Wing Lian
a159724e44
bump trl and accelerate for latest releases ( #1730 )
...
* bump trl and accelerate for latest releases
* ensure that the CI runs on new gh org
* drop kto_pair support since removed upstream
2024-07-10 11:15:44 -04:00
Charles Frye
8a20a7b711
document how to use share_strategy="no" ( #1653 ) [skip ci]
...
The literal value `no` is parsed in some YAML parsers to the boolean `False`, which fails Pydantic validation. To be sure that the value is parsed to the string `"no"`, the value should be enclosed in quotes. [Discussion on StackOverflow](https://stackoverflow.com/questions/53648244/specifying-the-string-value-yes-in-yaml ).
2024-05-24 14:15:44 -04:00
Wing Lian
367b2e879b
Switch to parallel FFD bin packing algorithm. ( #1619 )
...
* Switch to parallel FFD bin packing algorithm.
Add support for packing in a distributed context.
Add packing efficiency estimate back.
* revert changes to distributed code
* chore: lint
* fix config w new params for packing test
* add sample_packing_group_size and sample_packing_bin_size to cfg schema
* fix lamdbda function
* fix sampler/dataloader calculations for packing
---------
Co-authored-by: dsesclei <dave@sescleifer.com >
2024-05-23 17:32:14 -04:00
tpoisonooo
1ac899800b
docs(config.qmd): add loraplus example ( #1577 )
...
* Update qwen2-moe-lora.yaml
* feat(project): update
2024-05-06 14:05:28 +09:00
NanoCode012
1aeece6e24
chore(doc): clarify micro_batch_size ( #1579 ) [skip ci]
2024-05-01 00:33:53 +09:00
NanoCode012
d28ba2e405
feat(doc): Add example for pad_token ( #1535 )
2024-04-19 02:20:20 +09:00
Hamel Husain
86b7d22f35
Reorganize Docs ( #1468 )
2024-04-01 08:00:52 -07:00
Hamel Husain
629450cecd
Bootstrap Hosted Axolotl Docs w/Quarto ( #1429 )
...
* precommit
* mv styes.css
* fix links
2024-03-21 22:28:36 -07:00
