chaiml-ir-v3-4-dpo-irv1_64735

developer_uid: chai_backend_admin

submission_id: chaiml-ir-v3-4-dpo-irv1_64735_v1

model_name: training123

model_group: ChaiML/ir-v3-4-dpo_irv13

status: torndown

timestamp: 2025-11-17T13:01:29+00:00

num_battles: 6662

num_wins: 3448

celo_rating: 1294.17

family_friendly_score: 0.0

family_friendly_standard_error: 0.0

submission_type: basic

model_repo: ChaiML/ir-v3-4-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': 0.3207788174008219, 'latency_mean': 3.1173125672340394, 'latency_p50': 2.6655476093292236, 'latency_p90': 5.676704025268554}, {'batch_size': 2, 'throughput': 0.5752584956963676, 'latency_mean': 3.4733906352519988, 'latency_p50': 3.4705976247787476, 'latency_p90': 3.7770251512527464}, {'batch_size': 3, 'throughput': 0.7056521899141868, 'latency_mean': 4.233188940286636, 'latency_p50': 4.258459687232971, 'latency_p90': 4.531324434280395}, {'batch_size': 4, 'throughput': 0.8053927888369106, 'latency_mean': 4.950939860343933, 'latency_p50': 4.940038800239563, 'latency_p90': 5.6624651670455926}, {'batch_size': 5, 'throughput': 0.8696085170823694, 'latency_mean': 5.728296593427658, 'latency_p50': 5.722036719322205, 'latency_p90': 6.612297081947326}]

gpu_counts: {'NVIDIA L40S': 1}

display_name: training123

is_internal_developer: True

language_model: ChaiML/ir-v3-4-dpo_irv13-cosine

model_size: 24B

ranking_group: single

throughput_3p7s: 0.62

us_pacific_date: 2025-11-14

win_ratio: 0.5175622936055239

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': ['\n', 'User:', '</s>', 'You:'], '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-v3-4-dpo-irv1-64735-v1-mkmlizer
Waiting for job on chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer to finish
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   ██████   ██████  █████   ████  ████                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  ░░██████ ██████  ░░███   ███░  ░░███                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   ░███░█████░███   ░███  ███     ░███                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   ░███░░███ ░███   ░███████      ░███                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   ░███ ░░░  ░███   ░███░░███     ░███                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   ░███      ░███   ░███ ░░███    ░███                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║   █████     █████  █████ ░░████  █████                              ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  ░░░░░     ░░░░░  ░░░░░   ░░░░  ░░░░░                               ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Version: 0.30.2                                                    ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Features: FLYWHEEL, CUDA                                           ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Copyright 2023-2025 MK ONE TECHNOLOGIES Inc.                       ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  belonging to:                                                      ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ║                                                                     ║
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: Downloaded to shared memory in 97.705s
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: Checking if ChaiML/ir-v3-4-dpo_irv13-cosine already exists in ChaiML
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpcmhqlxct, device:0
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: quantized model in 41.540s
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: Processed model ChaiML/ir-v3-4-dpo_irv13-cosine in 139.246s
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: creating bucket guanaco-mkml-models
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/config.json
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/special_tokens_map.json
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/tokenizer_config.json
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/tokenizer.json
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/flywheel_model.1.safetensors
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-ir-v3-4-dpo-irv1-64735-v1/nvidia/flywheel_model.0.safetensors
chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:15, 23.10it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:08, 39.19it/s]
Loading 0:   5%|▍         | 18/363 [00:00<00:08, 41.98it/s]
Loading 0:   6%|▋         | 23/363 [00:00<00:10, 32.41it/s]
Loading 0:   9%|▉         | 32/363 [00:00<00:07, 44.66it/s]
Loading 0:  10%|█         | 38/363 [00:01<00:10, 30.45it/s]
Loading 0:  12%|█▏        | 42/363 [00:01<00:11, 27.86it/s]
Loading 0:  13%|█▎        | 48/363 [00:01<00:09, 33.05it/s]
Loading 0:  14%|█▍        | 52/363 [00:01<00:09, 31.92it/s]
Loading 0:  16%|█▌        | 57/363 [00:01<00:08, 35.11it/s]
Loading 0:  17%|█▋        | 61/363 [00:01<00:10, 29.49it/s]
Loading 0:  18%|█▊        | 65/363 [00:01<00:09, 31.06it/s]
Loading 0:  19%|█▉        | 70/363 [00:02<00:10, 28.64it/s]
Loading 0:  20%|██        | 74/363 [00:02<00:11, 26.01it/s]
Loading 0:  22%|██▏       | 79/363 [00:02<00:09, 30.11it/s]
Loading 0:  23%|██▎       | 83/363 [00:02<00:08, 31.41it/s]
Loading 0:  24%|██▍       | 87/363 [00:02<00:09, 27.86it/s]
Loading 0:  25%|██▌       | 92/363 [00:02<00:10, 27.07it/s]
Loading 0:  27%|██▋       | 99/363 [00:03<00:07, 34.56it/s]
Loading 0:  28%|██▊       | 103/363 [00:03<00:07, 33.01it/s]
Loading 0:  29%|██▉       | 107/363 [00:03<00:08, 28.76it/s]
Loading 0:  31%|███       | 112/363 [00:03<00:08, 31.03it/s]
Loading 0:  32%|███▏      | 116/363 [00:03<00:07, 31.58it/s]
Loading 0:  33%|███▎      | 120/363 [00:03<00:07, 33.46it/s]
Loading 0:  34%|███▍      | 124/363 [00:03<00:07, 32.07it/s]
Loading 0:  36%|███▌      | 129/363 [00:04<00:06, 35.34it/s]
Loading 0:  37%|███▋      | 133/363 [00:04<00:06, 33.01it/s]
Loading 0:  38%|███▊      | 138/363 [00:04<00:06, 36.18it/s]
Loading 0:  39%|███▉      | 142/363 [00:04<00:06, 33.13it/s]
Loading 0:  41%|████      | 149/363 [00:04<00:05, 39.52it/s]
Loading 0:  42%|████▏     | 154/363 [00:04<00:07, 27.22it/s]
Loading 0:  44%|████▎     | 158/363 [00:05<00:07, 25.69it/s]
Loading 0:  45%|████▌     | 165/363 [00:05<00:06, 32.82it/s]
Loading 0:  47%|████▋     | 169/363 [00:05<00:06, 31.99it/s]
Loading 0:  48%|████▊     | 174/363 [00:05<00:05, 35.07it/s]
Loading 0:  49%|████▉     | 178/363 [00:05<00:05, 33.30it/s]
Loading 0:  50%|█████     | 182/363 [00:05<00:05, 34.05it/s]
Loading 0:  52%|█████▏    | 187/363 [00:05<00:05, 31.71it/s]
Loading 0:  53%|█████▎    | 191/363 [00:06<00:05, 30.35it/s]
Loading 0:  54%|█████▎    | 195/363 [00:06<00:06, 27.16it/s]
Loading 0:  55%|█████▌    | 201/363 [00:19<02:13,  1.21it/s]
Loading 0:  56%|█████▌    | 203/363 [00:19<01:55,  1.39it/s]
Loading 0:  57%|█████▋    | 208/363 [00:19<01:14,  2.09it/s]
Loading 0:  58%|█████▊    | 211/363 [00:19<00:57,  2.65it/s]
Loading 0:  59%|█████▉    | 214/363 [00:19<00:43,  3.42it/s]
Loading 0:  60%|██████    | 218/363 [00:19<00:30,  4.82it/s]
Loading 0:  61%|██████    | 221/363 [00:19<00:23,  6.13it/s]
Loading 0:  62%|██████▏   | 224/363 [00:20<00:18,  7.44it/s]
Loading 0:  63%|██████▎   | 229/363 [00:20<00:12, 10.82it/s]
Loading 0:  64%|██████▍   | 232/363 [00:20<00:10, 12.85it/s]
Loading 0:  65%|██████▌   | 237/363 [00:20<00:07, 17.57it/s]
Loading 0:  66%|██████▋   | 241/363 [00:20<00:06, 19.99it/s]
Loading 0:  68%|██████▊   | 246/363 [00:20<00:04, 24.90it/s]
Loading 0:  69%|██████▉   | 250/363 [00:20<00:04, 26.04it/s]
Loading 0:  70%|███████   | 255/363 [00:20<00:03, 30.48it/s]
Loading 0:  71%|███████▏  | 259/363 [00:21<00:03, 30.44it/s]
Loading 0:  73%|███████▎  | 266/363 [00:21<00:02, 37.57it/s]
Loading 0:  75%|███████▍  | 271/363 [00:21<00:03, 27.23it/s]
Loading 0:  76%|███████▌  | 275/363 [00:21<00:03, 25.46it/s]
Loading 0:  78%|███████▊  | 282/363 [00:21<00:02, 32.60it/s]
Loading 0:  79%|███████▉  | 286/363 [00:21<00:02, 31.79it/s]
Loading 0:  80%|████████  | 291/363 [00:22<00:02, 35.07it/s]
Loading 0:  81%|████████▏ | 295/363 [00:22<00:02, 33.36it/s]
Loading 0:  82%|████████▏ | 299/363 [00:22<00:01, 34.27it/s]
Loading 0:  84%|████████▎ | 304/363 [00:22<00:01, 31.67it/s]
Loading 0:  85%|████████▍ | 308/363 [00:22<00:01, 30.20it/s]
Loading 0:  86%|████████▌ | 312/363 [00:22<00:01, 27.13it/s]
Loading 0:  88%|████████▊ | 318/363 [00:22<00:01, 32.77it/s]
Loading 0:  89%|████████▊ | 322/363 [00:23<00:01, 31.61it/s]
Loading 0:  90%|█████████ | 327/363 [00:23<00:01, 35.18it/s]
Loading 0:  91%|█████████ | 331/363 [00:23<00:00, 33.46it/s]
Loading 0:  92%|█████████▏| 335/363 [00:23<00:00, 34.36it/s]
Loading 0:  93%|█████████▎| 339/363 [00:23<00:00, 35.13it/s]
Loading 0:  94%|█████████▍| 343/363 [00:23<00:00, 21.09it/s]
Loading 0:  96%|█████████▌| 348/363 [00:24<00:00, 22.72it/s]
Loading 0:  98%|█████████▊| 355/363 [00:24<00:00, 30.43it/s]
Loading 0:  99%|█████████▉| 359/363 [00:24<00:00, 30.33it/s]
                                                            
Job chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer completed after 206.46s with status: succeeded
Stopping job with name chaiml-ir-v3-4-dpo-irv1-64735-v1-mkmlizer
Pipeline stage MKMLizer completed in 207.00s
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 chaiml-ir-v3-4-dpo-irv1-64735-v1
Waiting for inference service chaiml-ir-v3-4-dpo-irv1-64735-v1 to be ready
Inference service chaiml-ir-v3-4-dpo-irv1-64735-v1 ready after 150.73420071601868s
Pipeline stage MKMLDeployer completed in 151.53s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.8187222480773926s
Received healthy response to inference request in 2.1201136112213135s
Received healthy response to inference request in 2.2046728134155273s
Received healthy response to inference request in 2.5852694511413574s
Received healthy response to inference request in 2.8382863998413086s
5 requests
0 failed requests
5th percentile: 2.1370254516601563
10th percentile: 2.153937292098999
20th percentile: 2.1877609729766845
30th percentile: 2.2807921409606933
40th percentile: 2.4330307960510256
50th percentile: 2.5852694511413574
60th percentile: 2.6786505699157717
70th percentile: 2.7720316886901855
80th percentile: 2.8226350784301757
90th percentile: 2.8304607391357424
95th percentile: 2.8343735694885255
99th percentile: 2.837503833770752
mean time: 2.51341290473938
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 0.64s
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 0.69s
Shutdown handler de-registered
chaiml-ir-v3-4-dpo-irv1_64735_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 0.09s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-ir-v3-4-dpo-irv1-64735-v1-profiler
Waiting for inference service chaiml-ir-v3-4-dpo-irv1-64735-v1-profiler to be ready
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 6028.00s
Shutdown handler de-registered
chaiml-ir-v3-4-dpo-irv1_64735_v1 status is now torndown due to DeploymentManager action