fsdp2 w evals fixed upstream

.github/CONTRIBUTING.md

@@ -57,13 +57,6 @@ We welcome ideas for improvements and new features. To suggest an enhancement, o
 5. Push your branch to your fork on GitHub.
 6. Open a new pull request against the `main` branch of the axolotl repository. Include a clear and concise description of your changes, referencing any related issues.
 
-#### Skipping CI Checks
-
-You can skip certain CI checks by including specific keywords in your commit messages:
-
-- `[skip ci]` or `skip ci` - Skips all CI checks for that commit
-- `[skip-e2e]` or `skip-e2e` - Skips only end-to-end tests while running other CI checks
-
 ## Style Guidelines
 
 ### Code Style

TODO.md

@@ -0,0 +1,10 @@
+# todo list
+
+- [] Validation of parameters for combinations that won't work
+
+
+
+## things that are known not to work
+
+- FSDP offload and gradient_checkpointing - https://github.com/pytorch/pytorch/issues/82203
+- adamw_bnb_8bit doesn't play well with FSDP offload

requirements.txt

@@ -14,7 +14,7 @@ packaging==23.2
 
 huggingface_hub>=0.33.0
 peft==0.17.0
-transformers==4.55.0
+transformers @ git+https://github.com/vasqu/transformers@fix-fa-integration
 tokenizers>=0.21.1
 accelerate==1.10.0
 datasets==4.0.0

src/axolotl/cli/utils/train.py

@@ -67,12 +67,14 @@ def build_command(base_cmd: list[str], options: dict[str, Any]) -> list[str]:
 
 def generate_config_files(config: str, sweep: str | None) -> Iterator[tuple[str, bool]]:
     """
-    Generate list of configuration files to process. Yields a tuple of the configuration file name and a boolean indicating
-    whether this is a group of configurations (i.e., a sweep).
+    Generate list of configuration files to process.
 
     Args:
         config: Base configuration file
         sweep: Sweep configuration file
+
+    Yields:
+        Tuple of configuration file name and whether this is a group of configurations
     """
 
     if not sweep:

src/axolotl/monkeypatch/ring_attn/patch.py

@@ -18,9 +18,7 @@ from torch.distributed import DeviceMesh
 try:
     from transformers.modeling_flash_attention_utils import _flash_supports_window
 except ImportError:
-    from transformers.modeling_flash_attention_utils import (
-        _flash_supports_window_size as _flash_supports_window,
-    )
+    _flash_supports_window = True
 
 from axolotl.monkeypatch.utils import get_cu_seqlens_from_pos_ids
 from axolotl.utils.logging import get_logger

tests/monkeypatch/test_trainer_loss_calc.py

@@ -3,7 +3,6 @@
 import unittest
 
 from axolotl.monkeypatch.transformers.trainer_loss_calc import (
-    check_evaluation_loop_is_fsdp2_patchable,
     check_evaluation_loop_is_patchable,
     check_maybe_log_save_evaluate_is_patchable,
 )
@@ -20,7 +19,6 @@ class TestTrainerLossCalc(unittest.TestCase):
         the patched code changes upstream.
         """
         assert check_evaluation_loop_is_patchable()
-        assert check_evaluation_loop_is_fsdp2_patchable()
         assert check_maybe_log_save_evaluate_is_patchable()