shenzhi-wang-llama3-1-8b_9996

developer_uid: shiya

submission_id: shenzhi-wang-llama3-1-8b_9996_v1

model_name: Llama31-8B-Chinese-Chat

model_group: shenzhi-wang/Llama3.1-8B

status: torndown

timestamp: 2024-09-21T00:25:33+00:00

num_battles: 448875

num_wins: 204402

celo_rating: 1218.2

family_friendly_score: 0.0

submission_type: basic

model_repo: shenzhi-wang/Llama3.1-8B-Chinese-Chat

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.8635358881002675, 'latency_mean': 1.1579695224761963, 'latency_p50': 1.1676701307296753, 'latency_p90': 1.2737719058990478}, {'batch_size': 4, 'throughput': 1.828607131596326, 'latency_mean': 2.1793284714221954, 'latency_p50': 2.1853615045547485, 'latency_p90': 2.4472256183624266}, {'batch_size': 5, 'throughput': 1.9556202155499582, 'latency_mean': 2.543530591726303, 'latency_p50': 2.5640789270401, 'latency_p90': 2.8480534076690676}, {'batch_size': 8, 'throughput': 2.130423667939561, 'latency_mean': 3.7343252432346343, 'latency_p50': 3.761553406715393, 'latency_p90': 4.216649603843689}, {'batch_size': 10, 'throughput': 2.15216545426666, 'latency_mean': 4.617945420742035, 'latency_p50': 4.592981457710266, 'latency_p90': 5.312287020683288}, {'batch_size': 12, 'throughput': 2.1648363499372376, 'latency_mean': 5.490259439945221, 'latency_p50': 5.498508334159851, 'latency_p90': 6.1673398256301875}, {'batch_size': 15, 'throughput': 2.134514846708871, 'latency_mean': 6.9387581706047055, 'latency_p50': 6.9760377407073975, 'latency_p90': 7.77521014213562}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: Llama31-8B-Chinese-Chat

is_internal_developer: False

language_model: shenzhi-wang/Llama3.1-8B-Chinese-Chat

model_size: 8B

ranking_group: single

throughput_3p7s: 2.14

us_pacific_date: 2024-09-20

win_ratio: 0.45536507936507936

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 shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer
Waiting for job on shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer to finish
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║     _____            __           __                                ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║       /___/                                                         ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║                                                                     ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  https://mk1.ai                                                     ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║                                                                     ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  belonging to:                                                      ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║                                                                     ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  Chai Research Corp.                                                ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: ║                                                                     ║
shenzhi-wang-llama3-1-8b-9996-v1-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
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: Downloaded to shared memory in 100.240s
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpvj6ppaon, device:0
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: quantized model in 25.974s
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: Processed model shenzhi-wang/Llama3.1-8B-Chinese-Chat in 126.214s
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: creating bucket guanaco-mkml-models
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1/config.json
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1/special_tokens_map.json
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1/tokenizer_config.json
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1/tokenizer.json
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/shenzhi-wang-llama3-1-8b-9996-v1/flywheel_model.0.safetensors
shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 5/291 [00:00<00:08, 35.65it/s]
Loading 0:   5%|▍         | 14/291 [00:00<00:05, 47.52it/s]
Loading 0:   8%|▊         | 23/291 [00:00<00:05, 50.67it/s]
Loading 0:  11%|█         | 32/291 [00:00<00:04, 52.36it/s]
Loading 0:  14%|█▍        | 41/291 [00:00<00:04, 52.47it/s]
Loading 0:  17%|█▋        | 50/291 [00:00<00:04, 52.09it/s]
Loading 0:  20%|██        | 59/291 [00:01<00:04, 52.16it/s]
Loading 0:  23%|██▎       | 67/291 [00:01<00:03, 58.06it/s]
Loading 0:  25%|██▌       | 74/291 [00:01<00:03, 56.43it/s]
Loading 0:  27%|██▋       | 80/291 [00:01<00:03, 54.23it/s]
Loading 0:  30%|██▉       | 86/291 [00:01<00:05, 35.32it/s]
Loading 0:  32%|███▏      | 94/291 [00:01<00:04, 42.55it/s]
Loading 0:  34%|███▍      | 100/291 [00:02<00:04, 42.74it/s]
Loading 0:  36%|███▌      | 105/291 [00:02<00:04, 43.73it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:03, 49.59it/s]
Loading 0:  41%|████      | 118/291 [00:02<00:03, 47.80it/s]
Loading 0:  43%|████▎     | 124/291 [00:02<00:03, 48.70it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:03, 50.96it/s]
Loading 0:  47%|████▋     | 136/291 [00:02<00:03, 48.48it/s]
Loading 0:  49%|████▉     | 142/291 [00:02<00:03, 49.05it/s]
Loading 0:  51%|█████     | 149/291 [00:03<00:03, 47.06it/s]
Loading 0:  54%|█████▍    | 158/291 [00:03<00:02, 49.94it/s]
Loading 0:  57%|█████▋    | 166/291 [00:03<00:02, 55.73it/s]
Loading 0:  59%|█████▉    | 172/291 [00:03<00:02, 52.09it/s]
Loading 0:  61%|██████    | 178/291 [00:03<00:02, 53.99it/s]
Loading 0:  63%|██████▎   | 184/291 [00:03<00:02, 50.88it/s]
Loading 0:  65%|██████▌   | 190/291 [00:04<00:02, 35.07it/s]
Loading 0:  67%|██████▋   | 195/291 [00:04<00:02, 37.05it/s]
Loading 0:  69%|██████▉   | 202/291 [00:04<00:02, 43.70it/s]
Loading 0:  71%|███████▏  | 208/291 [00:04<00:01, 44.29it/s]
Loading 0:  73%|███████▎  | 213/291 [00:04<00:01, 43.61it/s]
Loading 0:  76%|███████▌  | 220/291 [00:04<00:01, 49.14it/s]
Loading 0:  78%|███████▊  | 226/291 [00:04<00:01, 46.93it/s]
Loading 0:  79%|███████▉  | 231/291 [00:04<00:01, 45.32it/s]
Loading 0:  82%|████████▏ | 238/291 [00:04<00:01, 49.99it/s]
Loading 0:  84%|████████▍ | 244/291 [00:05<00:01, 45.54it/s]
Loading 0:  86%|████████▌ | 249/291 [00:05<00:00, 45.25it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 51.19it/s]
Loading 0:  90%|█████████ | 262/291 [00:05<00:00, 49.43it/s]
Loading 0:  92%|█████████▏| 268/291 [00:05<00:00, 49.61it/s]
Loading 0:  94%|█████████▍| 274/291 [00:05<00:00, 51.12it/s]
Loading 0:  96%|█████████▌| 280/291 [00:05<00:00, 49.08it/s]
Loading 0:  98%|█████████▊| 285/291 [00:05<00:00, 46.51it/s]
Loading 0: 100%|█████████▉| 290/291 [00:11<00:00,  3.24it/s]
                                                            
Job shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer completed after 150.43s with status: succeeded
Stopping job with name shenzhi-wang-llama3-1-8b-9996-v1-mkmlizer
Pipeline stage MKMLizer completed in 151.44s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.08s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service shenzhi-wang-llama3-1-8b-9996-v1
Waiting for inference service shenzhi-wang-llama3-1-8b-9996-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
Inference service shenzhi-wang-llama3-1-8b-9996-v1 ready after 200.90104866027832s
Pipeline stage MKMLDeployer completed in 202.51s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.355759382247925s
Received healthy response to inference request in 1.318352460861206s
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
Received healthy response to inference request in 1.3301582336425781s
HTTPSConnectionPool(host='guanaco-submitter.chai-research.com', port=443): Read timed out. (read timeout=20)
Received unhealthy response to inference request!
Received healthy response to inference request in 2.5383267402648926s
5 requests
1 failed requests
5th percentile: 1.3207136154174806
10th percentile: 1.3230747699737548
20th percentile: 1.3277970790863036
30th percentile: 1.5717919349670408
40th percentile: 2.055059337615967
50th percentile: 2.5383267402648926
60th percentile: 2.8652997970581056
70th percentile: 3.192272853851318
80th percentile: 6.701780128479006
90th percentile: 13.393821620941164
95th percentile: 16.739842367172237
99th percentile: 19.416658964157104
mean time: 5.7256919860839846
%s, retrying in %s seconds...
Received healthy response to inference request in 2.5548086166381836s
Received healthy response to inference request in 1.733626127243042s
Received healthy response to inference request in 1.296489953994751s
Received healthy response to inference request in 1.3449771404266357s
Received healthy response to inference request in 1.3245129585266113s
5 requests
0 failed requests
5th percentile: 1.3020945549011231
Connection pool is full, discarding connection: %s. Connection pool size: %s
10th percentile: 1.307699155807495
20th percentile: 1.3189083576202392
30th percentile: 1.3286057949066161
40th percentile: 1.336791467666626
50th percentile: 1.3449771404266357
60th percentile: 1.5004367351531982
70th percentile: 1.6558963298797607
80th percentile: 1.8978626251220705
90th percentile: 2.226335620880127
95th percentile: 2.390572118759155
99th percentile: 2.521961317062378
mean time: 1.6508829593658447
Pipeline stage StressChecker completed in 46.20s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Pipeline stage TriggerMKMLProfilingPipeline completed in 5.95s
Shutdown handler de-registered
shenzhi-wang-llama3-1-8b_9996_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 0.14s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.13s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service shenzhi-wang-llama3-1-8b-9996-v1-profiler
Waiting for inference service shenzhi-wang-llama3-1-8b-9996-v1-profiler to be ready
Inference service shenzhi-wang-llama3-1-8b-9996-v1-profiler ready after 200.5324249267578s
Pipeline stage MKMLProfilerDeployer completed in 211.07s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/shenzhi-wang-llama3-c34d334e06f2f8b5ca7b0d353c8cf456-deplo9x6d8:/code/chaiverse_profiler_1726879039 --namespace tenant-chaiml-guanaco
kubectl exec -it shenzhi-wang-llama3-c34d334e06f2f8b5ca7b0d353c8cf456-deplo9x6d8 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726879039 && python profiles.py profile --best_of_n 8 --auto_batch 5 --batches 1,5,10,15,20,25,30,35,40,45,50,55,60,65,70,75,80,85,90,95,100,105,110,115,120,125,130,135,140,145,150,155,160,165,170,175,180,185,190,195 --samples 200 --input_tokens 1024 --output_tokens 64 --summary /code/chaiverse_profiler_1726879039/summary.json'
kubectl exec -it shenzhi-wang-llama3-c34d334e06f2f8b5ca7b0d353c8cf456-deplo9x6d8 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726879039/summary.json'
Pipeline stage MKMLProfilerRunner completed in 835.57s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service shenzhi-wang-llama3-1-8b-9996-v1-profiler is running
Tearing down inference service shenzhi-wang-llama3-1-8b-9996-v1-profiler
Service shenzhi-wang-llama3-1-8b-9996-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 16.45s
Shutdown handler de-registered
shenzhi-wang-llama3-1-8b_9996_v1 status is now inactive due to auto deactivation removed underperforming models
shenzhi-wang-llama3-1-8b_9996_v1 status is now torndown due to DeploymentManager action