arg-parser.cpp` has been updated to work consistently,
regardless of whether CURL or SSL support is available, and
now always points to `ggml.ai`.
The previous timeout test has been removed, but it can be
added back by providing a dedicated URL under `ggml.ai`.
common : simplify etag tracking by removing json (#16342)
The JSON parser is temporarily kept only for backward compatibility. It
reads the etag from old .json files to prevent unnecessary re-downloads
for existing users.
This legacy code can be removed in a future version.
Daniel Bevenius [Fri, 26 Sep 2025 15:34:42 +0000 (17:34 +0200)]
ggml : remove -dev suffix from release version (ggml/1355)
This commit removes the `-dev` suffix from the version string in
CMakeLists.txt and the release script. The version will now be
just be formatted as `MAJOR.MINOR.PATCH`.
ggml-backend : add root cause in error message if loading backend library fails (#16172)
This PR adds additional information to an error message when loading backend library via ld_load_library() fails. This helps spotting why backend library did not load (missing library, missing dependency or unresolved symbol etc.).
During prompt ingestion, prompt tokens are accepted into the sampler history (for repetition penalties). The conversation-mode path then appended `common_sampler_last(smpl)` to `assistant_ss` before any new token was sampled. At that point, "last" was a prompt-side token (e.g., an input prefix), so the assistant chat message began with an extra piece.
Fix: append to `assistant_ss` only for a newly sampled (non-EOG) token. This affects only chat message assembly (`assistant_ss` / `chat_msgs` / `common_chat_format_single`); terminal stdout is unchanged. Sampling order/logits are unchanged.
Jeff Bolz [Sun, 28 Sep 2025 06:38:37 +0000 (01:38 -0500)]
vulkan: 64-bit im2col (#16135)
* vulkan: 64-bit im2col
Add variants of the im2col shaders that use buffer_device_address/buffer_reference,
and use 64-bit address calculations. This is needed for large convolutions used in
stable-diffusion.cpp.
Jeff Bolz [Sun, 28 Sep 2025 01:36:34 +0000 (20:36 -0500)]
vulkan: handle mat_mul with A matrix > 4GB (#16176)
* vulkan: handle mat_mul with A matrix > 4GB
This change splits mat_mul operations with huge A matrix into chunks in the M
dimension. This works well for stable-diffusion use cases where the im2col
matrix has very large M.
Fix the order of setting the stride in mul_mm_cm2 - setting the dimension
clobbers the stride, so stride should be set after.
Jeff Bolz [Sat, 27 Sep 2025 20:43:39 +0000 (16:43 -0400)]
vulkan: support arbitrary KV dimension in flash attention (#16160)
The "Clamp" spec constant is already based on whether KV is a multiple of Bc,
so use that to control whether bounds checking is performed. Add bounds checking
to the scalar and coopmat1 paths. Coopmat2 didn't need any changes (the K/V
tensors are already optionally clamped, nothing else needed to be changed).
Jeff Bolz [Sat, 27 Sep 2025 10:36:11 +0000 (06:36 -0400)]
vulkan: support GET_ROWS for k-quants (#16235)
The dequantize functions are copy/pasted from mul_mm_funcs.comp with very few
changes - add a_offset and divide iqs by 2. It's probably possible to call
these functions from mul_mm_funcs and avoid the duplication, but I didn't go
that far in this change.
model : make minicpm embedding_scale, residual_scale and logit_scale optional with legacy defaults (#16273)
* minicpm: make GGUF scaling keys optional with legacy defaults
Older MiniCPM GGUFs do not include the scaling metadata keys (minicpm.embedding_scale, minicpm.residual_scale, minicpm.logit_scale). The loader currently treats these as required, so quantization fails with:
key not found in model: minicpm.embedding_scale
This change restores backward compatibility by treating these keys as optional in the loader and using the older MiniCPM scaling values:
When the GGUF provides the keys, their values override the defaults; otherwise the legacy defaults are used. Newer GGUFs that already include these keys are unaffected.
Signed-off-by: Aaron Teo <redacted>
* devops: fix test does not exist
Signed-off-by: Aaron Teo <redacted>
* devops: fix model not found llama-eval-callback
Signed-off-by: Aaron Teo <redacted>
* Fix q3_K dot product error in test-quantize-fns on s390x
Array q8bytes had only 4 elements allocated, but 8 elements accessed.
This lead to write out of bounds and later read of overwritten values out of bounds
and incorrect result.
Signed-off-by: Aaron Teo <redacted>
* devops: re-enable ppc64le for testing
Signed-off-by: Aaron Teo <redacted>
* devops: activate test-thread-safety for s390x
Aleksander Grygier [Fri, 26 Sep 2025 16:35:42 +0000 (18:35 +0200)]
Allow viewing conversations even when llama server is down (#16255)
* webui: allow viewing conversations and sending messages even if llama-server is down
- Cached llama.cpp server properties in browser localStorage on startup, persisting successful fetches and reloading them when refresh attempts fail so the chat UI continues to render while the backend is unavailable.
- Cleared the stored server properties when resetting the store to prevent stale capability data after cache-backed operation.
- Kept the original error-splash behavior when no cached props exist so fresh installs still surface a clear failure state instead of rendering stale data.
* webui: extend cached props fallback to offline errors
Treat connection failures (refused, DNS, timeout, fetch) the same way as
server 5xx so the warning banner shows up when cache is available, instead
of falling back to a full error screen.
* webui: Left the chat form enabled when a server warning is present so operators can keep sending messages
e.g., to restart the backend over llama-swap, even while cached /props data is in use
common : use cpp-httplib as a cURL alternative for downloads (#16185)
* vendor : update httplib
Signed-off-by: Adrien Gallouët <redacted>
* common : use cpp-httplib as a cURL alternative for downloads
The existing cURL implementation is intentionally left untouched to
prevent any regressions and to allow for safe, side-by-side testing by
toggling the `LLAMA_CURL` CMake option.
Signed-off-by: Adrien Gallouët <redacted>
* ggml : Bump to Windows 10
ci : create git tags for released docker images (#16008)
* ci : create git tags for released docker images
When releasing a docker image for build number X, we should also create
the corresponding git tag. This allows users to easily checkout the
corresponding source tree for given docker image.
Aman Gupta [Thu, 25 Sep 2025 14:35:05 +0000 (22:35 +0800)]
CUDA: add a fused top-K MoE kernel (#16130)
* CUDA: add a fused top-K MoE kernel
This kernel does the following:
1. softmax over the logits per token [n_experts, n_tokens]
2. argmax reduce over the top-k (n_experts_used) logits
3. write weights + ids to global memory
It is intended as fusion of softmax->top-k->get_rows pipeline for MoE models
* Refactor into ggml_cuda_should_use_topk_moe
* Review: Use better coalescing pattern, use WARP_SIZE, store logits into registers before
Daniel Bevenius [Thu, 25 Sep 2025 10:02:36 +0000 (12:02 +0200)]
model-conversion : add embedding prompt file support (#15871)
This commit adds support for passing a prompt file to the model
conversion targets/scripts. It also updates the logits.cpp to print out
embedding information in the same format as when running the original
embedding model.
The motivation for this is that it allows us to pass files of different
sizes when running the converted models and validating the logits.
This can be particularly important when testing the sliding window
functionality of models where the sequence length needs to exceed a
certain number of tokens to trigger the sliding window logic.
Daniel Bevenius [Thu, 25 Sep 2025 09:36:47 +0000 (11:36 +0200)]
server : add support for external server for tests (#16243)
This commit adds support for using an externally started llama-server
instance for the server tests. This can be enabled by setting the
DEBUG_EXTERNAL environment variable.
The motivation for this is to allow debugging of the server itself
when investigating a test failure. Instructions for how to do this are
added to the README.md file in the tests directory.
Use RPC_DEBUG environment variable to enable debug messages.
Add helper macro LOG_DBG() which does an early
check of the env var before calling GGML_LOG_DEBUG().
Make sure we log a debug message for every server function.
ggml : split graph allocations according to backend max buffer size (#15815)
* ggml : make gallocr respect the backend's max buffer size
* if the graph requires more memory than can fit into a single allocation, split it into multiple backend buffers
* vulkan: report the actual max allocation size in buffer type interface
* fix missing newline, apple-clang warning
* track size of individual chunks in ggml_dyn_tallocr and raise max chunks.
revert to use suballocation_block_size as max chunk size for vulkan.
* track (chunk, offset) pairs instead of "global" offsets through gallocr.
* simpler, don't need loops to map between local/global offsets
* touches more code
* fix dyn_tallocr_max_size and initialization
* fix memory leak when buffers are reused due to same buffer type appearing multiple times
* make vbuffer allocation follow the same logic as backend_buffer did before
* continue to use leftover unallocated space of previous chunks after a new one has been created
* treat free blocks of each chunk as separate list
* they're still allocated together, but start/end of each chunk is tracked, and allocate/free iterate over sub-ranges
* exhaust freed blocks of all chunks before considering their last blocks with unallocated space
* start with 0 chunks/blocks and create chunks as needed
* allow the last chunk to grow beyond max size
* refactor: move adding new free block and new chunk into separate functions
* allocate chunks individually with a separate free-blocks list for each one
* needs a bit more memory/allocations/indirections, but code is simpler
overview / pkg / ggml / sources / llama.cpp / log