unbabel-m-prometheus-14b

developer_uid: azuruce

submission_id: unbabel-m-prometheus-14b_v1

model_name: nis-best-config

model_group: Unbabel/M-Prometheus-14B

status: torndown

timestamp: 2025-04-11T20:40:22+00:00

num_battles: 6845

num_wins: 3014

celo_rating: 1230.12

family_friendly_score: 0.7644

family_friendly_standard_error: 0.006001543801389772

submission_type: basic

model_repo: Unbabel/M-Prometheus-14B

model_architecture: Qwen2ForCausalLM

model_num_parameters: 14769689600.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.4684389860796077, 'latency_mean': 2.134686167240143, 'latency_p50': 2.120203733444214, 'latency_p90': 2.3364407777786256}, {'batch_size': 3, 'throughput': 0.8648981461329157, 'latency_mean': 3.4595726561546325, 'latency_p50': 3.4676321744918823, 'latency_p90': 3.8121583223342896}, {'batch_size': 5, 'throughput': 1.0637580560327142, 'latency_mean': 4.6685516428947444, 'latency_p50': 4.645939350128174, 'latency_p90': 5.246165084838867}, {'batch_size': 6, 'throughput': 1.106975152175288, 'latency_mean': 5.379652343988418, 'latency_p50': 5.399034857749939, 'latency_p90': 6.156216382980347}, {'batch_size': 10, 'throughput': 1.2098203617934877, 'latency_mean': 8.219443836212157, 'latency_p50': 8.146091103553772, 'latency_p90': 9.195691013336182}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: nis-best-config

is_internal_developer: False

language_model: Unbabel/M-Prometheus-14B

model_size: 15B

ranking_group: single

throughput_3p7s: 0.92

us_pacific_date: 2025-04-11

win_ratio: 0.44032140248356466

generation_params: {'temperature': 0.45, 'top_p': 0.9, 'min_p': 0.05, 'top_k': 40, 'presence_penalty': 0.4, 'frequency_penalty': 0.4, 'stopping_words': ['<|im_end|>', '\n'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '<|system|>{memory}\n', 'prompt_template': '', 'bot_template': '<|im_start|>assistant\n{message}<|im_end|>\n', 'user_template': '<|im_start|>user\nYou:{message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n', '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 unbabel-m-prometheus-14b-v1-mkmlizer
Waiting for job on unbabel-m-prometheus-14b-v1-mkmlizer to finish
unbabel-m-prometheus-14b-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
unbabel-m-prometheus-14b-v1-mkmlizer: ║     _____            __           __                                ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║       /___/                                                         ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║                                                                     ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  Version: 0.12.8                                                    ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  https://mk1.ai                                                     ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║                                                                     ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  belonging to:                                                      ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║                                                                     ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  Chai Research Corp.                                                ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║  Expiration: 2025-04-15 23:59:59                                    ║
unbabel-m-prometheus-14b-v1-mkmlizer: ║                                                                     ║
unbabel-m-prometheus-14b-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
unbabel-m-prometheus-14b-v1-mkmlizer: Downloaded to shared memory in 44.863s
unbabel-m-prometheus-14b-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpulg_yr1h, device:0
unbabel-m-prometheus-14b-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
unbabel-m-prometheus-14b-v1-mkmlizer: quantized model in 38.019s
unbabel-m-prometheus-14b-v1-mkmlizer: Processed model Unbabel/M-Prometheus-14B in 82.882s
unbabel-m-prometheus-14b-v1-mkmlizer: creating bucket guanaco-mkml-models
unbabel-m-prometheus-14b-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
unbabel-m-prometheus-14b-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/config.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/added_tokens.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/added_tokens.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/special_tokens_map.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/tokenizer_config.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/merges.txt s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/merges.txt
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/vocab.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/vocab.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/tokenizer.json
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/flywheel_model.1.safetensors
unbabel-m-prometheus-14b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/unbabel-m-prometheus-14b-v1/flywheel_model.0.safetensors
unbabel-m-prometheus-14b-v1-mkmlizer: 
Loading 0:   0%|          | 0/579 [00:00<?, ?it/s]
Loading 0:   3%|▎         | 15/579 [00:00<00:04, 135.09it/s]
Loading 0:   5%|▌         | 29/579 [00:00<00:04, 123.93it/s]
Loading 0:   8%|▊         | 47/579 [00:00<00:03, 147.07it/s]
Loading 0:  11%|█         | 63/579 [00:00<00:03, 150.40it/s]
Loading 0:  14%|█▍        | 81/579 [00:00<00:05, 97.37it/s] 
Loading 0:  16%|█▌        | 94/579 [00:00<00:04, 101.32it/s]
Loading 0:  18%|█▊        | 107/579 [00:00<00:04, 105.41it/s]
Loading 0:  21%|██        | 120/579 [00:01<00:04, 108.50it/s]
Loading 0:  23%|██▎       | 136/579 [00:01<00:03, 119.62it/s]
Loading 0:  26%|██▌       | 151/579 [00:01<00:03, 125.58it/s]
Loading 0:  29%|██▊       | 166/579 [00:01<00:03, 123.56it/s]
Loading 0:  31%|███       | 179/579 [00:01<00:03, 118.83it/s]
Loading 0:  33%|███▎      | 192/579 [00:01<00:05, 74.61it/s] 
Loading 0:  36%|███▌      | 207/579 [00:01<00:04, 86.42it/s]
Loading 0:  38%|███▊      | 220/579 [00:02<00:03, 93.95it/s]
Loading 0:  40%|████      | 233/579 [00:02<00:03, 100.72it/s]
Loading 0:  43%|████▎     | 251/579 [00:02<00:02, 118.72it/s]
Loading 0:  46%|████▌     | 267/579 [00:02<00:02, 122.37it/s]
Loading 0:  49%|████▊     | 281/579 [00:02<00:02, 118.63it/s]
Loading 0:  51%|█████▏    | 297/579 [00:02<00:02, 94.37it/s] 
Loading 0:  53%|█████▎    | 308/579 [00:02<00:02, 91.66it/s]
Loading 0:  56%|█████▌    | 325/579 [00:03<00:02, 108.72it/s]
Loading 0:  59%|█████▊    | 339/579 [00:03<00:02, 115.94it/s]
Loading 0:  61%|██████    | 353/579 [00:03<00:01, 116.39it/s]
Loading 0:  64%|██████▍   | 373/579 [00:03<00:01, 137.56it/s]
Loading 0:  67%|██████▋   | 388/579 [00:03<00:01, 131.06it/s]
Loading 0:  70%|██████▉   | 405/579 [00:03<00:01, 100.73it/s]
Loading 0:  72%|███████▏  | 417/579 [00:03<00:01, 94.76it/s] 
Loading 0:  75%|███████▌  | 435/579 [00:03<00:01, 108.24it/s]
Loading 0:  77%|███████▋  | 448/579 [00:04<00:01, 111.68it/s]
Loading 0:  80%|███████▉  | 461/579 [00:04<00:01, 115.89it/s]
Loading 0:  83%|████████▎ | 479/579 [00:04<00:00, 131.54it/s]
Loading 0:  85%|████████▌ | 495/579 [00:04<00:00, 136.64it/s]
Loading 0:  88%|████████▊ | 512/579 [00:22<00:00, 136.64it/s]
Loading 0:  89%|████████▊ | 513/579 [00:27<00:27,  2.38it/s] 
Loading 0:  90%|████████▉ | 521/579 [00:27<00:20,  2.86it/s]
Loading 0:  92%|█████████▏| 534/579 [00:27<00:11,  3.97it/s]
Loading 0:  95%|█████████▍| 550/579 [00:27<00:04,  5.89it/s]
Loading 0:  98%|█████████▊| 565/579 [00:27<00:01,  8.39it/s]
                                                            
Job unbabel-m-prometheus-14b-v1-mkmlizer completed after 105.85s with status: succeeded
Stopping job with name unbabel-m-prometheus-14b-v1-mkmlizer
Pipeline stage MKMLizer completed in 106.39s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.19s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service unbabel-m-prometheus-14b-v1
Waiting for inference service unbabel-m-prometheus-14b-v1 to be ready
Inference service unbabel-m-prometheus-14b-v1 ready after 120.81824851036072s
Pipeline stage MKMLDeployer completed in 121.32s
run pipeline stage %s
Running pipeline stage StressChecker
HTTPConnectionPool(host='guanaco-submitter.guanaco-backend.k2.chaiverse.com', port=80): Read timed out. (read timeout=20)
Received unhealthy response to inference request!
Received healthy response to inference request in 2.5930349826812744s
Received healthy response to inference request in 2.196054458618164s
Received healthy response to inference request in 2.355677604675293s
Received healthy response to inference request in 2.1200356483459473s
5 requests
1 failed requests
5th percentile: 2.1352394104003904
10th percentile: 2.150443172454834
20th percentile: 2.180850696563721
30th percentile: 2.2279790878295898
40th percentile: 2.2918283462524416
50th percentile: 2.355677604675293
60th percentile: 2.4506205558776855
70th percentile: 2.545563507080078
80th percentile: 6.099495410919193
90th percentile: 13.112416267395021
95th percentile: 16.61887669563293
99th percentile: 19.424045038223266
mean time: 5.8780279636383055
%s, retrying in %s seconds...
Received healthy response to inference request in 1.6930506229400635s
Received healthy response to inference request in 1.9879770278930664s
Received healthy response to inference request in 2.1207313537597656s
Received healthy response to inference request in 2.145670175552368s
Received healthy response to inference request in 1.8692877292633057s
5 requests
0 failed requests
5th percentile: 1.728298044204712
10th percentile: 1.7635454654693603
20th percentile: 1.8340403079986571
30th percentile: 1.8930255889892578
40th percentile: 1.9405013084411622
50th percentile: 1.9879770278930664
60th percentile: 2.0410787582397463
70th percentile: 2.0941804885864257
80th percentile: 2.125719118118286
90th percentile: 2.135694646835327
95th percentile: 2.140682411193848
99th percentile: 2.1446726226806643
mean time: 1.9633433818817139
Pipeline stage StressChecker completed in 41.88s
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 0.71s
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 0.75s
Shutdown handler de-registered
unbabel-m-prometheus-14b_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 2805.28s
Shutdown handler de-registered
unbabel-m-prometheus-14b_v1 status is now inactive due to auto deactivation removed underperforming models
unbabel-m-prometheus-14b_v1 status is now torndown due to DeploymentManager action