* provide static workspace for cuBLAS handles
* account for concurrent streams when using GGML_CUDA_GRAPH_OPT
* drop cublas_handle overloads and remove direct cublasSetStream calls
* Update ggml/src/ggml-cuda/common.cuh
---------
Co-authored-by: Oliver Simons <osimons@nvidia.com>
* backend: propagate buffer usage in meta backend
* ggml-meta: make sure to call init_tensor for all new tensors
* meta: remove explicit check for meta backend in ggml_backend_meta_get_split_state
I can't seem to reproduce the original failure in the latest code.
* hexagon: fix FA HMX queue ordering in the pipelined path
* hexagon: double buffer D matrix, store diagonal tile only
* format code
* align the indentation
* opencl: port fused ssm_scan kernel (Mamba-2, d_state in {128, 256})
Fold the fused per-token SSM_SCAN recurrent step from opencl/gdn-qwen36-35b
onto the unified base. Previously SSM_SCAN fell back to CPU here; now scalar-A
Mamba-2 with d_state in {128,256}, all-f32, runs on GPU. Other shapes (incl.
Mamba-1 element-wise A) still fall back. test-backend-ops -o SSM_SCAN passes on
Adreno X2-90. opt-out via GGML_OPENCL_DISABLE_SSM_SCAN=1.
* opencl: cleanup
* opencl: require K == 1
---------
Co-authored-by: Li He <lih@qti.qualcomm.com>
* ggml-cpu: gate __fp16 on __ARM_FP16_FORMAT_IEEE
__ARM_NEON only signals NEON availability. The __fp16 type also needs
the IEEE half format, implied on AArch64 but selected with
-mfp16-format=ieee on 32 bit Arm, where the compiler otherwise rejects
the type.
The guard keeps every toolchain that provides the type on the same code
and sends that one configuration to the generic lookup path.
* ggml-cpu: gate the NEON+FMA block on __ARM_FP16_FORMAT_IEEE
Both halves of the F16 section dereference __fp16, so armv7 with
neon-vfpv4 hits the same unknown type error. Without the IEEE
format the configuration now falls back to the scalar path.
Address review from @JonathanC-ARM
* vulkan : dequant q8_0 KV once in coopmat1
Assisted-by: Claude (Opus 4.8)
* vulkan : fall back instead of aborting when FA scratch exceeds maxStorageBufferRange
* vulkan : require KV-cache layout in FA dequant path
Assisted-by: Claude (Opus 4.8)
* vulkan : skip FA dequant path on coopmat2
Assisted-by: Claude (Opus 4.8)
* tests : add contiguously-allocated quant K/V FA tests
Assisted-by: Claude (Opus 4.8)
* vulkan : trim comments
* vulkan : tighten permutation checks for FA path
* vulkan : set prealloc_x_need_sync after the FA dispatch
* vulkan : exclude Intel Xe1 from FA dequant path
* add params
* cpu kernel
* metal kernel
* add test backend ops
* gate other backends
* ggml: (cuda) support ggml_rope_set_offset (llama/27121)
* rm cuda supports_op guard, fix webgpu clang-format
* ggml: support ggml_rope_set_offset on vulkan (llama/27344)
* ggml: support ggml_rope_set_offset on vulkan
* remove inplace optimization
* vulkan: tiled transpose for 0<->2 permuted CONT
-ggml_vk_get_cpy_pipeline only routed to the tiled shared-memory transpose
shader when dim1 was the innermost dimension, i.e. ggml_transpose (a 0<->1
swap). A 0<->2 swap -- ggml_cont(ggml_permute(x, 2, 1, 0, 3)) -- fell back to
the generic per-element strided copy, whose source reads stride by ne0*ne1
elements: one cache line per lane.
-DeepSeek-V4's lightning indexer performs exactly that permute on a
[n_kv, n_tokens, n_head] tensor. On Vulkan/RADV gfx1151 it ran at ~1-9 GB/s of
a ~200 GB/s part and accounted for 43% of total prefill time.
-Add copy_transpose_02.comp, mirroring copy_transpose.comp but tiling over dst
dims (0, 2) with dims 1 and 3 as the batch, so reads walk src dim2 and writes
walk dst dim0 -- both contiguous. The selection condition additionally requires
a non-contiguous source and a contiguous destination so it cannot take cases
the contiguous-copy shader already handles.
-test-backend-ops only exercised ggml_transpose for CONT, so the strided path
was untested. Add test_cont_permute covering (2,1,0,3), (1,2,0,3) and (0,2,1,3)
over f32/f16 at tile-aligned, tile-unaligned and large shapes. The large shapes
are in the eval set rather than only in perf because perf mode does not verify
results.
-Measured on gfx1151, ne=[n_kv,64,64,1], perm=(2,1,0,3), f32:
n_kv=1024: 9.08 -> 579.85 GB/s
n_kv=1280: 20.03 -> 153.71 GB/s
n_kv=2048: 7.11 -> 91.68 GB/s
n_kv=2304: 16.24 -> 86.49 GB/s
-The ~2.2x penalty previously seen at power-of-two n_kv (destination-stride
aliasing) is gone. End to end, DeepSeek-V4-Flash IQ3_XXS prefill on a 9k-token
prompt goes from 56.33 t/s to 103.74 t/s (+84%).
-Note: at n_tokens=512 a single slow-path dispatch takes ~273 ms and looping it
in perf mode can trip the GPU watchdog, so the perf cases use n_tokens=64.
* tests: fold test_cont_permute into test_cont, add L2-exceeding perf shapes
Review feedback: test_cont gains a permute parameter ({0,0,0,0} = none),
matching test_mul_mat's pattern, and the separate struct is gone. Perf
adds [n_kv, 512, 64, 1] variants (~0.5 GB per run) that exceed GPU L2,
since the 64-token shapes fit in cache on large parts and read above
memory bandwidth.
* tests: trim perf-case comment to the two-line summary
* vulkan: trim comments on the 0<->2 transpose path
Drop the shader file header, the read/write block comments and the
rationale prose in the CONT test cases. Keep the tile-shape and
bank-conflict notes and the permute parameter documentation.
---------
Co-authored-by: Kevin Hopper <no-reply@maestro.press>
Kernel is a port of `ggml-cuda/fwht.cu`
(us/run, median):
```
m x n x k GEMM FWHT speedup
64 x 1 x 64 10.20 2.93 3.48x
64 x 2048 x 64 10.75 2.71 3.97x
128 x 1 x 128 10.33 2.88 3.59x
128 x 32 x 128 9.20 2.77 3.33x
128 x 2048 x 128 16.46 2.76 5.95x
256 x 1 x 256 10.19 2.77 3.68x
256 x 2048 x 256 16.69 3.41 4.89x
512 x 2048 x 512 54.16 12.89 4.20x
```
* added check for nullptr for wctx
Signed-off-by: GodRishUniverse <risagarw@amd.com>
* Apply the update for failed message in tests/test-vad-full.cpp
Co-authored-by: Daniel Bevenius <daniel.bevenius@gmail.com>
---------
Signed-off-by: GodRishUniverse <risagarw@amd.com>
Co-authored-by: Daniel Bevenius <daniel.bevenius@gmail.com>
max_len defaults to 60 in the server, and that only takes effect on the token
timestamps path. Until v1.8.3 token timestamps were limited to verbose_json, so
the default never reached other formats. #3679 removed that condition, which
left every response wrapped at 60 characters, on a token boundary rather than a
word one.
Resolve the default after the other parameters and enable it only for
verbose_json, or when max_len or split_on_word was asked for. Clients passing
max_len with any format still get wrapping.
Nine spelling mistakes across the main readme and seven example readmes:
geneated -> generated, recieved -> received, difinitions -> definitions,
continously -> continuously, ect -> etc, and "lasted"/"recommanded" ->
"latest"/"recommended" in the CANN section.
Found with codespell; each one read in context first. Docs only.
Assisted-by: Claude (Anthropic)
This commit moves the close-issue.yml file to the .github/workflows/
directory.
The motivation for this is that it is currently in the root of the repo
and is not currently active. Moving it to the workflows directory will
make it active and allow it to run.
This commit removes package.json from git and adds package.json to the
gitignore.
This file is generated by CMake configure and will be updated and
overwritten causing a lot of noise in the git history.
It have also been updated manually as part of our release process which
will not be needed either.
* cmake : update semver handling to be consistent with ggml/llama.cpp
This commit modifies the semantic version handling to be consistent with
how llama.cpp and ggml handle semver.
This commit introdues a new example named test-cmake which is intended
to be used to test the cmake configuration and installation.
* ci : update release workflow to be consistent with llama.cpp
work in progress...
* ci : fix if statement in release.yml
* ci : comment out all but one build in release.yml
This is just for testing and this commit should not be included in the
main PR later.
* ci : use DEPLOY_KEY_RELEASE
This commit updates the release and make-release workflows to use the
DEPLOY_KEY_RELEASE secret. Two github ruleset have been imported.
* ci : add github rulesets for releases
These were retrived from llama.cpp and then imported into my fork for
testing. If all works well they will be imported into whisper.cpp
upstream as well.
* fix move artifacts step
* examples : use FetchContent for llama.cpp in talk-llama
This commit updated the example talk-llama to remove the vendored
llama.cpp and instead use FetchContent to pull it in from the
upstream repo.
* ci: add GGML_NATIVE=OFF to build-clang.yml
This commit disables native CPU instructions from the ubuntu-22-clang
job.
The motivation for this is that currently it is possible that the
running compiling llama.cpp (via ccache) might have support for cpu
instructions that are not available on the target runner.
Refs: https://github.com/ggml-org/whisper.cpp/actions/runs/32224048267/job/95980031403?pr=3996
* ci : add missing GGML_NATIVE=OFF to jobs
* ci : add attestation for signed release artifacts
This commit add attenstions of artifacts to the release workflow.
After building the artifacts can be verified with the following command:
```console
$ curl -sSL -o whisper-bin-ubuntu-x64.tar.gz \
https://github.com/danbev/whisper.cpp/releases/download/b4947/whisper-bin-ubuntu-x64.tar.gz
$ gh attestation verify --repo danbev/whisper.cpp whisper-bin-ubuntu-x64.tar.gz
Loaded digest sha256:722a6812263195d7ee2192b57fc64a6d6b09a6cdf2f55a152f793db27a651e31 for file://whisper-bin-ubuntu-x64.tar.gz
Loaded 1 attestation from GitHub API
The following policy criteria will be enforced:
- Predicate type must match:................ https://slsa.dev/provenance/v1
- Source Repository Owner URI must match:... https://github.com/danbev
- Source Repository URI must match:......... https://github.com/danbev/whisper.cpp
- Subject Alternative Name must match regex: (?i)^https://github\.com/danbev/whisper\.cpp/
- OIDC Issuer must match:................... https://token.actions.githubusercontent.com
✓ Verification succeeded!
The following 1 attestation matched the policy criteria
- Attestation #1
- Build repo:..... danbev/whisper.cpp
- Build workflow:. .github/workflows/release.yml@refs/heads/master
- Signer repo:.... danbev/whisper.cpp
- Signer workflow: .github/workflows/release.yml@refs/heads/master
```
* cmake : add WHISPER_USE_SYSTEM_LLAMA option [no ci]
This commit adds a new CMake option WHISPER_USE_SYSTEM_LLAMA that allows
the talk-llama example to use a system-installed llama.cpp library.
Setting this will automatically also set WHISPER_USE_SYSTEM_GGML to ON
and the system ggml library will be used in addition to the system
llama.cpp.
* ci : remove unused ccache step
* Revert "ci : comment out all but one build in release.yml"
This reverts commit 24b56776e1.
* ci : set WHISPER_BUILD_IS_DEV=OFF in release.yml
* CUDA: MMVQ nwarps=8 for bs=1 for dense models on DGX Spark
Signed-off-by: ynankani <ynankani@nvidia.com>
* skip moe experts and allow others based on k geometry (allow only small idle tail)
Signed-off-by: ynankani <ynankani@nvidia.com>
* rename MMVQ DGX Spark params to GB10 and fix MSVC constexpr lambda capture
Signed-off-by: ynankani <ynankani@nvidia.com>
---------
Signed-off-by: ynankani <ynankani@nvidia.com>
Adjusts the thread/block count to be proportional to the size
of the quant, reducing under/over subscription.
Largest perf improvement is the q4_0 -> f32 path, with, on
a Arc 70, throughput goes from 20.21 GB/s to 158.19 GB/s
The rest of the quants are flat in performance uplift.
* Initial changes for Recurrent state rollback for nemotron for cpu and cuda
* Removing CPU RS rollback. Will enable it in subsequent PRs
* addition of test case
* Removing assert and calling runtime API to check if op is supported
* removing extra API and updating the call sites for K
* replace static cuda detection to runtime fused_op api
* address review comments and fallback when SSM rollback not supprted
* Adding changes for supporting RS-rollback in CPU. Also added test-backend-ops for cpu and cuda
* removing memory manipulation as rs rollback is now supported in CPU
* removing the static probe which is not needed now
* correcting the format
* address review comments
* enabling test for all the backends, unsupported backends will fallback to CPU
* Apply suggestions from code review
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
* choose different graph based on the result of fused_ssm_op is supported or not and also handled memory->n_rs_seq >1 case incase of op is not supported
* Support K > 1 in ssm_scan for all backends
* Fix CI Issues
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: Gaurav Garg <gaugarg@nvidia.com>
* OpenVINO backend: 1) enable gpt-oss moe on OV bk; 2) enable mxfp4 support
* OpenVINO backend: disable TOPK_MOE op test
* OpenVINO Backend: Add op FILL support
* OpenVINO backend: enable set rows with multi dims
* fix the name missmatch in setrow + view
* OpenVINO backend: enable op GGML_UNARY_OP_SIGMOID
* OpenVINO Backend: enable SQR & SQRT
* OpenVINO backend: 1) ensure unique node names for OpenVINO; 2) add org_src to recorde the src ggml tensor for OpenVINO dynamic shape infer
* OpenVINO backend: enable fallback for openVINO to CPU backend
* OpenVINO backend: fix accurace issue in gemma3n arch test
* fix mpt failed case
* OpenVINO backend: clean nodeinfo
* OpenVINO Backend: enable zero-size copy for view
* add concat ssm_conv in compute_dynamic_dim
enable qwen35
Fix after rebase
remove logging
* OpenVINO backend: disable EXP with FP32, which failed in op test. Root reason: the backend test initializes unary op inputs over a wide range, [-150, 150]. For FP32, exp(x) overflows around x ~= 88.7, so this test can randomly generate values right in or beyond the overflow region
* OpenVINO backend: fix CPY op test failed issue
* OpenVINO backend: fix GATED_DELTA_NET op test failed issue
* handle in-place op, handle qwen35 dynamic clearing of cache in cgraph
* handle qwen35 dynamic clearing of cache correctly
* Enable qwen35 dense multi seq
* Fix qwen35 9b gqa
* Fix after rebase
* Disable SOLVE_TRI
* openvino: fix NEOX RoPE accuracy on GPU stateful (mixed-rank Multiply)
In stateful mode the NEOX RoPE branch fed rank-3 data ([S, n_heads,
head_size]) into the Multiply against the rank-4 cos/sin tables
([1, S, 1, n_dims/2]). That mixed-rank broadcast is miscomputed by the
OpenVINO GPU plugin, corrupting the rotated Q/K and producing garbage
output (e.g. Phi-3-mini). Lift the data to rank-4 before the split/
Multiply so the operands are equal-rank, matching what the TYPE_NORMAL
branch already does. CPU and stateless paths are unaffected.
Phi-3-mini-Q4_K_M, wiki.test perplexity, GPU stateful:
before: PPL = 27120.43
after: PPL = 6.2263 (CPU reference: 6.2251)
* OpenVINO backend: 1) remove the unique name in llama.cpp; 2) add new ov name in ov bk; 3) fix issue in arch test & op test with latest code update
* OpenVINO Backenb: remove changes in llama.cpp
* Doc change (use x64 Native Tools Command Prompt for VS)
* Cleaner sentence
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
* OpenVINO Backend: cache key upgrade includes all src name
* OpenVINO Backend: enable llama arch test on ci
* OpenVINO Backend: move parameter node creating from decoder into translate
* OpenVINO Backend: create extra input ov node move from decoder to translate
* fix for op regression due to is_model_splitted
* openvino: fix CPY writeback for recurrent state rollback
Detect the rollback conv/gdn state writeback CPY nodes structurally
instead of by tensor name, since the rollback path in
build_conv_state does not call cb() and left the nodes unnamed. Add
per-node runtime offsets (rs_slot_begin_*, rs_src_begin_*) so the
cached IR handles any kv head, sequence count and snapshot slot for
both the conv state and the GDN state writeback.
Assisted-by: GitHub Copilot
* qwen35 moe
* optimize MoE expert aggregation with ReduceSum
* Skip GET_ROWS inaccurate test
* openvino: fallback dynamic MUL_MAT_ID shapes
* OpenVINO Backend: fix error in arch test model mpt
* fix error caused by cpy in arch test model kimi-linear
* OpenVINO Backend: fix error in arch test model minimax-m3
* openvino: fix GPU mul_mat_id op tests
* ggml-openvino: add GGML_OPENVINO_RELEASE_WEIGHTS to reclaim host weight RSS on GPU
The OpenVINO weight Constants are zero-copy views into host buffers
allocated by the backend (ggml_aligned_malloc, anonymous memory). On GPU
the plugin holds its own device copy after compile_model, so these host
pages are dead weight for inference. For a 1B Q4_K_M model this leaves
~850 MB of host RSS resident that the GPU path never reads again.
Add an opt-in GGML_OPENVINO_RELEASE_WEIGHTS mode that madvise(MADV_DONTNEED)s
the registered host weight buffers once the model is compiled, dropping
their resident pages while keeping the mappings valid (ggml still owns the
lifetime; tensors still point in). Measured steady-state RSS drops from
~1555 MB to ~710 MB on Llama-3.2-1B-Q4_K_M (Arc iGPU) with unchanged
throughput and correct output.
The GPU backend uses a single dynamic-shape model for both prefill and
decode, so a graph is compiled once and reused; the only event that forces
a recompile is clear_caches() on backend teardown. The change therefore:
- releases on the first cache-hit (model compiled, plugin has its copy);
- pins the compiled-model cache across backend teardown so a later
context reuses it instead of recompiling against the dropped pages;
- fails loud (GGML_ABORT) on a cache-miss recompile or on a second model
load, both of which would otherwise read zeroed weights or silently
reuse the wrong compiled graph.
Scope/limitations (all fail loud, never silently wrong): GPU only (the CPU
plugin reads the host Constants at inference time), one model per process,
and stable graph shapes. This reduces steady-state RSS, not the transient
compile-time peak. All changes are confined to the OpenVINO backend.
* ggml-openvino: stream weight requantization to cut the compile-time RSS peak
requantize_to_buffers() dequantized the entire tensor to a temporary
std::vector<float> of n_elements before requantizing. For token_embd.weight
(128256 x 2048) that transient is ~1 GB (1B model) / ~2 GB (8B), and it is
the single largest contributor to the OpenVINO compile-time memory peak --
it also fires twice for token_embd (once at load, once at graph build,
because token_embd is loaded via a CPU/mmap buffer and not cached as an OV
weight extra).
Stream the dequant instead: process a fixed window of complete rows
(CHUNK_ROWS=256) into a small scratch buffer and quantize/convert each chunk
straight into the output buffers. The transient F32 footprint is now
CHUNK_ROWS*ne0 floats regardless of tensor size.
quantize_q8_0/q8_1 gain an optional block_offset arg (default 0) so a chunk
writes its weights/scales/zp at the correct block. Streaming is applied to
the Q8_0_C / Q8_1_C / F16 targets (the large requant cases); the u4 (Q4_0)
path keeps the whole-array call because it packs two weights per byte with
running zp ORs, and a fallback handles any future target whose block size
does not divide a row.
Measured peak RSS (cold compile, GPU): 1B 2868 -> 1809 MB (-1.06 GB);
8B 11618 -> 9608 MB (-2.0 GB). Output verified unchanged
("capital of France is Paris"); throughput unchanged. Unlike
GGML_OPENVINO_RELEASE_WEIGHTS this reduces the transient peak, not just
steady-state, and needs no env flag. All changes confined to the OpenVINO
backend.
* ggml-openvino: avoid redundant token_embd requantization at compile
token_embd.weight is referenced twice in the graph path: as the GET_ROWS
embedding (a CPU/mmap-buffer tensor) it was re-extracted/re-requantized on
every weight-node build, and is_model_splitted() built a full (naive) set of
weight nodes just to test name membership — each requant is a ~1-2 GB F32
dequant of the 262M-element embedding.
Two changes:
- Add collect_weight_names(): a name-only collector for topology checks.
is_model_splitted() now uses it instead of create_weight_nodes(cgraph,
true), so the splitted-check no longer triggers any weight extraction.
- Memoize weight nodes built from non-OpenVINO buffers in a process-lifetime
cache keyed by tensor->data. These tensors have no OV buffer context to own
a cached extra, so without this they were rebuilt on every (re)compile;
prefill and decode graphs now share one build (verified: 2nd graph hits the
cache instead of re-requantizing).
Peak RSS is unchanged (the streaming-requant commit already removed the F32
transient); this removes redundant compile-time work. Output verified
unchanged ("capital of France is Paris"). Confined to the OpenVINO backend.
* ggml-openvino: gate compile-memory optimizations behind GGML_OPENVINO_REDUCE_COMPILE_MEM
The streaming requantization and the non-OpenVINO-buffer weight-node cache
(plus the name-only is_model_splitted path that pairs with it) are now opt-in
via GGML_OPENVINO_REDUCE_COMPILE_MEM. When unset, requantize_to_buffers()
fully materializes the F32 buffer and weights are rebuilt per compile exactly
as before; when set, the streaming path and the cross-compile weight cache
are used.
Default off keeps behavior identical to upstream unless explicitly enabled.
Verified: flag off -> peak RSS 2800 MB (original), flag on -> 1810 MB; output
"capital of France is Paris" in both modes. (GGML_OPENVINO_RELEASE_WEIGHTS,
added earlier, remains a separate opt-in for the steady-state release.)
* ggml-openvino: add frontend model cache (GGML_OPENVINO_MODEL_CACHE_DIR)
The plugin-level ov::cache_dir caches the compiled blob keyed by the OV
model, but producing that model still runs the full frontend every time:
weight requantization (incl. the large token_embd F32 transient) and the
ggml->OV graph conversion. This adds an opt-in frontend cache keyed off a
fingerprint computed directly from the ggml cgraph, so a hit imports a
previously exported CompiledModel and skips requant + convert + compile
entirely.
Key (model-cache.{h,cpp}) = 64-bit FNV-1a of: graph topology (n_nodes + per
node op/name), a sampled per-weight fingerprint (name/shape/type + bounded
head+tail byte sample), and blob-affecting config (device, flash-attn, rope
params, REDUCE_COMPILE_MEM/stateful flags, OpenVINO version). A sidecar
manifest stores every weight's fingerprint and is re-verified on load, so a
sampled-hash collision cannot cause a wrong-model hit (verified: two
different quantizations of the same model produce distinct cache entries).
Flow (dynamic single-model path only; split models defer to ov::cache_dir):
on a verified hit, core.import_model() restores the CompiledModel and a
lightweight decoder is built with a names-only weight map (membership is all
the decoder needs for I/O mapping; weights live in the imported model). On a
miss, compile as usual then export the blob (atomic temp+rename, manifest
written first). The frontend cache supersedes ov::cache_dir, so CACHE_DIR/
CACHE_MODE are stripped from the config used for the cached compile and the
import — a blob compiled with cache_dir set cannot be re-imported.
Measured 8B Q4_K_M (GPU): full requant+convert+compile 15.3s -> import 6.3s
(~2.4x faster compile phase). Output verified unchanged on cold and warm,
standalone and combined with REDUCE_COMPILE_MEM + RELEASE_WEIGHTS. Default
off; confined to the OpenVINO backend.
* ggml-openvino: harden frontend model cache correctness
The frontend model cache imports a previously exported CompiledModel keyed by a fingerprint of the ggml graph, weights, and blob-affecting config. The original key covered device, stateful execution, REDUCE_COMPILE_MEM, RoPE params, OpenVINO version, topology, and sampled weights, but missed runtime/frontend toggles that can change the lowered graph or the I/O binding contract. That made it possible to reuse a blob produced under a different OpenVINO backend configuration.
Add a small extra-config helper for the dynamic model-cache path and fold in the effective values of GGML_OPENVINO_DISABLE_KV_SLICE and GGML_OPENVINO_MANUAL_GQA_ATTN. MANUAL_GQA_ATTN is keyed by the behavior that actually takes effect: an explicit env value wins, otherwise GPU defaults to enabled and other devices default to disabled. This matches flash_attn_ext lowering and avoids unnecessary cache splits for equivalent configurations while separating genuinely different attention graphs.
DISABLE_KV_SLICE is also included because it changes the KV-cache tensor shape/output binding strategy used around imported models. Even when weights and graph topology are identical, switching this flag should not inherit a CompiledModel cache entry created for a different binding mode.
Also make cache artifact publication cleaner: write manifest.tmp and blob.tmp, publish the blob first, and publish the manifest last. Cache hits already require both blob and a verified manifest, so making the manifest the final visible artifact avoids leaving an apparently complete manifest for a failed or interrupted blob export. Temporary files are removed on the handled failure paths.
While touching this path, fix the indentation of the non-imported compile branch so the cache miss flow is easier to review. Behavior is otherwise unchanged: verified hits still import, misses still create weights, convert, compile, export, and create the infer request normally.
* ggml-openvino: add memory optimization umbrella switch
Add GGML_OPENVINO_MEMORY_OPTIMIZE as a single opt-in switch for the OpenVINO backend memory-saving paths. The existing fine-grained GGML_OPENVINO_REDUCE_COMPILE_MEM and GGML_OPENVINO_RELEASE_WEIGHTS variables remain supported and explicitly override the umbrella switch when set, so users can still bisect or disable one side of the optimization independently.
Centralize the policy in ggml_openvino_reduce_compile_mem_enabled() and ggml_openvino_release_weights_enabled(device). The umbrella switch enables compile-memory reductions everywhere REDUCE_COMPILE_MEM is used today: streaming requantization, non-OV weight-node caching, split-model weight-name collection, and the frontend model-cache fingerprint. On GPU it also enables host weight-buffer release unless GGML_OPENVINO_RELEASE_WEIGHTS is explicitly set.
Keep host weight release GPU-only because it relies on the plugin holding its own device copy after compile_model. Update the fail-fast diagnostic and comments to mention GGML_OPENVINO_MEMORY_OPTIMIZE, so users who enable the umbrella switch get accurate guidance if a later cache-miss recompile would read released host weight pages.
* ggml-openvino: rename compiled model cache env
Rename the frontend export/import cache environment variable from GGML_OPENVINO_MODEL_CACHE_DIR to GGML_OPENVINO_COMPILED_MODEL_CACHE_DIR. The cache stores blobs produced by ov::CompiledModel::export_model() and restores them with core.import_model(), so the new name distinguishes it from GGML_OPENVINO_CACHE_DIR, which configures OpenVINO plugin-level ov::cache_dir.
Update the registered env var, the cache-directory lookup, and comments around the frontend compiled-model cache. The old GGML_OPENVINO_MODEL_CACHE_DIR name is removed rather than kept as a fallback so there is a single spelling for the new option.
* docs: document OpenVINO memory optimization env vars
Add runtime configuration entries for the newly recognized OpenVINO environment variables.
Document GGML_OPENVINO_COMPILED_MODEL_CACHE_DIR as the frontend compiled-model cache used to export and import compiled blobs for matching single-graph models.
Document GGML_OPENVINO_MEMORY_OPTIMIZE as the umbrella switch, including how GGML_OPENVINO_REDUCE_COMPILE_MEM and the GPU-only GGML_OPENVINO_RELEASE_WEIGHTS override or inherit from it.
* ggml-openvino: fix Qwen3VL crash and deepstack correctness bug
1. GGML_OP_PAD was missing from compute_node_dynamic_dims(), causing a crash
on decode for models that pad the token embedding (n_embd -> n_embd_inp).
PAD never reorders/merges dims, so it keeps the same dynamic dim index as
its source.
2. process_view_input_new() chained VIEW inputs through src[0] (the
immediate op-graph parent) using offsets treated as relative to that
parent. But ggml_tensor::view_offs is always absolute from the true root
allocation (ggml collapses VIEW-of-VIEW chains internally). For the
per-layer deepstack view ("embd (view)", whose src[0] is "embd" - itself
an already-narrowed, zero-offset VIEW of the padded root, with the SAME
ggml shape as the deepstack view but a different absolute offset), this
caused an out-of-bounds re-slice that silently fell back to returning the
wrong (already-resolved sibling) tensor. In practice every deepstack ADD
ended up adding the real base token embedding into the residual stream
instead of zero, corrupting generation ("Hello my name is 1000000..."
instead of coherent text). Fixed by detecting this pattern (same shape as
the immediate src, different absolute offset) and re-slicing directly
from the untouched root tensor using the innermost view's absolute
offset.
Also adds a GGML_OPENVINO_DEBUG_NODE=<name1>,<name2>,... env var that attaches
extra debug Result nodes for arbitrary intermediate tensors, without binding
them to any ggml buffer (avoiding the risk of reading a ggml buffer that has
since been overwritten by a later in-place op). This was instrumental in
diagnosing bug #2 above and is left in as a general-purpose debugging aid.
* ggml-openvino: fix IMROPE inp_pos padding for NPU static shapes
IMROPE's inp_pos tensor packs 4 stacked t/h/w/e position planes into
ne[0] = 4*n_tokens instead of one value per token. On NPU's static-shape
path, inp_pos was padded/shaped as if it held a single plane, which
interleaved padding across the 4 planes and desynced later reshapes
from the rest of the (chunk_size-wide) graph.
- add GgmlOvDecoder::get_inp_pos_n_planes() to detect IMROPE's 4-plane layout
- get_graph_input_shape(): size inp_pos as n_planes * chunk_size (prefill)
or n_planes (decode) instead of assuming 1 value per token
- get_ov_input_tensor_static_prefill(): pad each plane to chunk_size
independently instead of one flat block
- get_ov_input_tensor_static_decode(): copy n_planes contiguous values
instead of asserting/copying a single scalar
* disable test-llama-archs tests.
* openvino: gate fallback with env var
* Revert changes in test-llama-archs
* Apply editor config
* reject CPY with quantized destination as unsupported
---------
Co-authored-by: Xuejun <Xuejun.Zhai@intel.com>
Co-authored-by: Mustafa Cavus <mustafa.cavus@intel.com>
Co-authored-by: virajwad <84867530+virajwad@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: suryasidd <surya.siddharth.pemmaraju@intel.com>
Co-authored-by: Mustafa Cavus <mustafacavus@intel.com>
Co-authored-by: Ravi Panchumarthy <ravi.panchumarthy@intel.com>
* metal: add TQ2_0 support
Add support for the GGML_TYPE_TQ2_0 (ternary, 2 bits per element) type in
the Metal backend.
Assisted-by: llama.cpp:DeepSeek-v4-Flash-0731
* cont : optimize mul_mv kernel
- float ops over integer ops
- precalculate sums
- hoist coef out of the inner loop
- contiguous y loads
llama.cpp:DeepSeek-v4-Flash-0731
Alexander Heisler
Markus Tavenrath
Hongqiang Wang
Max Krasnyansky
Yiwei Shao
Pascal
Nathanw1014
Jetson Tan
Titaniumtown
Xuan-Son Nguyen
Georgi Gerganov
Kevin Hopper
Masashi Yoshimura
Aman Gupta
Fred N. Garvin
Rishabh Agarwal
Daniel Bevenius
Kayvan Zahiri
vk
ynankani
Mario Limonciello
Neo Zhang
Jiang, Fish
Masato Nakasaka
lnigam
Mendy Berger
Zijun Yu
jinzihao
Ian Faust