try with METAL_XCODE=1 for tinygrad metal

04f2aa2a65d446119bc3ebd66c635ff87d2843af · 2024-07-20 12:16:54 -0700 · Alex Cheema

Files touched

M .github/workflows/test.yml
M exo/inference/test_inference_engine.py

Diff

commit 04f2aa2a65d446119bc3ebd66c635ff87d2843af
Author: Alex Cheema <alexcheema123@gmail.com>
Date:   Sat Jul 20 12:16:54 2024 -0700

    try with METAL_XCODE=1 for tinygrad metal
---
 .github/workflows/test.yml             |  2 +-
 exo/inference/test_inference_engine.py | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index 94466607..8c87f320 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -23,7 +23,7 @@ jobs:
         pip install .
 
     - name: Run tests
-      run: python3 -m exo.inference.test_inference_engine
+      run: METAL_XCODE=1 python3 -m exo.inference.test_inference_engine
 
   discovery_integration_test:
     runs-on: macos-latest
diff --git a/exo/inference/test_inference_engine.py b/exo/inference/test_inference_engine.py
index 2cdba8b6..f5e7d863 100644
--- a/exo/inference/test_inference_engine.py
+++ b/exo/inference/test_inference_engine.py
@@ -31,8 +31,8 @@ asyncio.run(test_inference_engine(
 ))
 
 # TODO: Waiting on https://github.com/tinygrad/tinygrad/issues/5549
-# asyncio.run(test_inference_engine(
-#     TinygradDynamicShardInferenceEngine(),
-#     TinygradDynamicShardInferenceEngine(),
-#     "llama3-8b-sfr",
-# ))
+asyncio.run(test_inference_engine(
+    TinygradDynamicShardInferenceEngine(),
+    TinygradDynamicShardInferenceEngine(),
+    "llama3-8b-sfr",
+))

← d2ed4c2a disable tinygrad infernece engine test waiting Waiting on ht · back to Exo · cleaner node info = a2a7ca1f →