update

Author: Yunnglin

cookbook/client/server/megatron/run.sh

@@ -1,6 +1,5 @@
 export RAY_ROTATION_MAX_BYTES=1024
 export RAY_ROTATION_BACKUP_COUNT=1
-CUDA_VISIBLE_DEVICES=0,1,2,3 ray start --head --port=6379 --num-gpus=4 --disable-usage-stats --include-dashboard=false
-CUDA_VISIBLE_DEVICES=4,5,6,7 ray start --address=127.0.0.1:6379 --num-gpus=4
+CUDA_VISIBLE_DEVICES=0,1,2,3 ray start --head --port=6379 --num-gpus=4 --disable-usage-stats --temp-dir=/dashscope/caches/application/ray_logs
 CUDA_VISIBLE_DEVICES="" ray start --address=127.0.0.1:6379 --num-gpus=0
 python server.py

cookbook/client/server/megatron/server.py

@@ -15,7 +15,7 @@
 
 # Resolve the path to server_config.yaml relative to this script's location
 file_dir = os.path.abspath(os.path.dirname(__file__))
-config_path = os.path.join(file_dir, 'server_config_4b.yaml')
+config_path = os.path.join(file_dir, 'server_config.yaml')
 
 # Launch the Twinkle server — this call blocks until the server is shut down
 launch_server(config_path=config_path)

cookbook/client/server/megatron/start_monitor.sh

@@ -0,0 +1 @@
+./prometheus --config.file=/dashscope/caches/application/ray_logs/session_latest/metrics/prometheus/prometheus.yml