rirv938-grpo-to-compare_30616

developer_uid: rirv938

submission_id: rirv938-grpo-to-compare_30616_v1

model_name: rirv938-grpo-to-compare_30616_v1

model_group: rirv938/grpo_to_compare_

status: torndown

timestamp: 2025-04-25T23:32:01+00:00

num_battles: 9021

num_wins: 4533

celo_rating: 1290.74

family_friendly_score: 0.5611999999999999

family_friendly_standard_error: 0.0070178994008178825

submission_type: basic

model_repo: rirv938/grpo_to_compare_to_dpo_10k_cp624_g16_merged

model_architecture: MistralForCausalLM

model_num_parameters: 24096691200.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.5301833495658825, 'latency_mean': 1.8860855531692504, 'latency_p50': 1.886960744857788, 'latency_p90': 2.079413318634033}, {'batch_size': 3, 'throughput': 1.076411796465262, 'latency_mean': 2.775205113887787, 'latency_p50': 2.7804938554763794, 'latency_p90': 3.0558087110519407}, {'batch_size': 5, 'throughput': 1.3837848566480195, 'latency_mean': 3.6000614082813263, 'latency_p50': 3.612500309944153, 'latency_p90': 4.024900460243225}, {'batch_size': 6, 'throughput': 1.4894904392120756, 'latency_mean': 3.9993526375293733, 'latency_p50': 3.970721960067749, 'latency_p90': 4.511315250396729}, {'batch_size': 8, 'throughput': 1.6214322968822192, 'latency_mean': 4.890961202383042, 'latency_p50': 4.878976464271545, 'latency_p90': 5.555827951431274}, {'batch_size': 10, 'throughput': 1.7193130128079412, 'latency_mean': 5.7686170852184295, 'latency_p50': 5.750702857971191, 'latency_p90': 6.514082551002502}]

gpu_counts: {'NVIDIA A100-SXM4-80GB': 1}

display_name: rirv938-grpo-to-compare_30616_v1

is_internal_developer: True

language_model: rirv938/grpo_to_compare_to_dpo_10k_cp624_g16_merged

model_size: 24B

ranking_group: single

throughput_3p7s: 1.42

us_pacific_date: 2025-04-25

win_ratio: 0.5024941802460925

generation_params: {'temperature': 0.9, 'top_p': 0.9, 'min_p': 0.2, 'top_k': 80, 'presence_penalty': 0.5, 'frequency_penalty': 0.5, 'stopping_words': ['You:', '###', '\n', '</s>'], '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 rirv938-grpo-to-compare-30616-v1-mkmlizer
Waiting for job on rirv938-grpo-to-compare-30616-v1-mkmlizer to finish
rirv938-grpo-to-compare-30616-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║     _____            __           __                                ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║       /___/                                                         ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║                                                                     ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  Version: 0.12.8                                                    ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  https://mk1.ai                                                     ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║                                                                     ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  belonging to:                                                      ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║                                                                     ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  Chai Research Corp.                                                ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ║                                                                     ║
rirv938-grpo-to-compare-30616-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rirv938-grpo-to-compare-30616-v1-mkmlizer: Downloaded to shared memory in 167.576s
rirv938-grpo-to-compare-30616-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpfhb2dhw5, device:0
rirv938-grpo-to-compare-30616-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rirv938-grpo-to-compare-30616-v1-mkmlizer: quantized model in 82.994s
rirv938-grpo-to-compare-30616-v1-mkmlizer: Processed model rirv938/grpo_to_compare_to_dpo_10k_cp624_g16_merged in 250.571s
rirv938-grpo-to-compare-30616-v1-mkmlizer: creating bucket guanaco-mkml-models
rirv938-grpo-to-compare-30616-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rirv938-grpo-to-compare-30616-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/config.json
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/special_tokens_map.json
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/tokenizer_config.json
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/tokenizer.json
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/flywheel_model.1.safetensors
rirv938-grpo-to-compare-30616-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rirv938-grpo-to-compare-30616-v1/flywheel_model.0.safetensors
rirv938-grpo-to-compare-30616-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|          | 3/363 [00:00<01:24,  4.25it/s]
Loading 0:   1%|          | 4/363 [00:01<01:55,  3.10it/s]
Loading 0:   1%|▏         | 5/363 [00:01<02:34,  2.31it/s]
Loading 0:   2%|▏         | 9/363 [00:02<01:02,  5.68it/s]
Loading 0:   3%|▎         | 11/363 [00:02<00:52,  6.76it/s]
Loading 0:   4%|▎         | 13/363 [00:02<01:14,  4.73it/s]
Loading 0:   4%|▍         | 14/363 [00:03<01:36,  3.60it/s]
Loading 0:   4%|▍         | 15/363 [00:04<02:04,  2.81it/s]
Loading 0:   5%|▌         | 19/363 [00:04<01:03,  5.44it/s]
Loading 0:   6%|▌         | 21/363 [00:04<01:12,  4.75it/s]
Loading 0:   6%|▌         | 22/363 [00:05<01:29,  3.80it/s]
Loading 0:   6%|▋         | 23/363 [00:05<01:50,  3.08it/s]
Loading 0:   7%|▋         | 27/363 [00:06<00:57,  5.85it/s]
Loading 0:   8%|▊         | 29/363 [00:06<00:46,  7.21it/s]
Loading 0:   9%|▉         | 32/363 [00:06<00:35,  9.29it/s]
Loading 0:   9%|▉         | 34/363 [00:06<00:43,  7.65it/s]
Loading 0:  10%|▉         | 36/363 [00:07<00:55,  5.91it/s]
Loading 0:  11%|█         | 39/363 [00:07<00:43,  7.39it/s]
Loading 0:  11%|█▏        | 41/363 [00:08<00:53,  6.01it/s]
Loading 0:  13%|█▎        | 46/363 [00:08<00:32,  9.81it/s]
Loading 0:  14%|█▍        | 50/363 [00:08<00:25, 12.50it/s]
Loading 0:  14%|█▍        | 52/363 [00:08<00:31,  9.93it/s]
Loading 0:  15%|█▍        | 54/363 [00:09<00:40,  7.65it/s]
Loading 0:  16%|█▌        | 57/363 [00:09<00:34,  8.85it/s]
Loading 0:  16%|█▋        | 59/363 [00:10<00:48,  6.33it/s]
Loading 0:  18%|█▊        | 64/363 [00:10<00:30,  9.91it/s]
Loading 0:  19%|█▊        | 68/363 [00:10<00:23, 12.44it/s]
Loading 0:  19%|█▉        | 70/363 [00:10<00:29,  9.78it/s]
Loading 0:  20%|█▉        | 72/363 [00:11<00:41,  7.02it/s]
Loading 0:  21%|██        | 75/363 [00:11<00:37,  7.78it/s]
Loading 0:  21%|██        | 77/363 [00:11<00:36,  7.75it/s]
Loading 0:  23%|██▎       | 82/363 [00:11<00:23, 12.22it/s]
Loading 0:  24%|██▎       | 86/363 [00:12<00:17, 15.52it/s]
Loading 0:  25%|██▍       | 89/363 [00:12<00:21, 12.85it/s]
Loading 0:  25%|██▌       | 91/363 [00:12<00:21, 12.70it/s]
Loading 0:  26%|██▌       | 95/363 [00:12<00:16, 16.46it/s]
Loading 0:  27%|██▋       | 99/363 [00:12<00:13, 19.81it/s]
Loading 0:  28%|██▊       | 102/363 [00:13<00:14, 18.47it/s]
Loading 0:  29%|██▉       | 105/363 [00:13<00:18, 13.96it/s]
Loading 0:  29%|██▉       | 107/363 [00:13<00:19, 13.23it/s]
Loading 0:  30%|███       | 109/363 [00:13<00:19, 13.07it/s]
Loading 0:  31%|███       | 111/363 [00:13<00:18, 13.93it/s]
Loading 0:  31%|███       | 113/363 [00:14<00:21, 11.41it/s]
Loading 0:  33%|███▎      | 118/363 [00:14<00:14, 17.49it/s]
Loading 0:  34%|███▎      | 122/363 [00:14<00:11, 20.99it/s]
Loading 0:  34%|███▍      | 125/363 [00:14<00:15, 15.05it/s]
Loading 0:  35%|███▌      | 128/363 [00:14<00:14, 15.94it/s]
Loading 0:  36%|███▌      | 130/363 [00:15<00:16, 13.90it/s]
Loading 0:  36%|███▋      | 132/363 [00:15<00:17, 13.46it/s]
Loading 0:  37%|███▋      | 136/363 [00:15<00:12, 17.59it/s]
Loading 0:  39%|███▊      | 140/363 [00:15<00:10, 21.25it/s]
Loading 0:  39%|███▉      | 143/363 [00:15<00:14, 14.92it/s]
Loading 0:  40%|███▉      | 145/363 [00:16<00:15, 14.07it/s]
Loading 0:  40%|████      | 147/363 [00:16<00:14, 14.86it/s]
Loading 0:  41%|████      | 149/363 [00:16<00:17, 12.10it/s]
Loading 0:  42%|████▏     | 154/363 [00:16<00:11, 18.28it/s]
Loading 0:  44%|████▎     | 158/363 [00:16<00:09, 21.93it/s]
Loading 0:  44%|████▍     | 161/363 [00:16<00:12, 15.60it/s]
Loading 0:  45%|████▌     | 164/363 [00:17<00:12, 16.42it/s]
Loading 0:  46%|████▌     | 167/363 [00:17<00:14, 13.15it/s]
Loading 0:  47%|████▋     | 172/363 [00:17<00:10, 18.48it/s]
Loading 0:  48%|████▊     | 176/363 [00:17<00:08, 21.93it/s]
Loading 0:  49%|████▉     | 179/363 [00:18<00:11, 16.07it/s]
Loading 0:  50%|█████     | 182/363 [00:18<00:10, 16.89it/s]
Loading 0:  51%|█████     | 185/363 [00:18<00:13, 13.34it/s]
Loading 0:  52%|█████▏    | 190/363 [00:18<00:09, 18.60it/s]
Loading 0:  53%|█████▎    | 194/363 [00:18<00:07, 22.03it/s]
Loading 0:  54%|█████▍    | 197/363 [00:19<00:10, 16.19it/s]
Loading 0:  55%|█████▌    | 200/363 [00:19<00:09, 16.83it/s]
Loading 0:  55%|█████▌    | 200/363 [00:30<00:09, 16.83it/s]
Loading 0:  55%|█████▌    | 201/363 [00:39<06:16,  2.32s/it]
Loading 0:  56%|█████▌    | 203/363 [00:39<04:47,  1.80s/it]
Loading 0:  57%|█████▋    | 207/363 [00:39<02:47,  1.07s/it]
Loading 0:  58%|█████▊    | 210/363 [00:40<01:56,  1.31it/s]
Loading 0:  59%|█████▊    | 213/363 [00:40<01:23,  1.79it/s]
Loading 0:  59%|█████▉    | 215/363 [00:40<01:07,  2.20it/s]
Loading 0:  60%|█████▉    | 217/363 [00:40<00:53,  2.74it/s]
Loading 0:  60%|██████    | 219/363 [00:40<00:40,  3.52it/s]
Loading 0:  61%|██████    | 221/363 [00:41<00:34,  4.15it/s]
Loading 0:  62%|██████▏   | 226/363 [00:41<00:18,  7.45it/s]
Loading 0:  63%|██████▎   | 230/363 [00:41<00:12, 10.36it/s]
Loading 0:  64%|██████▍   | 233/363 [00:41<00:13,  9.91it/s]
Loading 0:  65%|██████▌   | 236/363 [00:41<00:11, 11.35it/s]
Loading 0:  66%|██████▌   | 239/363 [00:42<00:12, 10.29it/s]
Loading 0:  67%|██████▋   | 244/363 [00:42<00:08, 14.86it/s]
Loading 0:  68%|██████▊   | 248/363 [00:42<00:06, 18.12it/s]
Loading 0:  69%|██████▉   | 251/363 [00:42<00:07, 14.17it/s]
Loading 0:  70%|██████▉   | 254/363 [00:42<00:07, 14.87it/s]
Loading 0:  71%|███████   | 257/363 [00:43<00:08, 12.04it/s]
Loading 0:  72%|███████▏  | 262/363 [00:43<00:06, 16.69it/s]
Loading 0:  73%|███████▎  | 266/363 [00:43<00:04, 19.78it/s]
Loading 0:  74%|███████▍  | 269/363 [00:43<00:06, 14.45it/s]
Loading 0:  75%|███████▍  | 272/363 [00:44<00:05, 15.17it/s]
Loading 0:  76%|███████▌  | 275/363 [00:44<00:07, 12.13it/s]
Loading 0:  77%|███████▋  | 280/363 [00:44<00:04, 16.73it/s]
Loading 0:  78%|███████▊  | 284/363 [00:44<00:04, 19.44it/s]
Loading 0:  79%|███████▉  | 287/363 [00:45<00:05, 14.11it/s]
Loading 0:  80%|███████▉  | 290/363 [00:45<00:04, 14.90it/s]
Loading 0:  80%|████████  | 292/363 [00:45<00:05, 13.46it/s]
Loading 0:  81%|████████  | 294/363 [00:45<00:05, 13.09it/s]
Loading 0:  82%|████████▏ | 298/363 [00:45<00:03, 17.08it/s]
Loading 0:  83%|████████▎ | 302/363 [00:45<00:02, 20.51it/s]
Loading 0:  84%|████████▍ | 305/363 [00:46<00:04, 14.06it/s]
Loading 0:  85%|████████▍ | 307/363 [00:46<00:04, 13.51it/s]
Loading 0:  85%|████████▌ | 309/363 [00:46<00:03, 14.44it/s]
Loading 0:  86%|████████▌ | 311/363 [00:46<00:04, 12.01it/s]
Loading 0:  87%|████████▋ | 316/363 [00:46<00:02, 18.45it/s]
Loading 0:  88%|████████▊ | 320/363 [00:47<00:01, 22.40it/s]
Loading 0:  89%|████████▉ | 323/363 [00:47<00:02, 15.18it/s]
Loading 0:  90%|████████▉ | 326/363 [00:47<00:02, 15.79it/s]
Loading 0:  91%|█████████ | 329/363 [00:47<00:02, 12.26it/s]
Loading 0:  92%|█████████▏| 334/363 [00:48<00:01, 17.41it/s]
Loading 0:  93%|█████████▎| 338/363 [00:48<00:01, 21.02it/s]
Loading 0:  94%|█████████▍| 341/363 [00:48<00:01, 15.08it/s]
Loading 0:  95%|█████████▍| 344/363 [00:48<00:01, 16.02it/s]
Loading 0:  96%|█████████▌| 347/363 [00:49<00:01, 13.07it/s]
Loading 0:  97%|█████████▋| 352/363 [00:49<00:00, 18.32it/s]
Loading 0:  98%|█████████▊| 356/363 [00:49<00:00, 21.86it/s]
Loading 0:  99%|█████████▉| 360/363 [00:57<00:01,  1.57it/s]
Loading 0: 100%|██████████| 363/363 [00:57<00:00,  2.03it/s]
                                                            
Job rirv938-grpo-to-compare-30616-v1-mkmlizer completed after 278.93s with status: succeeded
Stopping job with name rirv938-grpo-to-compare-30616-v1-mkmlizer
Pipeline stage MKMLizer completed in 279.59s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.17s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rirv938-grpo-to-compare-30616-v1
Waiting for inference service rirv938-grpo-to-compare-30616-v1 to be ready
Inference service rirv938-grpo-to-compare-30616-v1 ready after 140.72149300575256s
Pipeline stage MKMLDeployer completed in 141.40s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.5358331203460693s
Received healthy response to inference request in 1.844144344329834s
Received healthy response to inference request in 1.873417615890503s
Received healthy response to inference request in 1.8361003398895264s
Received healthy response to inference request in 2.222938299179077s
5 requests
0 failed requests
5th percentile: 1.8377091407775878
10th percentile: 1.8393179416656493
20th percentile: 1.8425355434417725
30th percentile: 1.8499989986419678
40th percentile: 1.8617083072662353
50th percentile: 1.873417615890503
60th percentile: 2.0132258892059327
70th percentile: 2.153034162521362
80th percentile: 2.285517263412476
90th percentile: 2.4106751918792724
95th percentile: 2.4732541561126706
99th percentile: 2.5233173274993894
mean time: 2.062486743927002
Pipeline stage StressChecker completed in 11.66s
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 0.87s
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 0.74s
Shutdown handler de-registered
rirv938-grpo-to-compare_30616_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
%s, retrying in %s seconds...
Evaluating %s Family Friendly Score with %s threads
Pipeline stage OfflineFamilyFriendlyScorer completed in 4739.92s
Shutdown handler de-registered
rirv938-grpo-to-compare_30616_v1 status is now inactive due to auto deactivation removed underperforming models
rirv938-grpo-to-compare_30616_v1 status is now torndown due to DeploymentManager action