stark2000s-ultra-du-v1-4

developer_uid: stark2000s

submission_id: stark2000s-ultra-du-v1-4_v2

model_name: stark2000s-ultra-du-v1-4_v1

model_group: stark2000s/ultra_du-v1.4

status: torndown

timestamp: 2024-11-10T10:13:43+00:00

num_battles: 18754

num_wins: 8869

celo_rating: 1221.19

family_friendly_score: 0.5568

family_friendly_standard_error: 0.007025293730514049

submission_type: basic

model_repo: stark2000s/ultra_du-v1.4

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.8608808070085988, 'latency_mean': 1.1615428447723388, 'latency_p50': 1.1602816581726074, 'latency_p90': 1.2908814191818236}, {'batch_size': 3, 'throughput': 1.5257811959169505, 'latency_mean': 1.961249829530716, 'latency_p50': 1.964179277420044, 'latency_p90': 2.2000192403793335}, {'batch_size': 5, 'throughput': 1.7053073578172988, 'latency_mean': 2.916597949266434, 'latency_p50': 2.9361852407455444, 'latency_p90': 3.2293442726135253}, {'batch_size': 6, 'throughput': 1.7533637055869165, 'latency_mean': 3.3972862803936006, 'latency_p50': 3.4084078073501587, 'latency_p90': 3.80097382068634}, {'batch_size': 8, 'throughput': 1.8510825317200263, 'latency_mean': 4.282130318880081, 'latency_p50': 4.285045742988586, 'latency_p90': 4.863341450691223}, {'batch_size': 10, 'throughput': 1.8361261572217518, 'latency_mean': 5.384067444801331, 'latency_p50': 5.414209485054016, 'latency_p90': 6.147931361198426}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: stark2000s-ultra-du-v1-4_v1

is_internal_developer: False

language_model: stark2000s/ultra_du-v1.4

model_size: 8B

ranking_group: single

throughput_3p7s: 1.81

us_pacific_date: 2024-11-10

win_ratio: 0.47291244534499305

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-ultra-du-v1-4-v2-mkmlizer
Waiting for job on stark2000s-ultra-du-v1-4-v2-mkmlizer to finish
stark2000s-ultra-du-v1-4-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║     _____            __           __                                ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║       /___/                                                         ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║                                                                     ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  Version: 0.11.33                                                   ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  https://mk1.ai                                                     ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║                                                                     ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  belonging to:                                                      ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║                                                                     ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  Chai Research Corp.                                                ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ║                                                                     ║
stark2000s-ultra-du-v1-4-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Failed to get response for submission sao10k-mn-12b-lyra-v4a1_v3: ('http://chaiml-llama-8b-pairwis-8189-v27-predictor.tenant-chaiml-guanaco.k2.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:37734->127.0.0.1:8080: read: connection reset by peer\n')
stark2000s-ultra-du-v1-4-v2-mkmlizer: Downloaded to shared memory in 35.847s
stark2000s-ultra-du-v1-4-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp3wso_obz, device:0
stark2000s-ultra-du-v1-4-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
stark2000s-ultra-du-v1-4-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-ultra-du-v1-4-v2-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
Connection pool is full, discarding connection: %s. Connection pool size: %s
stark2000s-ultra-du-v1-4-v2-mkmlizer: quantized model in 26.969s
stark2000s-ultra-du-v1-4-v2-mkmlizer: Processed model stark2000s/ultra_du-v1.4 in 62.817s
stark2000s-ultra-du-v1-4-v2-mkmlizer: creating bucket guanaco-mkml-models
stark2000s-ultra-du-v1-4-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
stark2000s-ultra-du-v1-4-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2
stark2000s-ultra-du-v1-4-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2/config.json
stark2000s-ultra-du-v1-4-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2/special_tokens_map.json
stark2000s-ultra-du-v1-4-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2/tokenizer_config.json
stark2000s-ultra-du-v1-4-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2/tokenizer.json
stark2000s-ultra-du-v1-4-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/stark2000s-ultra-du-v1-4-v2/flywheel_model.0.safetensors
stark2000s-ultra-du-v1-4-v2-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:05, 47.71it/s]
Loading 0:   5%|▌         | 16/291 [00:00<00:04, 67.02it/s]
Loading 0:   9%|▊         | 25/291 [00:00<00:04, 60.18it/s]
Loading 0:  12%|█▏        | 34/291 [00:00<00:04, 63.36it/s]
Loading 0:  15%|█▍        | 43/291 [00:00<00:03, 62.26it/s]
Loading 0:  18%|█▊        | 52/291 [00:00<00:04, 59.42it/s]
Loading 0:  21%|██        | 61/291 [00:01<00:03, 60.58it/s]
Loading 0:  24%|██▍       | 70/291 [00:01<00:03, 64.61it/s]
Loading 0:  27%|██▋       | 79/291 [00:01<00:03, 66.63it/s]
Loading 0:  30%|██▉       | 86/291 [00:02<00:11, 18.50it/s]
Loading 0:  32%|███▏      | 94/291 [00:02<00:08, 23.46it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:06, 29.84it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 36.83it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:04, 42.35it/s]
Loading 0:  45%|████▍     | 130/291 [00:03<00:03, 49.91it/s]
Loading 0:  48%|████▊     | 139/291 [00:03<00:02, 57.30it/s]
Loading 0:  51%|█████     | 148/291 [00:03<00:02, 63.70it/s]
Loading 0:  54%|█████▍    | 157/291 [00:03<00:02, 59.55it/s]
Loading 0:  57%|█████▋    | 166/291 [00:03<00:02, 62.28it/s]
Loading 0:  60%|██████    | 175/291 [00:03<00:01, 67.45it/s]
Loading 0:  63%|██████▎   | 184/291 [00:03<00:01, 69.82it/s]
Loading 0:  66%|██████▌   | 192/291 [00:04<00:04, 20.25it/s]
Loading 0:  68%|██████▊   | 198/291 [00:05<00:03, 23.82it/s]
Loading 0:  70%|███████   | 204/291 [00:05<00:03, 27.49it/s]
Loading 0:  73%|███████▎  | 211/291 [00:05<00:02, 31.11it/s]
Loading 0:  76%|███████▌  | 220/291 [00:05<00:01, 38.66it/s]
Loading 0:  79%|███████▊  | 229/291 [00:05<00:01, 47.24it/s]
Loading 0:  82%|████████▏ | 238/291 [00:05<00:00, 54.69it/s]
Loading 0:  85%|████████▍ | 247/291 [00:05<00:00, 58.24it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 60.36it/s]
Loading 0:  91%|█████████ | 265/291 [00:06<00:00, 64.05it/s]
Loading 0:  94%|█████████▍| 274/291 [00:06<00:00, 68.35it/s]
Loading 0:  97%|█████████▋| 283/291 [00:06<00:00, 70.84it/s]
Loading 0: 100%|██████████| 291/291 [00:11<00:00,  5.00it/s]
                                                            
Job stark2000s-ultra-du-v1-4-v2-mkmlizer completed after 85.32s with status: succeeded
Stopping job with name stark2000s-ultra-du-v1-4-v2-mkmlizer
Pipeline stage MKMLizer completed in 85.82s
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-ultra-du-v1-4-v2
Waiting for inference service stark2000s-ultra-du-v1-4-v2 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
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-ultra-du-v1-4-v2 ready after 171.1546287536621s
Pipeline stage MKMLDeployer completed in 171.70s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.800619125366211s
Received healthy response to inference request in 1.2375273704528809s
Received healthy response to inference request in 1.2820947170257568s
Received healthy response to inference request in 1.3356223106384277s
Received healthy response to inference request in 1.2953481674194336s
5 requests
0 failed requests
5th percentile: 1.246440839767456
10th percentile: 1.2553543090820312
20th percentile: 1.2731812477111817
30th percentile: 1.2847454071044921
40th percentile: 1.290046787261963
50th percentile: 1.2953481674194336
60th percentile: 1.3114578247070312
70th percentile: 1.3275674819946288
80th percentile: 1.4286216735839845
90th percentile: 1.6146203994750976
95th percentile: 1.7076197624206542
99th percentile: 1.7820192527770995
mean time: 1.390242338180542
Pipeline stage StressChecker completed in 8.19s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%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
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 5.65s
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 2.36s
Shutdown handler de-registered
stark2000s-ultra-du-v1-4_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 2822.20s
Shutdown handler de-registered
stark2000s-ultra-du-v1-4_v2 status is now inactive due to auto deactivation removed underperforming models
stark2000s-ultra-du-v1-4_v2 status is now torndown due to DeploymentManager action