gradient accumulation tests, embeddings w pad_token fix, smaller models (#2059)

* add more test cases for gradient accumulation and fix zero3

* swap out for smaller model

* fix missing return

* fix missing pad_token in config

* support concurrency for multigpu testing

* cast empty deepspeed to empty string for zero3 check

* fix temp_dir as fixture so parametrize works properly

* fix test file for multigpu evals

* don't use default

* don't use default for fsdp_state_dict_type

* don't use llama tokenizer w smollm

* also automatically cancel multigpu for concurrency

tests/e2e/conftest.py

@@ -0,0 +1,16 @@
+"""
+shared pytest fixtures
+"""
+import shutil
+import tempfile
+
+import pytest
+
+
+@pytest.fixture
+def temp_dir():
+    # Create a temporary directory
+    _temp_dir = tempfile.mkdtemp()
+    yield _temp_dir
+    # Clean up the directory after the test
+    shutil.rmtree(_temp_dir)