submission_id: rica40325-alignment-first_v4
developer_uid: rica40325
alignment_samples: 13519
alignment_score: 0.28161526657934255
best_of: 16
celo_rating: 1197.66
display_name: rica40325-alignment-first_v1
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': 2.0, 'top_p': 0.95, 'min_p': 0.05, 'top_k': 80, '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/alignment-first
latencies: [{'batch_size': 1, 'throughput': 0.896928474117526, 'latency_mean': 1.1148475360870362, 'latency_p50': 1.1164132356643677, 'latency_p90': 1.2513904571533203}, {'batch_size': 4, 'throughput': 1.7912859707148705, 'latency_mean': 2.2230956518650054, 'latency_p50': 2.2181501388549805, 'latency_p90': 2.4894498586654663}, {'batch_size': 5, 'throughput': 1.8618964206430488, 'latency_mean': 2.670583585500717, 'latency_p50': 2.6692804098129272, 'latency_p90': 3.0036890029907224}, {'batch_size': 8, 'throughput': 2.0000440321125805, 'latency_mean': 3.973714334964752, 'latency_p50': 3.9161770343780518, 'latency_p90': 4.483904719352722}, {'batch_size': 10, 'throughput': 2.0346037980335825, 'latency_mean': 4.863650344610214, 'latency_p50': 4.835874915122986, 'latency_p90': 5.68966052532196}, {'batch_size': 12, 'throughput': 2.0179374370951253, 'latency_mean': 5.862228792905808, 'latency_p50': 5.872660279273987, 'latency_p90': 6.725356268882751}, {'batch_size': 15, 'throughput': 2.018982679147122, 'latency_mean': 7.280075047016144, 'latency_p50': 7.439767956733704, 'latency_p90': 7.985824060440064}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: rica40325/alignment-firs
model_name: rica40325-alignment-first_v1
model_num_parameters: 8030261248.0
model_repo: rica40325/alignment-first
model_size: 8B
num_battles: 13518
num_wins: 5912
propriety_score: 0.7285223367697594
propriety_total_count: 1164.0
ranking_group: single
status: inactive
submission_type: basic
throughput_3p7s: 1.99
timestamp: 2024-09-08T10:44:26+00:00
us_pacific_date: 2024-09-08
win_ratio: 0.43734280218967303
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-alignment-first-v4-mkmlizer
Waiting for job on rica40325-alignment-first-v4-mkmlizer to finish
rica40325-alignment-first-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rica40325-alignment-first-v4-mkmlizer: ║     _____            __           __                                ║
rica40325-alignment-first-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rica40325-alignment-first-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rica40325-alignment-first-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rica40325-alignment-first-v4-mkmlizer: ║       /___/                                                         ║
rica40325-alignment-first-v4-mkmlizer: ║                                                                     ║
rica40325-alignment-first-v4-mkmlizer: ║  Version: 0.10.1                                                    ║
rica40325-alignment-first-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rica40325-alignment-first-v4-mkmlizer: ║  https://mk1.ai                                                     ║
rica40325-alignment-first-v4-mkmlizer: ║                                                                     ║
rica40325-alignment-first-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
rica40325-alignment-first-v4-mkmlizer: ║  belonging to:                                                      ║
rica40325-alignment-first-v4-mkmlizer: ║                                                                     ║
rica40325-alignment-first-v4-mkmlizer: ║  Chai Research Corp.                                                ║
rica40325-alignment-first-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rica40325-alignment-first-v4-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
rica40325-alignment-first-v4-mkmlizer: ║                                                                     ║
rica40325-alignment-first-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rica40325-alignment-first-v4-mkmlizer: Downloaded to shared memory in 21.629s
rica40325-alignment-first-v4-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpdzhqd1ng, device:0
rica40325-alignment-first-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rica40325-alignment-first-v4-mkmlizer: quantized model in 25.168s
rica40325-alignment-first-v4-mkmlizer: Processed model rica40325/alignment-first in 46.797s
rica40325-alignment-first-v4-mkmlizer: creating bucket guanaco-mkml-models
rica40325-alignment-first-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rica40325-alignment-first-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rica40325-alignment-first-v4
rica40325-alignment-first-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rica40325-alignment-first-v4/config.json
rica40325-alignment-first-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rica40325-alignment-first-v4/special_tokens_map.json
rica40325-alignment-first-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rica40325-alignment-first-v4/tokenizer_config.json
rica40325-alignment-first-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rica40325-alignment-first-v4/tokenizer.json
rica40325-alignment-first-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rica40325-alignment-first-v4/flywheel_model.0.safetensors
rica40325-alignment-first-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:05, 54.33it/s]
Loading 0:   8%|▊         | 22/291 [00:00<00:03, 86.84it/s]
Loading 0:  12%|█▏        | 34/291 [00:00<00:02, 87.27it/s]
Loading 0:  17%|█▋        | 49/291 [00:00<00:02, 96.34it/s]
Loading 0:  20%|██        | 59/291 [00:00<00:02, 94.59it/s]
Loading 0:  24%|██▍       | 70/291 [00:00<00:02, 88.49it/s]
Loading 0:  29%|██▊       | 83/291 [00:01<00:07, 26.87it/s]
Loading 0:  31%|███       | 90/291 [00:02<00:06, 30.43it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:04, 40.39it/s]
Loading 0:  40%|███▉      | 115/291 [00:02<00:03, 48.87it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:02, 60.70it/s]
Loading 0:  48%|████▊     | 139/291 [00:02<00:02, 63.74it/s]
Loading 0:  52%|█████▏    | 150/291 [00:02<00:01, 72.67it/s]
Loading 0:  55%|█████▍    | 160/291 [00:02<00:01, 71.90it/s]
Loading 0:  58%|█████▊    | 169/291 [00:02<00:01, 75.57it/s]
Loading 0:  63%|██████▎   | 184/291 [00:03<00:01, 85.14it/s]
Loading 0:  67%|██████▋   | 194/291 [00:04<00:03, 26.33it/s]
Loading 0:  70%|██████▉   | 203/291 [00:04<00:02, 32.15it/s]
Loading 0:  74%|███████▎  | 214/291 [00:04<00:01, 39.36it/s]
Loading 0:  79%|███████▊  | 229/291 [00:04<00:01, 52.01it/s]
Loading 0:  83%|████████▎ | 241/291 [00:04<00:00, 59.32it/s]
Loading 0:  88%|████████▊ | 256/291 [00:04<00:00, 69.92it/s]
Loading 0:  92%|█████████▏| 268/291 [00:04<00:00, 74.43it/s]
Loading 0:  97%|█████████▋| 283/291 [00:05<00:00, 82.73it/s]
                                                            

Connection pool is full, discarding connection: %s. Connection pool size: %s
Job rica40325-alignment-first-v4-mkmlizer completed after 74.91s with status: succeeded
Stopping job with name rica40325-alignment-first-v4-mkmlizer
Pipeline stage MKMLizer completed in 76.39s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rica40325-alignment-first-v4
Waiting for inference service rica40325-alignment-first-v4 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
Inference service rica40325-alignment-first-v4 ready after 170.5515637397766s
Pipeline stage MKMLDeployer completed in 172.56s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.254585027694702s
Received healthy response to inference request in 2.134403705596924s
Received healthy response to inference request in 1.8879153728485107s
Received healthy response to inference request in 1.2221324443817139s
Received healthy response to inference request in 2.144540786743164s
5 requests
0 failed requests
5th percentile: 1.3552890300750733
10th percentile: 1.4884456157684327
20th percentile: 1.7547587871551513
30th percentile: 1.9372130393981934
40th percentile: 2.0358083724975584
50th percentile: 2.134403705596924
60th percentile: 2.1384585380554197
70th percentile: 2.142513370513916
80th percentile: 2.166549634933472
90th percentile: 2.210567331314087
95th percentile: 2.2325761795043944
99th percentile: 2.2501832580566408
mean time: 1.928715467453003
Pipeline stage StressChecker completed in 10.71s
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 5.35s
Shutdown handler de-registered
rica40325-alignment-first_v4 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 rica40325-alignment-first-v4-profiler
Waiting for inference service rica40325-alignment-first-v4-profiler to be ready
Inference service rica40325-alignment-first-v4-profiler ready after 150.35045528411865s
Pipeline stage MKMLProfilerDeployer completed in 150.72s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/rica40325-alignment-8ab3e2a5c862785f9a5158b1b6ea328f-deploq89j9:/code/chaiverse_profiler_1725792726 --namespace tenant-chaiml-guanaco
kubectl exec -it rica40325-alignment-8ab3e2a5c862785f9a5158b1b6ea328f-deploq89j9 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725792726 && 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_1725792726/summary.json'
kubectl exec -it rica40325-alignment-8ab3e2a5c862785f9a5158b1b6ea328f-deploq89j9 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725792726/summary.json'
Pipeline stage MKMLProfilerRunner completed in 846.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-alignment-first-v4-profiler is running
Tearing down inference service rica40325-alignment-first-v4-profiler
Service rica40325-alignment-first-v4-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.70s
Shutdown handler de-registered
rica40325-alignment-first_v4 status is now inactive due to auto deactivation removed underperforming models