Commit graph

  • bba4b66a81
    concat: Handle SYCL exceptions Akarshan Biswas 2025-02-02 18:34:04 +05:30
  • 1ccfaaedbb
    Add sum to backend hpp Akarshan Biswas 2025-02-02 18:24:45 +05:30
  • d31c62d758
    norm: add try catch sycl exception Akarshan Biswas 2025-02-02 18:20:44 +05:30
  • 5c05a3eedc
    Move sum and sum rows to a separate file Akarshan Biswas 2025-02-02 18:16:41 +05:30
  • eb466d733a
    pool2d: move to a separate file Akarshan Biswas 2025-02-02 17:49:19 +05:30
  • 4db56d6ed2
    im2col: add try catch block and move wrapper function from ggml-sycl.cpp Akarshan Biswas 2025-02-02 17:37:48 +05:30
  • ba79258a2b
    Add spaces to end of files Akarshan Biswas 2025-02-02 12:16:36 +05:30
  • ddc5e428f2
    clamp: move to a separate file Akarshan Biswas 2025-02-02 12:08:22 +05:30
  • 0c319bf721
    DUP: move to cpy.cpp, set debug logs and adjust include Akarshan Biswas 2025-02-02 11:48:03 +05:30
  • 927925ffe2
    scale: move to a separate file Akarshan Biswas 2025-02-02 11:34:37 +05:30
  • 7f2d24fdca
    rope: add try catch sycl exception and debug log Akarshan Biswas 2025-02-02 11:21:36 +05:30
  • 8e86732cf2
    diagmask: move to a separate file Akarshan Biswas 2025-02-01 19:33:52 +05:30
  • 98f5fd2fd1
    getrows: move to a separate file Akarshan Biswas 2025-02-01 19:08:42 +05:30
  • 04d8b038b8
    Add back split buffer type checks Akarshan Biswas 2025-02-01 12:06:58 +05:30
  • 7d8d689d39
    eltwise: add back split buffer type checks Akarshan Biswas 2025-02-01 11:45:18 +05:30
  • ecacff3f6e
    CPY: move to a separate file Akarshan Biswas 2025-02-01 11:21:09 +05:30
  • a16b6b7681
    eltwise: sort includes Akarshan Biswas 2025-02-01 10:59:28 +05:30
  • aaf9ed070d
    Add spaces Akarshan Biswas 2025-02-01 09:49:42 +05:30
  • 3a346592b8
    argsort: add a space at the end of file Akarshan Biswas 2025-02-01 09:48:29 +05:30
  • 51bedb847e
    argmax: move missing function to file and fix function name Akarshan Biswas 2025-02-01 09:44:30 +05:30
  • a153f1972d
    ggml_sycl_compute_forward: fixup function calling names and remove comments Akarshan Biswas 2025-02-01 09:40:44 +05:30
  • 5288bd5896
    Argsort: move to a separate file Akarshan Biswas 2025-02-01 09:37:29 +05:30
  • 95a09ab505
    ARGMAX: move to a separate file Akarshan Biswas 2025-02-01 09:22:25 +05:30
  • fa7c4d86f3
    Fix GGML_SYCL_DEBUG in kernels in other files Akarshan Biswas 2025-01-31 21:13:28 +05:30
  • e1326a7897
    binbcast: add try catch sycl::exception Akarshan Biswas 2025-01-31 20:51:12 +05:30
  • 108be39dfe
    binbcast: move to a separate file Akarshan Biswas 2025-01-31 20:10:44 +05:30
  • 957c11b2cf
    binbcast: use void pointer to prevent intermediate type conversions Akarshan Biswas 2025-01-31 18:30:29 +05:30
  • 2d72bd94b0
    SYCL: remove ggml_sycl_op_flatten function Akarshan Biswas 2025-01-30 19:46:34 +05:30
  • 7fd0ae588b ggml-cpu : add chunking support to mul_mat_id slaren 2025-02-05 01:34:35 +01:00
  • 1f1f06aa26
    Merge branch 'master' into r1-toolcall Olivier Chafik 2025-02-05 01:10:45 +00:00
  • 9f4cc8f8d3
    sync: minja (#11641) b4641 Olivier Chafik 2025-02-05 01:00:12 +00:00
  • fd08255d0d
    CUDA: non-contiguous (RMS) norm support (#11659) b4640 Johannes Gäßler 2025-02-04 22:21:42 +01:00
  • 1f9316651f whitespace fix Michal Moskal 2025-02-04 11:40:21 -08:00
  • 73befb5306 change VS Community link from DE to EN Michal Moskal (from Dev Box) 2025-02-04 11:31:46 -08:00
  • 274c0ecd84 add build instructions for windows and update script link Michal Moskal (from Dev Box) 2025-02-04 11:31:28 -08:00
  • 7d8c170cf5 setup windows linking for llguidance; thanks @phil-scott-78 Michal Moskal (from Dev Box) 2025-02-04 11:31:04 -08:00
  • df11fb7033
    Add information for Podman as well as Docker Daniel J Walsh 2025-02-04 09:27:18 -05:00
  • 36c2f388cc Merge branch 'master' into llamacli-tools Mason M 2025-02-04 15:05:43 -04:00
  • 34370803fb Pass template group to common_chat_apply_template Mason M 2025-02-04 15:00:58 -04:00
  • 3ec9fd4b77
    HIP: force max threads per block to be 1024 (#11621) b4639 fxzjshm 2025-02-05 02:18:38 +08:00
  • 5d60cebbcc Update test_tool_call.py Olivier Chafik 2025-02-04 17:48:29 +00:00
  • 3962fc1a79
    server : add try..catch to places not covered by set_exception_handler (#11620) Xuan-Son Nguyen 2025-02-04 18:25:42 +01:00
  • 4e8beb0c53 Reset tools when empty string provided Mason M 2025-02-04 13:21:28 -04:00
  • 1bef571f6a
    arg : list RPC devices first when using --list-devices (#11655) b4637 Radoslav Gerganov 2025-02-04 18:16:20 +02:00
  • 8ef9a5ae14 try CI fix Johannes Gäßler 2025-02-04 17:05:42 +01:00
  • a0402ad95d try CI fix Johannes Gäßler 2025-02-04 17:03:38 +01:00
  • 421e1f0cbd Merge branch 'master' into ci-write-less-ccache Olivier Chafik 2025-02-04 15:58:26 +00:00
  • 178ad4e8c9 Merge branch 'master' into cuda-releases Olivier Chafik 2025-02-04 15:57:59 +00:00
  • 933f7a186e Merge branch 'master' into r1-toolcall Olivier Chafik 2025-02-04 15:56:25 +00:00
  • cd16957f7d Add variant include Mason M 2025-02-04 11:54:48 -04:00
  • 57d170fb98 try CI fix Johannes Gäßler 2025-02-04 16:50:00 +01:00
  • af7fead947 Update ggml/src/ggml-cuda/norm.cu Johannes Gäßler 2025-02-04 15:42:29 +01:00
  • 386c52cfeb CUDA: non-contiguous (RMS) norm support Johannes Gäßler 2025-02-04 14:11:53 +01:00
  • db288b60cb
    tool-call: command r7b fix for normal responses (#11608) b4636 Olivier Chafik 2025-02-04 15:48:53 +00:00
  • becf9b4003 add tool-choice parameter Mason M 2025-02-04 11:40:29 -04:00
  • 988702168a CUDA: support for mat. mul. with ne03 != ne13 Johannes Gäßler 2025-02-03 14:11:22 +01:00
  • 8777473a43 Remove white trailing sami 2025-02-04 21:46:39 +07:00
  • b2d17287aa update readme section about common model tool call formats Olivier Chafik 2025-02-04 14:27:38 +00:00
  • c31dd2aac2
    build : fix llama.pc Adrien Gallouët 2025-02-04 14:07:04 +00:00
  • bddb1efc1f common : change longest common subsequence to substring [no ci] Daniel Bevenius 2025-02-04 14:10:30 +01:00
  • 2758680299 Merge branch 'master' into c7rb-fix Olivier Chafik 2025-02-04 12:54:18 +00:00
  • f4eb1b3854 Add support for multithread LHS conversion Charles Xu 2025-02-04 13:38:27 +01:00
  • f415d61ed1 arg : list RPC devices first when using --list-devices Radoslav Gerganov 2025-02-04 14:15:11 +02:00
  • 8af6664d6b
    Merge e0580f9d66 into 106045e7bb Sigbjørn Skjæret 2025-02-04 13:11:39 +01:00
  • 39c1d8163b return thoughts in reasoning_content field Olivier Chafik 2025-02-04 11:37:09 +00:00
  • 106045e7bb
    readme : add llm_client Rust crate to readme bindings (#11628) Shelby Jenkins 2025-02-04 05:20:55 -06:00
  • f117d84b48
    swift : fix llama-vocab api usage (#11645) b4634 Jhen-Jie Hong 2025-02-04 19:15:24 +08:00
  • 534c46b53c
    metal : use residency set for other platforms (#11648) b4633 Jhen-Jie Hong 2025-02-04 19:07:18 +08:00
  • 387a1598ca
    authors : update Georgi Gerganov 2025-02-04 13:04:10 +02:00
  • 7c9e0ca520
    sync : ggml b4631 Georgi Gerganov 2025-02-04 12:59:21 +02:00
  • 8f8290ada9
    cmake: Add ability to pass in GGML_BUILD_NUMBER (ggml/1096) Christian Kastner 2025-02-04 00:17:15 +01:00
  • d1b66910c5 r1: revert making <|tool▁calls▁begin|> optional as somehow sampling triggers us on "<|tool▁call▁begin|><", which is already invalid per the grammar ochafik 2025-02-04 10:38:03 +00:00
  • 0db9881285 Fix r1 grammar since we made <|tool▁calls▁begin|> optional (triggering on just <|tool▁call▁begin|> for 7B's sake) ochafik 2025-02-04 10:30:10 +00:00
  • b5b117fa1c Merge branch 'sync-minja-4' into r1-toolcall ochafik 2025-02-04 09:45:27 +00:00
  • 18f15b104c swift : revert changes Jhen-Jie Hong 2025-02-04 16:14:34 +08:00
  • b34aedd558
    ci : do not stale-close roadmap issues Georgi Gerganov 2025-02-04 09:30:42 +02:00
  • b0569130c5 metal : use residency set for other platforms Jhen-Jie Hong 2025-02-04 15:25:21 +08:00
  • b7a0c02658 server: (webui) use Map() to store file content dannyl1u 2025-02-03 23:18:11 -08:00
  • 8d721dcca8 server: (webui) file upload and pdf parsing dannyl1u 2025-02-03 22:45:52 -08:00
  • c31bd273ef batched.swift : fix vocab api usage Jhen-Jie Hong 2025-02-04 14:42:39 +08:00
  • b7fafb7f2a Add script to convert Janus encoder to GGUF format and update requirements ravenouse 2025-02-04 06:40:38 +00:00
  • 6b9e659cfb swiftui : fix vocab api usage Jhen-Jie Hong 2025-02-04 14:12:49 +08:00
  • 20806cf16f Added quantization for visual projector * Added README * Fixed the clip quantize implementation in the file sami 2025-02-04 12:58:42 +07:00
  • 21f207156f Update chat.cpp ochafik 2025-02-04 05:16:23 +00:00
  • 23bce61a2f Added README * Also allows no --image argument cli to the qwen2vl-cli sami 2025-02-04 12:13:57 +07:00
  • 59ad593a95
    HIP: force max threads per block to be 1024 fxzjshm 2025-02-03 22:33:38 +08:00
  • bbf98f1c4d squash! server : use httplib status codes Daniel Bevenius 2025-02-04 06:04:23 +01:00
  • 9779626803 squash! server : use httplib status codes [no ci] Daniel Bevenius 2025-02-04 06:03:32 +01:00
  • 438ce0b8a1 fix test-chat ochafik 2025-02-04 04:51:36 +00:00
  • 1f5ec59809 ensure deepseek r1 thoughts parsed even w/o tool calls ochafik 2025-02-04 04:48:08 +00:00
  • b6e14a4101 fix mistral expectation ochafik 2025-02-04 04:26:49 +00:00
  • d44eb95c67 tool-call: ensure we don't return content when there are tool calls / warn ochafik 2025-02-04 04:18:49 +00:00
  • 812544ab8b server: check that content is null when we get tool_calls ochafik 2025-02-04 04:14:15 +00:00
  • d43e4f6c22 Merge branch 'sync-minja-4' into r1-toolcall ochafik 2025-02-04 04:05:02 +00:00
  • f12e3507f7 Update chat.cpp ochafik 2025-02-04 04:02:18 +00:00
  • 56a14ddc83 fix mistral chat test: need empty tokens ochafik 2025-02-04 04:01:35 +00:00
  • b1527292b6 Update test-chat.cpp ochafik 2025-02-04 03:56:03 +00:00
  • 09caa63451 sync: minja ochafik 2025-02-04 03:52:59 +00:00
  • 86994db697 fix spaces ochafik 2025-02-04 03:47:52 +00:00
  • 78b47bb0e9 fix test_calc_result ochafik 2025-02-04 03:46:26 +00:00