stark2000s-ult-du-v1-5

developer_uid: stark2000s

submission_id: stark2000s-ult-du-v1-5_v2

model_name: stark2000s-ult-du-v1-5_v1

model_group: stark2000s/ult-du-v1.5

status: inactive

timestamp: 2024-11-28T13:31:29+00:00

num_battles: 24034

num_wins: 10451

celo_rating: 1218.46

family_friendly_score: 0.5682

family_friendly_standard_error: 0.0070049805138915275

submission_type: basic

model_repo: stark2000s/ult-du-v1.5

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.9088346055389582, 'latency_mean': 1.1002529215812684, 'latency_p50': 1.1022155284881592, 'latency_p90': 1.2361827611923217}, {'batch_size': 4, 'throughput': 1.7996637219790244, 'latency_mean': 2.2105026972293853, 'latency_p50': 2.2012699842453003, 'latency_p90': 2.4875208377838134}, {'batch_size': 5, 'throughput': 1.8823258622313044, 'latency_mean': 2.639392646551132, 'latency_p50': 2.6350561380386353, 'latency_p90': 2.973614001274109}, {'batch_size': 8, 'throughput': 2.0423083170783207, 'latency_mean': 3.883461289405823, 'latency_p50': 3.861005187034607, 'latency_p90': 4.376500058174133}, {'batch_size': 10, 'throughput': 2.0389841507039588, 'latency_mean': 4.864599982500076, 'latency_p50': 4.8783698081970215, 'latency_p90': 5.51254608631134}, {'batch_size': 12, 'throughput': 2.0176086806454383, 'latency_mean': 5.8653270268440245, 'latency_p50': 5.818166136741638, 'latency_p90': 6.922132968902588}, {'batch_size': 15, 'throughput': 2.0116820567335623, 'latency_mean': 7.309472028017044, 'latency_p50': 7.426189541816711, 'latency_p90': 8.142833995819093}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: stark2000s-ult-du-v1-5_v1

is_internal_developer: False

language_model: stark2000s/ult-du-v1.5

model_size: 8B

ranking_group: single

throughput_3p7s: 2.04

us_pacific_date: 2024-11-28

win_ratio: 0.4348423067321295

generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, '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': 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 stark2000s-ult-du-v1-5-v2-mkmlizer
Waiting for job on stark2000s-ult-du-v1-5-v2-mkmlizer to finish
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
stark2000s-ult-du-v1-5-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
stark2000s-ult-du-v1-5-v2-mkmlizer: ║     _____            __           __                                ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║       /___/                                                         ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║                                                                     ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  Version: 0.11.12                                                   ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  https://mk1.ai                                                     ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║                                                                     ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  belonging to:                                                      ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║                                                                     ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  Chai Research Corp.                                                ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ║                                                                     ║
stark2000s-ult-du-v1-5-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
stark2000s-ult-du-v1-5-v2-mkmlizer: Downloaded to shared memory in 36.195s
stark2000s-ult-du-v1-5-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpniqwdkdb, device:0
stark2000s-ult-du-v1-5-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
stark2000s-ult-du-v1-5-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/mk1/flywheel/functional/loader.py:55: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
stark2000s-ult-du-v1-5-v2-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
stark2000s-ult-du-v1-5-v2-mkmlizer: quantized model in 26.251s
stark2000s-ult-du-v1-5-v2-mkmlizer: Processed model stark2000s/ult-du-v1.5 in 62.446s
stark2000s-ult-du-v1-5-v2-mkmlizer: creating bucket guanaco-mkml-models
stark2000s-ult-du-v1-5-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
stark2000s-ult-du-v1-5-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2
stark2000s-ult-du-v1-5-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2/config.json
stark2000s-ult-du-v1-5-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2/special_tokens_map.json
stark2000s-ult-du-v1-5-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2/tokenizer_config.json
stark2000s-ult-du-v1-5-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2/tokenizer.json
stark2000s-ult-du-v1-5-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/stark2000s-ult-du-v1-5-v2/flywheel_model.0.safetensors
stark2000s-ult-du-v1-5-v2-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 6/291 [00:00<00:04, 58.79it/s]
Loading 0:   4%|▍         | 13/291 [00:00<00:04, 64.25it/s]
Loading 0:   8%|▊         | 22/291 [00:00<00:03, 75.77it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:03, 80.75it/s]
Loading 0:  14%|█▍        | 41/291 [00:00<00:02, 87.54it/s]
Loading 0:  17%|█▋        | 50/291 [00:00<00:02, 83.40it/s]
Loading 0:  20%|██        | 59/291 [00:00<00:02, 85.00it/s]
Loading 0:  23%|██▎       | 68/291 [00:00<00:02, 85.50it/s]
Loading 0:  26%|██▋       | 77/291 [00:00<00:02, 85.84it/s]
Loading 0:  30%|██▉       | 86/291 [00:02<00:10, 20.07it/s]
Loading 0:  33%|███▎      | 95/291 [00:02<00:07, 26.34it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:05, 32.08it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 38.30it/s]
Loading 0:  42%|████▏     | 123/291 [00:02<00:03, 49.66it/s]
Loading 0:  45%|████▌     | 132/291 [00:02<00:02, 56.75it/s]
Loading 0:  48%|████▊     | 141/291 [00:02<00:02, 63.26it/s]
Loading 0:  52%|█████▏    | 150/291 [00:02<00:02, 66.51it/s]
Loading 0:  55%|█████▍    | 159/291 [00:03<00:01, 71.14it/s]
Loading 0:  58%|█████▊    | 168/291 [00:03<00:01, 74.84it/s]
Loading 0:  61%|██████    | 177/291 [00:03<00:01, 78.69it/s]
Loading 0:  64%|██████▍   | 186/291 [00:03<00:01, 79.66it/s]
Loading 0:  67%|██████▋   | 195/291 [00:04<00:04, 20.28it/s]
Loading 0:  69%|██████▉   | 202/291 [00:04<00:03, 24.57it/s]
Loading 0:  73%|███████▎  | 211/291 [00:04<00:02, 31.53it/s]
Loading 0:  76%|███████▌  | 220/291 [00:04<00:01, 38.79it/s]
Loading 0:  79%|███████▊  | 229/291 [00:04<00:01, 46.48it/s]
Loading 0:  82%|████████▏ | 238/291 [00:05<00:00, 54.01it/s]
Loading 0:  85%|████████▍ | 247/291 [00:05<00:00, 61.18it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 64.82it/s]
Loading 0:  91%|█████████ | 265/291 [00:05<00:00, 68.08it/s]
Loading 0:  94%|█████████▍| 274/291 [00:05<00:00, 73.27it/s]
Loading 0:  98%|█████████▊| 286/291 [00:05<00:00, 76.49it/s]
                                                            
Job stark2000s-ult-du-v1-5-v2-mkmlizer completed after 83.77s with status: succeeded
Stopping job with name stark2000s-ult-du-v1-5-v2-mkmlizer
Pipeline stage MKMLizer completed in 84.29s
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 stark2000s-ult-du-v1-5-v2
Waiting for inference service stark2000s-ult-du-v1-5-v2 to be ready
Connection pool is full, discarding connection: %s. Connection pool size: %s
Inference service stark2000s-ult-du-v1-5-v2 ready after 130.64597415924072s
Pipeline stage MKMLDeployer completed in 131.17s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.7056829929351807s
Received healthy response to inference request in 1.346862554550171s
Received healthy response to inference request in 1.768157720565796s
Received healthy response to inference request in 1.3469610214233398s
Received healthy response to inference request in 1.1186308860778809s
5 requests
0 failed requests
5th percentile: 1.1642772197723388
10th percentile: 1.209923553466797
20th percentile: 1.301216220855713
30th percentile: 1.3468822479248046
40th percentile: 1.3469216346740722
50th percentile: 1.3469610214233398
60th percentile: 1.490449810028076
70th percentile: 1.6339385986328123
80th percentile: 1.7181779384613036
90th percentile: 1.7431678295135498
95th percentile: 1.755662775039673
99th percentile: 1.7656587314605714
mean time: 1.4572590351104737
Pipeline stage StressChecker completed in 8.52s
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 2.03s
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 2.32s
Shutdown handler de-registered
stark2000s-ult-du-v1-5_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
Pipeline stage OfflineFamilyFriendlyScorer completed in 2921.46s
Shutdown handler de-registered
stark2000s-ult-du-v1-5_v2 status is now inactive due to auto deactivation removed underperforming models