chaiml-exp-grpo-cp312-_36146

developer_uid: NischayDnk

submission_id: chaiml-exp-grpo-cp312-_36146_v28

model_name: chaiml-exp-grpo-cp312-_36146_v28

model_group: ChaiML/exp_grpo_cp312_96

status: torndown

timestamp: 2025-09-11T21:48:51+00:00

num_battles: 6561

num_wins: 3714

celo_rating: 1305.82

family_friendly_score: 0.518

family_friendly_standard_error: 0.0070664842743757665

submission_type: basic

model_repo: ChaiML/exp_grpo_cp312_96ff_b6_r2_less_ff_merged

model_architecture: MistralForCausalLM

model_num_parameters: 24096691200.0

best_of: 7

max_input_tokens: 1280

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.42466702256327254, 'latency_mean': 2.3546642112731933, 'latency_p50': 2.3505442142486572, 'latency_p90': 2.596436047554016}, {'batch_size': 3, 'throughput': 0.8759633763158822, 'latency_mean': 3.4185243642330168, 'latency_p50': 3.4144139289855957, 'latency_p90': 3.8057310342788697}, {'batch_size': 5, 'throughput': 1.1418263006556835, 'latency_mean': 4.363410176038742, 'latency_p50': 4.3832350969314575, 'latency_p90': 4.87439374923706}, {'batch_size': 6, 'throughput': 1.2311389783058786, 'latency_mean': 4.834949328899383, 'latency_p50': 4.832029461860657, 'latency_p90': 5.456196045875549}, {'batch_size': 8, 'throughput': 1.3603613609059202, 'latency_mean': 5.847718056440353, 'latency_p50': 5.85704505443573, 'latency_p90': 6.720157718658447}, {'batch_size': 10, 'throughput': 1.4669970374561188, 'latency_mean': 6.7939623260498045, 'latency_p50': 6.795201897621155, 'latency_p90': 7.877695560455322}]

gpu_counts: {'NVIDIA L40S': 1}

display_name: chaiml-exp-grpo-cp312-_36146_v28

is_internal_developer: False

language_model: ChaiML/exp_grpo_cp312_96ff_b6_r2_less_ff_merged

model_size: 24B

ranking_group: single

throughput_3p7s: 0.97

us_pacific_date: 2025-09-11

win_ratio: 0.5660722450845908

generation_params: {'temperature': 0.9, 'top_p': 0.95, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.45, 'frequency_penalty': 0.45, 'stopping_words': ['\n'], 'max_input_tokens': 1280, 'best_of': 7, 'max_output_tokens': 64}

formatter: {'memory_template': "{bot_name}'s Persona: {memory}\n####\n", '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-exp-grpo-cp312-36146-v28-mkmlizer
Waiting for job on chaiml-exp-grpo-cp312-36146-v28-mkmlizer to finish
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║                                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   ██████   ██████  █████   ████  ████                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  ░░██████ ██████  ░░███   ███░  ░░███                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   ░███░█████░███   ░███  ███     ░███                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   ░███░░███ ░███   ░███████      ░███                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   ░███ ░░░  ░███   ░███░░███     ░███                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   ░███      ░███   ░███ ░░███    ░███                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║   █████     █████  █████ ░░████  █████                              ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  ░░░░░     ░░░░░  ░░░░░   ░░░░  ░░░░░                               ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║                                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Version: 0.30.2                                                    ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Features: FLYWHEEL, CUDA                                           ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Copyright 2023-2025 MK ONE TECHNOLOGIES Inc.                       ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║                                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  belonging to:                                                      ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║                                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ║                                                                     ║
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Failed to get response for submission mistralai-mistral-nem_93303_v569: ('http://mistralai-mistral-nem-93303-v569-predictor.tenant-chaiml-guanaco.k2.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: Downloaded to shared memory in 100.417s
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: Checking if ChaiML/exp_grpo_cp312_96ff_b6_r2_less_ff_merged already exists in ChaiML
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpgusn59u5, device:0
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: quantized model in 50.117s
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: Processed model ChaiML/exp_grpo_cp312_96ff_b6_r2_less_ff_merged in 150.535s
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: creating bucket guanaco-mkml-models
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/special_tokens_map.json
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/config.json
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/tokenizer_config.json
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/tokenizer.json
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/flywheel_model.1.safetensors
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-exp-grpo-cp312-36146-v28/nvidia/flywheel_model.0.safetensors
chaiml-exp-grpo-cp312-36146-v28-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|          | 4/363 [00:00<00:15, 23.56it/s]
Loading 0:   2%|▏         | 7/363 [00:00<00:18, 19.58it/s]
Loading 0:   3%|▎         | 11/363 [00:00<00:14, 24.73it/s]
Loading 0:   4%|▍         | 14/363 [00:00<00:22, 15.76it/s]
Loading 0:   4%|▍         | 16/363 [00:00<00:22, 15.16it/s]
Loading 0:   6%|▌         | 21/363 [00:01<00:17, 19.77it/s]
Loading 0:   7%|▋         | 24/363 [00:01<00:19, 17.18it/s]
Loading 0:   8%|▊         | 30/363 [00:01<00:13, 24.90it/s]
Loading 0:   9%|▉         | 34/363 [00:01<00:15, 20.62it/s]
Loading 0:  10%|█         | 37/363 [00:01<00:18, 18.04it/s]
Loading 0:  11%|█         | 40/363 [00:02<00:18, 17.72it/s]
Loading 0:  12%|█▏        | 43/363 [00:02<00:18, 17.59it/s]
Loading 0:  13%|█▎        | 48/363 [00:02<00:13, 23.54it/s]
Loading 0:  14%|█▍        | 51/363 [00:02<00:14, 21.06it/s]
Loading 0:  15%|█▍        | 54/363 [00:02<00:19, 15.96it/s]
Loading 0:  16%|█▌        | 58/363 [00:03<00:17, 17.70it/s]
Loading 0:  17%|█▋        | 61/363 [00:03<00:16, 18.03it/s]
Loading 0:  18%|█▊        | 65/363 [00:03<00:13, 21.99it/s]
Loading 0:  19%|█▉        | 69/363 [00:03<00:13, 21.29it/s]
Loading 0:  20%|█▉        | 72/363 [00:03<00:17, 16.17it/s]
Loading 0:  21%|██        | 76/363 [00:04<00:16, 17.80it/s]
Loading 0:  22%|██▏       | 79/363 [00:04<00:15, 18.22it/s]
Loading 0:  23%|██▎       | 85/363 [00:04<00:10, 25.50it/s]
Loading 0:  25%|██▍       | 89/363 [00:04<00:14, 18.74it/s]
Loading 0:  25%|██▌       | 92/363 [00:04<00:14, 18.42it/s]
Loading 0:  27%|██▋       | 98/363 [00:04<00:10, 25.11it/s]
Loading 0:  28%|██▊       | 102/363 [00:05<00:11, 22.66it/s]
Loading 0:  29%|██▉       | 105/363 [00:05<00:14, 18.34it/s]
Loading 0:  30%|██▉       | 108/363 [00:05<00:17, 14.55it/s]
Loading 0:  31%|███       | 112/363 [00:05<00:15, 16.16it/s]
Loading 0:  31%|███▏      | 114/363 [00:06<00:16, 15.49it/s]
Loading 0:  33%|███▎      | 119/363 [00:06<00:11, 21.15it/s]
Loading 0:  34%|███▍      | 123/363 [00:06<00:11, 20.86it/s]
Loading 0:  35%|███▍      | 126/363 [00:06<00:14, 15.80it/s]
Loading 0:  36%|███▌      | 130/363 [00:06<00:13, 17.54it/s]
Loading 0:  37%|███▋      | 133/363 [00:07<00:13, 17.65it/s]
Loading 0:  38%|███▊      | 138/363 [00:07<00:09, 23.40it/s]
Loading 0:  39%|███▉      | 141/363 [00:07<00:10, 21.17it/s]
Loading 0:  40%|███▉      | 144/363 [00:07<00:13, 16.07it/s]
Loading 0:  41%|████      | 148/363 [00:07<00:11, 18.00it/s]
Loading 0:  42%|████▏     | 151/363 [00:07<00:11, 18.35it/s]
Loading 0:  43%|████▎     | 157/363 [00:08<00:08, 25.67it/s]
Loading 0:  44%|████▍     | 161/363 [00:08<00:10, 19.43it/s]
Loading 0:  45%|████▌     | 164/363 [00:08<00:10, 19.75it/s]
Loading 0:  46%|████▌     | 167/363 [00:08<00:12, 15.54it/s]
Loading 0:  48%|████▊     | 174/363 [00:08<00:07, 23.69it/s]
Loading 0:  49%|████▉     | 178/363 [00:09<00:08, 20.58it/s]
Loading 0:  50%|████▉     | 181/363 [00:09<00:10, 17.91it/s]
Loading 0:  51%|█████     | 184/363 [00:09<00:10, 17.68it/s]
Loading 0:  52%|█████▏    | 187/363 [00:09<00:09, 18.06it/s]
Loading 0:  53%|█████▎    | 192/363 [00:09<00:07, 23.82it/s]
Loading 0:  54%|█████▎    | 195/363 [00:10<00:07, 21.36it/s]
Loading 0:  55%|█████▍    | 198/363 [00:10<00:10, 16.10it/s]
Loading 0:  55%|█████▌    | 200/363 [00:23<00:10, 16.10it/s]
Loading 0:  55%|█████▌    | 201/363 [00:23<03:18,  1.23s/it]
Loading 0:  56%|█████▌    | 203/363 [00:23<02:41,  1.01s/it]
Loading 0:  57%|█████▋    | 207/363 [00:23<01:40,  1.55it/s]
Loading 0:  58%|█████▊    | 211/363 [00:24<01:05,  2.32it/s]
Loading 0:  59%|█████▉    | 214/363 [00:24<00:50,  2.95it/s]
Loading 0:  60%|█████▉    | 217/363 [00:24<00:39,  3.74it/s]
Loading 0:  61%|██████    | 220/363 [00:24<00:29,  4.78it/s]
Loading 0:  61%|██████    | 222/363 [00:24<00:25,  5.49it/s]
Loading 0:  63%|██████▎   | 227/363 [00:25<00:15,  8.92it/s]
Loading 0:  64%|██████▎   | 231/363 [00:25<00:12, 10.94it/s]
Loading 0:  64%|██████▍   | 234/363 [00:25<00:12, 10.36it/s]
Loading 0:  66%|██████▌   | 238/363 [00:25<00:09, 12.53it/s]
Loading 0:  66%|██████▋   | 241/363 [00:25<00:08, 13.63it/s]
Loading 0:  68%|██████▊   | 246/363 [00:25<00:06, 18.89it/s]
Loading 0:  69%|██████▊   | 249/363 [00:26<00:06, 18.37it/s]
Loading 0:  69%|██████▉   | 252/363 [00:26<00:07, 14.38it/s]
Loading 0:  71%|███████   | 256/363 [00:26<00:06, 16.13it/s]
Loading 0:  71%|███████▏  | 259/363 [00:26<00:06, 16.71it/s]
Loading 0:  73%|███████▎  | 264/363 [00:26<00:04, 22.45it/s]
Loading 0:  74%|███████▎  | 267/363 [00:27<00:04, 20.09it/s]
Loading 0:  74%|███████▍  | 270/363 [00:27<00:06, 15.21it/s]
Loading 0:  75%|███████▌  | 274/363 [00:27<00:05, 16.81it/s]
Loading 0:  76%|███████▋  | 277/363 [00:27<00:05, 17.16it/s]
Loading 0:  78%|███████▊  | 282/363 [00:27<00:03, 22.88it/s]
Loading 0:  79%|███████▊  | 285/363 [00:28<00:03, 20.78it/s]
Loading 0:  79%|███████▉  | 288/363 [00:28<00:04, 15.28it/s]
Loading 0:  80%|████████  | 292/363 [00:28<00:04, 16.88it/s]
Loading 0:  81%|████████▏ | 295/363 [00:28<00:03, 17.19it/s]
Loading 0:  83%|████████▎ | 300/363 [00:28<00:02, 22.80it/s]
Loading 0:  83%|████████▎ | 303/363 [00:29<00:02, 20.82it/s]
Loading 0:  84%|████████▍ | 306/363 [00:29<00:03, 15.49it/s]
Loading 0:  85%|████████▌ | 310/363 [00:29<00:03, 17.04it/s]
Loading 0:  86%|████████▌ | 313/363 [00:29<00:02, 17.04it/s]
Loading 0:  88%|████████▊ | 318/363 [00:29<00:01, 22.69it/s]
Loading 0:  88%|████████▊ | 321/363 [00:30<00:02, 20.43it/s]
Loading 0:  89%|████████▉ | 324/363 [00:30<00:02, 15.32it/s]
Loading 0:  90%|█████████ | 328/363 [00:30<00:02, 16.88it/s]
Loading 0:  91%|█████████ | 331/363 [00:30<00:01, 17.14it/s]
Loading 0:  93%|█████████▎| 336/363 [00:30<00:01, 22.89it/s]
Loading 0:  93%|█████████▎| 339/363 [00:31<00:01, 20.77it/s]
Loading 0:  94%|█████████▍| 342/363 [00:31<00:01, 15.39it/s]
Loading 0:  95%|█████████▌| 346/363 [00:31<00:01, 16.99it/s]
Loading 0:  96%|█████████▌| 349/363 [00:31<00:00, 17.32it/s]
Loading 0:  98%|█████████▊| 355/363 [00:31<00:00, 24.47it/s]
Loading 0:  99%|█████████▉| 359/363 [00:32<00:00, 13.50it/s]
Loading 0: 100%|█████████▉| 362/363 [00:32<00:00, 13.16it/s]
                                                            
Job chaiml-exp-grpo-cp312-36146-v28-mkmlizer completed after 235.78s with status: succeeded
Stopping job with name chaiml-exp-grpo-cp312-36146-v28-mkmlizer
Pipeline stage MKMLizer completed in 236.31s
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-exp-grpo-cp312-36146-v28
Waiting for inference service chaiml-exp-grpo-cp312-36146-v28 to be ready
Inference service chaiml-exp-grpo-cp312-36146-v28 ready after 40.23371410369873s
Pipeline stage MKMLDeployer completed in 40.77s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.6392431259155273s
Received healthy response to inference request in 2.3443596363067627s
Received healthy response to inference request in 2.4713213443756104s
Received healthy response to inference request in 2.2004685401916504s
Received healthy response to inference request in 2.1092100143432617s
5 requests
0 failed requests
5th percentile: 2.1274617195129393
10th percentile: 2.1457134246826173
20th percentile: 2.182216835021973
30th percentile: 2.2292467594146728
40th percentile: 2.286803197860718
50th percentile: 2.3443596363067627
60th percentile: 2.395144319534302
70th percentile: 2.445929002761841
80th percentile: 2.5049057006835938
90th percentile: 2.5720744132995605
95th percentile: 2.605658769607544
99th percentile: 2.6325262546539308
mean time: 2.3529205322265625
Pipeline stage StressChecker completed in 13.02s
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 0.61s
Shutdown handler de-registered
chaiml-exp-grpo-cp312-_36146_v28 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 2937.37s
Shutdown handler de-registered
chaiml-exp-grpo-cp312-_36146_v28 status is now inactive due to auto deactivation removed underperforming models
chaiml-exp-grpo-cp312-_36146_v28 status is now torndown due to DeploymentManager action