Commit History

mypy wandb ignore (#572)
c6d870b
unverified

winglian commited on

remove columns after tokenizing for pretraining (#571)
1157950
unverified

winglian commited on

set auto for other params that hf trainer sets for ds. include zero1 json (#570)
3b18c96
unverified

winglian commited on

fix save_steps so it doesn't get duplicated (#567)
3fbde76
unverified

winglian commited on

Model parallel (#538)
f6060a6
unverified

winglian commited on

let hf trainer handle torch compile (#516)
a4e1bb6
unverified

winglian tmm1 commited on

improve how we setup eval/save strategies and steps (#547)
36e53c7
unverified

winglian commited on

gracefully handle length feature used for group by (#565)
e7aa7b1
unverified

winglian commited on

add optimization for group-by-len (#563)
e5bb22a
unverified

winglian commited on

check for the existence of the default accelerate config that can create headaches (#561)
fdb777b
unverified

winglian commited on

fix wandb so mypy doesn't complain (#562)
bf08044
unverified

winglian commited on

Add training callback to send predictions to WandB table (#521)
5b67ea9
unverified

Glavin001 commited on

Fix pretraining with iterable/streaming Dataset (#556)
2f586d1
unverified

Jan Philipp Harries Jan Philipp Harries commited on

document that packaging needs to be installed before flash-attn (#559)
9845c5e
unverified

winglian commited on

fix the sed command to replace the version w the tag
772cd87

winglian commited on

add long_description for pypi push (#555)
6c5fbe6
unverified

winglian commited on

replace tags, build dist for pypi publish (#553)
bcbc959
unverified

winglian commited on

ergonomic update to optimizer config doc (#548)
6d57f2f
unverified

The Objective Dad commited on

pypi on tag push (#552)
20ed4c1
unverified

winglian commited on

remove with section, doesn't seem to work (#551)
c5dedb1
unverified

winglian commited on

publish to pypi workflow on tagged release (#549)
b56503d
unverified

winglian commited on

fix for quant config from model (#540)
a94f9cb
unverified

winglian commited on

Update requirements.txt (#543)
c1921c9
unverified

dongxiaolong commited on

workaround for md5 variations (#533)
0b4cf5b
unverified

winglian commited on

add git environment variables to compose: avoid checkout failure error 128 on build (#534)
78ee2cd
unverified

SlapDrone commited on

update readme to point to direct link to runpod template, cleanup install instrucitons (#532)
34c0a86
unverified

winglian commited on

Adding NCCL Timeout Guide (#536)
5e2d8a4
unverified

The Objective Dad commited on

Early stopping metric (#537)
e30f1e3
unverified

winglian commited on

recommend padding when using sample packing (#531)
3437149
unverified

winglian commited on

log rank too (#527)
245c5c4
unverified

winglian commited on

misc fixes/improvements (#513)
a546ca2
unverified

winglian commited on

Add support for GPTQ using native transformers/peft (#468)
3355706
unverified

winglian commited on

Merge pull request #520 from bdashore3/sharegpt-fixes
daa4fac
unverified

mhenrichsen commited on

reorg a bit
fc8766e

tmm1 commited on

use flash_attn rmsnorm when available (#526)
72a6fe1
unverified

tmm1 commited on

use flash_attn xentropy when available (#525)
5fe30b1
unverified

tmm1 commited on

move is_llama_derived_model into normalize_config (#524)
44454ae
unverified

tmm1 commited on

No gather single gpu (#523)
09f1543
unverified

winglian commited on

Prompters: ShareGPT: Allow for custom system prompts
995557b

kingbri commited on

fix: bad dtype for full finetune (#504)
1991946
unverified

Maxime winglian commited on

Fix(doc): Inform Windows users to use WSL/docker (#518)
f51c9c5
unverified

Nanobit commited on

log supervised token count (#448)
7710e81
unverified

winglian commited on

Debug tokenization output: Add ability to output text only (no tokens), and/or specify num samples to see (#511)
48434be
unverified

Tom Jobbins commited on

Added advanced DDP args (#515)
396a7a7
unverified

Jan Philipp Harries Jan Philipp Harries commited on

split train from other cli options (#503)
b21e4a2
unverified

winglian commited on

Changed Bench Eval to report metrics correctly by split. Added total accuracy and renamed previously used bench_accuracy to bench_average_accuracy. (#512)
42f9642
unverified

Alpay Ariyak commited on

drop empty tokenized rows too (#509)
c56b450
unverified

winglian commited on

set zero3 optimizer betas to auto so they inherit from HF trainer config (#507)
1e07c16
unverified

tmm1 commited on

add eval benchmark callback (#441)
7657632
unverified

winglian commited on

customizable ascii art (#506)
548787d
unverified

winglian commited on