submission_id: jic062-dpo-v1-5-nemo-c500_v1
developer_uid: chace9580
best_of: 8
celo_rating: 1240.54
display_name: jic062-dpo-v1-5-nemo-c500_v1
family_friendly_score: 0.0
formatter: {'memory_template': '[INST]system\n{memory}[/INST]\n', 'prompt_template': '[INST]user\n{prompt}[/INST]\n', 'bot_template': '[INST]assistant\n{bot_name}: {message}[/INST]\n', 'user_template': '[INST]user\n{user_name}: {message}[/INST]\n', 'response_template': '[INST]assistant\n{bot_name}:', 'truncate_by_message': False}
generation_params: {'temperature': 0.75, 'top_p': 1.0, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '\n\n', '\nYou:', '[/INST]', '<|im_end|>', '</s>'], 'max_input_tokens': 512, 'best_of': 8, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: False
language_model: jic062/dpo-v1.5-Nemo-c500
latencies: [{'batch_size': 1, 'throughput': 0.7008830494705103, 'latency_mean': 1.4266752088069916, 'latency_p50': 1.4362400770187378, 'latency_p90': 1.5783403873443604}, {'batch_size': 3, 'throughput': 1.325439571827967, 'latency_mean': 2.2617265677452085, 'latency_p50': 2.2787580490112305, 'latency_p90': 2.479091191291809}, {'batch_size': 5, 'throughput': 1.5728334811323406, 'latency_mean': 3.169850291013718, 'latency_p50': 3.176653265953064, 'latency_p90': 3.499379110336304}, {'batch_size': 6, 'throughput': 1.6042207914090985, 'latency_mean': 3.7194345355033875, 'latency_p50': 3.7271474599838257, 'latency_p90': 4.213351416587829}, {'batch_size': 8, 'throughput': 1.5871593868736065, 'latency_mean': 5.016155124902725, 'latency_p50': 5.028848886489868, 'latency_p90': 5.668827390670776}, {'batch_size': 10, 'throughput': 1.5239310585117556, 'latency_mean': 6.52420462012291, 'latency_p50': 6.6006598472595215, 'latency_p90': 7.351833200454712}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: MistralForCausalLM
model_group: jic062/dpo-v1.5-Nemo-c50
model_name: jic062-dpo-v1-5-nemo-c500_v1
model_num_parameters: 12772070400.0
model_repo: jic062/dpo-v1.5-Nemo-c500
model_size: 13B
num_battles: 10097
num_wins: 4865
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 1.61
timestamp: 2024-09-15T18:06:42+00:00
us_pacific_date: 2024-09-15
win_ratio: 0.48182628503515895
Resubmit model
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 jic062-dpo-v1-5-nemo-c500-v1-mkmlizer
Waiting for job on jic062-dpo-v1-5-nemo-c500-v1-mkmlizer to finish
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║     _____            __           __                                ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║       /___/                                                         ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  https://mk1.ai                                                     ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  belonging to:                                                      ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  Chai Research Corp.                                                ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s. Connection pool size: %s
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: Downloaded to shared memory in 51.586s
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
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpo7mr7xb6, device:0
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: quantized model in 35.876s
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: Processed model jic062/dpo-v1.5-Nemo-c500 in 87.463s
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: creating bucket guanaco-mkml-models
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-dpo-v1-5-nemo-c500-v1
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-dpo-v1-5-nemo-c500-v1/config.json
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-dpo-v1-5-nemo-c500-v1/special_tokens_map.json
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-dpo-v1-5-nemo-c500-v1/tokenizer_config.json
jic062-dpo-v1-5-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-dpo-v1-5-nemo-c500-v1/tokenizer.json
Job jic062-dpo-v1-5-nemo-c500-v1-mkmlizer completed after 116.7s with status: succeeded
Stopping job with name jic062-dpo-v1-5-nemo-c500-v1-mkmlizer
Pipeline stage MKMLizer completed in 123.52s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.09s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service jic062-dpo-v1-5-nemo-c500-v1
Waiting for inference service jic062-dpo-v1-5-nemo-c500-v1 to be ready
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', '')
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 jic062-dpo-v1-5-nemo-c500-v1 ready after 181.1396095752716s
Pipeline stage MKMLDeployer completed in 187.38s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
run pipeline stage %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
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.4673280715942383s
Received healthy response to inference request in 2.167800188064575s
Received healthy response to inference request in 2.0284295082092285s
Received healthy response to inference request in 2.622598886489868s
Received healthy response to inference request in 2.3382294178009033s
5 requests
0 failed requests
5th percentile: 2.0563036441802978
10th percentile: 2.084177780151367
20th percentile: 2.139926052093506
30th percentile: 2.201886034011841
40th percentile: 2.270057725906372
50th percentile: 2.3382294178009033
60th percentile: 2.4519772052764894
70th percentile: 2.565724992752075
80th percentile: 2.7915447235107425
90th percentile: 3.12943639755249
95th percentile: 3.298382234573364
99th percentile: 3.4335389041900632
mean time: 2.5248772144317626
Pipeline stage StressChecker completed in 18.05s
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 4.81s
Shutdown handler de-registered
jic062-dpo-v1-5-nemo-c500_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 0.12s
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 jic062-dpo-v1-5-nemo-c500-v1-profiler
Waiting for inference service jic062-dpo-v1-5-nemo-c500-v1-profiler to be ready
Inference service jic062-dpo-v1-5-nemo-c500-v1-profiler ready after 170.37543272972107s
Pipeline stage MKMLProfilerDeployer completed in 170.73s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/jic062-dpo-v1-5-nemo3ce566ff9c2d9467dc6da217865591df-deplonn5l6:/code/chaiverse_profiler_1726424160 --namespace tenant-chaiml-guanaco
kubectl exec -it jic062-dpo-v1-5-nemo3ce566ff9c2d9467dc6da217865591df-deplonn5l6 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726424160 && python profiles.py profile --best_of_n 8 --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_1726424160/summary.json'
kubectl exec -it jic062-dpo-v1-5-nemo3ce566ff9c2d9467dc6da217865591df-deplonn5l6 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726424160/summary.json'
Pipeline stage MKMLProfilerRunner completed in 950.82s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service jic062-dpo-v1-5-nemo-c500-v1-profiler is running
Tearing down inference service jic062-dpo-v1-5-nemo-c500-v1-profiler
Service jic062-dpo-v1-5-nemo-c500-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 7.28s
Shutdown handler de-registered
jic062-dpo-v1-5-nemo-c500_v1 status is now inactive due to auto deactivation removed underperforming models
jic062-dpo-v1-5-nemo-c500_v1 status is now torndown due to DeploymentManager action