force mlx inference engine in github workflow, where it defaults to tinygrad because it's running on 'model': 'Apple Virtual Machine 1', 'chip': 'Apple M1 (Virtual)'

628d8679b01d2883779d6e991a011123d1a7e61b · 2024-07-26 21:46:41 -0700 · Alex Cheema

Files touched

M .github/workflows/test.yml

Diff

commit 628d8679b01d2883779d6e991a011123d1a7e61b
Author: Alex Cheema <alexcheema123@gmail.com>
Date:   Fri Jul 26 21:46:41 2024 -0700

    force mlx inference engine in github workflow, where it defaults to tinygrad because it's running on 'model': 'Apple Virtual Machine 1', 'chip': 'Apple M1 (Virtual)'
---
 .github/workflows/test.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
index 8c339c7a..7971965e 100644
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@@ -117,11 +117,11 @@ jobs:
         ls ~/.cache/huggingface/hub/models--mlx-community--Meta-Llama-3-8B-Instruct-4bit/**/* || true
 
         # Start first instance
-        DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --listen-port 5678 --broadcast-port 5679 --chatgpt-api-port 8000 --chatgpt-api-response-timeout-secs 900 > output1.log 2>&1 &
+        DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine mlx --listen-port 5678 --broadcast-port 5679 --chatgpt-api-port 8000 --chatgpt-api-response-timeout-secs 900 > output1.log 2>&1 &
         PID1=$!
 
         # Start second instance
-        DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --listen-port 5679 --broadcast-port 5678 --chatgpt-api-port 8001 --chatgpt-api-response-timeout-secs 900 > output2.log 2>&1 &
+        DEBUG_DISCOVERY=9 DEBUG=9 python3 main.py --inference-engine mlx --listen-port 5679 --broadcast-port 5678 --chatgpt-api-port 8001 --chatgpt-api-response-timeout-secs 900 > output2.log 2>&1 &
         PID2=$!
 
         # Wait for discovery

← e856d7f7 log chatgpt integration test output from each process on git · back to Exo · check processes in github workflow 67a1aaa8 →