stark2000s-ult-du-gpt-v1-6

developer_uid: stark2000s

submission_id: stark2000s-ult-du-gpt-v1-6_v1

model_name: stark2000s-ult-du-gpt-v1-6_v1

model_group: stark2000s/ult-du-gpt-v1

status: torndown

timestamp: 2024-11-29T01:32:44+00:00

num_battles: 12944

num_wins: 5663

celo_rating: 1212.44

family_friendly_score: 0.5734

family_friendly_standard_error: 0.006994461237293406

submission_type: basic

model_repo: stark2000s/ult-du-gpt-v1.6

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.8569514793464521, 'latency_mean': 1.1668605172634126, 'latency_p50': 1.1713427305221558, 'latency_p90': 1.285763931274414}, {'batch_size': 4, 'throughput': 1.7892454705765277, 'latency_mean': 2.226517599821091, 'latency_p50': 2.2507976293563843, 'latency_p90': 2.458768844604492}, {'batch_size': 5, 'throughput': 1.9356612572493919, 'latency_mean': 2.573856483697891, 'latency_p50': 2.58963143825531, 'latency_p90': 2.852812099456787}, {'batch_size': 8, 'throughput': 2.1045088367661253, 'latency_mean': 3.7759424376487734, 'latency_p50': 3.8025803565979004, 'latency_p90': 4.244400930404663}, {'batch_size': 10, 'throughput': 2.1618089549403714, 'latency_mean': 4.586739747524262, 'latency_p50': 4.578598260879517, 'latency_p90': 5.127688074111938}, {'batch_size': 12, 'throughput': 2.2090408699505706, 'latency_mean': 5.3797574234008785, 'latency_p50': 5.320847511291504, 'latency_p90': 6.036277508735656}, {'batch_size': 15, 'throughput': 2.2305249216822887, 'latency_mean': 6.635617233514786, 'latency_p50': 6.634029030799866, 'latency_p90': 7.396101760864258}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: stark2000s-ult-du-gpt-v1-6_v1

is_internal_developer: False

language_model: stark2000s/ult-du-gpt-v1.6

model_size: 8B

ranking_group: single

throughput_3p7s: 2.11

us_pacific_date: 2024-11-28

win_ratio: 0.4375

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': 1024, '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': 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-gpt-v1-6-v1-mkmlizer
Waiting for job on stark2000s-ult-du-gpt-v1-6-v1-mkmlizer to finish
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║     _____            __           __                                ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║       /___/                                                         ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║                                                                     ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  Version: 0.11.12                                                   ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  https://mk1.ai                                                     ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║                                                                     ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  belonging to:                                                      ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║                                                                     ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  Chai Research Corp.                                                ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ║                                                                     ║
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: Downloaded to shared memory in 41.294s
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp3wxhhz7s, device:0
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
stark2000s-ult-du-gpt-v1-6-v1-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-gpt-v1-6-v1-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: quantized model in 26.168s
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: Processed model stark2000s/ult-du-gpt-v1.6 in 67.462s
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: creating bucket guanaco-mkml-models
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1/config.json
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1/special_tokens_map.json
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1/tokenizer_config.json
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1/tokenizer.json
Connection pool is full, discarding connection: %s. Connection pool size: %s
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/stark2000s-ult-du-gpt-v1-6-v1/flywheel_model.0.safetensors
stark2000s-ult-du-gpt-v1-6-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 4/291 [00:00<00:07, 39.69it/s]
Loading 0:   4%|▍         | 13/291 [00:00<00:04, 69.18it/s]
Loading 0:   8%|▊         | 22/291 [00:00<00:03, 78.14it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:03, 77.31it/s]
Loading 0:  14%|█▎        | 40/291 [00:00<00:03, 76.81it/s]
Loading 0:  17%|█▋        | 49/291 [00:00<00:03, 80.06it/s]
Loading 0:  20%|█▉        | 58/291 [00:00<00:02, 82.04it/s]
Loading 0:  23%|██▎       | 67/291 [00:00<00:02, 82.48it/s]
Loading 0:  26%|██▌       | 76/291 [00:00<00:02, 80.50it/s]
Loading 0:  29%|██▉       | 85/291 [00:02<00:10, 20.31it/s]
Loading 0:  33%|███▎      | 97/291 [00:02<00:06, 28.21it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 39.66it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:03, 45.50it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:03, 52.14it/s]
Loading 0:  49%|████▉     | 142/291 [00:02<00:02, 59.11it/s]
Loading 0:  54%|█████▍    | 157/291 [00:02<00:01, 70.19it/s]
Loading 0:  57%|█████▋    | 166/291 [00:03<00:01, 71.91it/s]
Loading 0:  60%|██████    | 175/291 [00:03<00:01, 75.34it/s]
Loading 0:  64%|██████▍   | 187/291 [00:04<00:04, 23.43it/s]
Loading 0:  67%|██████▋   | 196/291 [00:04<00:03, 28.95it/s]
Loading 0:  70%|███████   | 205/291 [00:04<00:02, 34.73it/s]
Loading 0:  74%|███████▎  | 214/291 [00:04<00:01, 41.11it/s]
Loading 0:  77%|███████▋  | 223/291 [00:04<00:01, 48.57it/s]
Loading 0:  80%|███████▉  | 232/291 [00:04<00:01, 56.04it/s]
Loading 0:  83%|████████▎ | 241/291 [00:05<00:00, 62.77it/s]
Loading 0:  86%|████████▌ | 250/291 [00:05<00:00, 66.93it/s]
Loading 0:  89%|████████▉ | 259/291 [00:05<00:00, 67.59it/s]
Loading 0:  92%|█████████▏| 268/291 [00:05<00:00, 72.58it/s]
Loading 0:  95%|█████████▌| 277/291 [00:05<00:00, 77.00it/s]
Loading 0:  98%|█████████▊| 286/291 [00:05<00:00, 79.99it/s]
                                                            
Job stark2000s-ult-du-gpt-v1-6-v1-mkmlizer completed after 93.93s with status: succeeded
Stopping job with name stark2000s-ult-du-gpt-v1-6-v1-mkmlizer
Pipeline stage MKMLizer completed in 94.45s
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-gpt-v1-6-v1
Waiting for inference service stark2000s-ult-du-gpt-v1-6-v1 to be ready
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
Inference service stark2000s-ult-du-gpt-v1-6-v1 ready after 130.4908730983734s
Pipeline stage MKMLDeployer completed in 131.07s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0528616905212402s
Received healthy response to inference request in 1.3004696369171143s
Received healthy response to inference request in 1.5496594905853271s
Received healthy response to inference request in 1.163337230682373s
Received healthy response to inference request in 1.575197458267212s
5 requests
0 failed requests
5th percentile: 1.1907637119293213
10th percentile: 1.2181901931762695
20th percentile: 1.273043155670166
30th percentile: 1.3503076076507567
40th percentile: 1.449983549118042
50th percentile: 1.5496594905853271
60th percentile: 1.559874677658081
70th percentile: 1.570089864730835
80th percentile: 1.6707303047180178
90th percentile: 1.861795997619629
95th percentile: 1.9573288440704344
99th percentile: 2.033755121231079
mean time: 1.5283051013946534
Pipeline stage StressChecker completed in 8.98s
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 2.21s
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 2.02s
Shutdown handler de-registered
stark2000s-ult-du-gpt-v1-6_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 stark2000s-ult-du-gpt-v1-6-v1-profiler
Waiting for inference service stark2000s-ult-du-gpt-v1-6-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
Pipeline stage OfflineFamilyFriendlyScorer completed in 2645.66s
Shutdown handler de-registered
stark2000s-ult-du-gpt-v1-6_v1 status is now inactive due to auto deactivation removed underperforming models
stark2000s-ult-du-gpt-v1-6_v1 status is now torndown due to DeploymentManager action
stark2000s-ult-du-gpt-v1-6_v1 status is now torndown due to DeploymentManager action
admin requested tearing down of wespro-dr-rp-samantha-l3-8b_v1
stark2000s-ult-du-gpt-v1-6_v1 status is now torndown due to DeploymentManager action
Shutdown handler not registered because Python interpreter is not running in the main thread