add ipex

huggingface · Dec 27, 2023 · 9164d9a · 9164d9a
1 parent 1d3f6c1
commit 9164d9a
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 1 deletion.
diff --git a/examples/textual_inversion/README.md b/examples/textual_inversion/README.md
@@ -94,6 +94,13 @@ to a number larger than one, *e.g.*:
 --num_vectors 5
 ```
 
+**CPU**: If you run on Intel Gen 4th Xeon (and later), use ipex and bf16 will get a significant acceleration.
+You need to add `--mixed_precision="bf16"` and `--use_ipex` in the command and install the following package:
+```
+pip install torch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2 --index-url https://download.pytorch.org/whl/cpu
+pip install intel-extension-for-pytorch==2.0.0
+```
+
 The saved textual inversion vectors will then be larger in size compared to the default case.
 
 ### Inference

diff --git a/examples/textual_inversion/textual_inversion.py b/examples/textual_inversion/textual_inversion.py
@@ -349,7 +349,7 @@ def parse_args():
         action="store_true",
         help=(
             "Whether or not to use ipex to accelerate the training process,"
-            "requires Intel Gen 3rd Xeon (and later) or Intel XPU (PVC)"
+            "requires Intel Gen 3rd Xeon (and later)"
         ),
     )
     parser.add_argument(