submission_id: rica40325-9-15-sftdpo_v1
developer_uid: rica40325
best_of: 16
celo_rating: 1267.11
display_name: rica40325-9-15-sftdpo_v1
family_friendly_score: 0.0
formatter: {'memory_template': "{bot_name}'s Persona: {memory}\n####\n", 'prompt_template': '{prompt}\n<START>\n', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: False
language_model: rica40325/9-15-sftdpo
latencies: [{'batch_size': 1, 'throughput': 0.8879871408033594, 'latency_mean': 1.1260458552837371, 'latency_p50': 1.1427234411239624, 'latency_p90': 1.2432791233062743}, {'batch_size': 4, 'throughput': 1.8001177978830583, 'latency_mean': 2.2116847121715546, 'latency_p50': 2.207395911216736, 'latency_p90': 2.4787020444869996}, {'batch_size': 5, 'throughput': 1.8739046714735121, 'latency_mean': 2.657439855337143, 'latency_p50': 2.6615357398986816, 'latency_p90': 3.022435688972473}, {'batch_size': 8, 'throughput': 2.0141142343009513, 'latency_mean': 3.9411645770072936, 'latency_p50': 3.9832788705825806, 'latency_p90': 4.434662675857544}, {'batch_size': 10, 'throughput': 2.005491540592479, 'latency_mean': 4.938176889419555, 'latency_p50': 4.9737643003463745, 'latency_p90': 5.545082449913025}, {'batch_size': 12, 'throughput': 2.036518189007474, 'latency_mean': 5.810856127738953, 'latency_p50': 5.7276307344436646, 'latency_p90': 7.059406018257141}, {'batch_size': 15, 'throughput': 2.0203722383094194, 'latency_mean': 7.282128564119339, 'latency_p50': 7.398141384124756, 'latency_p90': 8.14166476726532}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: rica40325/9-15-sftdpo
model_name: rica40325-9-15-sftdpo_v1
model_num_parameters: 8030261248.0
model_repo: rica40325/9-15-sftdpo
model_size: 8B
num_battles: 12296
num_wins: 6398
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 2.01
timestamp: 2024-09-16T15:24:43+00:00
us_pacific_date: 2024-09-16
win_ratio: 0.5203318152244633
Download Preference Data
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 rica40325-9-15-sftdpo-v1-mkmlizer
Waiting for job on rica40325-9-15-sftdpo-v1-mkmlizer to finish
rica40325-9-15-sftdpo-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rica40325-9-15-sftdpo-v1-mkmlizer: ║ _____ __ __ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ /___/ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ Version: 0.10.1 ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ https://mk1.ai ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ The license key for the current software has been verified as ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ belonging to: ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ Chai Research Corp. ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
rica40325-9-15-sftdpo-v1-mkmlizer: ║ ║
rica40325-9-15-sftdpo-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rica40325-9-15-sftdpo-v1-mkmlizer: Downloaded to shared memory in 30.676s
rica40325-9-15-sftdpo-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpzbo1fm0j, device:0
rica40325-9-15-sftdpo-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rica40325-9-15-sftdpo-v1-mkmlizer: quantized model in 26.130s
rica40325-9-15-sftdpo-v1-mkmlizer: Processed model rica40325/9-15-sftdpo in 56.806s
rica40325-9-15-sftdpo-v1-mkmlizer: creating bucket guanaco-mkml-models
rica40325-9-15-sftdpo-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rica40325-9-15-sftdpo-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1
rica40325-9-15-sftdpo-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1/config.json
rica40325-9-15-sftdpo-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1/special_tokens_map.json
rica40325-9-15-sftdpo-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1/tokenizer_config.json
rica40325-9-15-sftdpo-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1/tokenizer.json
rica40325-9-15-sftdpo-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rica40325-9-15-sftdpo-v1/flywheel_model.0.safetensors
rica40325-9-15-sftdpo-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 2%|▏ | 6/291 [00:00<00:04, 59.56it/s] Loading 0: 4%|▍ | 13/291 [00:00<00:04, 65.66it/s] Loading 0: 9%|▊ | 25/291 [00:00<00:03, 72.98it/s] Loading 0: 12%|█▏ | 34/291 [00:00<00:03, 69.90it/s] Loading 0: 15%|█▍ | 43/291 [00:00<00:03, 75.24it/s] Loading 0: 20%|█▉ | 58/291 [00:00<00:02, 86.39it/s] Loading 0: 24%|██▍ | 70/291 [00:00<00:02, 83.54it/s] Loading 0: 27%|██▋ | 79/291 [00:01<00:02, 80.92it/s] Loading 0: 30%|███ | 88/291 [00:02<00:09, 21.94it/s] Loading 0: 33%|███▎ | 97/291 [00:02<00:06, 27.82it/s] Loading 0: 38%|███▊ | 112/291 [00:02<00:04, 39.47it/s] Loading 0: 42%|████▏ | 123/291 [00:02<00:03, 48.73it/s] Loading 0: 45%|████▌ | 132/291 [00:02<00:02, 54.27it/s] Loading 0: 48%|████▊ | 141/291 [00:02<00:02, 57.55it/s] Loading 0: 52%|█████▏ | 150/291 [00:02<00:02, 60.96it/s] Loading 0: 54%|█████▍ | 158/291 [00:03<00:02, 64.99it/s] Loading 0: 58%|█████▊ | 169/291 [00:03<00:01, 68.22it/s] Loading 0: 61%|██████ | 178/291 [00:03<00:01, 66.42it/s] Loading 0: 64%|██████▍ | 187/291 [00:04<00:04, 21.64it/s] Loading 0: 67%|██████▋ | 196/291 [00:04<00:03, 27.31it/s] Loading 0: 73%|███████▎ | 211/291 [00:04<00:02, 39.07it/s] Loading 0: 76%|███████▌ | 220/291 [00:04<00:01, 45.71it/s] Loading 0: 79%|███████▉ | 230/291 [00:04<00:01, 54.32it/s] Loading 0: 83%|████████▎ | 241/291 [00:05<00:00, 58.38it/s] Loading 0: 86%|████████▌ | 250/291 [00:05<00:00, 60.01it/s] Loading 0: 89%|████████▉ | 259/291 [00:05<00:00, 63.15it/s] Loading 0: 92%|█████████▏| 269/291 [00:05<00:00, 71.12it/s] Loading 0: 96%|█████████▌| 278/291 [00:05<00:00, 73.65it/s] Loading 0: 99%|█████████▊| 287/291 [00:05<00:00, 40.86it/s]
Job rica40325-9-15-sftdpo-v1-mkmlizer completed after 86.94s with status: succeeded
Stopping job with name rica40325-9-15-sftdpo-v1-mkmlizer
Pipeline stage MKMLizer completed in 90.04s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.28s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rica40325-9-15-sftdpo-v1
Waiting for inference service rica40325-9-15-sftdpo-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
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
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
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 rica40325-9-15-sftdpo-v1 ready after 171.80312585830688s
Pipeline stage MKMLDeployer completed in 175.28s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.00215220451355s
Received healthy response to inference request in 2.179192543029785s
Received healthy response to inference request in 2.2801177501678467s
Received healthy response to inference request in 1.6981580257415771s
Received healthy response to inference request in 2.654038429260254s
5 requests
0 failed requests
5th percentile: 1.7589568614959716
10th percentile: 1.8197556972503661
20th percentile: 1.9413533687591553
30th percentile: 2.0375602722167967
40th percentile: 2.108376407623291
50th percentile: 2.179192543029785
60th percentile: 2.21956262588501
70th percentile: 2.2599327087402346
80th percentile: 2.354901885986328
90th percentile: 2.504470157623291
95th percentile: 2.5792542934417724
99th percentile: 2.6390816020965575
mean time: 2.1627317905426025
Pipeline stage StressChecker completed in 14.55s
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 7.40s
Shutdown handler de-registered
rica40325-9-15-sftdpo_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 0.13s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service rica40325-9-15-sftdpo-v1-profiler
Waiting for inference service rica40325-9-15-sftdpo-v1-profiler to be ready
Inference service rica40325-9-15-sftdpo-v1-profiler ready after 170.49553298950195s
Pipeline stage MKMLProfilerDeployer completed in 170.87s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/rica40325-9-15-sftdpo-v1-profiler-predictor-00001-deploymef2k8g:/code/chaiverse_profiler_1726500786 --namespace tenant-chaiml-guanaco
kubectl exec -it rica40325-9-15-sftdpo-v1-profiler-predictor-00001-deploymef2k8g --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726500786 && python profiles.py profile --best_of_n 16 --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_1726500786/summary.json'
kubectl exec -it rica40325-9-15-sftdpo-v1-profiler-predictor-00001-deploymef2k8g --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726500786/summary.json'
Pipeline stage MKMLProfilerRunner completed in 854.04s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-9-15-sftdpo-v1-profiler is running
Tearing down inference service rica40325-9-15-sftdpo-v1-profiler
Service rica40325-9-15-sftdpo-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 12.92s
Shutdown handler de-registered
rica40325-9-15-sftdpo_v1 status is now inactive due to auto deactivation removed underperforming models
rica40325-9-15-sftdpo_v1 status is now torndown due to DeploymentManager action