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

Commit graph

  • 9d4365df40
         nix: add cuda, use a symlinked toolkit instead for cmake Green Sky 2023-09-18 15:48:24 +02:00
  • 51a7cf5c6e
         examples : fix RoPE defaults to match PR #3240 (#3315) b1269 Cebtenzzre 2023-09-23 05:28:50 -04:00
  • 60098a176b update colab model Concedo 2023-09-23 16:30:40 +08:00
  • 49caf47003 examples : fix RoPE defaults to match PR #3240 Cebtenzzre 2023-09-23 00:57:02 -04:00
  • a2c43f8aa2
         Update build.yml Eve 2023-09-23 04:36:16 +00:00
  • bfc696fcc4 update lite, update ver Concedo 2023-09-23 12:35:23 +08:00
  • 45b4697376
         Update build.yml Eve 2023-09-23 04:27:31 +00:00
  • 4913506896
         try to fix freebsd Eve 2023-09-23 04:19:02 +00:00
  • bedb92b603
         scripts : use /usr/bin/env in shebang (#3313) Kevin Ji 2023-09-22 23:52:23 -04:00
  • 05c5adaadc
         windows syntax Eve 2023-09-23 03:45:44 +00:00
  • ee5ddbad4b
         scripts : use /usr/bin/env in shebang Kevin Ji 2023-09-22 20:38:47 -07:00
  • 1e2fec1249
         automatically get thread count Eve 2023-09-23 03:37:52 +00:00
  • bd2500db36 Merge branch 'master' into concedo_experimental Concedo 2023-09-23 10:51:34 +08:00
  • a64d182b8b sched yield fix again Concedo 2023-09-23 10:44:41 +08:00
  • 1f9e36c733 minor lite fixes Concedo 2023-09-23 09:37:49 +08:00
  • 3469684177 add documentation about --threads-batch to the READMEs slaren 2023-09-23 02:43:02 +02:00
  • a6084cc719 add n_threads_batch to system info, refactor to get_system_info() slaren 2023-09-23 02:29:39 +02:00
  • 25f4e523f1
         Update build.yml Eve 2023-09-23 00:17:18 +00:00
  • 55eb46b966
         Update build.yml Eve 2023-09-22 23:08:24 +00:00
  • ad1e2a7d6d
         Update build.yml Eve 2023-09-22 23:02:44 +00:00
  • b38e1dc03c
         windows Eve 2023-09-22 22:58:11 +00:00
  • 537ff64a61
         mac and linux threads Eve 2023-09-22 22:19:31 +00:00
  • da05205af6
         cancel optimization when specified number of epochs is completed xaedes 2023-09-22 21:00:46 +02:00
  • 9145c87acc
         update finetune README.md xaedes 2023-09-22 20:54:00 +02:00
  • 758c46cf7d
         add export-lora build dependency to llama xaedes 2023-09-22 20:20:13 +02:00
  • 904c19bebd
         remove unused struct tensor_info from export-lora xaedes 2023-09-22 20:19:12 +02:00
  • d38260be0f
         add export-lora build in Makefile xaedes 2023-09-22 20:04:07 +02:00
  • b91e3dd2ee
         remove trailing whitespace xaedes 2023-09-22 20:01:53 +02:00
  • 0ede0f4434
         add export-lora program xaedes 2023-09-22 19:56:13 +02:00
  • de4e27904d clear reader copy on new gen Concedo 2023-09-23 00:13:19 +08:00
  • 92fb8ab5a7 remove low-vram option slaren 2023-09-22 17:34:02 +02:00
  • ebf6d5f126
         Update ggml.h Hua Jiang 2023-09-21 18:41:37 -07:00
  • ceb18e44db remove kv_size(), cuda scratch fixes slaren 2023-09-22 03:18:53 +02:00
  • 82bff97a7c
         Create dependabot.yml Alon 2023-09-22 02:44:59 +03:00
  • d41b53ca7b fix mpi build slaren 2023-09-22 00:58:45 +02:00
  • 7f95379295 move n_threads to llama_context_params, add n_threads_batch slaren 2023-09-22 00:00:55 +02:00
  • a7d13ac15b Merge branch 'master' of https://github.com/ggerganov/llama.cpp into clang-warnings Cebtenzzre 2023-09-21 17:47:43 -04:00
  • 4d5fe73449 better option descriptions in help Cebtenzzre 2023-09-21 15:41:06 -04:00
  • 96f6dcdeae llama-bench : keep the same model between tests when possible slaren 2023-09-21 21:41:05 +02:00
  • 19bb74e74c restore backwards compatiblity with *.rope.scale_linear Cebtenzzre 2023-09-21 15:10:39 -04:00
  • bc9d3e3971
         Update README.md (#3289) Lee Drake 2023-09-21 13:00:24 -06:00
  • e0b120c3ca address review comments Cebtenzzre 2023-09-21 15:00:08 -04:00
  • aff20ab0c5
         Update README.md Lee Drake 2023-09-21 12:51:46 -06:00
  • 2534c587a9
         remove semicolon Cebtenzzre 2023-09-21 14:48:06 -04:00
  • f28e4953a8 fix freq_base/scale default to model value slaren 2023-09-21 20:20:42 +02:00
  • 39f4afac20 fix metal build slaren 2023-09-21 20:15:22 +02:00
  • 36b904e200
         ggml-opencl.cpp: Make private functions static (#3300) b1266 shibe2 2023-09-21 22:10:26 +04:00
  • 8845160058
         simple : add README.md Georgi Gerganov 2023-09-21 20:10:14 +02:00
  • 5a3369d8e8
         llama : llama.h formatting + comments Georgi Gerganov 2023-09-21 19:51:32 +02:00
  • cf1f80596d llama.cpp : split llama_context_params into model and context params slaren 2023-09-21 18:13:06 +02:00
  • 81381a0ce7 ggml-opencl.cpp: Make private functions static shibe2 2023-09-21 20:40:56 +04:00
  • c746914615 Try to fix Baichuan2 models by using vocab size in config.json KerfuffleV2 2023-09-21 09:51:26 -06:00
  • 14295922f9 updated ver, updated lite (+1 squashed commits) Concedo 2023-09-21 17:40:35 +08:00
  • 324f3403d5
         zig : fix for updated c lib (#3259) Edward Taylor 2023-09-21 21:08:20 +12:00
  • f56c418ab0
         embedding : update README.md (#3224) yuiseki 2023-09-21 17:57:40 +09:00
  • 8185710a80
         CUDA: use only 1 thread if fully offloaded (#2915) b1263 Johannes Gäßler 2023-09-21 10:43:53 +02:00
  • 9d3514a2a6
         vvhg-code-infill (#1) vvhg1 2023-09-21 10:36:17 +02:00
  • 55c283c5aa Release the requested thread pool resource 2 Zongfu ZF3 Qu 2023-09-21 14:37:24 +08:00
  • fe788c45c8 don't hardcode max_pos_emb Cebtenzzre 2023-09-21 00:01:48 -04:00
  • 048ef5e4bf Release the requested thread pool resource Zongfu ZF3 Qu 2023-09-21 11:40:47 +08:00
  • 43eaf06a2f llama : fix C compatibility Cebtenzzre 2023-09-20 23:29:08 -04:00
  • 56abb9a406 fix convert scripts Cebtenzzre 2023-09-20 22:25:21 -04:00
  • 904d4edfa1 llama : store YaRN parameters in GGUF Cebtenzzre 2023-09-14 13:26:10 -04:00
  • dc26a0dd32 llama : simplify use of context params Cebtenzzre 2023-09-20 21:33:33 -04:00
  • a06c72924c Merge branch 'master' of https://github.com/ggerganov/llama.cpp into ntkv2 Cebtenzzre 2023-09-20 21:04:21 -04:00
  • 281b26e647 convert : reduce unnecessary variables in Params Cebtenzzre 2023-09-06 13:00:04 -04:00
  • dcb058ce5d YaRN : fix missing parameter in CUDA impl Cebtenzzre 2023-09-05 14:17:50 -04:00
  • cf731d5648 YaRN : avoid NaN if unused betas are zero Cebtenzzre 2023-09-05 14:14:05 -04:00
  • 826269adc5 ggml : increase GGML_MAX_OP_PARAMS Cebtenzzre 2023-09-04 22:15:59 -04:00
  • b5ced4fb63 Merge branch 'master' of https://github.com/ggerganov/llama.cpp into ntkv2-fix Cebtenzzre 2023-09-04 21:42:34 -04:00
  • 46bc744722 New model conversions richardr1126 2023-09-20 19:16:21 -06:00
  • 7cdc3eaa76 Produces garbage output Phillip Kravtsov 2023-09-20 17:11:33 -07:00
  • 1310dbed55
         my eyes must be getting bad... Eve 2023-09-21 00:15:43 +00:00
  • 169c622850
         alternate implementation Eve 2023-09-21 00:02:18 +00:00
  • 58bb7d5631 Makefile: move unused logits flags where they don't interfere w/ targets ochafik 2023-09-21 00:52:24 +01:00
  • 67a24930e9 Update README.md Lee Drake 2023-09-20 17:47:50 -06:00
  • f43c97d4e9
         Merge branch 'ggerganov:master' into master Richard Roberson 2023-09-20 17:22:58 -06:00
  • 55d49b26bd Merge remote-tracking branch 'origin/master' into skip-unused-2 ochafik 2023-09-21 00:16:13 +01:00
  • 191f11c010 ggml_tensor: update the structure comments. Hua Jiang 2023-09-20 12:51:50 -07:00
  • 7eb41179ed
         readme : update hot topics Georgi Gerganov 2023-09-20 20:48:22 +03:00
  • b2debf65f2
         parallel : add disabled experimental batch chunking in powers of two Georgi Gerganov 2023-09-20 20:14:05 +03:00
  • a5661d7e71
         llama : allow gguf RoPE keys to be overridden with defaults (#3240) b1261 Cebtenzzre 2023-09-20 12:12:47 -04:00
  • d38d59cc91 Merge branch 'master' of https://github.com/ggerganov/llama.cpp into clang-warnings Cebtenzzre 2023-09-20 12:11:05 -04:00
  • a6b74764c7 compiler version detection Cebtenzzre 2023-09-19 13:49:50 -04:00
  • ded9b43cad
         parallel : fix cases where the input prompts can overflow the batch Georgi Gerganov 2023-09-20 19:09:25 +03:00
  • 65c2c1c5ab
         benchmark-matmult : do not use integer abs() on a float (#3277) b1260 Cebtenzzre 2023-09-20 12:06:08 -04:00
  • 2dda63a4eb add tensor split field Concedo 2023-09-20 22:46:47 +08:00
  • c9d5b549c7 benchmark-matmult : fix missing 'static' specifiers Cebtenzzre 2023-09-20 10:38:12 -04:00
  • b530a55b1d benchmark-matmult : do not use integer abs() on a float Cebtenzzre 2023-09-20 10:37:32 -04:00
  • ee1d670cc6 parallel : fix bug (extra BOS) + smaller token_prev array Georgi Gerganov 2023-09-20 17:32:21 +03:00
  • 80834daecf
         flake : Restore default package's buildInputs (#3262) kang 2023-09-20 22:48:22 +09:00
  • 712b8423f6 class.py changes Concedo 2023-09-20 21:27:49 +08:00
  • 1be2b8c19b
         ggml : revert change to ggml_cpy, add ggml_cont_Nd instead (#3275) slaren 2023-09-20 15:12:51 +02:00
  • b63cf223c9 add queue info Concedo 2023-09-20 21:07:21 +08:00
  • 0eb52cf6c2 Merge branch 'master' into concedo_experimental Concedo 2023-09-20 21:01:34 +08:00
  • 006e87cb56 requirements txt Concedo 2023-09-20 21:00:23 +08:00
  • 0fd462fdd6 ggml : revert change to ggml_cpy, add ggml_cont_Nd instead slaren 2023-09-20 14:46:21 +02:00
  • a40f2b656f
         CI: FreeBSD fix (#3258) b1258 Alon 2023-09-20 15:06:36 +03:00
  • 2f3a46fccf
         train : make KQ_pos memory buffer permanent via dummy scale op Georgi Gerganov 2023-09-20 14:14:50 +03:00
  • 54206962c7
         llama : disable MPI for now Georgi Gerganov 2023-09-20 14:06:41 +03:00