Shutdown handler not registered because Python interpreter is not running in the main thread
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLizer
Starting job with name sanchuanhehe-adjusted-co-5560-v1-mkmlizer
Waiting for job on sanchuanhehe-adjusted-co-5560-v1-mkmlizer to finish
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ _____ __ __ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ /___/ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ Version: 0.10.1 ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ https://mk1.ai ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ The license key for the current software has been verified as ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ belonging to: ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ Chai Research Corp. ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ║ ║
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: Downloaded to shared memory in 32.513s
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpa2harlgf, device:0
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: quantized model in 17.615s
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: Processed model sanchuanhehe/adjusted_config_6_le in 50.129s
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: creating bucket guanaco-mkml-models
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1/config.json
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1/tokenizer_config.json
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1/special_tokens_map.json
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1/tokenizer.model
sanchuanhehe-adjusted-co-5560-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/sanchuanhehe-adjusted-co-5560-v1/tokenizer.json
Job sanchuanhehe-adjusted-co-5560-v1-mkmlizer completed after 75.52s with status: succeeded
Stopping job with name sanchuanhehe-adjusted-co-5560-v1-mkmlizer
Pipeline stage MKMLizer completed in 77.10s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.10s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service sanchuanhehe-adjusted-co-5560-v1
Waiting for inference service sanchuanhehe-adjusted-co-5560-v1 to be ready
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
Inference service sanchuanhehe-adjusted-co-5560-v1 ready after 181.90998578071594s
Pipeline stage MKMLDeployer completed in 182.36s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.817850112915039s
Received healthy response to inference request in 0.7523891925811768s
Received healthy response to inference request in 1.0450901985168457s
Received healthy response to inference request in 0.7373971939086914s
Received healthy response to inference request in 0.5563156604766846s
5 requests
0 failed requests
5th percentile: 0.592531967163086
10th percentile: 0.6287482738494873
20th percentile: 0.70118088722229
30th percentile: 0.7403955936431885
40th percentile: 0.7463923931121826
50th percentile: 0.7523891925811768
60th percentile: 0.8694695949554443
70th percentile: 0.9865499973297118
80th percentile: 1.1996421813964846
90th percentile: 1.5087461471557617
95th percentile: 1.6632981300354002
99th percentile: 1.7869397163391112
mean time: 0.9818084716796875
Pipeline stage StressChecker completed in 6.88s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 6.93s
Shutdown handler de-registered
sanchuanhehe-adjusted-co_5560_v1 status is now deployed due to DeploymentManager action
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Skipping teardown as no inference service was successfully deployed
Pipeline stage MKMLProfilerDeleter completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service sanchuanhehe-adjusted-co-5560-v1-profiler
Waiting for inference service sanchuanhehe-adjusted-co-5560-v1-profiler to be ready
Inference service sanchuanhehe-adjusted-co-5560-v1-profiler ready after 180.43903470039368s
Pipeline stage MKMLProfilerDeployer completed in 180.81s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/sanchuanhehe-adjuste10c23eaeb8a0fdaad217051fa3928a45-deplo7q5sq:/code/chaiverse_profiler_1726578870 --namespace tenant-chaiml-guanaco
kubectl exec -it sanchuanhehe-adjuste10c23eaeb8a0fdaad217051fa3928a45-deplo7q5sq --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726578870 && python profiles.py profile --best_of_n 1 --auto_batch 5 --batches 1,5,10,15,20,25,30,35,40,45,50,55,60,65,70,75,80,85,90,95,100,105,110,115,120,125,130,135,140,145,150,155,160,165,170,175,180,185,190,195 --samples 200 --input_tokens 512 --output_tokens 64 --summary /code/chaiverse_profiler_1726578870/summary.json'
kubectl exec -it sanchuanhehe-adjuste10c23eaeb8a0fdaad217051fa3928a45-deplo7q5sq --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726578870/summary.json'
Pipeline stage MKMLProfilerRunner completed in 448.46s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service sanchuanhehe-adjusted-co-5560-v1-profiler is running
Tearing down inference service sanchuanhehe-adjusted-co-5560-v1-profiler
Service sanchuanhehe-adjusted-co-5560-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.37s
Shutdown handler de-registered
sanchuanhehe-adjusted-co_5560_v1 status is now inactive due to auto deactivation removed underperforming models
sanchuanhehe-adjusted-co_5560_v1 status is now torndown due to DeploymentManager action