chaiml-ir-v3-2-dpo-irv13

developer_uid: chai_backend_admin

submission_id: chaiml-ir-v3-2-dpo-irv13_v1

model_name: training123

model_group: ChaiML/ir-v3-2-dpo_irv13

status: torndown

timestamp: 2025-11-17T09:21:33+00:00

num_battles: 5686

num_wins: 2974

celo_rating: 1303.35

family_friendly_score: 0.43379999999999996

family_friendly_standard_error: 0.007008816733229654

submission_type: basic

model_repo: ChaiML/ir-v3-2-dpo_irv13

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': 0.2481170566968863, 'latency_mean': 4.030254775285721, 'latency_p50': 4.045350074768066, 'latency_p90': 5.954110240936279}, {'batch_size': 2, 'throughput': 0.5699053236291588, 'latency_mean': 3.507862628698349, 'latency_p50': 3.527121901512146, 'latency_p90': 3.7795716047286985}, {'batch_size': 3, 'throughput': 0.7070352552806866, 'latency_mean': 4.230080199241638, 'latency_p50': 4.209761738777161, 'latency_p90': 4.817786192893982}, {'batch_size': 4, 'throughput': 0.8024457447000684, 'latency_mean': 4.96455686211586, 'latency_p50': 4.9418556690216064, 'latency_p90': 5.738644385337829}, {'batch_size': 5, 'throughput': 0.8610193128297288, 'latency_mean': 5.778368763923645, 'latency_p50': 5.768812537193298, 'latency_p90': 6.520901679992676}]

gpu_counts: {'NVIDIA L40S': 1}

display_name: training123

is_internal_developer: True

language_model: ChaiML/ir-v3-2-dpo_irv13

model_size: 24B

ranking_group: single

throughput_3p7s: 0.27

us_pacific_date: 2025-11-14

win_ratio: 0.5230390432641576

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>', 'You:', '\n', '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-v3-2-dpo-irv13-v1-mkmlizer
Waiting for job on chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer to finish
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   ██████   ██████  █████   ████  ████                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  ░░██████ ██████  ░░███   ███░  ░░███                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   ░███░█████░███   ░███  ███     ░███                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   ░███░░███ ░███   ░███████      ░███                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   ░███ ░░░  ░███   ░███░░███     ░███                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   ░███      ░███   ░███ ░░███    ░███                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║   █████     █████  █████ ░░████  █████                              ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  ░░░░░     ░░░░░  ░░░░░   ░░░░  ░░░░░                               ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Version: 0.30.2                                                    ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Features: FLYWHEEL, CUDA                                           ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Copyright 2023-2025 MK ONE TECHNOLOGIES Inc.                       ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  belonging to:                                                      ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: Downloaded to shared memory in 66.593s
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: Checking if ChaiML/ir-v3-2-dpo_irv13 already exists in ChaiML
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp9lq3kwgn, device:0
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: quantized model in 41.711s
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: Processed model ChaiML/ir-v3-2-dpo_irv13 in 108.304s
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: creating bucket guanaco-mkml-models
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/config.json
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/special_tokens_map.json
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/tokenizer_config.json
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/tokenizer.json
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/flywheel_model.1.safetensors
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-2-dpo-irv13-v1/nvidia/flywheel_model.0.safetensors
chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:16, 22.24it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:09, 38.67it/s]
Loading 0:   5%|▍         | 18/363 [00:00<00:08, 42.26it/s]
Loading 0:   6%|▋         | 23/363 [00:00<00:10, 33.14it/s]
Loading 0:   9%|▉         | 32/363 [00:00<00:07, 45.87it/s]
Loading 0:  10%|█         | 38/363 [00:01<00:10, 31.40it/s]
Loading 0:  12%|█▏        | 43/363 [00:01<00:10, 29.68it/s]
Loading 0:  13%|█▎        | 48/363 [00:01<00:09, 32.91it/s]
Loading 0:  14%|█▍        | 52/363 [00:01<00:09, 31.87it/s]
Loading 0:  16%|█▌        | 57/363 [00:01<00:08, 35.74it/s]
Loading 0:  17%|█▋        | 62/363 [00:01<00:08, 35.48it/s]
Loading 0:  19%|█▊        | 68/363 [00:01<00:07, 40.78it/s]
Loading 0:  20%|██        | 73/363 [00:02<00:10, 26.42it/s]
Loading 0:  22%|██▏       | 79/363 [00:02<00:08, 31.81it/s]
Loading 0:  23%|██▎       | 84/363 [00:02<00:08, 31.42it/s]
Loading 0:  24%|██▍       | 88/363 [00:02<00:08, 30.94it/s]
Loading 0:  25%|██▌       | 92/363 [00:02<00:09, 28.19it/s]
Loading 0:  27%|██▋       | 99/363 [00:02<00:07, 35.81it/s]
Loading 0:  29%|██▊       | 104/363 [00:03<00:07, 35.76it/s]
Loading 0:  30%|██▉       | 108/363 [00:03<00:08, 29.18it/s]
Loading 0:  31%|███       | 113/363 [00:03<00:08, 28.17it/s]
Loading 0:  33%|███▎      | 120/363 [00:03<00:06, 35.43it/s]
Loading 0:  34%|███▍      | 124/363 [00:03<00:07, 33.99it/s]
Loading 0:  36%|███▌      | 129/363 [00:03<00:06, 36.83it/s]
Loading 0:  37%|███▋      | 133/363 [00:03<00:06, 34.85it/s]
Loading 0:  38%|███▊      | 138/363 [00:04<00:06, 37.41it/s]
Loading 0:  39%|███▉      | 142/363 [00:04<00:06, 33.44it/s]
Loading 0:  41%|████      | 149/363 [00:04<00:05, 39.32it/s]
Loading 0:  42%|████▏     | 154/363 [00:04<00:07, 27.71it/s]
Loading 0:  44%|████▎     | 158/363 [00:04<00:07, 25.63it/s]
Loading 0:  45%|████▌     | 165/363 [00:05<00:06, 32.80it/s]
Loading 0:  47%|████▋     | 169/363 [00:05<00:06, 31.76it/s]
Loading 0:  48%|████▊     | 174/363 [00:05<00:05, 34.95it/s]
Loading 0:  49%|████▉     | 178/363 [00:05<00:05, 33.08it/s]
Loading 0:  50%|█████     | 182/363 [00:05<00:05, 33.69it/s]
Loading 0:  52%|█████▏    | 187/363 [00:05<00:05, 31.06it/s]
Loading 0:  53%|█████▎    | 191/363 [00:05<00:05, 29.79it/s]
Loading 0:  54%|█████▎    | 195/363 [00:06<00:06, 26.67it/s]
Loading 0:  55%|█████▌    | 201/363 [00:18<02:13,  1.21it/s]
Loading 0:  56%|█████▌    | 203/363 [00:19<01:54,  1.39it/s]
Loading 0:  57%|█████▋    | 208/363 [00:19<01:13,  2.11it/s]
Loading 0:  58%|█████▊    | 212/363 [00:19<00:53,  2.82it/s]
Loading 0:  60%|██████    | 218/363 [00:19<00:32,  4.40it/s]
Loading 0:  61%|██████    | 222/363 [00:19<00:24,  5.74it/s]
Loading 0:  62%|██████▏   | 226/363 [00:19<00:19,  7.08it/s]
Loading 0:  63%|██████▎   | 230/363 [00:20<00:15,  8.71it/s]
Loading 0:  65%|██████▌   | 237/363 [00:20<00:09, 13.49it/s]
Loading 0:  66%|██████▋   | 241/363 [00:20<00:07, 15.58it/s]
Loading 0:  68%|██████▊   | 246/363 [00:20<00:05, 19.57it/s]
Loading 0:  69%|██████▉   | 250/363 [00:20<00:05, 21.32it/s]
Loading 0:  70%|███████   | 255/363 [00:20<00:04, 25.68it/s]
Loading 0:  71%|███████▏  | 259/363 [00:20<00:03, 26.66it/s]
Loading 0:  73%|███████▎  | 266/363 [00:21<00:03, 31.23it/s]
Loading 0:  74%|███████▍  | 270/363 [00:21<00:04, 20.58it/s]
Loading 0:  76%|███████▌  | 275/363 [00:21<00:04, 21.20it/s]
Loading 0:  78%|███████▊  | 282/363 [00:21<00:02, 27.01it/s]
Loading 0:  79%|███████▉  | 286/363 [00:21<00:02, 27.33it/s]
Loading 0:  80%|████████  | 291/363 [00:22<00:02, 31.10it/s]
Loading 0:  81%|████████▏ | 295/363 [00:22<00:02, 30.66it/s]
Loading 0:  82%|████████▏ | 299/363 [00:22<00:01, 32.08it/s]
Loading 0:  84%|████████▎ | 304/363 [00:22<00:01, 30.30it/s]
Loading 0:  85%|████████▍ | 308/363 [00:22<00:01, 29.67it/s]
Loading 0:  86%|████████▌ | 312/363 [00:22<00:01, 26.86it/s]
Loading 0:  88%|████████▊ | 318/363 [00:22<00:01, 32.57it/s]
Loading 0:  89%|████████▊ | 322/363 [00:23<00:01, 31.64it/s]
Loading 0:  90%|█████████ | 327/363 [00:23<00:01, 35.09it/s]
Loading 0:  91%|█████████ | 331/363 [00:23<00:00, 33.47it/s]
Loading 0:  92%|█████████▏| 335/363 [00:23<00:00, 33.99it/s]
Loading 0:  93%|█████████▎| 339/363 [00:23<00:00, 34.39it/s]
Loading 0:  94%|█████████▍| 343/363 [00:23<00:00, 20.83it/s]
Loading 0:  96%|█████████▌| 348/363 [00:24<00:00, 22.40it/s]
Loading 0:  98%|█████████▊| 355/363 [00:24<00:00, 30.08it/s]
Loading 0:  99%|█████████▉| 359/363 [00:24<00:00, 29.88it/s]
Loading 0: 100%|██████████| 363/363 [00:24<00:00, 30.68it/s]
                                                            
Job chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer completed after 185.29s with status: succeeded
Stopping job with name chaiml-ir-v3-2-dpo-irv13-v1-mkmlizer
Pipeline stage MKMLizer completed in 186.01s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.18s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-ir-v3-2-dpo-irv13-v1
Waiting for inference service chaiml-ir-v3-2-dpo-irv13-v1 to be ready
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', '')
Failed to get response for submission chaiml-7b07-69d4-linear-w01_v7: ('http://guanaco-model-mesh-load-balancer.model-mesh.k2.chaiverse.com/models/chaiml-7b07-69d4-linear-w01_v7/predict', '{"detail":"1 validation error for RuntimeResponse\\npredictions\\n  Field required [type=missing, input_value={\'detail\': \\"503, message=...linear-w01_v7/predict\'\\"}, input_type=dict]\\n    For further information visit https://errors.pydantic.dev/2.11/v/missing"}')
Inference service chaiml-ir-v3-2-dpo-irv13-v1 ready after 151.20743131637573s
Pipeline stage MKMLDeployer completed in 152.03s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.8279459476470947s
Received healthy response to inference request in 2.7587523460388184s
Received healthy response to inference request in 2.215569496154785s
Received healthy response to inference request in 2.1303064823150635s
Received healthy response to inference request in 2.386362314224243s
5 requests
0 failed requests
5th percentile: 2.147359085083008
10th percentile: 2.164411687850952
20th percentile: 2.1985168933868406
30th percentile: 2.249728059768677
40th percentile: 2.31804518699646
50th percentile: 2.386362314224243
60th percentile: 2.5353183269500734
70th percentile: 2.6842743396759032
80th percentile: 2.772591066360474
90th percentile: 2.8002685070037843
95th percentile: 2.8141072273254393
99th percentile: 2.8251782035827637
mean time: 2.463787317276001
Pipeline stage StressChecker completed in 13.80s
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 1.41s
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 0.77s
Shutdown handler de-registered
chaiml-ir-v3-2-dpo-irv13_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
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-v3-2-dpo-irv13-v1-profiler
Waiting for inference service chaiml-ir-v3-2-dpo-irv13-v1-profiler to be ready
Inference service chaiml-ir-v3-2-dpo-irv13-v1-profiler ready after 151.10342383384705s
Pipeline stage MKMLProfilerDeployer completed in 151.78s
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-v3-2-dpo-irv13-v1-profiler-predictor-00001-deplog2g6n:/code/chaiverse_profiler_1763106726 --namespace tenant-chaiml-guanaco
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-2-dpo-irv13-v1-profiler-predictor-00001-deplog2g6n --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1763106726 && 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_1763106726/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-v3-2-dpo-irv13-v1-profiler-predictor-00001-deplog2g6n:/code/chaiverse_profiler_1763107882 --namespace tenant-chaiml-guanaco
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-2-dpo-irv13-v1-profiler-predictor-00001-deplog2g6n --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1763107882 && 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_1763107882/summary.json'
kubectl --kubeconfig /code/guanaco/guanaco_services/resources/kchai_coreweave_us_east_04a.yaml exec -it chaiml-ir-v3-2-dpo-irv13-v1-profiler-predictor-00001-deplog2g6n --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1763107882/summary.json'
Pipeline stage MKMLProfilerRunner completed in 3085.17s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service chaiml-ir-v3-2-dpo-irv13-v1-profiler is running
Tearing down inference service chaiml-ir-v3-2-dpo-irv13-v1-profiler
Service chaiml-ir-v3-2-dpo-irv13-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 0.65s
Shutdown handler de-registered
chaiml-ir-v3-2-dpo-irv13_v1 status is now inactive due to auto deactivation removed underperforming models
chaiml-ir-v3-2-dpo-irv13_v1 status is now torndown due to DeploymentManager action