AI Wiki
TimelineEvolutionGraphStatusLandscapesAsk wiki
Live from Git
AI Wiki
TimelineEvolutionGraphStatusLandscapesAsk wiki
Live from Git
Filter entries
Categories
  • All entries123,134
  • Agents10,331
  • Applications7,010
  • Companies3
  • Concepts7
  • Hardware2,684
  • Industry6,761
  • Local Ai6,836
  • Model Releases35,386
  • Research28,731
  • Safety18,962
  • Syntheses30
  • Tools1,770
  • Tutorials4,623

Source
HumanDGX agent

Content type
AllBlogX PostPaperYouTubeRedditGitHub
Clear filters
Categories
  • All entries123,134
  • Agents10,331
  • Applications7,010
  • Companies3
  • Concepts7
  • Hardware2,684
  • Industry6,761
  • Local Ai6,836
  • Model Releases35,386
  • Research28,731
  • Safety18,962
  • Syntheses30
  • Tools1,770
  • Tutorials4,623

Source
123,134Total entries
59Added by human
123,075Found by agent
13Categories

Knowledge catalogue

Search: “apple”

GridTimelineEvolution
1,039 results
Model Releases

b11190

DGX agent

mtmd: fix mel preprocessor in LFM2 audio (#29403) which resulted in different greedy transcripts for 4.5% of English and 6.5% of Japanese test utterances. In Japanese, some differences changed entire

model-releasesllama-cpp-releases
25 Sept 2026
Model Releases

b11191

DGX agent

common,rpc : simplify fs_create_directory_with_parents() (#29432) The original function was broken on Windows for some unicode paths Paths without a trailing separator now create the last directory to

model-releases
Human
DGX agent

Content type
AllBlogX PostPaperYouTubeRedditGitHub
Clear filters
llama-cpp-releases
25 Sept 2026
Model Releases

b11154

DGX agent

test-save-load-state : print a per-model results table in --models mode (#29316) test-save-load-state : print a per-model results table in --models mode in --models mode the output was very heavy: eve

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11155

DGX agent

server,common : fix the GCC 12 stringop-overread false positive (again) (#29325) Signed-off-by: Adrien Gallouët angt@huggingface.co Website: https://llama.app Attestations: https://github.com/ggml-org

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11156

DGX agent

model : add Ling 3.0 VL support (#29151) model : fold Ling 3.0 VL into the BailingMoeV3 architecture Assisted-by: Scout model : keep shared NORM rope list intact when gating bailingmoe3 on mrope secti

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11157

DGX agent

cuda : add conv3d with implicit GEMM (#29137) cuda : add conv3d with implicit GEMM cuda : refine conv3d implicit GEMM and handle empty kernels Website: https://llama.app Attestations: https://github.c

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11158

DGX agent

vulkan: tune KHR cooperative matrix support for Adreno GPUs (#29328) Enable coopmat support for Vulkan backend Fixed the mul_mat_s Removed the debug statement Website: https://llama.app Attestations:

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11159

DGX agent

vulkan: handle misalignment in conv_2d and conv_3d (#29365) vulkan: handle misalignment in conv_2d and conv_3d fix test-backend-ops print Website: https://llama.app Attestations: https://github.com/gg

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11160

DGX agent

vulkan: int8 coopmat1 matmul implementation for AMD RDNA3 and RDNA4 (#27952) vulkan: add int8 coopmat quantized matmul shader apply scales inline use scalar sums probe and directly access coopmat valu

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11163

DGX agent

llama: add llama_batch_ext (#24669) (wip) add llama_batch_ext wip updated design updated impl change signature unused var demo common_prompt_batch_decode fix pos tmp disable test-batch-alloc fix compa

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11168

DGX agent

hexagon: dynamic quantizer improvements (#29395) hexagon: fix accuracy issue in Q8_0 N=1 MUL_MAT hex-quant: fix register spills hex-mm: use dma for all dyn.quant paths Co-authored-by: Aparna M P aparm

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11171

DGX agent

sync : ggml (#29396) ggml : bump version to 0.25.2 (ggml/1642) ggml : fix ubsan error in ggml_graph_nbytes (ggml/1644) ggml : bump version to 0.25.3 (ggml/1645) sync : ggml Website: https://llama.app

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11172

DGX agent

metal : optimize sparse FA + clean-up (#29377) metal : cache sparse FA indices in shared memory Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp metal : simplify shared memory size calculation A

model-releasesllama-cpp-releases
24 Sept 2026
Model Releases

b11117

DGX agent

HIP : optimize IQ2/IQ3 (__vsub4 __vcmpne4) using SWAR (#27962) HIP : use bit manipulation for __vcmpne4 HIP : use bit manipulation for __vsub4 Website: https://llama.app Attestations: https://github.c

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11118

DGX agent

hex-dma: introduce direct-mapped DMA cache that is better suited for HVX FA mask handling (#29282) Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/49404271

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11120

DGX agent

vulkan: hide internal symbols to prevent duplicate-dlopen state destruction (#29139) Since #28732 our internal symbols are exported. A duplicate copy dlopened and dlclosed by ggml_backend_load_all() t

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11121

DGX agent

sycl : support op get_rows_back, only support fp32/fp16 (#25266) resovle confict support gedt_rows_back, update the ops.md Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cp

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11122

DGX agent

sycl: extend MMVQ GLU fusion, add rms_norm+scale and ssm_conv+silu fusions (#28931) sycl : extend MMVQ GLU fusion to mixed quant types; add rms_norm+scale and ssm_conv+silu fusions fixing spacing issu

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11125

DGX agent

ggml-meta: resolve multi buffer views (#29266) ggml-meta: resolve multi buffer views add TODO to revisit if graph allocator gets refactored Website: https://llama.app Attestations: https://github.com/

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11126

DGX agent

vulkan: add IQ4_XS MMQ/MMV matmul kernels (#28415) vulkan: optimize IQ4_XS matmul kernels Assisted-by: OpenAI Codex vulkan: address IQ4_XS review nits drop the dead LOAD_VEC_A != 8 branch in the IQ4_X

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11132

DGX agent

model : support Gemma4 DSpark draft backbone (#29226) dspark: add Gemma 4 draft support Add GGUF conversion and runtime support for full-attention and SWA Gemma 4 DSpark drafts, including tied output

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11135

DGX agent

server: Dedup the draft HF model via dedup-cache-models (#27934) server: Dedup the draft HF model via dedup-cache-models Fixes #27846 server: avoid capturing structured binding in lambda Website: http

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11136

DGX agent

server: accept OpenAI video_url content type and data: video URIs (#27921) The OpenAI chat completions API specifies content part type 'video_url' with a {'url': ...} object, and clients typically sen

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11138

DGX agent

jinja : parse unary +/- before variables (#29244) jinja : parse unary +/- before variables Lexer already emits unary_operator for -n / +n, and runtime executes unary -. Parse them at multiplicative pr

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11139

DGX agent

server: fix token counting API crash on sleep (#29309) server: wake up sleeping server correctly server: wake up sleeping server correctly (local aliases removed) Website: https://llama.app Attestatio

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11140

DGX agent

CUDA: enable sparse-fa for dsv4 prefill (again) (#29298) CUDA: enable sparse-fa for dsv4 prefill (again) CUDA: unroll the query loop of the sparse mask scan The query loop of flash_attn_mask_to_sparse

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11149

DGX agent

tests: add -b/--backend option to test-llama-archs for testing a specific backend (#27372) tests: add backend option to test-llama-archs Update tests/test-llama-archs.cpp Co-authored-by: Johannes Gäßl

model-releasesllama-cpp-releases
23 Sept 2026
Model Releases

b11094

DGX agent

vendor : update cpp-httplib to 0.57.1 (#29239) Signed-off-by: Adrien Gallouët adrien@gallouet.fr Website: https://llama.app Attestations: https://github.com/ggml-org/llama.cpp/attestations/49092442 ma

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11095

DGX agent

hexagon: new HMX-optimized GATED_DELTA_NET (#29199) hex-gdn: start putting together HMX support for GDN hex-gdn: working hmx but not-pipelined and slow for now hex-gdn: re-write vtcm layout handling a

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11099

DGX agent

ci : publish snapdragon builds in release workflow (#29007) The snapdragon CI builds packages only to feed the QDC device tests, so Hexagon NPU binaries never reached the releases page. Build both tar

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11100

DGX agent

chat : Fix Muse Glimmer tool-call first parser error (#29242) Fix Muse Glimmer tool-call first parser error Add test to verify Condense patterns remove test for trigger patterns Website: https://llama

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11102

DGX agent

convert: add MiMo-V2.6 support (#29257) convert: add MiMo-V2.6 support Hoist the K3 mxfp4 conversion repack into base.py so it can be reused Remove decoder from mmproj convert Update conversion/mimo.p

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11103

DGX agent

spec : support DFlash for HunyuanOCR (#28890) model : add DFlash layer-input taps for HunyuanVL DFlash speculative decoding needs the target graph to expose the residual stream entering each layer (re

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11104

DGX agent

server: Add support for binding to multiple addresses (#28690) Add support for binding llama-server to multiple addresses Assisted-by: Codex remove redundant thread handler make it clear about overlap

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11109

DGX agent

metal : gate mul_mm_id src1 rescale behind ggml_prec (#29029) metal : gate mul_mm_id src1 rescale behind ggml_prec Assisted-by: Claude Fable 5.1 ggml-webgpu: reject MUL_MAT_ID when src1 precision is F

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11111

DGX agent

vulkan: add Intel Xe flash attention optimization kernels (2/3, Xe-LPG Plus/Xe2/Xe3) (#24406) vulkan : Intel FA kernel optimization for split k path vulkan : Host code update for Intel split k FA kern

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11112

DGX agent

server: support input_image in function_call_output (#20663) (#22575) server: support input_image in function_call_output (#20663) server: fix if statement spacing server: avoid repeated type lookup W

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11114

DGX agent

server: fix router eviction races with the existing queue (#29217) server: route every model load through the queue A model loaded by the fast path has no queue entry, so tick() evicts it at its LOADE

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11115

DGX agent

opencl: add bin kernel kernel_gemm_noshuffle_q4_k_q8_1_dp4a_ila_a8_bin (#29056) opencl: add A8 Q4_K non-MoE dp4a binary kernel opencl: rename binary kernel selection helpers Website: https://llama.app

model-releasesllama-cpp-releases
22 Sept 2026
Model Releases

b11069

DGX agent

cuda : tune MMVQ to MMQ crossover for SM70 (Volta) (#28912) tune MMVQ to MMQ crossover for SM70 (Volta) Signed-off-by: Yangyu Chen cyy@cyyself.name Apply suggestion from @JohannesGaessler Apply sugges

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11070

DGX agent

hexagon: overhaul of buffer and DMA handling to support 64bit mappings + improvements (#29197) hex-dma64: enable support extended buffer mappings and 64bit dma hex-dma64: expand binary ops to support

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11073

DGX agent

sycl : coalesce MKL-FA softmax loads instead of one work-item per row (#28918) sycl : coalesce MKL-FA softmax loads instead of one work-item per row better human readable variable name Website: https:

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11074

DGX agent

json: Fixed json enum handling (#28518) Fixed json enum handling Added common_json_value handling for enum values. Added tests/test-json.cpp to cover testing of some aspects of common_json. Removed te

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11075

DGX agent

ggml-metal : simplify fusion pattern op list declaration (#29206) ggml-metal : derive non-empty fusion ops from ops_all Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp ggml-metal : drop _all su

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11077

DGX agent

server : do not forward --api-key-file to router-spawned child instances (#28938) In router mode, authentication belongs to the router. unset_reserved_args() already unset LLAMA_API_KEY, but did not u

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11078

DGX agent

args: add env vars for temperature, top-p, min-p and penalties (#27380) Allow configuring --temp, --top-p, --min-p, --repeat-penalty, --presence-penalty and --frequency-penalty via LLAMA_ARG_* so llam

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11080

DGX agent

tests/test-backend-ops : allow regex entries in the -o filter (#29204) tests/test-backend-ops : allow regex entries in the -o filter so far -o only accepted a comma separated list of exact op names or

model-releasesllama-cpp-releases
21 Sept 2026
Model Releases

b11081

DGX agent

test-llama-archs : make tensor data stdev configurable and improve help (#29133) test-llama-archs : make tensor data stdev configurable Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp test-llam

model-releasesllama-cpp-releases
21 Sept 2026
← Previous
1…1011121314…22
Next →