Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1959 commits
Select commit Hold shift + click to select a range
922f53e
vulkan: fix CI llvmpipe by allowing dim 3 broadcast in MUL_MAT (#31239)
ServeurpersoCom Oct 27, 2026
8542e94
tests : speed up tensor data generation and model saving (#31145)
ggerganov Oct 27, 2026
7ecb57f
rpc : catch std::length_error in recv_msg (#31204)
sohaibharraoui Oct 27, 2026
27b7a8a
llama : build the k-pool sparse selection independently of the kq_mas…
am17an Oct 27, 2026
e86f5fd
cmake: WASI doesn't support spawning (#30751)
MendyBerger Oct 27, 2026
19cce69
model : add LiquidAI/LFM3-1.7B and LiquidAI/LFM3-VL-2B (#31252)
tdakhran Oct 27, 2026
6c44591
snapdragon: switch QDC artifact upload and log download to v2 API (#3…
mengshengwu Oct 27, 2026
eb7818c
common : support deferred tools and tool_reference content parts (#31…
tdakhran Oct 27, 2026
19ff817
model : add LiquidAI/LFM3-Audio-2B and LiquidAI/LFM3-Omni-2B (#31257)
tdakhran Oct 27, 2026
338fce6
HIP/CUDA: remove pragma unroll from loops in conv2d where itteration …
IMbackK Oct 27, 2026
a919179
tests: fix CI riscv timeout in test-llama-context (#31265)
ServeurpersoCom Oct 27, 2026
1b4554e
server: expose the reasoning effort levels of the chat template (#31271)
ServeurpersoCom Oct 27, 2026
d05706b
model : add LiquidAI/d1-2B and LiquidAI/d1-omni-2B (#31275)
tdakhran Oct 27, 2026
8ca4060
server : add realtime conversation and transcription on /v1/realtime …
tdakhran Oct 28, 2026
6886628
ui : Search dialog (#31173)
allozaur Oct 28, 2026
b1756b2
ui: Fix draft for new message disappearing after using "Add System Me…
allozaur Oct 28, 2026
02d502f
ui: Support OGG audio attachments by converting them to WAV (#31237)
allozaur Oct 28, 2026
f4d8af6
ui : show save and reset settings buttons only when relevant (#31248)
allozaur Oct 28, 2026
a71c00f
chat : stop qwen3-coder reasoning and content before template separat…
7konigsberg Oct 28, 2026
2864937
vulkan: clamp A prefetch to end_k in coopmat1 int8 matmul (#30283)
SelfRef Oct 28, 2026
a7af149
cuda : add top-k MoE fusion for 384 experts (#31288)
developer0hye Oct 28, 2026
f1a7dea
tests : adjust constants to reduce irrelevant warnings (#31254)
ggerganov Oct 28, 2026
e486a4b
ui: build the reasoning menu from the template effort levels (#31273)
ServeurpersoCom Oct 28, 2026
ec941ee
ggml : bump version to v.28.0 (ggml/1676)
ggerganov Oct 28, 2026
5d7793d
sync : ggml
ggerganov Oct 28, 2026
d1eb0de
server : support tool calls in /v1/realtime (#31287)
tdakhran Oct 28, 2026
d4f1b18
mtmd: enable video input support for LFM3 model family (#31283)
tdakhran Oct 28, 2026
8642541
llama.cpp : bump version to 0.8.0 (#31309)
ggerganov Oct 28, 2026
28cb47c
gguf-py: fix np_roundf dropping the sign (#31310)
ngxson Oct 28, 2026
f5222f2
cuda: fuse MoE (MUL_MAT_ID) gate/up and GLU in MMQ (#31188)
anujj Oct 28, 2026
39febd2
sycl: work-group argmax for short TOP_K rows (#31110)
newjordan Oct 28, 2026
604cc43
codeowners : refresh teams and maintainer PRs link (#31316)
ggerganov Oct 28, 2026
838093f
llama : remove simple batch splitting for DSV4 (#31300)
developer0hye Oct 28, 2026
ff0d396
ci(s390x): switch to Ubuntu 26.04 runners (#31317)
taronaeo Oct 28, 2026
c02b5c6
ggml-cpu: SVE 256-bit Implementation of mxfp4 gemv kernel (#28141)
abhijain1204fujitsu Oct 28, 2026
ac8e169
ggml-cpu: SVE 256 bit implementation of gemm_q6_K_8x8_q8_K kernel (#2…
anubhavfujitsu Oct 28, 2026
c3df4b7
Speed up IQ4 quant creation with pre-computed halfway points (#31201)
bartowski1182 Oct 28, 2026
5220213
hexagon: add conv2d support (#30995)
tboinovski1 Oct 28, 2026
74a4fb0
opencl: add tiled q4_k GEMM for adreno (#30412)
wanghqc Oct 28, 2026
cf24317
common, mtmd: bound control vector layer indices and M-RoPE position …
ServeurpersoCom Oct 28, 2026
872e428
ggml-hrx: add AMD ROCm HRX native ggml backend
stellaraccident Aug 17, 2026
43207c7
ggml-hrx: squash development through 8108322b7
AaronStGeorge Oct 7, 2026
70f80c1
ggml-hrx: adopt canonical Loom divisibility predicates (#139)
AaronStGeorge Oct 8, 2026
ab276e7
hrx: accept empty MoE expert views when matching graphs (#140)
AaronStGeorge Oct 12, 2026
91ea4b4
ggml-hrx: prove packet bounds before view linearization (#141)
AaronStGeorge Oct 13, 2026
1e26135
ggml-hrx: size fused router launch for target expert tiles (#146)
AaronStGeorge Oct 14, 2026
eacfd9b
ggml-hrx: adopt exclusively owned Loom link-index builders (#147)
AaronStGeorge Oct 14, 2026
e3e643d
ggml-hrx: reuse block metadata in dense and paired tiled matmul (#148)
AaronStGeorge Oct 15, 2026
fad4e3d
ggml-hrx: reuse IQ3_S metadata and Q5 loads (#150)
AaronStGeorge Oct 21, 2026
3ac8825
ggml-hrx: unify MoE partition descriptors and cover all experts (#153)
AaronStGeorge Oct 21, 2026
3456e1f
hrx: publish decode attention output before Q8 packing (#154)
AaronStGeorge Oct 23, 2026
292331d
ggml-hrx: fix suite dispatch expectations and concat bounds (#156)
AaronStGeorge Oct 26, 2026
b518403
ggml-hrx: normalize test names and CTest invocation (#157)
AaronStGeorge Oct 27, 2026
6c66e46
ggml-hrx: fix concat launch bounds and recurrent MTP snapshots (#155)
AaronStGeorge Oct 28, 2026
6851a42
ggml-hrx: adapt to latest upstream and HRX interfaces
AaronStGeorge Oct 28, 2026
cc8c508
hrx: preserve caller inputs during metadata initialization (#162)
AaronStGeorge Oct 30, 2026
2120f74
ggml-hrx: restore correctness after upstream interface rebase (#163)
AaronStGeorge Oct 30, 2026
76707b4
hrx: keep contiguous F32 duplicates on the device (#164)
AaronStGeorge Oct 30, 2026
f6d215d
ggml-hrx: add opt-in compilation diagnostics (#165)
AaronStGeorge Nov 1, 2026
930e43f
ggml-hrx: materialize weights on demand and keep them resident (#160)
AaronStGeorge Nov 3, 2026
f200667
ggml-hrx: fence global F32 output before Qwen Q8 packing (#174)
AaronStGeorge Nov 3, 2026
373507a
ggml-hrx: decode IQ3_XXS routed expert pairs (#175)
AaronStGeorge Nov 3, 2026
59f141f
ggml-hrx: honor explicit RoPE frequency factors in Qwen attention (#179)
AaronStGeorge Nov 5, 2026
0a4ac56
ggml-hrx: support Mellum routed experts and sliding-window attention …
AaronStGeorge Nov 5, 2026
62eb8dc
ggml-hrx: support Gemma routed IDs and expert scales (#180)
AaronStGeorge Nov 5, 2026
be33535
Revert "ggml-hrx: support Gemma routed IDs and expert scales (#180)" …
AaronStGeorge Nov 5, 2026
5acec43
ggml-hrx: support Gemma routed IDs and expert scales (#183)
AaronStGeorge Nov 9, 2026
9bf5871
tests: account for routed SwiGLU codebook binding
AaronStGeorge Nov 9, 2026
f9319d7
WIP, NOT MERGEABLE: AMD's rebased pin 9bf5871 merged into 1bit/hrx-vu…
Nov 11, 2026
c1552e0
2138 merge: keep both sides in CommandProgramBinding (graph_input + m…
bcloud Nov 11, 2026
cef4e64
2138 merge: resolve the mechanical hunks three-way, leave real confli…
bcloud Nov 11, 2026
aab5078
2138 merge: NOTE the heuristic's known failure and the compile-before…
bcloud Nov 11, 2026
278be43
2138 merge: keep the shared ROPE predicate, with AMD's new n_offs gua…
bcloud Nov 11, 2026
acc0bda
2138 merge: flash-attention gate matcher takes AMD's call site, our e…
bcloud Nov 11, 2026
f728610
2138 merge: mul-mat-id kernel refs are additive, except the ones noth…
bcloud Nov 11, 2026
922075b
2138 merge: two additive headers unioned; the aggregate-init hazard a…
bcloud Nov 11, 2026
ace442a
2138 merge: prepared-command-program-cache and host-memory unions; AM…
bcloud Nov 11, 2026
a187b23
2138 merge: measured the silent-loss class across both-modified files…
bcloud Nov 11, 2026
fdde152
2138 merge: restore set_input_qsa -- its caller survived, its definit…
bcloud Nov 11, 2026
33af29e
2138 merge: NOTE the accurate statement of the set_input_qsa break an…
bcloud Nov 11, 2026
fb45749
2138 merge: CMakeLists.txt header union -- this was the gate on any c…
bcloud Nov 11, 2026
55187cc
2138 merge: restore tools/parser, which our tools/CMakeLists.txt stil…
bcloud Nov 11, 2026
a4ddff1
2138 merge: src/CMakeLists.txt source list = AMD's variable plus our …
bcloud Nov 11, 2026
7b14ced
2138 merge: server target takes both prefill and realtime sources
bcloud Nov 11, 2026
35c62be
2138 merge: ggml-backend.cpp unions -- the file that blocked the firs…
bcloud Nov 11, 2026
9fdcee2
2138 merge: close ggml_backend_sched_set_placement_observer -- the co…
bcloud Nov 11, 2026
ed66d45
2138 merge: resolve the two corpus manifests structurally, not textually
bcloud Nov 11, 2026
0cebcb5
2138 merge: restore the 176 files AMD's pin has and the merge dropped
bcloud Nov 11, 2026
08d3ac0
2138 merge: corpus manifests take AMD's pair whole -- union was the w…
bcloud Nov 11, 2026
8b40b98
2138 merge: two compile errors from earlier resolutions, both predict…
bcloud Nov 11, 2026
7537200
2138 merge: restore the workload-specialization vocabulary AMD's regi…
bcloud Nov 11, 2026
2efe5d0
2138 merge: restore the tiled/skinny pair-input SWIGLU refs AMD's reg…
bcloud Nov 11, 2026
60d40fa
2138 merge: correct my own progress metric and the get-rows closure e…
bcloud Nov 11, 2026
73e55a1
2138 merge: gguf-py takes the union at entry level, structurally not …
bcloud Nov 11, 2026
ddde5dd
2138 merge: 18 conversion files -- register-name union plus upstream'…
bcloud Nov 11, 2026
c48eb0a
2138 merge: the last five conversion files -- apply our change to the…
bcloud Nov 11, 2026
1a712c2
2138 merge: arch/cparams/context unions, and the duplicate the per-hu…
bcloud Nov 11, 2026
037c500
2138 merge: the merged session format is a THIRD format, so it gets a…
bcloud Nov 11, 2026
90887e1
2138 merge: ssm-conv's abort message names the union of the cases, be…
bcloud Nov 11, 2026
d6934de
2138 merge: llama-vocab's pre-type enum takes AMD's renumbered list, …
bcloud Nov 11, 2026
e182a1b
2138 merge: hparams unions; llama-model.h left marked because my scri…
bcloud Nov 11, 2026
9c93ab3
2138 merge: llama-model.h done by hand, and the brace mystery resolved
bcloud Nov 11, 2026
0915a30
2138 merge: llama-bench README lists the union of the flags that stil…
bcloud Nov 11, 2026
55c1aa5
2138 merge: llama-bench keeps both flags, aligned per struct
bcloud Nov 11, 2026
d0cc6a9
2138 merge: server.cpp -- theirs at the conflict, our device-lost ins…
bcloud Nov 11, 2026
f64dd03
2138 merge: model-saver keeps our arch-refusal function, our typed ac…
bcloud Nov 11, 2026
5f982fd
2138 merge: model-loader -- take theirs twice for a superset, ours on…
bcloud Nov 11, 2026
71ffd71
2138 merge: kv-cache takes AMD's by_order generalization; the buffer-…
bcloud Nov 11, 2026
520a85f
2138 merge: CORRECTION -- restore the kv-cache buffer-ownership confl…
bcloud Nov 11, 2026
1b9c49c
2138 merge: llama-graph keeps OUR clamp expansion and takes AMD's WID…
bcloud Nov 11, 2026
25a33f7
2138 merge: llama-graph takes AMD's matcher refactor (callers demand …
bcloud Nov 11, 2026
d5bef19
2138 merge: llama-context -- four unions, one composition, and one fa…
bcloud Nov 12, 2026
5f540e8
2138 merge: qwen35 takes AMD's hdmd lookup, deepseek4 keeps our out-p…
bcloud Nov 12, 2026
bad4b6d
2138 merge: flint takes the per-layer expert FFN length, because the …
bcloud Nov 12, 2026
63dcb44
2138 merge: the rename-form markers, handled only where a side is empty
bcloud Nov 12, 2026
28d4ed8
2138 merge: test-backend-ops takes BOTH sides' test cases; class-defi…
bcloud Nov 12, 2026
c718cf2
2138 merge: test-backend-ops class hunks resolved by the member that …
bcloud Nov 12, 2026
6f73677
2138 merge: test-llama-archs takes the arch-chain unions, leaves the …
bcloud Nov 12, 2026
70c1f98
2138 merge: hrx-backend-test keeps the kernel-name literals, because …
bcloud Nov 12, 2026
9541667
2138 merge: hrx-backend-test fully resolved -- the signature decided …
bcloud Nov 12, 2026
bc53e1f
2138 merge: tests/CMakeLists registers both sides' tests, deduped by …
bcloud Nov 12, 2026
eaa52f4
2138 merge: llama-model resolved -- header-declared API wins, and one…
bcloud Nov 12, 2026
da4f1f4
2138 merge: dsv4 cache -- header decides no_alloc/comp_il_src, our co…
bcloud Nov 12, 2026
a6d2d88
2138 merge: dsv4 state write takes the comp_il_src rename; the mirror…
bcloud Nov 12, 2026
825affc
2138 merge: dsv4 mirror refill keeps our BOUNDED copy -- version 2 st…
bcloud Nov 12, 2026
0b6cc63
2138 merge: dflash takes the expert FF length, the call form, and COM…
bcloud Nov 12, 2026
1279e03
2138 merge: dflash output projection takes AMD's output_s form -- our…
bcloud Nov 12, 2026
82be9b4
2138 merge: dflash layer loop takes AMD's FFN half -- ours was the pr…
bcloud Nov 12, 2026
6e5d816
2138 merge: dflash conv taps keep OUR strided-view term -- AMD's pad …
bcloud Nov 12, 2026
dbdde59
2138 merge: dflash attn input takes AMD's attn_norm -- the third manu…
bcloud Nov 12, 2026
6a8b74f
2138 merge: models.h -- two classes unioned, a tag rename settled by …
bcloud Nov 12, 2026
47d354c
2138 merge: models.h duplicate class resolved by unioning the graph i…
bcloud Nov 12, 2026
2cb7cf7
2138 merge: models.h qwen4exp declares BOTH QSA representations, beca…
bcloud Nov 12, 2026
15eea94
2138 merge: qwen4exp 44 -> 28 hunks by a liveness signal, applied to …
bcloud Nov 12, 2026
fb48667
2138 merge: qwen4exp inject-param hunks follow the models.h decision;…
bcloud Nov 12, 2026
9e091ca
2138 merge: build_hc_mix is now ONE definition in the form the header…
bcloud Nov 12, 2026
8515865
2138 merge: qwen4exp build_hc_mix migration complete -- three hunks t…
bcloud Nov 12, 2026
99998c4
2138 merge: qwen4exp 22 -> 12 hunks by FORWARD-USAGE triage, and a wa…
bcloud Nov 12, 2026
f7a95e2
2138 merge: the first COMPILE of the merge found two things triage co…
bcloud Nov 12, 2026
a3a0e5c
2138 merge: models.h compile-clean, and the brace-delta rule that exp…
bcloud Nov 12, 2026
033648a
2138 merge: Hadamard members renamed to the LIVE names (rot/inv), whi…
bcloud Nov 12, 2026
ac38c20
2138 merge: four duplicate/stale declarations removed by compile, and…
bcloud Nov 12, 2026
f622d39
2138 merge: compile matrix 4 -> 8 PASS; dflash and deepseek4 now build
bcloud Nov 12, 2026
5aca51e
2138 merge: llama-model.cpp 4 defects fixed, 1 left -- each verified …
bcloud Nov 12, 2026
2c865ec
2138 merge: llama-model.cpp PASSES -- the dsv4 cache call sites fixed…
bcloud Nov 12, 2026
8d2c678
2138 merge: 10 of 12 TUs compile clean; the two failures are only the…
bcloud Nov 12, 2026
deeb468
2138 merge: deepseek41's compat-layer is_embd hunks keep OURS -- the …
bcloud Nov 12, 2026
2b17907
2138 merge: deepseek41 engram hunks keep OUR member, n_ff_exp takes t…
bcloud Nov 12, 2026
66f1cdd
2138 merge: deepseek41 17 -> 7 hunks -- comments unioned, compat rope…
bcloud Nov 12, 2026
d694c19
2138 merge: deepseek41.cpp COMPILES -- 24 hunks to zero, every one by…
bcloud Nov 12, 2026
bd8feeb
2138 merge: qwen4exp 12 -> 9 hunks -- the two QSA bodies are separate…
bcloud Nov 12, 2026
536c9df
2138 merge: note that qwen4exp's loader hunk is a composition (load_b…
bcloud Nov 12, 2026
a15cbfb
2138 merge: the load_block COMPOSITION is done -- ours' loop body bec…
bcloud Nov 12, 2026
a5902d4
2138 merge: NOTE -- qwen4exp hunks 2/3 sit inside kpool's set_input, …
bcloud Nov 12, 2026
3c33315
2138 merge: NOTE -- the exact hand split qwen4exp's final 6 hunks need
bcloud Nov 12, 2026
9b396bb
2138 merge: qwen4exp 6 -> 3 markers by resolving the three hunks inde…
bcloud Nov 12, 2026
1745991
2138 merge: llama-graph composes -- AMD's 3-path embedding block with…
bcloud Nov 12, 2026
dbfd917
2138 merge: server-context 8 -> 4 hunks, and two duplicate declaratio…
bcloud Nov 12, 2026
16cfa05
2138 merge: server-context.cpp CLEAN -- 8 hunks, and our engine#941 h…
bcloud Nov 12, 2026
88d88b3
2138 merge: hrx-loom-jit-test takes the suite-runner main
bcloud Nov 12, 2026
f27c14d
2138 merge: tests-enabled configure and generate now succeed
bcloud Nov 12, 2026
cc9db07
2138 merge: NOTE -- hrx-loom-jit-test has TWO mains and two undefined…
bcloud Nov 12, 2026
704d91c
2138 merge: hrx-loom-jit-test taken from AMD's pin -- it had the help…
bcloud Nov 12, 2026
3328382
2138 merge: NOTE -- test-backend-ops' structural damage comes from a …
bcloud Nov 12, 2026
c536611
2138 merge: test-backend-ops -- the damaged class region rebuilt VERB…
bcloud Nov 12, 2026
6bc33ed
2138 merge: NOTE -- test-backend-ops' remaining damage is pinned to a…
bcloud Nov 12, 2026
dcd60b9
2138 merge: NOTE -- three instruments measured on test-backend-ops, t…
bcloud Nov 12, 2026
5ee5ed5
2138 merge: NOTE -- string-aware scanner localises test-backend-ops' …
bcloud Nov 12, 2026
a272140
2138 merge: test-backend-ops.cpp COMPILES -- two swallowed loop close…
bcloud Nov 12, 2026
d8da0b9
2138 merge: hrx-backend-test -- our context_capture block restored fr…
bcloud Nov 12, 2026
ea2c799
2138 merge: NOTE -- four test failures collapse to ONE owner item (ac…
bcloud Nov 12, 2026
6a44b68
2138 merge: NOTE -- qwen4exp's last three hunks need BOTH classes, no…
bcloud Nov 12, 2026
d6732db
2138 merge: NOTE -- qwen4exp's split is a REORDER: each class' field …
bcloud Nov 12, 2026
63dac51
2138 merge: NOTE -- qwen4exp needs a PORT, not text assembly (my thre…
bcloud Nov 12, 2026
d398ab8
2138 merge: kv-cache composed -- and my documented "ownership conflic…
bcloud Nov 12, 2026
01a5a3b
2138 merge: memory-recurrent composed + the container's 3-line API, a…
bcloud Nov 12, 2026
5c9d1df
2138 merge: test-llama-archs -- the pin's file, plus our DEEPSEEK41 f…
bcloud Nov 12, 2026
b498885
2138 merge: ggml-vulkan unioned -- ours' 616-line block plus AMD's 2-…
bcloud Nov 12, 2026
843c992
2138 merge: hrx-ops-test taken from the pin; the two missing library …
bcloud Nov 12, 2026
1377475
2138 merge: NOTE -- the dispatch gap is a helper-chain port; my "two …
bcloud Nov 12, 2026
bc3faf8
2138 merge: restore AMD's DispatchActivationConsumerSource enum (firs…
bcloud Nov 12, 2026
f85e323
2138 merge: restore DispatchMatchContext::registry -- the field I del…
bcloud Nov 12, 2026
0fb3456
2138 merge: port AMD's generated-resource lookup into command-plan.h …
bcloud Nov 12, 2026
86bb7e0
2138 merge: NOTE -- the activation-publication chain, measured (five …
bcloud Nov 12, 2026
2ce4c91
2138 merge: hrx-ops-test.cpp COMPILES -- the activation-publication s…
bcloud Nov 12, 2026
2ba0403
2138 merge: hrx-backend-test -- 32 of AMD's dropped test definitions …
bcloud Nov 12, 2026
a74532e
2138 merge: hrx-backend-test -- template helper + testing_suite.h inc…
bcloud Nov 12, 2026
0a7662e
2138 merge: NOTE -- a near-miss worth recording: content-index slicin…
bcloud Nov 12, 2026
c645800
2138 merge: hrx-backend-test.cpp COMPILES -- duplicates deduped, frag…
bcloud Nov 12, 2026
6f1c799
2138 merge: NOTE -- hrx-loom-jit-test's failure is a symptom of the c…
bcloud Nov 12, 2026
aee1b68
2138 merge: decide the KernelExecutableCache API -- ours, because the…
bcloud Nov 12, 2026
ea86b5c
2138 merge: NOTE -- hrx-loom-jit-test needs per-call-site adaptation;…
bcloud Nov 12, 2026
33db9a1
2138 merge: hrx-loom-jit-test CLEAN -- API adaptation done per call s…
bcloud Nov 12, 2026
80a73fb
2138 merge: qwen4exp resolves by following the body, and it was AMD's…
bcloud Nov 13, 2026
ca96711
2138 merge: 15 rename-form hunks resolve by file identity, because th…
bcloud Nov 13, 2026
9af10bf
ggml-cpu: SET_ROWS belongs to one thread-count arm, not two (first no…
bcloud Nov 13, 2026
f0cb908
2138 merge: drop 7 pre-#118 flat-layout kernel twins my own restore r…
bcloud Nov 13, 2026
2d2e279
2138 merge: keep the IQ3S table decode, and take AMD's tiled/skinny s…
bcloud Nov 13, 2026
15fadd1
2138 merge: dispatch-get-rows.cpp is AMD-only, and the merge had trun…
bcloud Nov 13, 2026
7d0d135
2138 merge: two dispatch files that are AMD-only, restored from the p…
bcloud Nov 13, 2026
79818f5
REVERT of my own "AMD-only" reasoning: 7a1f2b1 was a broken ref, and …
bcloud Nov 13, 2026
638ff5d
2138 merge: dispatch-mul-mat-id.cpp -- AMD's file, verified this time…
bcloud Nov 13, 2026
65143ac
2138 merge: dispatch-get-rows.cpp resolved by union -- and the "heade…
bcloud Nov 13, 2026
a253158
2138 merge: rope-set-rows union (3 hunks) + a second ours-kernel entr…
bcloud Nov 13, 2026
ece6b80
2138 merge: dispatch-registry -- two unrelated blocks that collided, …
bcloud Nov 13, 2026
f7b78f0
2138 merge: moe-router -- our descriptor layout generalisation and th…
bcloud Nov 13, 2026
ba3c44b
2138 merge: NOTE -- host-memory's blocker now has a name: HostWeightS…
bcloud Nov 13, 2026
bb854ae
2138 merge: host-memory resolved -- the "runtime header generation" w…
bcloud Nov 13, 2026
a794806
2138 merge: NOTE on prepared-command-program-cache and gated-mul-mat …
bcloud Nov 13, 2026
ed4d5e4
2138 merge: gated-mul-mat resolves to our hunk and NOTHING else
bcloud Nov 13, 2026
1516237
2138 merge: restore our kernel exports to the manifest (50), undo my …
bcloud Nov 13, 2026
56a1814
2138 merge: the six held-back kernels go in -- the invariant was a fi…
bcloud Nov 13, 2026
baa0837
2138 merge: port our exports into all four kernel manifests -- the di…
bcloud Nov 13, 2026
8db9abf
2138 merge: NOTE -- dispatch layer complete; graph-diagnostics' singl…
bcloud Nov 13, 2026
2de8db0
2138 merge: graph-diagnostics -- AMD added a variant alternative and …
bcloud Nov 13, 2026
49d1a34
2138 merge: NOTE -- the runtime remainder is a design difference, des…
bcloud Nov 13, 2026
4cb0799
2138 merge: NOTE -- command-program-executor measured; the runtime gr…
bcloud Nov 13, 2026
39e70b9
2138 merge: NOTE -- command-program-executor measured to a subsystem,…
bcloud Nov 13, 2026
8d35ef1
2138 merge: graph-program-cache resolved -- the hunk boundary cut BOT…
bcloud Nov 13, 2026
adfcf7e
2138 merge: graph-executor resolved -- three missing pieces, and one …
bcloud Nov 13, 2026
3080ca9
2138 merge: prepared-command-program-cache resolved -- two implementa…
bcloud Nov 13, 2026
9c3ca7b
2138 merge: NOTE -- command-program-executor's recipe, derived to the…
bcloud Nov 13, 2026
7aaa893
2138 merge: NOTE -- command-program-executor reduced to four named items
bcloud Nov 13, 2026
919b21d
2138 merge: command-program-executor resolved -- the four named items…
bcloud Nov 13, 2026
7b50b52
2138 merge: ZERO conflict markers -- speculative resolved, plus the c…
bcloud Nov 13, 2026
9807e36
2138 merge: post-marker build tail -- models, ggml export, and five t…
bcloud Nov 13, 2026
2f3079f
2138 merge: NOTE -- the last build blocker is a fork in the chat/sche…
bcloud Nov 13, 2026
98e14e5
2138 merge: NOTE -- the chat/schema blocker measured from both sides;…
bcloud Nov 13, 2026
f45ba14
2138 merge: NOTE -- the chat blocker is one type alias, measured from…
bcloud Nov 13, 2026
5cd8a77
2138 merge: NOTE -- chat.cpp's option (b) measured to four items; thr…
bcloud Nov 13, 2026
e8f309c
2138 merge: NOTE -- the JSON bridge measured: right direction, 3 edit…
bcloud Nov 13, 2026
d422939
2138 merge: NOTE -- the chat.cpp port measured 74 -> 16 errors, with …
bcloud Nov 13, 2026
02c68cd
merge AMD 9bf5871 into 1bit/hrx-vulkan-patched: 0 conflict markers; l…
Nov 13, 2026
31d1a00
2138 merge: OUR dispatch-mul-mat-id.cpp per ORCH 11-13 10:10Z -- AMD'…
bcloud Nov 13, 2026
b0e1598
2138 merge: peg-parser -- restore AMD's gbnf_format_rule_name (the me…
bcloud Nov 13, 2026
b5ebe8e
merge AMD 9bf5871: 0 markers, our dispatch-mul-mat-id, peg rule-name …
Nov 13, 2026
b6e1167
2138 merge: take the current 1bit/hrx-vulkan-patched (29 commits thro…
Nov 13, 2026
4934172
2138 merge: dispatch-mul-mat-id.cpp is our line's file
Nov 13, 2026
1d56dca
tests: test-hrx-ops.cpp is back (our line's HRX op suite, 238 cases)
Nov 13, 2026
d97ec31
2138 merge: build loom-benchmark only where the hrx-system source has…
Nov 13, 2026
cba9b38
2138 merge: the kernel corpus builds in binary form with the held hrx…
Nov 13, 2026
6208950
2138 merge: llama, llama-common, mtmd and llama-bench compile
Nov 13, 2026
5f9e540
2138 merge: bring in b5ebe8eb0 (the branch moved while this was built)
Nov 13, 2026
2c178f4
2138 merge: json-schema-to-grammar.cpp is AMD's file again
Nov 13, 2026
7cd31c0
2138 merge: the HRX runtime, dispatch and manifest fixes that pass on…
Nov 13, 2026
616c7bf
tests: test-hrx-ops flags the hand-built graph node COMPUTE in two ca…
Nov 13, 2026
c06e741
2138 merge: restore the gguf-context saver ctor and its ownership fla…
Nov 13, 2026
ed02b25
2138 merge: drop MERGE-2138-NOTE.txt, a WIP working note that describ…
Nov 13, 2026
c6806ab
2138 merge: rope token_capacity is a power-of-two CLASS again, not th…
Nov 13, 2026
e1e503e
2138 merge: standalone RoPE is Dynamic-specialised, which is what mak…
Nov 13, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
5 changes: 3 additions & 2 deletions .devops/cann.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -145,7 +145,7 @@ ENTRYPOINT ["/app/tools.sh"]
# ==============================================================================
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

ENTRYPOINT [ "/app/llama-cli" ]

Expand All @@ -155,8 +155,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

HEALTHCHECK --interval=5m CMD [ "curl", "-f", "http://localhost:8080/health" ]

Expand Down
5 changes: 3 additions & 2 deletions .devops/cpu.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -104,7 +104,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -114,8 +114,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
5 changes: 3 additions & 2 deletions .devops/cuda.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -123,8 +123,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
18 changes: 12 additions & 6 deletions .devops/intel.Dockerfile
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
ARG ONEAPI_VERSION=2025.3.3-0-devel-ubuntu24.04
ARG ONEAPI_VERSION=2026.1.1-devel-ubuntu24.04
ARG BUILD_DATE=N/A
ARG APP_VERSION=N/A
ARG APP_REVISION=N/A
Expand All @@ -21,6 +21,11 @@ RUN LLAMA_BUILD_NUMBER="$APP_VERSION" npm run build

FROM docker.io/intel/deep-learning-essentials:$ONEAPI_VERSION AS build

RUN echo "Download oneDNN installation package" && \
wget https://registrationcenter-download.intel.com/akdlm/IRC_NAS/94c3dbac-0852-45be-a57d-21c204cada3e/intel-onednn-2026.0.2.46_offline.sh -O intel-onednn_offline.sh && \
echo "Install oneDNN components" && \
bash intel-onednn_offline.sh -s -a --silent --eula accept

ARG GGML_SYCL_F16=ON
ARG LEVEL_ZERO_VERSION=1.28.2
ARG LEVEL_ZERO_UBUNTU_VERSION=u24.04
Expand All @@ -45,7 +50,7 @@ RUN if [ "${GGML_SYCL_F16}" = "ON" ]; then \
fi && \
echo "Building with dynamic libs" && \
cmake -B build -DGGML_NATIVE=OFF -DGGML_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON -DLLAMA_BUILD_TESTS=OFF ${OPT_SYCL_F16} && \
cmake --build build --config Release -j$(nproc)
cmake --build build --config Release -j$((($(nproc)+1)/2))

RUN mkdir -p /app/lib && \
find build -name "*.so*" -exec cp -P {} /app/lib \;
Expand All @@ -59,7 +64,7 @@ RUN mkdir -p /app/full \
&& cp requirements.txt /app/full \
&& cp .devops/tools.sh /app/full/tools.sh

FROM docker.io/intel/deep-learning-essentials:$ONEAPI_VERSION AS base
FROM docker.io/intel/oneapi-toolkit:$ONEAPI_VERSION AS base

ARG BUILD_DATE=N/A
ARG APP_VERSION=N/A
Expand Down Expand Up @@ -102,7 +107,7 @@ RUN mkdir /tmp/neo/ && cd /tmp/neo/ \
&& dpkg --install *.deb

RUN apt-get update \
&& apt-get install -y libgomp1 curl ffmpeg \
&& apt-get install -y libgomp1 curl ffmpeg openssl libssl-dev \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \
Expand Down Expand Up @@ -141,7 +146,7 @@ ENTRYPOINT ["/app/tools.sh"]
FROM base AS light

COPY --from=build /app/lib/ /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -151,9 +156,10 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/lib/ /app
COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
20 changes: 13 additions & 7 deletions .devops/musa.Dockerfile
Original file line number Diff line number Diff line change
@@ -1,10 +1,9 @@
ARG UBUNTU_VERSION=22.04
# This needs to generally match the container host's environment.
ARG MUSA_VERSION=rc4.3.0
# Target the MUSA build image
ARG BASE_MUSA_DEV_CONTAINER=docker.io/mthreads/musa:${MUSA_VERSION}-devel-ubuntu${UBUNTU_VERSION}-amd64
ARG BASE_MUSA_DEV_CONTAINER=registry.mthreads.com/mcconline/musa_sdk:5.2.0-devel-ubuntu${UBUNTU_VERSION}-s5000

ARG BASE_MUSA_RUN_CONTAINER=docker.io/mthreads/musa:${MUSA_VERSION}-runtime-ubuntu${UBUNTU_VERSION}-amd64
ARG BASE_MUSA_RUN_CONTAINER=registry.mthreads.com/mcconline/musa_sdk:5.2.0-runtime-ubuntu${UBUNTU_VERSION}-s5000

ARG BUILD_DATE=N/A
ARG APP_VERSION=N/A
Expand Down Expand Up @@ -37,7 +36,10 @@ RUN apt-get update && \
python3-pip \
git \
libssl-dev \
libgomp1
libgomp1 \
musa-mualg-5-2 \
musa-muthrust-5-2 \
libmthreads-compute

WORKDIR /app

Expand Down Expand Up @@ -80,13 +82,16 @@ LABEL org.opencontainers.image.created=$BUILD_DATE \
org.opencontainers.image.source=$IMAGE_SOURCE

RUN apt-get update \
&& apt-get install -y libgomp1 curl ffmpeg \
&& apt-get install -y libgomp1 curl ffmpeg libmthreads-compute \
&& apt autoremove -y \
&& apt clean -y \
&& rm -rf /tmp/* /var/tmp/* \
&& find /var/cache/apt/archives /var/lib/apt/lists -not -name lock -type f -delete \
&& find /var/cache -type f -delete

# The MUSA runtime image does not register its library directory
RUN echo "/usr/local/musa/lib" > /etc/ld.so.conf.d/musa-runtime.conf && ldconfig

COPY --from=build /app/lib/ /app

### Full
Expand Down Expand Up @@ -115,7 +120,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -125,8 +130,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
12 changes: 6 additions & 6 deletions .devops/nix/package.nix
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@
]
&& blas.meta.available,
useCuda ? config.cudaSupport,
useMetalKit ? stdenv.isAarch64 && stdenv.isDarwin,
useMetalKit ? stdenv.hostPlatform.isAarch64 && stdenv.hostPlatform.isDarwin,
# Increases the runtime closure size by ~700M
useMpi ? false,
useRocm ? config.rocmSupport,
Expand Down Expand Up @@ -92,7 +92,7 @@ let

cudaBuildInputs = with cudaPackages; [
cuda_cudart
cuda_cccl # <nv/target>
cccl # <nv/target>
libcublas
];

Expand Down Expand Up @@ -166,7 +166,7 @@ effectiveStdenv.mkDerivation (finalAttrs: {
# `xcrun` is used find the path of the Metal compiler, which is varible
# and not on $PATH
# see https://github.com/ggml-org/llama.cpp/pull/6118 for discussion
__noChroot = effectiveStdenv.isDarwin && useMetalKit && precompileMetalShaders;
__noChroot = effectiveStdenv.hostPlatform.isDarwin && useMetalKit && precompileMetalShaders;

nativeBuildInputs =
[
Expand All @@ -181,10 +181,10 @@ effectiveStdenv.mkDerivation (finalAttrs: {
autoAddDriverRunpath
]
++ optionals (effectiveStdenv.hostPlatform.isGnu && enableStatic) [ glibc.static ]
++ optionals (effectiveStdenv.isDarwin && useMetalKit && precompileMetalShaders) [ xcrunHost ];
++ optionals (effectiveStdenv.hostPlatform.isDarwin && useMetalKit && precompileMetalShaders) [ xcrunHost ];

buildInputs =
optionals effectiveStdenv.isDarwin darwinBuildInputs
optionals effectiveStdenv.hostPlatform.isDarwin darwinBuildInputs
++ optionals useCuda cudaBuildInputs
++ optionals useMpi [ mpi ]
++ optionals useRocm rocmBuildInputs
Expand Down Expand Up @@ -245,7 +245,7 @@ effectiveStdenv.mkDerivation (finalAttrs: {

# Configurations that are known to result in build failures. Can be
# overridden by importing Nixpkgs with `allowBroken = true`.
broken = (useMetalKit && !effectiveStdenv.isDarwin);
broken = (useMetalKit && !effectiveStdenv.hostPlatform.isDarwin);

description = "Inference of LLaMA model in pure C/C++${descriptionSuffix}";
homepage = "https://github.com/ggml-org/llama.cpp/";
Expand Down
24 changes: 14 additions & 10 deletions .devops/openvino.Dockerfile
Original file line number Diff line number Diff line change
@@ -1,18 +1,18 @@
ARG OPENVINO_VERSION_MAJOR=2026.2.1
ARG OPENVINO_VERSION_FULL=2026.2.1.21919.ede283a88e3
ARG OPENVINO_VERSION_MAJOR=2026.4.1
ARG OPENVINO_VERSION_FULL=2026.4.1.22982.07f9c262b05
ARG UBUNTU_VERSION=24.04

# Intel GPU driver versions. https://github.com/intel/compute-runtime/releases
ARG IGC_VERSION=v2.36.3
ARG IGC_VERSION_FULL=2_2.36.3+21719
ARG COMPUTE_RUNTIME_VERSION=26.22.38646.4
ARG COMPUTE_RUNTIME_VERSION_FULL=26.22.38646.4-0
ARG IGC_VERSION=v2.41.5
ARG IGC_VERSION_FULL=2_2.41.5+22716
ARG COMPUTE_RUNTIME_VERSION=26.35.39758.10
ARG COMPUTE_RUNTIME_VERSION_FULL=26.35.39758.10-0
ARG IGDGMM_VERSION=22.10.0

# Intel NPU driver versions. https://github.com/intel/linux-npu-driver/releases
ARG NPU_DRIVER_VERSION=v1.33.0
ARG NPU_DRIVER_FULL=v1.33.0.20260529-26625960453
ARG LIBZE1_VERSION=1.27.0-1~24.04~ppa2
ARG NPU_DRIVER_VERSION=v1.38.0
ARG NPU_DRIVER_FULL=v1.38.0.20260910-34487311128
ARG LIBZE1_VERSION=1.32.0-1~24.04~ppa1

# Optional proxy build arguments
ARG http_proxy=
Expand Down Expand Up @@ -90,6 +90,9 @@ RUN bash -c "source ${OpenVINO_DIR}/setupvars.sh && \
cmake -B build/ReleaseOV -G Ninja \
-DCMAKE_BUILD_TYPE=Release \
-DLLAMA_BUILD_TESTS=OFF \
-DGGML_NATIVE=OFF \
-DGGML_BACKEND_DL=ON \
-DGGML_CPU_ALL_VARIANTS=ON \
-DGGML_OPENVINO=ON && \
cmake --build build/ReleaseOV --parallel "

Expand Down Expand Up @@ -170,7 +173,7 @@ RUN --mount=type=cache,target=/var/cache/intel-npu,sharing=locked \
fi; \
DEB=/var/cache/intel-npu/libze1_${LIBZE1_VERSION}_amd64.deb; \
if [ ! -f "$DEB" ]; then \
wget -q -O "$DEB" https://snapshot.ppa.launchpadcontent.net/kobuk-team/intel-graphics/ubuntu/20260324T100000Z/pool/main/l/level-zero-loader/libze1_${LIBZE1_VERSION}_amd64.deb; \
wget -q -O "$DEB" https://snapshot.ppa.launchpadcontent.net/kobuk-team/intel-graphics/ubuntu/20260830T100000Z/pool/main/l/level-zero-loader/libze1_${LIBZE1_VERSION}_amd64.deb; \
fi; \
mkdir /tmp/npu/ && cd /tmp/npu/ && tar -xf "$TGZ" && cp "$DEB" .; \
apt-get update; \
Expand Down Expand Up @@ -224,6 +227,7 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app/

Expand Down
6 changes: 3 additions & 3 deletions .devops/rocm.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -57,7 +57,6 @@ COPY --from=web /app/tools/ui/dist tools/ui/dist
RUN HIPCXX="$(hipconfig -l)/clang" HIP_PATH="$(hipconfig -R)" \
cmake -S . -B build \
-DGGML_HIP=ON \
-DGGML_HIP_ROCWMMA_FATTN=ON \
-DAMDGPU_TARGETS="$ROCM_DOCKER_ARCH" \
-DGGML_BACKEND_DL=ON -DGGML_CPU_ALL_VARIANTS=ON \
-DCMAKE_BUILD_TYPE=Release -DLLAMA_BUILD_TESTS=OFF \
Expand Down Expand Up @@ -127,7 +126,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -137,8 +136,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
1 change: 1 addition & 0 deletions .devops/s390x.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -133,6 +133,7 @@ ENTRYPOINT [ "/llama.cpp/bin/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

WORKDIR /llama.cpp/bin

Expand Down
5 changes: 3 additions & 2 deletions .devops/vulkan.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -107,7 +107,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -117,8 +117,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
5 changes: 3 additions & 2 deletions .devops/zendnn.Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -97,7 +97,7 @@ ENTRYPOINT ["/app/tools.sh"]
### Light, CLI only
FROM base AS light

COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app
COPY --from=build /app/full/llama /app/full/llama-cli /app/full/llama-completion /app/

WORKDIR /app

Expand All @@ -107,8 +107,9 @@ ENTRYPOINT [ "/app/llama-cli" ]
FROM base AS server

ENV LLAMA_ARG_HOST=0.0.0.0
ENV LLAMA_ARG_PORT=8080

COPY --from=build /app/full/llama /app/full/llama-server /app
COPY --from=build /app/full/llama /app/full/llama-server /app/

WORKDIR /app

Expand Down
2 changes: 1 addition & 1 deletion .ecrc
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
{
"Exclude": ["^\\.gitmodules$", "stb_image\\.h"],
"Exclude": ["^\\.gitmodules$", "stb_image\\.h", "examples/test-cmake/build/", "examples/test-cmake/build-subdir/"],
"Disable": {
"IndentSize": true
}
Expand Down
3 changes: 3 additions & 0 deletions .editorconfig
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 +45,9 @@ insert_final_newline = unset
trim_trailing_whitespace = unset
insert_final_newline = unset

[vendor/**.patch]
trim_trailing_whitespace = unset

[tools/ui/**]
indent_style = unset
indent_size = unset
Expand Down
2 changes: 2 additions & 0 deletions .flake8
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,8 @@ exclude =
# Do not traverse examples and tools
examples,
tools,
# Do not traverse .venv directories
.venv*,
# Do not include package initializers
__init__.py,
# No need to traverse our git directory
Expand Down
Loading