transformers v5 upgrade (#3272)

* Prepare for transformers v5 upgrade

* fix hf cli

* update for hf hub changes

* fix tokenizer apply_chat_template args

* remap include_tokens_per_second

* fix tps

* handle migration for warmup

* use latest hf hub

* Fix scan -> ls

* fix import

* fix for renaming of mistral common tokenizer -> backend

* update for fixed tokenziation for llama

* Skip phi35 tests for now

* remove mistral patch fixed upstream in huggingface/transformers#41439

* use namespacing for patch

* don't rely on sdist for e2e tests for now

* run modal ci without waiting too

* Fix dep for ci

* fix imports

* Fix fp8 check

* fsdp2 fixes

* fix version handling

* update fsdp version tests for new v5 behavior

* Fail multigpu tests after 3 failures

* skip known v5 broken tests for now and cleanup

* bump deps

* unmark skipped test

* re-enable test_fsdp_qlora_prequant_packed test

* increase multigpu ci timeout

* skip broken gemma3 test

* reduce timout back to original 120min now that the hanging test is skipped

* fix for un-necessary collator for pretraining with bsz=1

* fix: safe_serialization deprecated in transformers v5 rc01 (#3318)

* torch_dtype deprecated

* load model in float32 for consistency with tests

* revert some test fixtures back

* use hf cache ls instead of scan

* don't strip fsdp_version

more fdsp_Version fixes for v5
fix version in fsdp_config
fix aliasing
fix fsdp_version check
check fsdp_version is 2 in both places

* Transformers v5 rc2 (#3347)

* bump dep

* use latest fbgemm, grab model config as part of fixture, un-skip test

* import AutoConfig

* don't need more problematic autoconfig when specifying config.json manually

* add fixtures for argilla ultrafeedback datasets

* download phi4-reasoning

* fix arg

* update tests for phi fast tokenizer changes

* use explicit model types for gemma3

---------

Co-authored-by: Wing Lian <wing@axolotl.ai>

* fix: AutoModelForVision2Seq -> AutoModelForImageTextToText

* chore: remove duplicate

* fix: attempt fix gemma3 text mode

* chore: lint

* ga release of v5

* need property setter for name_or_path for mistral tokenizer

* vllm not compatible with transformers v5

* setter for chat_template w mistral too

---------

Co-authored-by: NanoCode012 <nano@axolotl.ai>
Co-authored-by: salman <salman.mohammadi@outlook.com>

tests/e2e/utils.py

@@ -167,6 +167,13 @@ def require_hopper(test_case):
     return unittest.skipUnless(is_hopper(), "test requires h100/hopper GPU")(test_case)
 
 
+def supports_fp8(test_case):
+    compute_capability = torch.cuda.get_device_capability()
+    return unittest.skipUnless(
+        compute_capability >= (9, 0), "test requires h100 or newer GPU"
+    )(test_case)
+
+
 def check_tensorboard(
     temp_run_dir: str,
     tag: str,
@@ -193,21 +200,10 @@ def check_model_output_exists(temp_dir: str, cfg: DictDefault) -> None:
     """
     helper function to check if a model output file exists after training
 
-    checks based on adapter or not and if safetensors saves are enabled or not
+    checks based on adapter or not (always safetensors in Transformers V5)
     """
 
-    if cfg.save_safetensors:
-        if not cfg.adapter:
-            assert (Path(temp_dir) / "model.safetensors").exists()
-        else:
-            assert (Path(temp_dir) / "adapter_model.safetensors").exists()
+    if not cfg.adapter:
+        assert (Path(temp_dir) / "model.safetensors").exists()
     else:
-        # check for both, b/c in trl, it often defaults to saving safetensors
-        if not cfg.adapter:
-            assert (Path(temp_dir) / "pytorch_model.bin").exists() or (
-                Path(temp_dir) / "model.safetensors"
-            ).exists()
-        else:
-            assert (Path(temp_dir) / "adapter_model.bin").exists() or (
-                Path(temp_dir) / "adapter_model.safetensors"
-            ).exists()
+        assert (Path(temp_dir) / "adapter_model.safetensors").exists()