chaiml-ir-v3-3-dpo-irv1_78559

developer_uid: chai_backend_admin

submission_id: chaiml-ir-v3-3-dpo-irv1_78559_v1

model_name: training123

model_group: ChaiML/ir-v3-3-dpo_irv13

status: torndown

timestamp: 2025-11-17T12:26:15+00:00

num_battles: 6948

num_wins: 3724

celo_rating: 1305.38

family_friendly_score: 0.44799999999999995

family_friendly_standard_error: 0.007032723512267491

submission_type: basic

model_repo: ChaiML/ir-v3-3-dpo_irv13-cosine

model_architecture: MistralForCausalLM

model_num_parameters: 24096691200.0

best_of: 8

max_input_tokens: 2048

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.26788041318435113, 'latency_mean': 3.7329055964946747, 'latency_p50': 2.8257442712783813, 'latency_p90': 5.798909926414489}, {'batch_size': 2, 'throughput': 0.5660692497240388, 'latency_mean': 3.5247074341773987, 'latency_p50': 3.5213634967803955, 'latency_p90': 3.8781903743743897}, {'batch_size': 3, 'throughput': 0.700462986253364, 'latency_mean': 4.272955170869827, 'latency_p50': 4.283927321434021, 'latency_p90': 4.72576630115509}, {'batch_size': 4, 'throughput': 0.7908405796116611, 'latency_mean': 5.0447333824634555, 'latency_p50': 5.029462456703186, 'latency_p90': 5.699724054336547}, {'batch_size': 5, 'throughput': 0.845455517051371, 'latency_mean': 5.885792486667633, 'latency_p50': 5.91865336894989, 'latency_p90': 6.711664891242981}]

gpu_counts: {'NVIDIA L40S': 1}

display_name: training123

is_internal_developer: True

language_model: ChaiML/ir-v3-3-dpo_irv13-cosine

model_size: 24B

ranking_group: single

throughput_3p7s: 0.27

us_pacific_date: 2025-11-14

win_ratio: 0.5359815774323546

generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['</s>', '\n', 'You:', 'User:'], 'max_input_tokens': 2048, 'best_of': 8, 'max_output_tokens': 64}

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': True}

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-ir-v3-3-dpo-irv1-78559-v1-mkmlizer
Waiting for job on chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer to finish
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   ██████   ██████  █████   ████  ████                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  ░░██████ ██████  ░░███   ███░  ░░███                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   ░███░█████░███   ░███  ███     ░███                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   ░███░░███ ░███   ░███████      ░███                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   ░███ ░░░  ░███   ░███░░███     ░███                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   ░███      ░███   ░███ ░░███    ░███                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║   █████     █████  █████ ░░████  █████                              ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  ░░░░░     ░░░░░  ░░░░░   ░░░░  ░░░░░                               ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Version: 0.30.2                                                    ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Features: FLYWHEEL, CUDA                                           ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Copyright 2023-2025 MK ONE TECHNOLOGIES Inc.                       ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  belonging to:                                                      ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: Downloaded to shared memory in 62.691s
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: Checking if ChaiML/ir-v3-3-dpo_irv13-cosine already exists in ChaiML
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp3hd9j3a0, device:0
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: quantized model in 42.095s
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: Processed model ChaiML/ir-v3-3-dpo_irv13-cosine in 104.787s
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: creating bucket guanaco-mkml-models
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/config.json
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/special_tokens_map.json
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/tokenizer_config.json
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/tokenizer.json
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/flywheel_model.0.safetensors
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-3-dpo-irv1-78559-v1/nvidia/flywheel_model.1.safetensors
chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:16, 22.32it/s]
Loading 0:   3%|▎         | 12/363 [00:00<00:08, 40.85it/s]
Loading 0:   5%|▍         | 18/363 [00:00<00:08, 41.31it/s]
Loading 0:   6%|▋         | 23/363 [00:00<00:10, 32.39it/s]
Loading 0:   9%|▉         | 32/363 [00:00<00:07, 44.24it/s]
Loading 0:  10%|█         | 38/363 [00:01<00:10, 29.87it/s]
Loading 0:  12%|█▏        | 42/363 [00:01<00:11, 27.18it/s]
Loading 0:  13%|█▎        | 48/363 [00:01<00:09, 32.42it/s]
Loading 0:  14%|█▍        | 52/363 [00:01<00:09, 31.12it/s]
Loading 0:  16%|█▌        | 57/363 [00:01<00:08, 34.72it/s]
Loading 0:  17%|█▋        | 61/363 [00:01<00:09, 32.49it/s]
Loading 0:  18%|█▊        | 65/363 [00:01<00:08, 33.47it/s]
Loading 0:  19%|█▉        | 70/363 [00:02<00:11, 26.23it/s]
Loading 0:  20%|██        | 74/363 [00:02<00:12, 23.40it/s]
Loading 0:  21%|██▏       | 78/363 [00:02<00:10, 26.33it/s]
Loading 0:  23%|██▎       | 82/363 [00:02<00:13, 21.46it/s]
Loading 0:  23%|██▎       | 85/363 [00:02<00:13, 20.86it/s]
Loading 0:  24%|██▍       | 88/363 [00:03<00:15, 17.49it/s]
Loading 0:  25%|██▌       | 92/363 [00:03<00:15, 17.55it/s]
Loading 0:  27%|██▋       | 99/363 [00:03<00:10, 25.59it/s]
Loading 0:  28%|██▊       | 103/363 [00:03<00:10, 25.81it/s]
Loading 0:  29%|██▉       | 107/363 [00:03<00:10, 24.61it/s]
Loading 0:  31%|███       | 112/363 [00:04<00:09, 27.63it/s]
Loading 0:  32%|███▏      | 116/363 [00:04<00:08, 28.96it/s]
Loading 0:  33%|███▎      | 120/363 [00:04<00:07, 31.38it/s]
Loading 0:  34%|███▍      | 124/363 [00:04<00:07, 30.31it/s]
Loading 0:  36%|███▌      | 129/363 [00:04<00:06, 33.95it/s]
Loading 0:  37%|███▋      | 133/363 [00:04<00:07, 31.91it/s]
Loading 0:  38%|███▊      | 138/363 [00:04<00:06, 35.48it/s]
Loading 0:  39%|███▉      | 142/363 [00:04<00:06, 33.13it/s]
Loading 0:  41%|████      | 149/363 [00:05<00:05, 39.73it/s]
Loading 0:  42%|████▏     | 154/363 [00:05<00:07, 27.33it/s]
Loading 0:  44%|████▎     | 158/363 [00:05<00:08, 24.92it/s]
Loading 0:  45%|████▌     | 165/363 [00:05<00:06, 32.07it/s]
Loading 0:  47%|████▋     | 169/363 [00:05<00:06, 31.28it/s]
Loading 0:  48%|████▊     | 174/363 [00:05<00:05, 34.41it/s]
Loading 0:  49%|████▉     | 178/363 [00:06<00:05, 32.54it/s]
Loading 0:  50%|█████     | 182/363 [00:06<00:05, 33.42it/s]
Loading 0:  52%|█████▏    | 187/363 [00:06<00:05, 30.90it/s]
Loading 0:  53%|█████▎    | 191/363 [00:06<00:05, 29.91it/s]
Loading 0:  54%|█████▎    | 195/363 [00:06<00:06, 26.86it/s]
Loading 0:  55%|█████▌    | 201/363 [00:19<02:12,  1.23it/s]
Loading 0:  56%|█████▌    | 203/363 [00:19<01:53,  1.41it/s]
Loading 0:  57%|█████▋    | 208/363 [00:19<01:12,  2.14it/s]
Loading 0:  58%|█████▊    | 211/363 [00:19<00:56,  2.70it/s]
Loading 0:  59%|█████▉    | 215/363 [00:20<00:39,  3.76it/s]
Loading 0:  60%|██████    | 219/363 [00:20<00:28,  5.02it/s]
Loading 0:  62%|██████▏   | 224/363 [00:20<00:19,  6.98it/s]
Loading 0:  63%|██████▎   | 229/363 [00:20<00:13,  9.60it/s]
Loading 0:  64%|██████▍   | 233/363 [00:20<00:10, 11.96it/s]
Loading 0:  66%|██████▌   | 238/363 [00:20<00:08, 15.26it/s]
Loading 0:  67%|██████▋   | 242/363 [00:20<00:06, 18.14it/s]
Loading 0:  68%|██████▊   | 246/363 [00:21<00:06, 18.89it/s]
Loading 0:  69%|██████▊   | 249/363 [00:21<00:05, 19.26it/s]
Loading 0:  70%|███████   | 255/363 [00:21<00:04, 25.47it/s]
Loading 0:  71%|███████▏  | 259/363 [00:21<00:03, 26.41it/s]
Loading 0:  72%|███████▏  | 263/363 [00:21<00:03, 26.00it/s]
Loading 0:  74%|███████▎  | 267/363 [00:21<00:03, 24.32it/s]
Loading 0:  74%|███████▍  | 270/363 [00:22<00:04, 21.09it/s]
Loading 0:  76%|███████▌  | 275/363 [00:22<00:03, 23.00it/s]
Loading 0:  78%|███████▊  | 282/363 [00:22<00:02, 30.86it/s]
Loading 0:  79%|███████▉  | 286/363 [00:22<00:02, 30.26it/s]
Loading 0:  80%|████████  | 291/363 [00:22<00:02, 33.61it/s]
Loading 0:  81%|████████▏ | 295/363 [00:22<00:02, 32.50it/s]
Loading 0:  82%|████████▏ | 299/363 [00:22<00:01, 33.29it/s]
Loading 0:  84%|████████▎ | 304/363 [00:23<00:01, 30.32it/s]
Loading 0:  85%|████████▍ | 308/363 [00:23<00:01, 29.62it/s]
Loading 0:  86%|████████▌ | 312/363 [00:23<00:01, 26.68it/s]
Loading 0:  88%|████████▊ | 318/363 [00:23<00:01, 32.27it/s]
Loading 0:  89%|████████▊ | 322/363 [00:23<00:01, 31.15it/s]
Loading 0:  90%|█████████ | 327/363 [00:23<00:01, 34.59it/s]
Loading 0:  91%|█████████ | 331/363 [00:23<00:00, 32.48it/s]
Loading 0:  92%|█████████▏| 335/363 [00:24<00:00, 33.43it/s]
Loading 0:  93%|█████████▎| 339/363 [00:24<00:00, 34.26it/s]
Loading 0:  94%|█████████▍| 343/363 [00:24<00:00, 20.62it/s]
Loading 0:  96%|█████████▌| 348/363 [00:24<00:00, 22.22it/s]
Loading 0:  98%|█████████▊| 355/363 [00:24<00:00, 29.80it/s]
Loading 0:  99%|█████████▉| 359/363 [00:24<00:00, 28.91it/s]
                                                            
Job chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer completed after 154.62s with status: succeeded
Stopping job with name chaiml-ir-v3-3-dpo-irv1-78559-v1-mkmlizer
Pipeline stage MKMLizer completed in 155.13s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.16s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-ir-v3-3-dpo-irv1-78559-v1
Waiting for inference service chaiml-ir-v3-3-dpo-irv1-78559-v1 to be ready
Retrying (%r) after connection broken by '%r': %s
Retrying (%r) after connection broken by '%r': %s
Retrying (%r) after connection broken by '%r': %s
Retrying (%r) after connection broken by '%r': %s
Retrying (%r) after connection broken by '%r': %s
Inference service chaiml-ir-v3-3-dpo-irv1-78559-v1 ready after 150.56500482559204s
Pipeline stage MKMLDeployer completed in 151.17s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.8431131839752197s
Received healthy response to inference request in 2.3539814949035645s
Received healthy response to inference request in 2.4880623817443848s
Received healthy response to inference request in 2.4698591232299805s
Received healthy response to inference request in 2.2066760063171387s
5 requests
0 failed requests
5th percentile: 2.236137104034424
10th percentile: 2.265598201751709
20th percentile: 2.324520397186279
30th percentile: 2.3771570205688475
40th percentile: 2.423508071899414
50th percentile: 2.4698591232299805
60th percentile: 2.477140426635742
70th percentile: 2.484421730041504
80th percentile: 2.559072542190552
90th percentile: 2.7010928630828857
95th percentile: 2.7721030235290525
99th percentile: 2.828911151885986
mean time: 2.4723384380340576
Pipeline stage StressChecker completed in 13.66s
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 0.69s
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 0.62s
Shutdown handler de-registered
chaiml-ir-v3-3-dpo-irv1_78559_v1 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 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.09s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler
Waiting for inference service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler to be ready
Inference service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler ready after 151.06487250328064s
Pipeline stage MKMLProfilerDeployer completed in 151.67s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/chaiml-ir-v3-3-dpo-i2b1d878029abfce89e48dc279ee63e01-deplojzzxc:/code/chaiverse_profiler_1763117154 --namespace tenant-chaiml-guanaco
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-3-dpo-i2b1d878029abfce89e48dc279ee63e01-deplojzzxc --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1763117154 && python profiles.py profile --best_of_n 8 --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 2048 --output_tokens 64 --summary /code/chaiverse_profiler_1763117154/summary.json'
%s, retrying in %s seconds...
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/chaiml-ir-v3-3-dpo-i2b1d878029abfce89e48dc279ee63e01-deplojzzxc:/code/chaiverse_profiler_1763118445 --namespace tenant-chaiml-guanaco
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-3-dpo-i2b1d878029abfce89e48dc279ee63e01-deplojzzxc --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1763118445 && python profiles.py profile --best_of_n 8 --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 2048 --output_tokens 64 --summary /code/chaiverse_profiler_1763118445/summary.json'
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-3-dpo-i2b1d878029abfce89e48dc279ee63e01-deplojzzxc --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1763118445/summary.json'
Pipeline stage MKMLProfilerRunner completed in 3170.22s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler is running
Tearing down inference service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler
Service chaiml-ir-v3-3-dpo-irv1-78559-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 0.71s
Shutdown handler de-registered
chaiml-ir-v3-3-dpo-irv1_78559_v1 status is now inactive due to auto deactivation removed underperforming models
chaiml-ir-v3-3-dpo-irv1_78559_v1 status is now torndown due to DeploymentManager action