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

Commit graph

  • f3590ad8d9
         remove trailing whitespace xaedes 2023-08-30 16:01:08 +02:00
  • aa4ad830e2 log.h is broken so disable it first Concedo 2023-08-30 21:58:54 +08:00
  • fc456edda6
         train-text-from-scratch can train (full finetune) gguf models xaedes 2023-08-30 15:57:17 +02:00
  • 3ef3cf8c3d llama_token_to_str -> llama_token_to_piece slaren 2023-08-30 15:52:53 +02:00
  • 44f5e2ad77 better n_threads slaren 2023-08-22 21:19:29 +02:00
  • fdd8bf5903 flush stdout every token slaren 2023-08-22 21:18:53 +02:00
  • 56ddd88d05 add llama2 chat example slaren 2023-08-22 19:23:39 +02:00
  • fb000f260a Update Windows CLBlast instructions Mason M 2023-08-30 10:40:52 -03:00
  • c235e8b4e6 Update Windows CLBlast instructions Mason M 2023-08-30 10:06:37 -03:00
  • a2a4eefa07 slight change to logits Concedo 2023-08-30 21:27:51 +08:00
  • e6b7158123
         replace custom data getters and setters by ggml functions xaedes 2023-08-30 15:19:26 +02:00
  • d487e0531f
         move gradient checkpointing code into ggml, new API function: xaedes 2023-08-30 15:21:10 +02:00
  • 804c78dcc9
         metal : minor indentations Georgi Gerganov 2023-08-30 15:48:23 +03:00
  • 2392b6725b
         use tensor->view_src instead of ggml_is_view and get_view_source xaedes 2023-08-30 14:46:12 +02:00
  • b1709f2d25
         Merge branch 'master' into finetune-lora xaedes 2023-08-30 13:28:29 +02:00
  • eec0490d86
         Merge branch 'master' into cleaner-gitignore alonfaraj 2023-08-30 14:10:01 +03:00
  • b532a69b2f
         convert.py : use dir name to name the llama Georgi Gerganov 2023-08-30 13:29:40 +03:00
  • 1301bd7e29 Fix to skip GPU offloading so falcon models work correctly Concedo 2023-08-30 18:26:41 +08:00
  • c90d135eb4
         examples : fix underscore in beam-search + .gitignore (close #2900) b1127 Georgi Gerganov 2023-08-30 12:52:46 +03:00
  • 0d1c706181
         gguf : add workflow for Pypi publishing (#2896) b1126 M. Yusuf Sarıgöz 2023-08-30 12:47:40 +03:00
  • 9509294420
         make : add test and update CI (#2897) b1125 alonfaraj 2023-08-30 12:42:51 +03:00
  • 687397d8be Expanding gitignore with additional entries for new binaries and explanatory comment Jiri Podivin 2023-08-30 11:10:55 +02:00
  • 417da653ac revert .log Alon Faraj 2023-08-30 12:08:31 +03:00
  • c72a87d9b2 remove all bin files without extension and replace with 2 lines Alon Faraj 2023-08-30 12:00:33 +03:00
  • d4c22a8b02 updated lite, added autorope config based on trained ctxlen, hotfix for falcon gpu broken Concedo 2023-08-30 16:50:55 +08:00
  • 35092fb547
         docs : add node-llama-cpp to README.md (#2885) Gilad S 2023-08-30 11:40:12 +03:00
  • 488e03200e Merge branch 'master' into gguf-publish-ci gguf-publish-ci M. Yusuf Sarıgöz 2023-08-30 11:34:55 +03:00
  • dc07dc492e
         convert : various script cleanups/fixes + merges and special token handling (#2842) Kerfuffle 2023-08-30 02:25:50 -06:00
  • 3303f38f34 fix trailing whitespace M. Yusuf Sarıgöz 2023-08-30 11:00:45 +03:00
  • 4d277cb563 gguf : add workflow for Pypi publishing M. Yusuf Sarıgöz 2023-08-30 10:56:41 +03:00
  • b8e572f6d3 gguf : add workflow for Pypi publishing M. Yusuf Sarıgöz 2023-08-30 10:52:06 +03:00
  • ad9ddcff6e
         llm.vim : stop generation at multiple linebreaks, bind to <F2> (#2879) chaihahaha 2023-08-30 14:50:55 +08:00
  • 8341a25957
         main : log file (#2748) b1121 staviq 2023-08-30 08:29:32 +02:00
  • 63df86754d
         Update api_like_OAI.py superchargez 2023-08-30 11:22:53 +05:00
  • 849408957c
         tests : add a C compliance test (#2848) b1120 Cebtenzzre 2023-08-30 02:20:26 -04:00
  • 90afd6dfad Merge branch 'master' of https://github.com/ggerganov/llama.cpp into ap-fim apaz-cli 2023-08-29 23:51:57 -05:00
  • 93753a87c9 Added FIM example. apaz-cli 2023-08-29 23:51:21 -05:00
  • bc84ba3964 add bias akawrykow 2023-08-29 21:29:45 -07:00
  • bca5d0c2de metal: template for mat-vec multiplication kernels lshzh-ww 2023-08-29 23:24:19 -04:00
  • 89495c0716 handle token unbanning over api Concedo 2023-08-30 10:51:49 +08:00
  • f2c02dd06d Merge branch 'master' into concedo_experimental Concedo 2023-08-30 10:51:28 +08:00
  • 52f8cc8f37 llama : fix bpe tokenize from byte opparco 2023-08-30 11:28:19 +09:00
  • d7bdfbdd78
         Update Makefile for misc amd gpu targetting (#407) YellowRoseCx 2023-08-29 20:54:15 -05:00
  • 999f1f879d add extra tensors akawrykow 2023-08-29 18:28:04 -07:00
  • 66a6dbf702 wqkv hack akawrykow 2023-08-29 18:06:35 -07:00
  • 370990abce support 1b model type akawrykow 2023-08-29 17:25:27 -07:00
  • ad45bb37f4 use vocab size from config.json akawrykow 2023-08-29 17:12:33 -07:00
  • 22c3522d78 use 4 space indentation akawrykow 2023-08-29 15:23:28 -07:00
  • 4380e52abc Update tensor map for falcon-rw akawrykow 2023-08-29 15:20:00 -07:00
  • de64f091c8 Skip qkv reshaping for non-parallel attention akawrykow 2023-08-29 15:13:04 -07:00
  • 89dc100653 replace _ prefix with _impl suffix staviq 2023-08-30 00:05:48 +02:00
  • 06abf8eeba
         ggml : add view_src and view_offs to ggml_tensor for views (#2874) b1119 slaren 2023-08-29 23:24:42 +02:00
  • 3b43f94c81
         Update Makefile alonfaraj 2023-08-30 00:20:40 +03:00
  • c03a243abf
         remove outdated references to -eps and -gqa from README (#2881) slaren 2023-08-29 23:17:34 +02:00
  • e276e4b606 Fix convert-falcon-hf-to-gguf.py for rw models akawrykow 2023-08-29 13:54:53 -07:00
  • 6310887983
         Update Makefile alonfaraj 2023-08-30 00:10:15 +03:00
  • bf70e27cd6
         fix check_gradient xaedes 2023-08-29 23:08:30 +02:00
  • fa3582f509
         Tell users attmepting to run perplexity with too few tokens to use more (#2882) b1117 Kawrakow 2023-08-29 23:55:45 +03:00
  • e37e69dcc3
         10X faster BPE tokenizer (#2876) b1116 Kawrakow 2023-08-29 23:55:03 +03:00
  • 5854f51188
         fix error message in ggml_allocr_alloc to display actual max_avail xaedes 2023-08-29 22:49:01 +02:00
  • 725a69ce40 docs: add node-llama-cpp to README.md Gilad S 2023-08-29 23:37:54 +03:00
  • f60f7d3640 review: if/else cleanup (2) staviq 2023-08-29 22:36:47 +02:00
  • 6fa208e157 review: if/else cleanup staviq 2023-08-29 22:13:03 +02:00
  • d77b74b41e Squish last type warning in gguf.py - yay! KerfuffleV2 2023-08-29 14:01:05 -06:00
  • 281245a48f
         Merge branch 'master' into finetune-lora xaedes 2023-08-29 21:47:28 +02:00
  • 8a96d4c2aa
         add missing argument 'int i0' to ggml_get_i32_nd & ggml_set_i32_nd header declarations xaedes 2023-08-29 21:24:37 +02:00
  • dd4e4bca09
         remove unused 'inplace' argument from ggml_compute_backward function xaedes 2023-08-29 21:20:44 +02:00
  • a76e66ac8d
         fix ggml_acc_or_set to return tensor of correct shape xaedes 2023-08-29 21:02:10 +02:00
  • b1aa26f718
         add sanity check to ggml_compute_backward, asserting the correct shape of gradients xaedes 2023-08-29 21:01:17 +02:00
  • 5fcfa7e49e
         increase test-grad0 context mem size to accommodate for bigger cgraph xaedes 2023-08-29 21:00:19 +02:00
  • 82c5247a20
         add ggml API functions ggml_unravel_index, ggml_get_i32_nd and its analogs for set and for f32 xaedes 2023-08-29 20:59:31 +02:00
  • ce005285aa convert.py: Set gpt2 as tokenizer model when using BPE KerfuffleV2 2023-08-29 12:01:59 -06:00
  • 5f0a4e971f
         avoid stack overflow of large cgraphs in test-grad0 xaedes 2023-08-29 19:59:41 +02:00
  • 794bb7ea42
         implement ggml_compute_forward_repeat_f16 xaedes 2023-08-29 19:59:14 +02:00
  • 721771203e
         remove exec Henri Vasserman 2023-08-29 20:56:12 +03:00
  • e28cf7e9ce
         update README.md xaedes 2023-08-29 19:38:23 +02:00
  • a6165dafcd
         remove trailing whitespace xaedes 2023-08-29 19:30:42 +02:00
  • 5813ac832f
         omit tokenization when training is disabled, only save llama lora adapter xaedes 2023-08-29 19:21:09 +02:00
  • 3edee3fa93 fix MSVC, formatting, FMT/VAL placeholders staviq 2023-08-29 19:20:16 +02:00
  • 58fa4dc870 Uniform args parsing and vocab only mode for convert examples KerfuffleV2 2023-08-29 10:47:47 -06:00
  • 3bf4f58589
         [Docker] fix tools.sh argument passing. Henri Vasserman 2023-08-29 19:43:27 +03:00
  • c6b8bdbca6 Tell users attmepting to run perplexity with too few tokens to use more Iwan Kawrakow 2023-08-29 19:35:02 +03:00
  • ebff3a14c3
         remove code to print data checksums which was used to verify correctness of new gguf code xaedes 2023-08-29 18:31:20 +02:00
  • 1425968ead
         remove old checkpoint save & load code xaedes 2023-08-29 18:30:16 +02:00
  • 6134ad4de7
         add python script to convert old finetune checkpoint files to gguf xaedes 2023-08-29 18:24:06 +02:00
  • 0564f4ed1f
         add load & save lora finetune checkpoints via gguf xaedes 2023-08-29 18:20:39 +02:00
  • 4a4051b8dd remove outdated references to -eps and -gqa from README slaren 2023-08-29 18:20:29 +02:00
  • 9fca82bee3 formatting slaren 2023-08-29 18:19:28 +02:00
  • 2ea133895c gguf: SpecialVocab: Actually set load_merges in object KerfuffleV2 2023-08-29 04:33:24 -06:00
  • 0c620ef63b convert-gptneox-hf-to-gguf and convert: Only handle merges for BPE tokenizer KerfuffleV2 2023-08-29 04:23:09 -06:00
  • 61911ca4db convert-falcon-hf-to-gguf: Support --vocab-only option, bail out if no tokenizer.json KerfuffleV2 2023-08-29 04:21:01 -06:00
  • 8534197f14 gguf: SpecialVocab: Fix issue with special token content not in a dict KerfuffleV2 2023-08-29 04:19:18 -06:00
  • 4a3d783d3e Second pass KerfuffleV2 2023-08-28 16:09:01 -06:00
  • f82aec99a4 First pass at implementing suggested changes KerfuffleV2 2023-08-28 08:56:57 -06:00
  • bb6b64d5e5 Use common special vocab handling in various conversion scripts KerfuffleV2 2023-08-27 19:59:01 -06:00
  • 120ed6453f convert: Fix type hint for special_token_types in SpecialVocab KerfuffleV2 2023-08-27 19:36:07 -06:00
  • 531746e953 gguf: Refactor tensor name mapping KerfuffleV2 2023-08-27 17:18:47 -06:00
  • c7b0952eb7 convert: Handle special tokens and merges in vocab only mode KerfuffleV2 2023-08-27 13:17:18 -06:00
  • ea43267819 Initial implementation of handling merges and special tokens KerfuffleV2 2023-08-27 12:31:24 -06:00
  • 795c0c6e9d Minor types cleanups. KerfuffleV2 2023-08-27 11:16:46 -06:00