|
|
8a2ef81d17
|
add copyright header
|
2024-07-20 22:57:16 +02:00 |
|
|
|
c38ac65d5b
|
various optimizations
|
2024-07-20 21:47:18 +02:00 |
|
|
|
0b39ba0843
|
Fix group_texts not grouping texts to a single length when the number of samples is less than the number of threads used
|
2024-05-08 22:37:42 +02:00 |
|
Carl Philipp Klemm
|
65482b55a6
|
actually print eval response
|
2024-05-07 15:27:59 +02:00 |
|
Carl Philipp Klemm
|
2ee4f294af
|
fix evaluate() invocation
|
2024-05-07 15:10:02 +02:00 |
|
Carl Philipp Klemm
|
68f748e99e
|
Fix mypy warnings
|
2024-05-07 14:53:28 +02:00 |
|
|
|
a74ef976e4
|
add support for huggingfacehub datasets and for specificying a prompt for eval
|
2024-05-07 00:23:12 +02:00 |
|
|
|
c33964371c
|
Inactive parameter quanitzation support
|
2024-04-07 19:15:42 +02:00 |
|
|
|
3fa1fc254f
|
add qunatized linear, refactor model for it soon to be addition
|
2024-03-23 21:38:27 +01:00 |
|
|
|
38a7f7cfc4
|
add gpu memory rebalanceing
|
2024-03-17 22:54:33 +01:00 |
|
|
|
5acb6809ed
|
wip refactor
|
2024-03-13 19:45:52 +01:00 |
|
|
|
11ea9eeaa7
|
working full training
|
2024-03-09 10:03:37 +01:00 |
|
|
|
7a47fcdcc0
|
Inital commit
|
2024-03-06 17:50:40 +01:00 |
|
