Commit graph

190 commits

Author SHA1 Message Date
Concedo
ad638285de Merge branch 'master' into concedo_experimental
# Conflicts:
#	Makefile
#	README.md
#	flake.lock
#	ggml-cuda.cu
#	llama.cpp
#	tests/test-backend-ops.cpp
#	tests/test-quantize-fns.cpp
2024-02-28 13:41:35 +08:00
Concedo
12b4c14847 updated lite (+1 squashed commits)
Squashed commits:

[0aafb3ad] updated lite
2024-02-27 23:34:41 +08:00
Concedo
1bcbd2e21b updated lite 2024-02-24 17:59:44 +08:00
Concedo
6181b46eef added nocertify mode 2024-02-19 16:05:17 +08:00
Concedo
a0440fd746 updated lite 2024-02-17 11:15:14 +08:00
Concedo
2338e95c6d updated lite 2024-02-16 23:08:58 +08:00
Concedo
1ae0a524ad added ai vision 2024-02-15 10:51:41 +08:00
Concedo
fdc54efcf0 updated lite 2024-02-12 17:35:52 +08:00
Concedo
038779af41 another fix for lite 2024-02-11 15:17:09 +08:00
Concedo
afa41c24f5 small fix lite (+1 squashed commits)
Squashed commits:

[f22db79a] updated lite
2024-02-10 22:31:19 +08:00
Concedo
0ec0055edc updated lite 2024-02-09 22:21:58 +08:00
Concedo
992eea71d7 fixes for vulkan multigpu 2024-02-09 14:42:27 +08:00
Concedo
de7be2f4e0 benchmarker done 2024-02-07 22:04:53 +08:00
Concedo
ea4b131310 updated lite 2024-02-04 16:53:16 +08:00
Concedo
504300784f updated lite 2024-02-03 21:11:06 +08:00
Concedo
71cc19e76d updated lite 2024-01-31 15:47:21 +08:00
Concedo
51fe7ac215 xtts fixes 2024-01-31 14:43:51 +08:00
Concedo
e989675b0d update lite 2024-01-30 19:12:30 +08:00
Concedo
f73de33f74 Merge branch 'master' into concedo_experimental
# Conflicts:
#	.github/workflows/build.yml
#	.github/workflows/docker.yml
#	CMakeLists.txt
#	Makefile
#	README.md
#	ci/README.md
#	ci/run.sh
#	flake.lock
#	ggml-metal.m
#	ggml-opencl.cpp
#	ggml-vulkan-shaders.hpp
#	ggml-vulkan.cpp
#	ggml-vulkan.h
#	ggml.c
#	ggml_vk_generate_shaders.py
#	llama.cpp
#	llama.h
#	pocs/vdot/vdot.cpp
#	tests/test-llama-grammar.cpp
#	tests/test-sampling.cpp
2024-01-29 23:12:09 +08:00
Concedo
af6f495c52 fix tts (+1 squashed commits)
Squashed commits:

[ae02f266] fix tts
2024-01-25 00:03:59 +08:00
Concedo
77fa26b2b1 added termux instructions 2024-01-24 20:12:46 +08:00
Concedo
346c1a97de fixed file select cancel, updated lite 2024-01-24 16:36:53 +08:00
Concedo
08236ccc97 better abort handling, added support for dynatemp exponent 2024-01-23 16:56:12 +08:00
Concedo
dc7bc0cb50 Merge commit '584d674be6' into concedo_experimental
# Conflicts:
#	.github/workflows/nix-flake-update.yml
#	Makefile
#	Package.swift
#	ggml-cuda.cu
#	tests/test-quantize-fns.cpp
2024-01-14 16:29:44 +08:00
Concedo
b9ad08af19 improved dynatemp wizard 2024-01-11 11:26:14 +08:00
Concedo
c6879f3fca Merge branch 'master' into concedo_experimental
# Conflicts:
#	README.md
2024-01-10 16:05:14 +08:00
Concedo
df20382206 sync slider 2024-01-09 16:50:23 +08:00
Concedo
5cc64ebb52 dynatemp wizard 2024-01-09 15:51:32 +08:00
kalomaze
123bff9a0f
Full DynaTemp implementation + UI (#600)
* move Dynatemp changes to new branch

* fix float header

* Properly reintroduce variable expert count

Controllable through experts.txt

* first pass at DynaTemp UI

Checkbox partial implemented, Min and Max Temp implemented

* DynaTemp UI Checkbox

Trigger DynaTemp on checkbox

* DynaTemp UI checkbox edition

Hell Yeah! DynaTemp!

* Remove greedy dynatemp

* Fix race condition caused by debug print

* Fixed broken presets and miro

Fixes broken presets and mirostat

* Remove debug function + HHI temp

Also removed unnecessary softmax double precision

* Fix whitespace (?) for generate function

* epic upstream renaming scheme fix

* fix stupid indents

* Other cleanup

Reintroduce unused rep pen function, move temp functions first before entropy dynamic temp

* Slight indent fix

* revert batch pyinstaller maker to mainline

and also delete experts.txt since adjustable routing is also being removed for the PR

* compact dynatemp into a single value dynatemp_range. This is a float which represents the allowed deviation from the min and max temperature when using dynatemp. Thus, if we want a value of dynatemp_min=0.3, dynatemp_max=0.5, then we would simply set temperature=0.4 and dynatemp_range=0.1. Functionally dynatemp would operate the same, but it would simplify usage and make it a single easy to adjust value.

---------

Co-authored-by: Alexander Abushady <aabushady214@gmail.com>
Co-authored-by: Concedo <39025047+LostRuins@users.noreply.github.com>
2024-01-06 11:13:16 +08:00
Concedo
24c3f3283a fixed numerical parsing for steps 2023-12-31 00:17:17 +08:00
Concedo
7ce32a151e updated lite, added background images and image gen support for custom step counts and cfg scales (+1 squashed commits)
Squashed commits:

[d9e9c591] updated lite, added background images and image gen support for custom step counts and cfg scales
2023-12-31 00:08:15 +08:00
Concedo
2d5d82e915 addlocate gpt_params on heap instead to avoid rare segfault 2023-12-28 11:48:21 +08:00
Concedo
5b2d93a1f8 updated lite and colab, added logit bias support to lite 2023-12-27 21:32:18 +08:00
Concedo
293395e0f5 Merge commit '708e179e85' into concedo_experimental
# Conflicts:
#	.github/workflows/docker.yml
2023-12-25 16:48:15 +08:00
Concedo
71a5afaab5 fixed incorrect localflag 2023-12-23 11:00:58 +08:00
Concedo
8823e8b06d added presence penalty into lite ui 2023-12-23 10:39:40 +08:00
Concedo
ec05230703 updated lite, up ver 2023-12-17 14:38:39 +08:00
Concedo
e447af669c Merge branch 'master' into concedo_experimental
# Conflicts:
#	Makefile
#	README.md
2023-12-13 21:09:47 +08:00
Concedo
930cdfb1ce updated lite, added patch that links to noscript mode 2023-12-08 16:53:30 +08:00
Concedo
0ca814e544 added minP preset 2023-12-03 11:18:03 +08:00
Concedo
a195cdeec8 fixed chub ai imports (+1 squashed commits)
Squashed commits:

[cdb74264] fixed chub ai imports
2023-11-30 18:07:56 +08:00
Concedo
ec1796bec1 updated lite 2023-11-27 14:04:53 +08:00
Concedo
eb42c73953 revert auto rope scaling for already-ropetuned models - just use their values 2023-11-24 14:20:36 +08:00
Concedo
dc4078c039 fixed segfault with all non-gguf models 2023-11-20 22:31:56 +08:00
Concedo
35ce2b054f typo fixes 2023-11-18 11:05:04 +08:00
Concedo
914e375602 support custom dalle urls 2023-11-15 18:37:50 +08:00
Concedo
f4ee91abbb improved estimation 2023-11-13 15:45:13 +08:00
Concedo
a00a32e049 fixed localflag 2023-11-11 10:21:45 +08:00
Concedo
e08e1bdc68 include opencl dll 2023-11-11 10:05:42 +08:00
Concedo
4b9a685f10 rename file comments to welcome 2023-11-11 01:15:35 +08:00