basic torchao fp8 mixed precision training (#2926)

* debug * debug * debug * revert unneeded change * add accelerator config to base trainer builder * add back accumulated_cache_size_limit setting * lint * accelerator constructor patch for single-GPU torch fp8 * lint * re-using existing fp8 code * lint * remove accelerate patch now fix in latest release * fix * docs * add fp8 + fsdp2 example * remove unused config * update config * smoke tests * add validator * add 2.7.0 guard for fsdp2 * fix * add config descriptions * add FSDP doc link * nit * set force_recompute_fp8_weight_in_bwd with enable_fsdp_float8_all_gather * better cfg for smoke tests * add test for accelerate patching * update fp8 validator
2025-07-22 16:27:47 -04:00
parent b86a1d47b0
commit 208fb7b8e7
11 changed files with 503 additions and 10 deletions
--- a/tests/e2e/integrations/test_fp8.py
+++ b/tests/e2e/integrations/test_fp8.py
@@ -0,0 +1,62 @@
+"""
+Simple end-to-end smoke tests for FP8 mixed precision training
+"""
+
+from axolotl.common.datasets import load_datasets
+from axolotl.train import train
+from axolotl.utils.config import normalize_config, validate_config
+from axolotl.utils.dict import DictDefault
+
+from tests.e2e.utils import check_model_output_exists, require_torch_2_7_0
+
+
+class FP8IntegrationTestCase:
+    """
+    e2e smoke tests for FP8 mixed precision training with Axolotl
+    """
+
+    @require_torch_2_7_0
+    def test_fp8_single_gpu_smoke(self, temp_dir):
+        """Smoke test for single GPU FP8 + torch.compile training"""
+        # pylint: disable=duplicate-code
+        cfg = DictDefault(
+            {
+                "base_model": "HuggingFaceTB/SmolLM2-135M",
+                "tokenizer_type": "AutoTokenizer",
+                "trust_remote_code": True,
+                "sequence_len": 512,
+                "val_set_size": 0.05,
+                "special_tokens": {
+                    "pad_token": "<|endoftext|>",
+                },
+                "datasets": [
+                    {
+                        "path": "mhenrichsen/alpaca_2k_test",
+                        "type": "alpaca",
+                    },
+                ],
+                "num_epochs": 1,
+                "max_steps": 3,  # Very short smoke test
+                "micro_batch_size": 1,
+                "gradient_accumulation_steps": 2,
+                "output_dir": temp_dir,
+                "learning_rate": 0.00001,
+                "optimizer": "adamw_torch_fused",
+                "lr_scheduler": "cosine",
+                "sdp_attention": True,
+                "pad_to_seq_len": True,
+                "sample_packing": True,
+                "fp8": True,
+                "torch_compile": True,
+                "save_safetensors": True,
+                "save_first_step": False,
+            }
+        )
+
+        # pylint: disable=duplicate-code
+        cfg = validate_config(cfg)
+        normalize_config(cfg)
+        dataset_meta = load_datasets(cfg=cfg)
+
+        train(cfg=cfg, dataset_meta=dataset_meta)
+        check_model_output_exists(temp_dir, cfg)