mikhail/llama.cpp-mtp-turboquant
1
0
Fork 0
Code Issues Pull Requests Actions 34 Packages Projects Releases Wiki Activity

Commit Graph

  • 9789e28459 debug : include LLAMA_POOLING_TYPE_UNSPECIFIED in pooling check (#18692) Daniel Bevenius 2026-01-11 16:34:41 +01:00
  • 84ae04f163 tests : refactor test-backend-sampler (#18753) Georgi Gerganov 2026-01-11 17:31:03 +02:00
  • 506bb6e010 model: try to improve Qwen3 Next (#18683) Xuan-Son Nguyen 2026-01-11 12:53:33 +01:00
  • 79456a690a readme : update UIs (#18751) thom-dev-fr 2026-01-11 12:46:50 +01:00
  • 28068af789 security: narrow down the scope of what we consider a vulnerability (#18752) Xuan-Son Nguyen 2026-01-11 12:23:36 +01:00
  • 707cbafcaa opencl: add SOFTPLUS op support (#18726) shaofeiqi 2026-01-10 21:57:44 -08:00
  • b137718878 test-backend-ops: fix mxfp4 tests on blackwell (#18736) Aman Gupta 2026-01-11 01:12:57 +08:00
  • d2ff4e23ac HIP: adjust RDNA3.5 MMQ kernel selction logic (#18666) Johannes Gäßler 2026-01-10 17:19:01 +01:00
  • 657a2e644b cmake : update blas logic (#18205) Perry Naseck 2026-01-10 11:00:54 -05:00
  • f307926482 server : adjust unified KV cache tests (#18716) Georgi Gerganov 2026-01-10 17:51:56 +02:00
  • 7fdc8c893d scripts : follow api redirects in pr2wt.sh (#18739) Sigbjørn Skjæret 2026-01-10 16:04:05 +01:00
  • 23f82f2420 preset: allow named remote preset (#18728) Xuan-Son Nguyen 2026-01-10 15:12:29 +01:00
  • 2656c0d265 docs(ggml): update backend ops (#18734) Aaron Teo 2026-01-10 18:48:17 +08:00
  • 600a366478 Corrected: changed s13 = src1->nb[3] instead of nb[2] (#18724) Michael Wand 2026-01-10 01:16:07 -08:00
  • ea23c15990 common : add --license to display embedded licenses (#18696) Adrien Gallouët 2026-01-10 09:46:24 +01:00
  • 9ac2693a30 server: fix n_cmpl not skipping processing prompt (#18663) Xuan-Son Nguyen 2026-01-10 00:00:41 +01:00
  • a61c8bc3bf mtmd: Add Gemma3n multimodal support with MobileNetV5 vision encoder (#18256) Simranjeet Singh 2026-01-09 22:42:38 +00:00
  • 593da7fa49 opencl: add EXPM1 op (#18704) shaofeiqi 2026-01-09 10:13:13 -08:00
  • 9e41884dce Updates to webgpu get_memory (#18707) Reese Levine 2026-01-09 08:17:18 -08:00
  • ec8fd7876b Webui/file upload (#18694) Pascal 2026-01-09 16:45:32 +01:00
  • a180ba78c7 cmake: only build cli when server is enabled (#18670) Asbjørn Olling 2026-01-09 16:43:26 +01:00
  • 53eb9435da server : fix timing of prompt/generation (#18713) Georgi Gerganov 2026-01-09 12:59:50 +02:00
  • d3435efc8a scripts : pr2wt.sh reset to remote head (#18695) Georgi Gerganov 2026-01-09 12:16:40 +02:00
  • f5f8812f7c server : use different seeds for child completions (#18700) Georgi Gerganov 2026-01-09 09:33:50 +02:00
  • 8ece3836b4 common: support remote preset (#18520) Xuan-Son Nguyen 2026-01-08 22:35:40 +01:00
  • 046d5fd44e llama: use host memory if device reports 0 memory (#18587) Aaron Teo 2026-01-09 05:34:56 +08:00
  • 480160d472 ggml-webgpu: Fix GGML_MEM_ALIGN to 8 for emscripten. (#18628) Masashi Yoshimura 2026-01-09 01:36:42 +09:00
  • 15bff84bf5 ggml webgpu: initial flashattention implementation (#18610) Reese Levine 2026-01-08 08:23:39 -08:00
  • 2524c26164 vulkan: fix push constant size for quantize_q8_1 (#18687) Jeff Bolz 2026-01-08 08:40:58 -06:00
  • cb14b06995 vulkan: optimize ssm_scan (#18630) Jeff Bolz 2026-01-08 08:16:54 -06:00
  • 55abc39355 vendor : update cpp-httplib to 0.30.0 (#18660) Adrien Gallouët 2026-01-08 13:53:54 +01:00
  • f2f6c88067 scripts : support chaining commands in pr2wt.sh (#18671) Georgi Gerganov 2026-01-08 13:40:23 +02:00
  • 945bf10627 metal : add MoE kernel specialization for ne20=5 (#18667) 도로로도로또 2026-01-08 19:37:45 +09:00
  • 64848deb18 llama-fit-params: free memory target per device (#18679) Johannes Gäßler 2026-01-08 10:07:58 +01:00
  • 9a5724dee2 ggml: add env var GGML_OP_OFFLOAD_MIN_BATCH (#18535) Doctor Shotgun 2026-01-08 01:03:21 -08:00
  • 9c142e3a2a model-conversion : add warn about transformers mismatch (#18691) Daniel Bevenius 2026-01-08 09:29:53 +01:00
  • df7fb92170 model-conversion : remove -st targets for converted model (#18689) Daniel Bevenius 2026-01-08 09:29:15 +01:00
  • 2038101bd9 llama : add use_direct_io flag for model loading (#18166) Julius Tischbein 2026-01-08 07:35:30 +01:00
  • 568371a726 opencl: add FILL op support (#18682) shaofeiqi 2026-01-07 22:04:50 -08:00
  • 5b8844ae53 scripts : fix repos cloned with .git extension (#18669) Sigbjørn Skjæret 2026-01-07 22:35:34 +01:00
  • 7e16fef085 convert : more variants of rope_theta config entries (#18668) Sigbjørn Skjæret 2026-01-07 22:34:51 +01:00
  • f5245b5e4e cuda : fix build on cuda 12.8 (#18672) Oliver Walsh 2026-01-07 21:32:44 +00:00
  • ae9f8df778 fix(docker): add missing libglvnd libraries to Vulkan image (#18664) R 2026-01-07 16:57:42 +01:00
  • 56d2fed2b3 tools : remove llama-run (#18661) Adrien Gallouët 2026-01-07 16:18:26 +01:00
  • 56426673cb scripts : add pr2wt.sh (#18644) Georgi Gerganov 2026-01-07 15:16:20 +02:00
  • bb77764c2d convert : clarify sentence-transformers-dense-modules help [no ci] (#18662) Daniel Bevenius 2026-01-07 13:18:53 +01:00
  • 9dfa8ee950 ci : run cann build unconditionally [no ci] (#18659) Sigbjørn Skjæret 2026-01-07 13:07:08 +01:00
  • ca4a8370bc vulkan: reject ops when a tensor is too large to allocate (#18646) Jeff Bolz 2026-01-07 05:03:32 -06:00
  • 03023296cf vulkan: Warptile tuning for Intel Xe2/Xe3 (#18178) virajwad 2026-01-07 02:59:47 -08:00
  • 8c77a04cc7 vulkan: more mul mat optimizations (#18533) Eve 2026-01-07 10:13:17 +00:00
  • ffba4f29e6 examples : add debug utility/example (#18464) Daniel Bevenius 2026-01-07 10:42:19 +01:00
  • 3333951d86 CANN: Fix rename for get_env (#18652) hipudding 2026-01-07 16:11:31 +08:00
  • 193ee38a1b CANN: Rename get_env to get_env_as_lowercase (#18624) Raul Torres 2026-01-07 02:01:25 +00:00
  • 95ea9e0861 Hexagon add support for f16/f32 flash attention, scale, set-rows and improve f16/32 matmul (#18611) Max Krasnyansky 2026-01-06 17:38:29 -08:00
  • ccbc84a537 mtmd: mtmd_audio_streaming_istft (#18645) Tarek Dakhran 2026-01-06 21:00:29 +01:00
  • 68b4d516c3 llama-params-fit: fix last devices with low VRAM (#18494) Johannes Gäßler 2026-01-06 20:02:30 +01:00
  • 24af22fc36 ggml : optimize cuda ssm_scan using warp-level reduction (#18505) Aadeshveer Singh 2026-01-06 23:54:34 +05:30
  • 07fbe19f1f arg: use CSV escape style for multiple-value args (#18643) Xuan-Son Nguyen 2026-01-06 17:51:08 +01:00
  • ea13cba850 vulkan: support buffer_from_host_ptr (#18467) Jeff Bolz 2026-01-06 10:37:07 -06:00
  • 090b137e56 ggml-cuda: refactor cuda graph usage (#18637) Aman Gupta 2026-01-06 23:48:45 +08:00
  • 968929528c mmq.cu: tune mmq/rocblas switching for RDNA (#18537) Beinsezii 2026-01-06 07:26:07 -08:00
  • 3d26a09dc7 server : add thinking content blocks to Anthropic Messages API (#18551) R 2026-01-06 16:17:13 +01:00
  • bd2a93d475 gguf-py : add requests to dependencies (#18629) Christian Kastner 2026-01-06 08:56:38 +01:00
  • e75ee11024 ggml : fix avx512bf16 build (#18623) Adrien Gallouët 2026-01-06 07:54:10 +01:00
  • da9b8d3300 CANN: Make valid_values variable static const (#18627) Raul Torres 2026-01-06 03:53:28 +00:00
  • e443fbcfa5 ggml webgpu: add CEIL operation support (#18605) nwyin 2026-01-05 13:38:57 -06:00
  • 73d284a250 model : add LFM2-ColBert-350M (#18607) Tarek Dakhran 2026-01-05 19:52:56 +01:00
  • df17a4c94f CUDA: fix FA FP16 accumulator overflow for Granite (#18614) Johannes Gäßler 2026-01-05 19:51:13 +01:00
  • 1871f0ba56 add YoutuVLForConditionalGeneration architectures (#18620) tt 2026-01-06 01:15:14 +08:00
  • f47edb8c19 ggml-cuda: check for srcs outside the cgraph (#18583) Aman Gupta 2026-01-05 22:46:36 +08:00
  • da143b9940 server : fix router child env in containerized environments (#18562) Vladislav Sayapin 2026-01-05 16:12:05 +03:00
  • f1768d8f03 vulkan: fix topk_moe_sigmoid_norm_bias failures in GLM-4.6 (#18582) Jeff Bolz 2026-01-05 04:51:39 -06:00
  • 2da64a2f8a models : fix backend assignment for Granite/Nemotron graphs (#18599) Georgi Gerganov 2026-01-05 12:34:23 +02:00
  • b37124d2d2 vulkan: handle quantize_q8_1 overflowing the max workgroup count (#18515) Jeff Bolz 2026-01-05 04:30:14 -06:00
  • eadc4184ca llama : refactor rope_freq_base/scale_swa conversion and init (#18553) Sigbjørn Skjæret 2026-01-05 09:14:04 +01:00
  • 67e3f6f601 CANN: add operator fusion support for ADD + RMS_NORM (#17512) Chenguang Li 2026-01-05 15:38:18 +08:00
  • 92ac1e016b doc: clarify that steps also apply to linux for opencl (#18002) Francisco Herrera 2026-01-04 23:39:25 -05:00
  • 8e3a761189 ci : init git lfs in every build for RISC-V (#18590) Ali Tariq 2026-01-05 06:18:33 +05:00
  • d3dce4e0a5 sampling : add support for backend sampling (#17004) Daniel Bevenius 2026-01-04 21:22:16 +01:00
  • 4974bf53cf model : mtmd : make input norm optional in LFM2-VL (#18594) Tarek Dakhran 2026-01-04 18:50:02 +01:00
  • 908a9e5a1e CUDA: disable cuda graph when using n-cpu-moe (#18593) Aman Gupta 2026-01-05 01:37:48 +08:00
  • 5126c41c1c ggml-cuda: remove unused params in ggml_cuda_graph (#18579) Aman Gupta 2026-01-05 01:37:09 +08:00
  • cef1d23c5a common/grammar : replace problematic backtracking regex [\s\S]* (#18342) Aldehir Rojas 2026-01-03 16:02:43 -06:00
  • c69c7ebc90 graph : fix graph reuse logic when n_pos_per_embd > 1 (#18566) Georgi Gerganov 2026-01-03 23:59:06 +02:00
  • e57f52334b ggml-cuda: fixes for concurrent streams (#18496) Aman Gupta 2026-01-03 23:15:01 +08:00
  • a554a1ecc7 context : fix reserve token padding to n_seqs (#18536) Georgi Gerganov 2026-01-03 15:45:34 +02:00
  • 0f2e42ca1d CUDA: only allocate FA tmp buffer if needed (#18564) Johannes Gäßler 2026-01-03 13:55:53 +01:00
  • 9dba9f5352 (Bugfix, ggml-cuda) Pool alloc count fix + small size computation type adjustment (#18559) pl752 2026-01-03 15:13:40 +05:00
  • bcfc8c3cec ggml-hexagon: optimize activation function (#18393) Shouyu 2026-01-03 00:24:24 -05:00
  • 18ddaea2ae vulkan: Optimize GGML_OP_CUMSUM (#18417) Jeff Bolz 2026-01-02 15:32:30 -06:00
  • 706e3f93a6 vulkan: Implement mmvq for iq1_s/iq1_m (#18450) Jeff Bolz 2026-01-02 13:19:04 -06:00
  • 5755e52d15 model : Maincoder-1B support (#18534) Prabod 2026-01-03 06:11:59 +11:00
  • f38de16341 metal : adjust extra size for FA buffer to avoid reallocations (#18545) Georgi Gerganov 2026-01-02 19:02:18 +02:00
  • af1e8e1a6c graph : reduce topology branching (#18548) Georgi Gerganov 2026-01-02 19:01:56 +02:00
  • d84a6a98be vocab : reduce debug logs about non-EOG control tokens (#18541) Georgi Gerganov 2026-01-02 16:17:33 +02:00
  • c6f0e832da rpc : use unordered_map::reserve and emplace (#18513) Chris Rohlf 2026-01-02 05:09:36 -05:00
  • e86f3c2221 cuda : fix copy of large tensors (ggml_nbytes <= INT_MAX assertion) (#18433) MeeMin 2026-01-02 04:54:20 +05:30
  • 169ee68ffb model : remove modern-bert iswa template (#18529) Sigbjørn Skjæret 2026-01-02 00:06:42 +01:00
  • ced765be44 model: support youtu-vl model (#18479) tt 2026-01-02 02:25:54 +08:00
  • 3ccccc83f7 Add conversion support for IQuestCoderForCausalLM (#18524) Piotr Wilkin (ilintar) 2026-01-01 18:45:55 +01:00