deps: cap transformers at 4.40.2 (#218)

* deps: pin transformers below v4.41d

Signed-off-by: Anh-Uong <[email protected]>

* remove deprecated requirements.yaml

Signed-off-by: Anh-Uong <[email protected]>

* update unit tests with old evaluation_strategy flag

Signed-off-by: Anh-Uong <[email protected]>

* set transformers upper bound to 4.40.2

- update eval flag in docs

Signed-off-by: Anh-Uong <[email protected]>

---------

Signed-off-by: Anh-Uong <[email protected]>

Author: anhuong

examples/prompt_tuning_twitter_complaints/README.md

@@ -51,7 +51,7 @@ tuning/sft_trainer.py  \
 --per_device_train_batch_size 1  \
 --per_device_eval_batch_size 1  \
 --gradient_accumulation_steps 1  \
---eval_strategy "no"  \
+--evaluation_strategy "no"  \
 --save_strategy "epoch"  \
 --learning_rate 1e-5  \
 --weight_decay 0.  \

pyproject.toml

@@ -28,7 +28,7 @@ classifiers=[
 dependencies = [
 "numpy>=1.26.4,<2.0",
 "accelerate>=0.20.3,<0.40",
-"transformers>=4.34.1,<5.0,!=4.38.2",
+"transformers>=4.34.1,<=4.40.2,!=4.38.2",
 "torch>=2.2.0,<3.0",
 "sentencepiece>=0.1.99,<0.3",
 "tokenizers>=0.13.3,<1.0",

tests/test_sft_trainer.py

@@ -304,7 +304,7 @@ def test_run_causallm_pt_with_validation():
     with tempfile.TemporaryDirectory() as tempdir:
         train_args = copy.deepcopy(TRAIN_ARGS)
         train_args.output_dir = tempdir
-        train_args.eval_strategy = "epoch"
+        train_args.evaluation_strategy = "epoch"
         data_args = copy.deepcopy(DATA_ARGS)
         data_args.validation_data_path = TWITTER_COMPLAINTS_DATA
 
@@ -317,7 +317,7 @@ def test_run_causallm_pt_with_validation_data_formatting():
     with tempfile.TemporaryDirectory() as tempdir:
         train_args = copy.deepcopy(TRAIN_ARGS)
         train_args.output_dir = tempdir
-        train_args.eval_strategy = "epoch"
+        train_args.evaluation_strategy = "epoch"
         data_args = copy.deepcopy(DATA_ARGS)
         data_args.validation_data_path = TWITTER_COMPLAINTS_DATA
         data_args.dataset_text_field = None