vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Commit graph

  • 717f0e5100
         tests : re-enable DeepSeek tokenizer tests Georgi Gerganov 2024-07-21 10:40:05 +03:00
  • 1d44fc0fbf
         tests : re-enable MPT tokenizer tests Georgi Gerganov 2024-07-21 10:38:48 +03:00
  • eeeffb84f4
         models : remove old stablelm vocab Georgi Gerganov 2024-07-21 10:38:06 +03:00
  • a6ece88d1b
         models : remove duplicated gpt-2 vocab Georgi Gerganov 2024-07-21 10:37:23 +03:00
  • f4600e653c handle regex Jason Stillerman 2024-07-21 03:07:33 -04:00
  • 689e38cacc
         Update src/llama.cpp Jason Stillerman 2024-07-21 02:49:10 -04:00
  • 7647916c92
         Update convert_hf_to_gguf_update.py Jason Stillerman 2024-07-21 02:48:50 -04:00
  • 6e15f9b51d
         Merge branch 'master' into merge-to-upstream-v2 Keke Han 2024-07-21 10:15:44 +08:00
  • 22f281aa16
         examples : Rewrite pydantic_models_to_grammar_examples.py (#8493) M-A 2024-07-20 22:09:17 -04:00
  • 328884f421
         gguf-py : fix some metadata name extraction edge cases (#8591) compilade 2024-07-20 21:58:49 -04:00
  • c69c63039c
         convert_hf : fix Gemma v1 conversion (#8597) compilade 2024-07-20 21:53:01 -04:00
  • cf206e036b flake.lock: Update github-actions[bot] 2024-07-21 00:20:07 +00:00
  • 28bd56fd13 Adding SmolLM Pre Tokenizer Jason Stillerman 2024-07-20 18:48:45 -04:00
  • 793a1cdc6d fix export-lora example ngxson 2024-07-21 00:45:00 +02:00
  • 23cf064e3b Replace 'codepoint_flags' with 'codepoint_categ' jaime-m-p 2024-07-20 23:28:05 +02:00
  • 2636cb6170 Decode unicode data categories jaime-m-p 2024-07-20 23:19:42 +02:00
  • 8f9f05bf6d Update unicode data jaime-m-p 2024-07-20 23:12:08 +02:00
  • ba4bbbd1ad Reimplement 'codepoint_flags' as 'codepoint_categ' jaime-m-p 2024-07-20 23:09:33 +02:00
  • 5ceab90b4d Store all unicode codepoint categories jaime-m-p 2024-07-20 23:04:23 +02:00
  • 3d16f647d1 Update bruteforce test: jaime-m-p 2024-07-20 22:57:59 +02:00
  • 1932a1b871 gguf-py : do not use title case for naming convention compilade/fix-metadata-name-extraction Francis Couture-Harpin 2024-07-20 16:47:43 -04:00
  • bf8e71b0c0 convert_lora : fix default filename Francis Couture-Harpin 2024-07-20 16:40:58 -04:00
  • 69c487f4ed
         CUDA: MMQ code deduplication + iquant support (#8495) b3428 Johannes Gäßler 2024-07-20 22:25:26 +02:00
  • a3d154b260 gguf-py : add more name metadata extraction tests Francis Couture-Harpin 2024-07-20 15:57:46 -04:00
  • 6515e787d1 mistral nemo inference support Douglas Hanley 2024-07-19 16:04:59 -05:00
  • f675874c14 examples : Rewrite pydantic_models_to_grammar_examples.py Marc-Antoine Ruel 2024-07-15 09:51:33 -04:00
  • 07283b1a90
         gguf : handle null name during init (#8587) b3427 Georgi Gerganov 2024-07-20 17:15:42 +03:00
  • 0282b71628 1 less parallel job for CI build Johannes Gäßler 2024-07-20 16:07:40 +02:00
  • 940362224d
         llama : add support for Tekken pre-tokenizer (#8579) b3426 Michael Coppola 2024-07-20 09:43:51 -04:00
  • 69b9945b44
         llama.swiftui: fix end of generation bug (#8268) b3425 Huifeng Ou 2024-07-20 09:09:37 -04:00
  • 271cc20c67
         minor : add space Georgi Gerganov 2024-07-20 16:08:58 +03:00
  • 30f122e7ae
         minor : fix trailing whitespace Georgi Gerganov 2024-07-20 16:07:15 +03:00
  • 096c3da8dd Merge branch 'embed_files' of https://github.com/katsu560/llama.cpp into embed_files katsu560 2024-07-20 19:03:38 +09:00
  • 562d4f2de4 delete commented line katsu560 2024-07-20 18:41:13 +09:00
  • b173c4e061 feat: update tensor name when bind to graph hongruichen 2024-07-20 16:57:51 +08:00
  • 5f3b1ae3b0 fix: try fix graph cache with append the tensors name hongruichen 2024-07-20 16:21:09 +08:00
  • 51f95d6980 fix: dimension could be wrong for tensor liked 1x1x8 hongruichen 2024-07-20 16:11:35 +08:00
  • c3776cacab
         gguf_dump.py: fix markddown kv array print (#8588) Brian 2024-07-20 17:35:25 +10:00
  • 27299463ae fix: try fix tensor type error hongruichen 2024-07-20 14:23:44 +08:00
  • 292a46906d change pr readme caitianchi 2024-07-20 14:45:19 +08:00
  • 1949847a9c gguf_dump.py: handle edge case about backticks on start or end of a string brian khuu 2024-07-20 16:17:01 +10:00
  • 28a00e5e6c fix: try fix QNN_GRAPH_ERROR_INVALID_OP_CONFIG hongruichen 2024-07-20 14:10:00 +08:00
  • c8ee1bccdd Fix Vulkan matmul tests compile errors 0cc4m/vulkan-fix-mm-tests 0cc4m 2024-07-20 08:01:18 +02:00
  • 50d55d601f gguf_dump.py: inline code markdown escape handler added brian khuu 2024-07-20 15:59:37 +10:00
  • 1679dcf47e fix: check all dimentions in can offload hongruichen 2024-07-19 22:56:00 +08:00
  • 1d37843094 gguf_dump.py: escape backticks inside of strings brian khuu 2024-07-20 15:26:37 +10:00
  • 923886f7ba gguf_dump.py: refactor kv array string handling brian khuu 2024-07-20 14:31:24 +10:00
  • e8c7a25c9e change variable name to is_done (variable name suggested by ggerganov) o 2024-07-19 23:50:33 -04:00
  • 130d396ec7
         Update gguf-py/scripts/gguf_dump.py Brian 2024-07-20 13:42:39 +10:00
  • 50d1a035f0 convert_hf : fix Gemma v1 not setting BOS and EOS tokens compilade/fix-convert-gemma-1-instruct Francis Couture-Harpin 2024-07-19 22:46:35 -04:00
  • dd5a0bfffc * Tekken pre-tokenizer no longer uses clean_up_tokenization_spaces * Updated chkhsh for Tekken tokenizer Michael Coppola 2024-07-19 19:09:08 -04:00
  • f0f71a5dc1 CUDA: MMQ code deduplication + iquant support Johannes Gäßler 2024-07-12 09:35:38 +02:00
  • 5a9cb57494 convert_hf : fix Gemma v1 conversion Francis Couture-Harpin 2024-07-19 16:57:48 -04:00
  • a083c6cf6d
         chore : Add gguf template entrypoint to scripts sub package and fix toml entry teleprint-me 2024-07-19 14:53:56 -04:00
  • fe883305c3
         chore : Add gguf_template to poetry scripts teleprint-me 2024-07-19 14:38:17 -04:00
  • 0de43fcc00
         chore : Set sentencepiece to 0.2.0 to match requirements.txt teleprint-me 2024-07-19 14:05:21 -04:00
  • 912e6fa5c6 gguf-py : more metadata edge cases fixes Francis Couture-Harpin 2024-07-19 13:46:41 -04:00
  • 2164c9deb3 gguf-py : fix some metadata name extraction edge cases Francis Couture-Harpin 2024-07-19 12:30:37 -04:00
  • 6875ace401
         Merge branch 'gguf-model-template' of github.com:teleprint-me/llama.cpp into gguf-model-template teleprint-me 2024-07-19 12:18:56 -04:00
  • 5481cec7c9
         Merge branch 'ggerganov:master' into gguf-model-template Austin 2024-07-19 12:20:33 -04:00
  • 27070de9af
         Merge branch 'master' of github.com:teleprint-me/llama.cpp into gguf-model-template teleprint-me 2024-07-19 12:16:56 -04:00
  • 56b8229937 llama : fix codeshell support hankeke303 2024-07-19 23:52:36 +08:00
  • 3db5058dd3 Merge branch 'master' into tokenizer-fixes jaime-m-p 2024-07-19 17:23:08 +02:00
  • 87e397d00b
         ggml : fix quant dot product with odd number of blocks (#8549) b3423 slaren 2024-07-19 17:17:27 +02:00
  • b1b5cc10b1 add function to convert qnn error into string hongruichen 2024-07-19 22:51:17 +08:00
  • 8cc26be77c
         ggml : fix odd blocks for ARM_NEON (#8556) Georgi Gerganov 2024-07-19 17:13:17 +03:00
  • 57b1d4f9eb
         convert-*.py: remove add_name from ChatGLMModel class (#8590) b3422 Brian 2024-07-20 00:04:38 +10:00
  • 5ff83bd3ae
         Merge branch 'sl/fix-iqnl-odd-blocks' into gg/fix-odd-blocks-arm Georgi Gerganov 2024-07-19 17:01:18 +03:00
  • d197545530
         llama : bump max layers from 256 to 512 (#8530) b3421 Georgi Gerganov 2024-07-19 16:50:47 +03:00
  • 31c11e14b7 convert-*.py: remove add_name from ChatGLMModel class brian khuu 2024-07-19 23:47:35 +10:00
  • d99a34b245 gguf_dump.py: fix markddown kv array print brian khuu 2024-07-19 23:23:46 +10:00
  • 8506e13940
         Merge branch 'ggerganov:master' into master Michael Coppola 2024-07-19 09:34:51 -04:00
  • be0cfb4175
         readme : fix server badge b3420 Georgi Gerganov 2024-07-19 14:34:55 +03:00
  • b57eb9ca4f
         ggml : add friendlier error message to fopen errors (#8575) b3419 Clint Herron 2024-07-19 07:05:45 -04:00
  • 32e9c41812
         ggml : fix sumf redefinition Georgi Gerganov 2024-07-19 13:59:38 +03:00
  • 38061254b9
         gguf : handle null name during init gg/gguf-fix-null-defer Georgi Gerganov 2024-07-19 13:45:00 +03:00
  • 447c08092d
         llama : fix order of pre-tokenizers Georgi Gerganov 2024-07-19 13:21:38 +03:00
  • f299aa98ec
         fix: typo of chatglm4 chat tmpl (#8586) b3418 Frank Mai 2024-07-19 17:44:41 +08:00
  • 7173ea5c08 fix: typo of chatglm4 chat tmpl thxCode 2024-07-19 17:10:21 +08:00
  • 3d0e4367d9
         convert-*.py: add general.name kv override (#8571) b3417 Brian 2024-07-19 17:51:51 +10:00
  • a607995f95 Reapply "tried fix the add node error 6005" hongruichen 2024-07-19 15:35:55 +08:00
  • 0153a23d3f fix support ops hongruichen 2024-07-19 15:22:23 +08:00
  • f45fbec8f4 Revert "tried fix the add node error 6005" hongruichen 2024-07-19 12:59:38 +08:00
  • ce3d09e5f2 tried fix the add node error 6005 hongruichen 2024-07-19 10:13:56 +08:00
  • 5959b14b06 fix llama-minicpmv-cli in cmake file caitianchi 2024-07-19 11:29:17 +08:00
  • 7fc85054bf llama : Added support for Tekken pre-tokenizer (#8577) Michael Coppola 2024-07-18 18:00:45 -04:00
  • d93f725d9d fix miss-matched parenthesis Stéphane du Hamel 2024-07-19 00:01:46 +02:00
  • a15ef8f8a0
         CUDA: fix partial offloading for ne0 % 256 != 0 (#8572) b3416 Johannes Gäßler 2024-07-18 23:48:47 +02:00
  • 5edd6ea8bc WIP for adding support for Tekken tokenizer needed for Mistral NeMo support. Clint Herron 2024-07-18 17:34:06 -04:00
  • 8784fcd5c8 CUDA: fix partial offloading for ne0 % 256 != 0 Johannes Gäßler 2024-07-18 18:55:41 +02:00
  • cc6a0f5275 ggml : fix iq4_nl dot product with odd number of blocks slaren 2024-07-17 23:11:05 +02:00
  • cd3ef015a5 Adding additional error information to most instances of fopen. Clint Herron 2024-07-18 13:18:56 -04:00
  • 0c5f49d507 Add additional error information when model files fail to load. Clint Herron 2024-07-18 13:11:16 -04:00
  • 02aea9eb19 Merge branch 'snapdragonxwin-fix1' of https://github.com/AndreasKunar/llama.cpp into snapdragonxwin-fix1 AndreasKunar 2024-07-18 18:46:07 +02:00
  • 6f2366210d cmake build for Windows clarifications AndreasKunar 2024-07-18 18:45:45 +02:00
  • 4f17468b9d Vulkan-shaders: attempt fix compilation on windows Stéphane du Hamel 2024-07-18 17:32:42 +02:00
  • 705b7ecf60
         cmake : install all ggml public headers (#8480) b3415 65a 2024-07-18 07:47:12 -07:00
  • 665f823748 fix op checker hongruichen 2024-07-18 20:26:05 +08:00
  • e833680ac6 convert-*.py: add general.name kv override brian khuu 2024-07-19 00:12:55 +10:00
  • 871644b04b
         Merge branch 'ggerganov:master' into snapdragonxwin-fix1 Andreas (Andi) Kunar 2024-07-18 15:17:31 +02:00