Wing Lian 493910559a Configurable embeddings upcast (#2621) ...

* fsdp embeddings should be float32 per comment

* patch peft to not upcast everything

* add tabs back to code check

* fix import

* add configurable option and fix check

* add check for dtypes

* move embeddings test to patch dir

* fix test

* fix comment and logic

2025-05-07 16:10:16 -04:00

..

lora_kernels

auto-enable lora kernels where possible (#2589)

2025-05-07 16:10:11 -04:00

__init__.py

attempt to also run e2e tests that needs gpus (#1070)

2024-01-09 21:23:23 -05:00

test_4d_multipack_llama.py

replace references to random 68m model w 135m smollm2 (#2570) [skip ci]

2025-04-28 10:08:07 -04:00

test_activation_checkpointing.py

add e2e smoke test for using activation/gradient checkpointing with offload (#2565)

2025-04-25 21:11:17 -04:00

test_cli_integrations.py

adding pre-commit auto-update GH action and bumping plugin versions (#2428)

2025-03-21 11:02:43 -04:00

test_fa_xentropy.py

feat: add config for optional parameters in a chat message (#2260)

2025-02-18 09:59:27 +07:00

test_falcon_samplepack.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00

test_fused_llama.py

replace references to random 68m model w 135m smollm2 (#2570) [skip ci]

2025-04-28 10:08:07 -04:00

test_llama_s2_attention.py

replace references to random 68m model w 135m smollm2 (#2570) [skip ci]

2025-04-28 10:08:07 -04:00

test_lora_llama_multipack.py

replace references to random 68m model w 135m smollm2 (#2570) [skip ci]

2025-04-28 10:08:07 -04:00

test_mistral_samplepack.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00

test_mixtral_samplepack.py

Sequence parallel training context manager (#2553)

2025-04-25 10:33:54 -04:00

test_model_patches.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00

test_peft_embeddings.py

Configurable embeddings upcast (#2621)

2025-05-07 16:10:16 -04:00

test_phi_multipack.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00

test_resume.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00

test_sp.py

don't use is_main_process during config validation (#2569)

2025-04-26 14:14:52 -04:00

test_unsloth_integration.py

adding pre-commit auto-update GH action and bumping plugin versions (#2428)

2025-03-21 11:02:43 -04:00

test_unsloth_qlora.py

make sure to validate the config before normalizing so defaults get set (#2554)

2025-04-24 13:01:43 -04:00