developer_uid: NischayDnk
submission_id: chaiml-qwen-dpo-exp1-da_53786_v2
model_name: chaiml-qwen-dpo-exp1-da_53786_v2
model_group: ChaiML/qwen-dpo-exp1-dat
status: torndown
timestamp: 2025-04-14T08:00:42+00:00
num_battles: 7844
num_wins: 3761
celo_rating: 1269.84
family_friendly_score: 0.5791999999999999
family_friendly_standard_error: 0.006981795757539746
submission_type: basic
model_repo: ChaiML/qwen-dpo-exp1-datav1noffsftfull
model_architecture: Qwen2ForCausalLM
model_num_parameters: 32763417600.0
best_of: 1
max_input_tokens: 512
max_output_tokens: 64
reward_model: default
latencies: [{'batch_size': 1, 'throughput': 0.29843543548867807, 'latency_mean': 3.350741503238678, 'latency_p50': 3.3543373346328735, 'latency_p90': 3.725955080986023}, {'batch_size': 3, 'throughput': 0.7111000616589608, 'latency_mean': 4.193629829883576, 'latency_p50': 4.20060122013092, 'latency_p90': 4.671200275421143}, {'batch_size': 5, 'throughput': 1.0483346995605287, 'latency_mean': 4.734107844829559, 'latency_p50': 4.685012936592102, 'latency_p90': 5.280592083930969}, {'batch_size': 6, 'throughput': 1.166771013565068, 'latency_mean': 5.093998324871063, 'latency_p50': 5.103861927986145, 'latency_p90': 5.749396657943725}, {'batch_size': 10, 'throughput': 1.5744410181813229, 'latency_mean': 6.255968832969666, 'latency_p50': 6.264776229858398, 'latency_p90': 7.182001161575317}]
gpu_counts: {'NVIDIA RTX A6000': 1}
display_name: chaiml-qwen-dpo-exp1-da_53786_v2
is_internal_developer: False
language_model: ChaiML/qwen-dpo-exp1-datav1noffsftfull
model_size: 33B
ranking_group: single
throughput_3p7s: 0.49
us_pacific_date: 2025-04-14
win_ratio: 0.47947475777664456
generation_params: {'temperature': 0.5, 'top_p': 0.95, 'min_p': 0.025, 'top_k': 60, 'presence_penalty': 0.4, 'frequency_penalty': 0.4, 'stopping_words': ['<|im_end|>', '\n'], 'max_input_tokens': 512, 'best_of': 1, 'max_output_tokens': 64}
formatter: {'memory_template': '<|system|>{memory}\n', 'prompt_template': '', 'bot_template': '<|im_start|>assistant\n{message}<|im_end|>\n', 'user_template': '<|im_start|>user\nYou:{message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n', '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-qwen-dpo-exp1-da-53786-v2-mkmlizer
Waiting for job on chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer to finish
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: Downloaded to shared memory in 84.076s
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpb74hfjy6, device:0
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Failed to get response for submission jellywibble-felix-black_21853_v1: HTTPConnectionPool(host='jellywibble-felix-black-21853-v1-predictor.tenant-chaiml-guanaco.k.chaiverse.com', port=80): Read timed out. (read timeout=12.0)
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: quantized model in 65.580s
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: Processed model ChaiML/qwen-dpo-exp1-datav1noffsftfull in 149.656s
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: creating bucket guanaco-mkml-models
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/config.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/added_tokens.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/added_tokens.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/special_tokens_map.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/tokenizer_config.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/merges.txt s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/merges.txt
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/vocab.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/vocab.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/tokenizer.json
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.2.safetensors s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/flywheel_model.2.safetensors
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/flywheel_model.1.safetensors
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-qwen-dpo-exp1-da-53786-v2/flywheel_model.0.safetensors
chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer: Loading 0: 0%| | 0/771 [00:00<?, ?it/s] Loading 0: 1%| | 5/771 [00:00<00:19, 38.43it/s] Loading 0: 2%|▏ | 16/771 [00:00<00:10, 73.41it/s] Loading 0: 4%|▎ | 27/771 [00:00<00:08, 87.03it/s] Loading 0: 5%|▍ | 36/771 [00:00<00:08, 86.53it/s] Loading 0: 6%|▌ | 46/771 [00:00<00:12, 56.26it/s] Loading 0: 7%|▋ | 53/771 [00:00<00:15, 47.10it/s] Loading 0: 8%|▊ | 64/771 [00:01<00:12, 58.79it/s] Loading 0: 10%|▉ | 75/771 [00:01<00:10, 67.82it/s] Loading 0: 11%|█ | 83/771 [00:01<00:09, 69.65it/s] Loading 0: 12%|█▏ | 91/771 [00:01<00:10, 67.23it/s] Loading 0: 14%|█▎ | 106/771 [00:01<00:10, 61.43it/s] Loading 0: 15%|█▍ | 113/771 [00:01<00:11, 58.38it/s] Loading 0: 16%|█▌ | 121/771 [00:01<00:11, 58.26it/s] Loading 0: 17%|█▋ | 132/771 [00:02<00:09, 68.61it/s] Loading 0: 18%|█▊ | 142/771 [00:02<00:08, 75.11it/s] Loading 0: 20%|█▉ | 152/771 [00:02<00:07, 79.81it/s] Loading 0: 21%|██ | 161/771 [00:02<00:08, 73.89it/s] Loading 0: 22%|██▏ | 169/771 [00:02<00:11, 51.79it/s] Loading 0: 23%|██▎ | 176/771 [00:02<00:11, 52.10it/s] Loading 0: 24%|██▍ | 184/771 [00:02<00:10, 57.84it/s] Loading 0: 25%|██▌ | 195/771 [00:03<00:08, 69.18it/s] Loading 0: 26%|██▋ | 203/771 [00:03<00:08, 70.44it/s] Loading 0: 27%|██▋ | 211/771 [00:03<00:08, 67.45it/s] Loading 0: 29%|██▉ | 226/771 [00:03<00:08, 61.79it/s] Loading 0: 30%|███ | 233/771 [00:03<00:10, 51.57it/s] Loading 0: 32%|███▏ | 244/771 [00:03<00:08, 62.31it/s] Loading 0: 33%|███▎ | 255/771 [00:03<00:07, 72.02it/s] Loading 0: 34%|███▍ | 264/771 [00:04<00:06, 75.47it/s] Loading 0: 35%|███▌ | 273/771 [00:04<00:06, 71.86it/s] Loading 0: 37%|███▋ | 286/771 [00:04<00:07, 61.77it/s] Loading 0: 38%|███▊ | 293/771 [00:04<00:09, 51.17it/s] Loading 0: 39%|███▉ | 303/771 [00:19<03:43, 2.09it/s] Loading 0: 40%|████ | 311/771 [00:19<02:43, 2.81it/s] Loading 0: 41%|████ | 318/771 [00:19<02:03, 3.66it/s] Loading 0: 43%|████▎ | 328/771 [00:20<01:22, 5.40it/s] Loading 0: 44%|████▍ | 338/771 [00:20<00:55, 7.77it/s] Loading 0: 45%|████▍ | 346/771 [00:20<00:43, 9.86it/s] Loading 0: 46%|████▌ | 353/771 [00:20<00:36, 11.33it/s] Loading 0: 47%|████▋ | 364/771 [00:20<00:24, 16.67it/s] Loading 0: 49%|████▊ | 375/771 [00:20<00:16, 23.34it/s] Loading 0: 50%|████▉ | 383/771 [00:21<00:13, 28.56it/s] Loading 0: 51%|█████ | 391/771 [00:21<00:11, 33.48it/s] Loading 0: 53%|█████▎ | 406/771 [00:21<00:09, 40.42it/s] Loading 0: 54%|█████▎ | 413/771 [00:21<00:09, 38.45it/s] Loading 0: 55%|█████▍ | 424/771 [00:21<00:07, 48.64it/s] Loading 0: 56%|█████▋ | 435/771 [00:21<00:05, 58.87it/s] Loading 0: 58%|█████▊ | 444/771 [00:21<00:05, 64.48it/s] Loading 0: 59%|█████▉ | 453/771 [00:22<00:04, 64.65it/s] Loading 0: 60%|██████ | 466/771 [00:22<00:05, 57.46it/s] Loading 0: 61%|██████▏ | 473/771 [00:22<00:06, 49.32it/s] Loading 0: 63%|██████▎ | 484/771 [00:22<00:04, 59.79it/s] Loading 0: 64%|██████▍ | 495/771 [00:22<00:03, 69.98it/s] Loading 0: 65%|██████▌ | 504/771 [00:22<00:03, 73.83it/s] Loading 0: 67%|██████▋ | 513/771 [00:22<00:03, 71.46it/s] Loading 0: 68%|██████▊ | 526/771 [00:23<00:03, 62.89it/s] Loading 0: 69%|██████▉ | 534/771 [00:23<00:04, 53.99it/s] Loading 0: 71%|███████ | 544/771 [00:23<00:03, 62.47it/s] Loading 0: 72%|███████▏ | 555/771 [00:23<00:02, 72.07it/s] Loading 0: 73%|███████▎ | 564/771 [00:23<00:02, 75.57it/s] Loading 0: 74%|███████▍ | 573/771 [00:23<00:02, 72.06it/s] Loading 0: 76%|███████▌ | 586/771 [00:24<00:02, 62.17it/s] Loading 0: 77%|███████▋ | 593/771 [00:24<00:03, 52.51it/s] Loading 0: 78%|███████▊ | 605/771 [00:24<00:02, 60.70it/s] Loading 0: 80%|███████▉ | 616/771 [00:24<00:02, 68.61it/s] Loading 0: 81%|████████▏ | 627/771 [00:39<01:02, 2.32it/s] Loading 0: 82%|████████▏ | 635/771 [00:39<00:44, 3.04it/s] Loading 0: 84%|████████▍ | 646/771 [00:40<00:29, 4.30it/s] Loading 0: 85%|████████▍ | 653/771 [00:40<00:22, 5.28it/s] Loading 0: 86%|████████▌ | 664/771 [00:40<00:13, 7.77it/s] Loading 0: 88%|████████▊ | 675/771 [00:40<00:08, 11.13it/s] Loading 0: 89%|████████▊ | 683/771 [00:40<00:06, 14.21it/s] Loading 0: 90%|████████▉ | 691/771 [00:40<00:04, 17.86it/s] Loading 0: 92%|█████████▏| 706/771 [00:41<00:02, 25.04it/s] Loading 0: 92%|█████████▏| 713/771 [00:41<00:02, 26.51it/s] Loading 0: 94%|█████████▍| 724/771 [00:41<00:01, 35.17it/s] Loading 0: 95%|█████████▌| 735/771 [00:41<00:00, 44.71it/s] Loading 0: 96%|█████████▋| 744/771 [00:41<00:00, 51.42it/s] Loading 0: 98%|█████████▊| 753/771 [00:41<00:00, 54.84it/s] Loading 0: 99%|█████████▉| 766/771 [00:49<00:01, 4.28it/s]
Job chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer completed after 186.32s with status: succeeded
Stopping job with name chaiml-qwen-dpo-exp1-da-53786-v2-mkmlizer
Pipeline stage MKMLizer completed in 187.02s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.20s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-qwen-dpo-exp1-da-53786-v2
Waiting for inference service chaiml-qwen-dpo-exp1-da-53786-v2 to be ready
Failed to get response for submission jellywibble-felix-black_21853_v1: HTTPConnectionPool(host='jellywibble-felix-black-21853-v1-predictor.tenant-chaiml-guanaco.k.chaiverse.com', port=80): Read timed out. (read timeout=12.0)
Inference service chaiml-qwen-dpo-exp1-da-53786-v2 ready after 120.46721649169922s
Pipeline stage MKMLDeployer completed in 121.02s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.545187473297119s
Received healthy response to inference request in 3.074899673461914s
Received healthy response to inference request in 3.1060328483581543s
Received healthy response to inference request in 3.041520833969116s
Received healthy response to inference request in 3.0615737438201904s
5 requests
0 failed requests
5th percentile: 3.045531415939331
10th percentile: 3.0495419979095457
20th percentile: 3.0575631618499757
30th percentile: 3.064238929748535
40th percentile: 3.069569301605225
50th percentile: 3.074899673461914
60th percentile: 3.0873529434204103
Failed to get response for submission jellywibble-felix-black_21853_v1: HTTPConnectionPool(host='jellywibble-felix-black-21853-v1-predictor.tenant-chaiml-guanaco.k.chaiverse.com', port=80): Read timed out. (read timeout=12.0)
70th percentile: 3.099806213378906
80th percentile: 3.193863773345947
90th percentile: 3.3695256233215334
95th percentile: 3.4573565483093263
99th percentile: 3.5276212882995606
mean time: 3.1658429145812987
Pipeline stage StressChecker completed in 17.02s
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 0.72s
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 0.67s
Shutdown handler de-registered
chaiml-qwen-dpo-exp1-da_53786_v2 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 5240.38s
Shutdown handler de-registered
chaiml-qwen-dpo-exp1-da_53786_v2 status is now inactive due to auto deactivation removed underperforming models
chaiml-qwen-dpo-exp1-da_53786_v2 status is now torndown due to DeploymentManager action