{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":612354784,"defaultBranch":"master","name":"llama.cpp","ownerLogin":"ggerganov","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2023-03-10T18:58:00.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/1991296?v=4","public":true,"private":false,"isOrgOwned":false},"refInfo":{"name":"","listCacheKey":"v0:1716657867.0","currentOid":""},"activityList":{"items":[{"before":"61a88a1da399be2207c8aa0a8a280dffc3f64887","after":"fc59407efea1d49a3d8338fd20fa38afbe06fdb5","ref":"refs/heads/compilade/refactor-kv-cache","pushedAt":"2024-05-25T17:57:00.000Z","pushType":"push","commitsCount":2,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"convert-hf : support Mini-Jamba conversion","shortMessageHtmlLink":"convert-hf : support Mini-Jamba conversion"}},{"before":null,"after":"80787c2a26c54998fff5f621e5aa7ae9866d0bfd","ref":"refs/heads/gg/ci-loongson","pushedAt":"2024-05-25T17:24:27.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"ggml : fix loongson compile warnings\n\nggml-ci","shortMessageHtmlLink":"ggml : fix loongson compile warnings"}},{"before":"93b9baee738dea83951d35879e4b9407a99796e9","after":"11f78c6a2d651e450330d1eee60de2415cac939f","ref":"refs/heads/compilade/lazier-moe-convert-hf","pushedAt":"2024-05-25T16:53:00.000Z","pushType":"push","commitsCount":140,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"convert-hf : adapt ArcticModel to use yield too","shortMessageHtmlLink":"convert-hf : adapt ArcticModel to use yield too"}},{"before":"3cbd23ed88c03a27e1eb6090ac4a8186ca9ac29a","after":"9588f196b1d7b21bdff013fcf958c249576b2619","ref":"refs/heads/master","pushedAt":"2024-05-25T12:22:39.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"train : change default FA argument (#7528)","shortMessageHtmlLink":"train : change default FA argument (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"6726351\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/discussions/7528\" data-hovercard-type=\"discussion\" data-hovercard-url=\"/ggerganov/llama.cpp/discussions/7528/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/discussions/7528\">#7528</a>)"}},{"before":"00c63907931bb08a0ed2b7e38cf44dd290143cb9","after":"3cbd23ed88c03a27e1eb6090ac4a8186ca9ac29a","ref":"refs/heads/master","pushedAt":"2024-05-25T09:30:42.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"labeler: added Apple Metal detector (+Kompute) (#7529)\n\n* labeler: added Apple Metal detector [no ci]\r\n\r\n* labeler: add Kompute to detector [no ci]","shortMessageHtmlLink":"labeler: added Apple Metal detector (+Kompute) (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2316528572\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7529\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7529/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7529\">#7529</a>)"}},{"before":"faa0e6979a11dcb731e9d778ad42ceaa0302015e","after":"00c63907931bb08a0ed2b7e38cf44dd290143cb9","ref":"refs/heads/master","pushedAt":"2024-05-25T09:04:03.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"main : don't print special tokens with --grammar (#6923)\n\n* main : don't print special tokens with --grammar\r\n\r\nThe CLI interface was recently changed to print special control tokens\r\nlike the </s> stop message one. This token shouldn't be printed if the\r\ngrammar flag was passed, unless the grammar specifies it, because that\r\nbreaks shell-scriptability.\r\n\r\n* main: use seperate stream for control characters\r\n\r\n* main: use dprintf and add --ctrl-token-no-out and --ctrl-token-fd-out\r\n\r\n* main: dprintf isn't part of the IEEE POSIX standard. Just use write().\r\n\r\n* main: remove --ctrl-token-fd-out in favor for fcntl() based detection\r\n\r\n* common.cpp: accidentally removed --interactive-first\r\n\r\n* main: only merge stdout and control token if not in conversation or grammar mode\r\n\r\n* main: rejig control token descriptor handling\r\n\r\n* main: must check pipe status on very top of program\r\n\r\n* main: renamed --no-special from  --ctrl-token-no-out and other refactoring\r\n\r\n* main: refactor ctrl_token_no_out --> no_special\r\n\r\n* llama: rename llama_token_is_control_token() to llama_token_is_control()\r\n\r\n* main: remove special token file descriptor feature (#5)\r\n\r\n---------\r\n\r\nCo-authored-by: Brian <mofosyne@gmail.com>","shortMessageHtmlLink":"main : don't print special tokens with --grammar (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2265336475\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/6923\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/6923/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/6923\">#6923</a>)"}},{"before":"9791f402580838d7f8543ae7bc633ef265e436f0","after":"faa0e6979a11dcb731e9d778ad42ceaa0302015e","ref":"refs/heads/master","pushedAt":"2024-05-25T08:42:31.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"ggml: aarch64: SVE kernels for q8_0_q8_0, q4_0_q8_0 vector dot (#7433)\n\n* Add SVE support for q4_0_q8_0 q8_0_q8_0\r\n\r\n* remove ifdef","shortMessageHtmlLink":"ggml: aarch64: SVE kernels for q8_0_q8_0, q4_0_q8_0 vector dot (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2307606346\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7433\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7433/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7433\">#7433</a>)"}},{"before":"902184dd3a9d6685e752b19027a48423742531db","after":"9791f402580838d7f8543ae7bc633ef265e436f0","ref":"refs/heads/master","pushedAt":"2024-05-25T08:11:33.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"android : module (#7502)\n\n* move ndk code to a new library\r\n\r\n* add gradle file","shortMessageHtmlLink":"android : module (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2313925237\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7502\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7502/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7502\">#7502</a>)"}},{"before":"57684331fc2d685f7d1f5775af0b9e47d1829833","after":"902184dd3a9d6685e752b19027a48423742531db","ref":"refs/heads/master","pushedAt":"2024-05-25T03:30:59.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"fix missing slash in `fs_get_cache_directory()` (#7503)\n\n* fix missing slash in fs_get_cache_directory()\r\n\r\n* use LOCALAPPDATA for fs_get_cache_directory()\r\n\r\n* better code style","shortMessageHtmlLink":"fix missing slash in <code>fs_get_cache_directory()</code> (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2313961225\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7503\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7503/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7503\">#7503</a>)"}},{"before":"0fd13e9473dbb7d767501c9dd64d11edefd93356","after":"61a88a1da399be2207c8aa0a8a280dffc3f64887","ref":"refs/heads/compilade/refactor-kv-cache","pushedAt":"2024-05-25T02:42:03.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"llama : fix BERT inference without KV cache","shortMessageHtmlLink":"llama : fix BERT inference without KV cache"}},{"before":"b83bab15a5d2a1e7807d09613a9b34309d86cfaa","after":"57684331fc2d685f7d1f5775af0b9e47d1829833","ref":"refs/heads/master","pushedAt":"2024-05-25T01:14:42.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"Make tokenize CLI tool have nicer command line arguments. (#6188)\n\n* Make tokenizer.cpp CLI tool nicer.\r\n\r\nBefore this commit, tokenize was a simple CLI tool like this:\r\n\r\n  tokenize MODEL_FILENAME PROMPT [--ids]\r\n\r\nThis simple tool loads the model, takes the prompt, and shows the tokens\r\nllama.cpp is interpreting.\r\n\r\nThis changeset makes the tokenize more sophisticated, and more useful\r\nfor debugging and troubleshooting:\r\n\r\n  tokenize [-m, --model MODEL_FILENAME]\r\n           [--ids]\r\n           [--stdin]\r\n           [--prompt]\r\n           [-f, --file]\r\n           [--no-bos]\r\n           [--log-disable]\r\n\r\nIt also behaves nicer on Windows now, interpreting and rendering Unicode\r\nfrom command line arguments and pipes no matter what code page the user\r\nhas set on their terminal.\r\n\r\n* style fix: strlen(str) == 0 --> *str == 0\r\n\r\n* Simplify tokenize.cpp; by getting rid of handling positional style arguments.\r\n\r\nIt must now be invoked with long --model, --prompt etc. arguments only.\r\nShortens the code.\r\n\r\n* tokenize.cpp: iostream header no longer required\r\n\r\n---------\r\n\r\nCo-authored-by: Georgi Gerganov <ggerganov@gmail.com>\r\nCo-authored-by: brian khuu <mofosyne@gmail.com>","shortMessageHtmlLink":"Make tokenize CLI tool have nicer command line arguments. (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2198743432\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/6188\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/6188/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/6188\">#6188</a>)"}},{"before":"d041d2ceaaf50e058622d92921b3e680ffa4e9e7","after":"b83bab15a5d2a1e7807d09613a9b34309d86cfaa","ref":"refs/heads/master","pushedAt":"2024-05-25T01:11:48.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"gguf-py : fix and simplify quantized shape round-trip (#7483)\n\n* gguf-py : fix and simplify quantized shape round-trip\r\n\r\n* gguf-py : remove unused import","shortMessageHtmlLink":"gguf-py : fix and simplify quantized shape round-trip (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2311891673\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7483\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7483/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7483\">#7483</a>)"}},{"before":"cbc743e6006349dde61fe214d56c2d6efa34828d","after":"0fd13e9473dbb7d767501c9dd64d11edefd93356","ref":"refs/heads/compilade/refactor-kv-cache","pushedAt":"2024-05-24T23:36:08.000Z","pushType":"push","commitsCount":20,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"Merge branch 'master' into compilade/refactor-kv-cache","shortMessageHtmlLink":"Merge branch 'master' into compilade/refactor-kv-cache"}},{"before":"7e13f19fb527b62ca87930841608b7369d86173a","after":"cbc743e6006349dde61fe214d56c2d6efa34828d","ref":"refs/heads/compilade/refactor-kv-cache","pushedAt":"2024-05-24T23:27:48.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"llama : support Jamba","shortMessageHtmlLink":"llama : support Jamba"}},{"before":"b7ec12ebf77160ed106e39935f96c371b1c6eba4","after":"7e13f19fb527b62ca87930841608b7369d86173a","ref":"refs/heads/compilade/refactor-kv-cache","pushedAt":"2024-05-24T20:21:57.000Z","pushType":"push","commitsCount":111,"pusher":{"login":"compilade","name":null,"path":"/compilade","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/113953597?s=80&v=4"},"commit":{"message":"llama : rethink recurrent state cell counts\n\n* llama : begin work on support for variable GQA\n\nThis will also be useful for Jamba if we consider the Mamba layers\nto have 0 KV heads.\n\n* llama : gracefully fail when not finding hybrid slot","shortMessageHtmlLink":"llama : rethink recurrent state cell counts"}},{"before":"150b835b96526d0c52cc0e6ff1f13a2d9a5cfd27","after":null,"ref":"refs/heads/update_flake_lock_action","pushedAt":"2024-05-24T15:59:10.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"philiptaron","name":"Philip Taron","path":"/philiptaron","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/43863?s=80&v=4"}},{"before":"27891f6db03de6e3fd5941983838c29bef253352","after":"d041d2ceaaf50e058622d92921b3e680ffa4e9e7","ref":"refs/heads/master","pushedAt":"2024-05-24T15:59:06.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"philiptaron","name":"Philip Taron","path":"/philiptaron","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/43863?s=80&v=4"},"commit":{"message":"flake.lock: Update (#7232)\n\nFlake lock file updates:\r\n\r\n• Updated input 'flake-parts':\r\n    'github:hercules-ci/flake-parts/e5d10a24b66c3ea8f150e47dfdb0416ab7c3390e?narHash=sha256-yzcRNDoyVP7%2BSCNX0wmuDju1NUCt8Dz9%2BlyUXEI0dbI%3D' (2024-05-02)\r\n  → 'github:hercules-ci/flake-parts/8dc45382d5206bd292f9c2768b8058a8fd8311d9?narHash=sha256-/GJvTdTpuDjNn84j82cU6bXztE0MSkdnTWClUCRub78%3D' (2024-05-16)\r\n• Updated input 'nixpkgs':\r\n    'github:NixOS/nixpkgs/63c3a29ca82437c87573e4c6919b09a24ea61b0f?narHash=sha256-4cPymbty65RvF1DWQfc%2BBc8B233A1BWxJnNULJKQ1EY%3D' (2024-05-02)\r\n  → 'github:NixOS/nixpkgs/4a6b83b05df1a8bd7d99095ec4b4d271f2956b64?narHash=sha256-%2BNpbZRCRisUHKQJZF3CT%2Bxn14ZZQO%2BKjxIIanH3Pvn4%3D' (2024-05-17)\r\n\r\nCo-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>","shortMessageHtmlLink":"flake.lock: Update (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2291113593\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7232\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7232/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7232\">#7232</a>)"}},{"before":"fbca2f27fc7fa9aa4a8ad0357478fdb908472908","after":"27891f6db03de6e3fd5941983838c29bef253352","ref":"refs/heads/master","pushedAt":"2024-05-24T13:47:56.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"docker.yml: disable light-intel and server-intel test (#7515)\n\n* docker.yml: disable light-intel test\r\n\r\n* docker.yml: disable server-intel test","shortMessageHtmlLink":"docker.yml: disable light-intel and server-intel test (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2315026788\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7515\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7515/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7515\">#7515</a>)"}},{"before":"0df0aa8e43c3378975269a51f9b876c8692e70da","after":"fbca2f27fc7fa9aa4a8ad0357478fdb908472908","ref":"refs/heads/master","pushedAt":"2024-05-24T12:31:14.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"fairydreaming","name":null,"path":"/fairydreaming","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/166155368?s=80&v=4"},"commit":{"message":"Add support for ArcticForCausalLM (#7020)\n\n* common : increase max number of experts to 128\r\n\r\n* common : add tensor LLM_TENSOR_FFN_NORM_EXPS for normalization before MoE that runs in parallel to attention + ffn\r\n\r\n* gguf-py : add architecture-specific block mappings that override selected general block mappings\r\n\r\n* convert-hf : add model conversion support for ArcticForCausalLM\r\n\r\n* convert-hf : use added_tokens_decoder from tokenizer_config.json to redefine tokens from SentencePiece model (only for ArcticForCausalLM)\r\n\r\n* llama : add inference support for LLM_ARCH_ARCTIC\r\n\r\n---------\r\n\r\nCo-authored-by: Stanisław Szymczyk <sszymczy@gmail.com>","shortMessageHtmlLink":"Add support for ArcticForCausalLM (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2273113105\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7020\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7020/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7020\">#7020</a>)"}},{"before":null,"after":"dd14d818e040f5b96d214371f26362aa0030d7f4","ref":"refs/heads/7507-main-intel-dockerfile","pushedAt":"2024-05-24T02:47:59.000Z","pushType":"branch_creation","commitsCount":0,"pusher":{"login":"mofosyne","name":"Brian","path":"/mofosyne","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/827793?s=80&v=4"},"commit":{"message":"Update main-intel.Dockerfile base image to 2024.1.0\n\nTo address https://github.com/ggerganov/llama.cpp/issues/7507 .\r\n\r\nSuspect intel discarded an old GPG key, so may need to update dockerfile tag.\r\n\r\nFound tag from: https://hub.docker.com/r/intel/oneapi-basekit/tags?page=&page_size=&ordering=&name=","shortMessageHtmlLink":"Update main-intel.Dockerfile base image to 2024.1.0"}},{"before":"74f33adf5f8b20b08fc5a6aa17ce081abe86ef2f","after":"0df0aa8e43c3378975269a51f9b876c8692e70da","ref":"refs/heads/master","pushedAt":"2024-05-24T02:06:56.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"airMeng","name":"Meng, Hengyu","path":"/airMeng","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39229107?s=80&v=4"},"commit":{"message":"add build shared lib in win release package (#7438)","shortMessageHtmlLink":"add build shared lib in win release package (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2308453819\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7438\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7438/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7438\">#7438</a>)"}},{"before":"27c3f2961a5ba21047a7204127ce5f8f9e28cbb5","after":"a458e6a2f0dc82e7aa0ad7a09399d69cd65da377","ref":"refs/heads/sycl-refactor","pushedAt":"2024-05-24T01:31:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"airMeng","name":"Meng, Hengyu","path":"/airMeng","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/39229107?s=80&v=4"},"commit":{"message":"backup","shortMessageHtmlLink":"backup"}},{"before":"1debe72737ea131cb52975da3d53ed3a835df3a6","after":"74f33adf5f8b20b08fc5a6aa17ce081abe86ef2f","ref":"refs/heads/master","pushedAt":"2024-05-23T14:43:39.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"readme : remove trailing space (#7469)","shortMessageHtmlLink":"readme : remove trailing space (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2310883791\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7469\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7469/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7469\">#7469</a>)"}},{"before":"007489e895bad02e4e54758bf0bdf2d6a4cdb7c1","after":"1debe72737ea131cb52975da3d53ed3a835df3a6","ref":"refs/heads/master","pushedAt":"2024-05-23T14:25:41.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"ggml : silence UB sanitizer error during iq2_xxs quantization (#0)","shortMessageHtmlLink":"ggml : silence UB sanitizer error during iq2_xxs quantization (#0)"}},{"before":"8b94e799dfa482adf63419df4905dc79b37e179f","after":"007489e895bad02e4e54758bf0bdf2d6a4cdb7c1","ref":"refs/heads/master","pushedAt":"2024-05-23T14:15:15.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ngxson","name":"Xuan Son Nguyen","path":"/ngxson","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/7702203?s=80&v=4"},"commit":{"message":"Fix phi3 chat template confusion with zephyr (#7449)\n\n* Fix phi3 template matching vs zephyr\r\n\r\n* Add regression test for new phi3 chat template\r\n\r\n* Implement review suggestions\r\n\r\n* Fix phi3 jinja test templates & match by <|end|>\r\n\r\n* Apply suggestion\r\n\r\nCo-authored-by: Xuan Son Nguyen <thichthat@gmail.com>\r\n\r\n* Add all phi3 template variants in tests\r\n\r\n* Remove unneeded message trimming\r\n\r\nCo-authored-by: Xuan Son Nguyen <thichthat@gmail.com>\r\n\r\n* Fix tests to not expect trimmed messages\r\n\r\n---------\r\n\r\nCo-authored-by: Xuan Son Nguyen <thichthat@gmail.com>","shortMessageHtmlLink":"Fix phi3 chat template confusion with zephyr (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2309386966\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7449\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7449/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7449\">#7449</a>)"}},{"before":"3015851c5ac7334fb544a23a70a284c117b87044","after":"8b94e799dfa482adf63419df4905dc79b37e179f","ref":"refs/heads/master","pushedAt":"2024-05-23T12:30:13.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"readme : add Bunny in supported models [no ci] (#7469)","shortMessageHtmlLink":"readme : add Bunny in supported models [no ci] (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2310883791\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7469\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7469/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7469\">#7469</a>)"}},{"before":"55ac3b7aeaf52f19786ed96e885d89521fc0f6c8","after":"3015851c5ac7334fb544a23a70a284c117b87044","ref":"refs/heads/master","pushedAt":"2024-05-23T12:29:26.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"llama : add getters for n_threads/n_threads_batch (#7464)\n\n* llama : add getters for n_threads/n_threads_batch\r\n\r\nThis commit adds two new functions to the llama API. The functions\r\ncan be used to get the number of threads used for generating a single\r\ntoken and the number of threads used for prompt and batch processing\r\n(multiple tokens).\r\n\r\nThe motivation for this is that we want to be able to get the number of\r\nthreads that the a context is using. The main use case is for a\r\ntesting/verification that the number of threads is set correctly.\r\n\r\nSigned-off-by: Daniel Bevenius <daniel.bevenius@gmail.com>\r\n\r\n* squash! llama : add getters for n_threads/n_threads_batch\r\n\r\nRename the getters to llama_n_threads and llama_n_threads_batch.\r\n\r\nSigned-off-by: Daniel Bevenius <daniel.bevenius@gmail.com>\r\n\r\n---------\r\n\r\nSigned-off-by: Daniel Bevenius <daniel.bevenius@gmail.com>","shortMessageHtmlLink":"llama : add getters for n_threads/n_threads_batch (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2310529302\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7464\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7464/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7464\">#7464</a>)"}},{"before":"d2bae455466102487c4c5fce15f37e750c7b5756","after":null,"ref":"refs/heads/gg/ggml-ci-pythia","pushedAt":"2024-05-23T12:28:15.000Z","pushType":"branch_deletion","commitsCount":0,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"}},{"before":"dacfcebd6022175848e978f82811a244f1033038","after":"55ac3b7aeaf52f19786ed96e885d89521fc0f6c8","ref":"refs/heads/master","pushedAt":"2024-05-23T12:28:14.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"ci : use Pythia models instead of OpenLlama (#7470)\n\n* ci : start using Pythia models over OpenLlama\r\n\r\nggml-ci\r\n\r\n* ci : disable q2_k ppl tests\r\n\r\n* ci : use convert-hf-to-gguf.py\r\n\r\n* ci : update gg_get_model\r\n\r\n* ci : fix convert outfile name\r\n\r\nggml-ci\r\n\r\n* llama : gptneox arch use F32 attn prec\r\n\r\nggml-ci","shortMessageHtmlLink":"ci : use Pythia models instead of OpenLlama (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2310939188\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7470\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7470/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7470\">#7470</a>)"}},{"before":"9b82476ee9e73065a759f8bcc4cf27ec7ab2ed8c","after":"dacfcebd6022175848e978f82811a244f1033038","ref":"refs/heads/master","pushedAt":"2024-05-23T12:12:43.000Z","pushType":"pr_merge","commitsCount":1,"pusher":{"login":"ggerganov","name":"Georgi Gerganov","path":"/ggerganov","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/1991296?s=80&v=4"},"commit":{"message":"readme : add GPT-NeoX + Pythia to the list of supported models (#7491)","shortMessageHtmlLink":"readme : add GPT-NeoX + Pythia to the list of supported models (<a class=\"issue-link js-issue-link\" data-error-text=\"Failed to load title\" data-id=\"2312758982\" data-permission-text=\"Title is private\" data-url=\"https://github.com/ggerganov/llama.cpp/issues/7491\" data-hovercard-type=\"pull_request\" data-hovercard-url=\"/ggerganov/llama.cpp/pull/7491/hovercard\" href=\"https://github.com/ggerganov/llama.cpp/pull/7491\">#7491</a>)"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEU9FDyAA","startCursor":null,"endCursor":null}},"title":"Activity · ggerganov/llama.cpp"}