chaiml-espresso-llama-24_9292

developer_uid: azuruce

submission_id: chaiml-espresso-llama-24_9292_v1

model_name: chaiml-espresso-llama-24_9292_v1

model_group: ChaiML/espresso_llama_24

status: inactive

timestamp: 2024-12-05T05:52:21+00:00

num_battles: 9303

num_wins: 4584

celo_rating: 1256.78

family_friendly_score: 0.585

family_friendly_standard_error: 0.00696814178960216

submission_type: basic

model_repo: ChaiML/espresso_llama_241204_albert_v2_sft_2epoch_128alpha

model_architecture: MistralForCausalLM

model_num_parameters: 22247282688.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.38342297562683353, 'latency_mean': 2.60800345659256, 'latency_p50': 2.5992847681045532, 'latency_p90': 2.8893085956573485}, {'batch_size': 2, 'throughput': 0.6012901246993432, 'latency_mean': 3.31583660364151, 'latency_p50': 3.3348242044448853, 'latency_p90': 3.68406445980072}, {'batch_size': 3, 'throughput': 0.7610269843512476, 'latency_mean': 3.932958686351776, 'latency_p50': 3.9108381271362305, 'latency_p90': 4.356282806396484}, {'batch_size': 4, 'throughput': 0.8726447356805713, 'latency_mean': 4.570032052993774, 'latency_p50': 4.5919729471206665, 'latency_p90': 5.011233997344971}, {'batch_size': 5, 'throughput': 0.9625095112064084, 'latency_mean': 5.1678644931316375, 'latency_p50': 5.15890634059906, 'latency_p90': 5.628885126113891}]

gpu_counts: {'NVIDIA RTX A6000': 1}

display_name: chaiml-espresso-llama-24_9292_v1

is_internal_developer: True

language_model: ChaiML/espresso_llama_241204_albert_v2_sft_2epoch_128alpha

model_size: 22B

ranking_group: single

throughput_3p7s: 0.71

us_pacific_date: 2024-12-04

win_ratio: 0.4927442760399871

generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|eot_id|>', '<|end_of_text|>', 'You:'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '', 'prompt_template': '', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}

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 chaiml-espresso-llama-24-9292-v1-mkmlizer
Waiting for job on chaiml-espresso-llama-24-9292-v1-mkmlizer to finish
chaiml-espresso-llama-24-9292-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║     _____            __           __                                ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║       /___/                                                         ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║                                                                     ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  Version: 0.11.12                                                   ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║                                                                     ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  belonging to:                                                      ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║                                                                     ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ║                                                                     ║
chaiml-espresso-llama-24-9292-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-espresso-llama-24-9292-v1-mkmlizer: Downloaded to shared memory in 73.590s
chaiml-espresso-llama-24-9292-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpj6phl8bk, device:0
chaiml-espresso-llama-24-9292-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-espresso-llama-24-9292-v1-mkmlizer: quantized model in 54.251s
chaiml-espresso-llama-24-9292-v1-mkmlizer: Processed model ChaiML/espresso_llama_241204_albert_v2_sft_2epoch_128alpha in 127.841s
chaiml-espresso-llama-24-9292-v1-mkmlizer: creating bucket guanaco-mkml-models
chaiml-espresso-llama-24-9292-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-espresso-llama-24-9292-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/config.json
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/special_tokens_map.json
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/tokenizer_config.json
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/tokenizer.json
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/flywheel_model.1.safetensors
chaiml-espresso-llama-24-9292-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-espresso-llama-24-9292-v1/flywheel_model.0.safetensors
chaiml-espresso-llama-24-9292-v1-mkmlizer: 
Loading 0:   0%|          | 0/507 [00:00<?, ?it/s]
Loading 0:   1%|          | 4/507 [00:00<00:12, 39.90it/s]
Loading 0:   2%|▏         | 8/507 [00:00<00:17, 28.10it/s]
Loading 0:   2%|▏         | 12/507 [00:00<00:16, 29.70it/s]
Loading 0:   3%|▎         | 16/507 [00:00<00:17, 28.62it/s]
Loading 0:   4%|▍         | 21/507 [00:00<00:15, 31.79it/s]
Loading 0:   5%|▍         | 25/507 [00:00<00:16, 29.93it/s]
Loading 0:   6%|▌         | 29/507 [00:00<00:14, 32.36it/s]
Loading 0:   7%|▋         | 33/507 [00:01<00:16, 28.63it/s]
Loading 0:   7%|▋         | 37/507 [00:01<00:15, 30.88it/s]
Loading 0:   8%|▊         | 41/507 [00:01<00:16, 28.07it/s]
Loading 0:   9%|▉         | 47/507 [00:01<00:13, 33.01it/s]
Loading 0:  10%|█         | 51/507 [00:01<00:14, 31.24it/s]
Loading 0:  11%|█         | 55/507 [00:02<00:21, 21.23it/s]
Loading 0:  11%|█▏        | 58/507 [00:02<00:20, 21.42it/s]
Loading 0:  12%|█▏        | 61/507 [00:02<00:19, 22.53it/s]
Loading 0:  13%|█▎        | 65/507 [00:02<00:19, 22.35it/s]
Loading 0:  14%|█▍        | 70/507 [00:02<00:16, 27.28it/s]
Loading 0:  15%|█▍        | 75/507 [00:02<00:14, 30.69it/s]
Loading 0:  16%|█▌        | 80/507 [00:02<00:14, 30.30it/s]
Loading 0:  17%|█▋        | 85/507 [00:02<00:12, 33.82it/s]
Loading 0:  18%|█▊        | 89/507 [00:03<00:14, 29.70it/s]
Loading 0:  19%|█▊        | 94/507 [00:03<00:14, 28.51it/s]
Loading 0:  19%|█▉        | 98/507 [00:03<00:25, 15.97it/s]
Loading 0:  20%|██        | 103/507 [00:04<00:25, 15.63it/s]
Loading 0:  21%|██        | 107/507 [00:04<00:24, 16.64it/s]
Loading 0:  22%|██▏       | 112/507 [00:04<00:19, 20.71it/s]
Loading 0:  23%|██▎       | 115/507 [00:04<00:22, 17.17it/s]
Loading 0:  23%|██▎       | 118/507 [00:04<00:21, 17.93it/s]
Loading 0:  24%|██▍       | 122/507 [00:05<00:19, 19.32it/s]
Loading 0:  25%|██▌       | 127/507 [00:05<00:15, 24.73it/s]
Loading 0:  26%|██▌       | 131/507 [00:05<00:15, 24.11it/s]
Loading 0:  27%|██▋       | 138/507 [00:05<00:11, 31.52it/s]
Loading 0:  28%|██▊       | 142/507 [00:05<00:11, 31.93it/s]
Loading 0:  29%|██▉       | 147/507 [00:05<00:10, 35.50it/s]
Loading 0:  30%|██▉       | 151/507 [00:05<00:10, 34.54it/s]
Loading 0:  31%|███       | 156/507 [00:05<00:09, 37.72it/s]
Loading 0:  32%|███▏      | 160/507 [00:06<00:09, 36.10it/s]
Loading 0:  32%|███▏      | 164/507 [00:07<00:39,  8.66it/s]
Loading 0:  33%|███▎      | 168/507 [00:07<00:30, 11.05it/s]
Loading 0:  34%|███▍      | 172/507 [00:08<00:40,  8.25it/s]
Loading 0:  35%|███▍      | 175/507 [00:08<00:35,  9.48it/s]
Loading 0:  35%|███▌      | 178/507 [00:08<00:33,  9.75it/s]
Loading 0:  36%|███▌      | 183/507 [00:09<00:33,  9.70it/s]
Loading 0:  36%|███▋      | 185/507 [00:09<00:33,  9.56it/s]
Loading 0:  37%|███▋      | 190/507 [00:09<00:22, 13.79it/s]
Loading 0:  38%|███▊      | 193/507 [00:09<00:21, 14.93it/s]
Loading 0:  39%|███▊      | 196/507 [00:09<00:18, 16.67it/s]
Loading 0:  39%|███▉      | 199/507 [00:10<00:16, 18.48it/s]
Loading 0:  40%|████      | 203/507 [00:10<00:15, 19.52it/s]
Loading 0:  41%|████      | 208/507 [00:10<00:11, 24.95it/s]
Loading 0:  42%|████▏     | 212/507 [00:10<00:12, 24.19it/s]
Loading 0:  43%|████▎     | 217/507 [00:10<00:09, 29.35it/s]
Loading 0:  44%|████▎     | 221/507 [00:11<00:23, 12.29it/s]
Loading 0:  44%|████▍     | 224/507 [00:11<00:24, 11.70it/s]
Loading 0:  45%|████▌     | 230/507 [00:11<00:18, 15.29it/s]
Loading 0:  46%|████▋     | 235/507 [00:12<00:13, 19.45it/s]
Loading 0:  47%|████▋     | 238/507 [00:12<00:24, 10.77it/s]
Loading 0:  48%|████▊     | 241/507 [00:13<00:23, 11.21it/s]
Loading 0:  49%|████▊     | 246/507 [00:13<00:17, 15.08it/s]
Loading 0:  49%|████▉     | 249/507 [00:13<00:26,  9.63it/s]
Loading 0:  50%|████▉     | 253/507 [00:14<00:22, 11.23it/s]
Loading 0:  51%|█████     | 257/507 [00:14<00:18, 13.35it/s]
Loading 0:  52%|█████▏    | 262/507 [00:14<00:13, 17.81it/s]
Loading 0:  52%|█████▏    | 266/507 [00:14<00:12, 19.69it/s]
Loading 0:  54%|█████▍    | 273/507 [00:14<00:08, 27.09it/s]
Loading 0:  55%|█████▍    | 277/507 [00:15<00:19, 11.72it/s]
Loading 0:  56%|█████▌    | 283/507 [00:15<00:14, 15.65it/s]
Loading 0:  57%|█████▋    | 287/507 [00:16<00:29,  7.58it/s]
Loading 0:  58%|█████▊    | 293/507 [00:17<00:20, 10.39it/s]
Loading 0:  59%|█████▉    | 298/507 [00:17<00:17, 12.02it/s]
Loading 0:  59%|█████▉    | 299/507 [00:32<00:17, 12.02it/s]
Loading 0:  59%|█████▉    | 300/507 [00:32<03:58,  1.15s/it]
Loading 0:  60%|█████▉    | 302/507 [00:32<03:18,  1.03it/s]
Loading 0:  61%|██████    | 307/507 [00:32<02:02,  1.63it/s]
Loading 0:  61%|██████    | 310/507 [00:32<01:33,  2.11it/s]
Loading 0:  62%|██████▏   | 314/507 [00:33<01:04,  3.01it/s]
Loading 0:  63%|██████▎   | 319/507 [00:33<00:41,  4.53it/s]
Loading 0:  64%|██████▍   | 324/507 [00:33<00:27,  6.54it/s]
Loading 0:  65%|██████▍   | 328/507 [00:33<00:21,  8.50it/s]
Loading 0:  65%|██████▌   | 332/507 [00:33<00:16, 10.88it/s]
Loading 0:  66%|██████▋   | 337/507 [00:33<00:11, 14.68it/s]
Loading 0:  67%|██████▋   | 341/507 [00:33<00:11, 14.66it/s]
Loading 0:  68%|██████▊   | 345/507 [00:33<00:09, 17.03it/s]
Loading 0:  69%|██████▉   | 349/507 [00:34<00:08, 19.45it/s]
Loading 0:  70%|██████▉   | 354/507 [00:34<00:06, 24.30it/s]
Loading 0:  71%|███████   | 358/507 [00:34<00:05, 26.28it/s]
Loading 0:  72%|███████▏  | 363/507 [00:34<00:04, 30.62it/s]
Loading 0:  72%|███████▏  | 367/507 [00:34<00:04, 31.12it/s]
Loading 0:  73%|███████▎  | 372/507 [00:34<00:03, 33.82it/s]
Loading 0:  74%|███████▍  | 376/507 [00:34<00:03, 32.91it/s]
Loading 0:  75%|███████▌  | 381/507 [00:34<00:03, 35.86it/s]
Loading 0:  76%|███████▌  | 385/507 [00:35<00:03, 35.54it/s]
Loading 0:  77%|███████▋  | 389/507 [00:35<00:03, 36.09it/s]
Loading 0:  78%|███████▊  | 393/507 [00:35<00:03, 34.99it/s]
Loading 0:  78%|███████▊  | 397/507 [00:35<00:04, 25.27it/s]
Loading 0:  79%|███████▉  | 401/507 [00:35<00:04, 24.50it/s]
Loading 0:  80%|████████  | 408/507 [00:35<00:03, 32.33it/s]
Loading 0:  81%|████████▏ | 412/507 [00:35<00:02, 31.69it/s]
Loading 0:  82%|████████▏ | 417/507 [00:36<00:02, 34.70it/s]
Loading 0:  83%|████████▎ | 421/507 [00:36<00:02, 33.99it/s]
Loading 0:  84%|████████▍ | 426/507 [00:36<00:02, 37.53it/s]
Loading 0:  85%|████████▍ | 430/507 [00:36<00:02, 36.83it/s]
Loading 0:  86%|████████▌ | 435/507 [00:36<00:01, 40.11it/s]
Loading 0:  87%|████████▋ | 440/507 [00:36<00:01, 40.72it/s]
Loading 0:  88%|████████▊ | 445/507 [00:36<00:01, 41.07it/s]
Loading 0:  89%|████████▉ | 451/507 [00:36<00:01, 40.75it/s]
Loading 0:  90%|████████▉ | 456/507 [00:39<00:07,  6.65it/s]
Loading 0:  91%|█████████ | 459/507 [00:39<00:07,  6.31it/s]
Loading 0:  91%|█████████▏| 463/507 [00:40<00:06,  6.65it/s]
Loading 0:  92%|█████████▏| 465/507 [00:40<00:06,  6.01it/s]
Loading 0:  93%|█████████▎| 470/507 [00:40<00:04,  8.83it/s]
Loading 0:  93%|█████████▎| 474/507 [00:41<00:03, 10.88it/s]
Loading 0:  95%|█████████▍| 481/507 [00:41<00:01, 16.76it/s]
Loading 0:  96%|█████████▌| 485/507 [00:41<00:01, 19.38it/s]
Loading 0:  97%|█████████▋| 490/507 [00:41<00:00, 23.66it/s]
Loading 0:  97%|█████████▋| 494/507 [00:41<00:00, 25.51it/s]
Loading 0:  98%|█████████▊| 499/507 [00:41<00:00, 29.04it/s]
Loading 0:  99%|█████████▉| 503/507 [00:41<00:00, 30.49it/s]
                                                            
Job chaiml-espresso-llama-24-9292-v1-mkmlizer completed after 153.95s with status: succeeded
Stopping job with name chaiml-espresso-llama-24-9292-v1-mkmlizer
Pipeline stage MKMLizer completed in 155.75s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.15s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-espresso-llama-24-9292-v1
Waiting for inference service chaiml-espresso-llama-24-9292-v1 to be ready
Inference service chaiml-espresso-llama-24-9292-v1 ready after 164.44255208969116s
Pipeline stage MKMLDeployer completed in 166.41s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.1250321865081787s
Received healthy response to inference request in 2.405688524246216s
Received healthy response to inference request in 2.3694283962249756s
Received healthy response to inference request in 2.8684210777282715s
Received healthy response to inference request in 2.295680284500122s
5 requests
0 failed requests
5th percentile: 2.310429906845093
10th percentile: 2.3251795291900637
20th percentile: 2.354678773880005
30th percentile: 2.3766804218292235
40th percentile: 2.3911844730377196
50th percentile: 2.405688524246216
60th percentile: 2.590781545639038
70th percentile: 2.77587456703186
80th percentile: 2.919743299484253
90th percentile: 3.022387742996216
95th percentile: 3.0737099647521973
99th percentile: 3.1147677421569826
mean time: 2.6128500938415526
Pipeline stage StressChecker completed in 14.75s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 2.40s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 2.32s
Shutdown handler de-registered
chaiml-espresso-llama-24_9292_v1 status is now deployed due to DeploymentManager action
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyScorer
Evaluating %s Family Friendly Score with %s threads
Pipeline stage OfflineFamilyFriendlyScorer completed in 3257.06s
Shutdown handler de-registered
chaiml-espresso-llama-24_9292_v1 status is now inactive due to auto deactivation removed underperforming models