jic062-dpo-v1-2-nemo-c500

developer_uid: chace9580

submission_id: jic062-dpo-v1-2-nemo-c500_v1

model_name: jic062-dpo-v1-2-nemo-c500_v1

model_group: jic062/dpo-v1.2-Nemo-c50

status: torndown

timestamp: 2024-09-11T04:03:24+00:00

num_battles: 10576

num_wins: 3958

celo_rating: 1155.93

family_friendly_score: 0.0

submission_type: basic

model_repo: jic062/dpo-v1.2-Nemo-c500

model_architecture: MistralForCausalLM

model_num_parameters: 12772070400.0

best_of: 8

max_input_tokens: 512

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.691243164979831, 'latency_mean': 1.4466006457805634, 'latency_p50': 1.4495887756347656, 'latency_p90': 1.6008866786956788}, {'batch_size': 3, 'throughput': 1.3317282778478805, 'latency_mean': 2.24319122672081, 'latency_p50': 2.2600873708724976, 'latency_p90': 2.4852947473526}, {'batch_size': 5, 'throughput': 1.5755177834451612, 'latency_mean': 3.1548821651935577, 'latency_p50': 3.192306876182556, 'latency_p90': 3.5483023881912232}, {'batch_size': 6, 'throughput': 1.6177754515659202, 'latency_mean': 3.687238829135895, 'latency_p50': 3.6781270503997803, 'latency_p90': 4.153065133094787}, {'batch_size': 8, 'throughput': 1.613034825735949, 'latency_mean': 4.923209717273712, 'latency_p50': 4.938737630844116, 'latency_p90': 5.638543725013733}, {'batch_size': 10, 'throughput': 1.5755705367320934, 'latency_mean': 6.304563139677048, 'latency_p50': 6.34716010093689, 'latency_p90': 7.336219143867493}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: jic062-dpo-v1-2-nemo-c500_v1

is_internal_developer: False

language_model: jic062/dpo-v1.2-Nemo-c500

model_size: 13B

ranking_group: single

throughput_3p7s: 1.63

us_pacific_date: 2024-09-10

win_ratio: 0.37424357034795763

generation_params: {'temperature': 0.75, 'top_p': 1.0, 'min_p': 0.1, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '\n\n', '\nYou:', '[/INST]', '<|im_end|>', '</s>'], 'max_input_tokens': 512, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '[INST]system\n{memory}[/INST]\n', 'prompt_template': '[INST]user\n{prompt}[/INST]\n', 'bot_template': '[INST]assistant\n{bot_name}: {message}[/INST]\n', 'user_template': '[INST]user\n{user_name}: {message}[/INST]\n', 'response_template': '[INST]assistant\n{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 jic062-dpo-v1-2-nemo-c500-v1-mkmlizer
Waiting for job on jic062-dpo-v1-2-nemo-c500-v1-mkmlizer to finish
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║     _____            __           __                                ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║       /___/                                                         ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  https://mk1.ai                                                     ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  belonging to:                                                      ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  Chai Research Corp.                                                ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Retrying (%r) after connection broken by '%r': %s
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: Downloaded to shared memory in 51.414s
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp7b6kwwj0, device:0
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: quantized model in 35.985s
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: Processed model jic062/dpo-v1.2-Nemo-c500 in 87.399s
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: creating bucket guanaco-mkml-models
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-dpo-v1-2-nemo-c500-v1
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-dpo-v1-2-nemo-c500-v1/config.json
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-dpo-v1-2-nemo-c500-v1/special_tokens_map.json
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-dpo-v1-2-nemo-c500-v1/tokenizer_config.json
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-dpo-v1-2-nemo-c500-v1/tokenizer.json
jic062-dpo-v1-2-nemo-c500-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:11, 31.24it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:06, 50.42it/s]
Loading 0:   5%|▌         | 19/363 [00:00<00:07, 45.43it/s]
Loading 0:   7%|▋         | 24/363 [00:00<00:07, 43.46it/s]
Loading 0:   9%|▊         | 31/363 [00:00<00:06, 49.69it/s]
Loading 0:  10%|█         | 37/363 [00:00<00:07, 44.93it/s]
Loading 0:  12%|█▏        | 42/363 [00:00<00:07, 43.15it/s]
Loading 0:  13%|█▎        | 49/363 [00:01<00:06, 48.13it/s]
Loading 0:  15%|█▍        | 54/363 [00:01<00:06, 48.12it/s]
Loading 0:  17%|█▋        | 60/363 [00:01<00:06, 47.20it/s]
Loading 0:  18%|█▊        | 65/363 [00:01<00:09, 31.30it/s]
Loading 0:  20%|█▉        | 72/363 [00:01<00:07, 38.11it/s]
Loading 0:  21%|██▏       | 78/363 [00:01<00:07, 39.45it/s]
Loading 0:  23%|██▎       | 83/363 [00:01<00:06, 40.47it/s]
Loading 0:  25%|██▍       | 90/363 [00:02<00:06, 45.32it/s]
Loading 0:  26%|██▌       | 95/363 [00:02<00:05, 45.03it/s]
Loading 0:  28%|██▊       | 100/363 [00:02<00:06, 37.78it/s]
Loading 0:  29%|██▉       | 106/363 [00:02<00:06, 41.86it/s]
Loading 0:  31%|███       | 112/363 [00:02<00:05, 44.14it/s]
Loading 0:  32%|███▏      | 117/363 [00:02<00:06, 39.81it/s]
Loading 0:  34%|███▎      | 122/363 [00:02<00:05, 40.89it/s]
Loading 0:  35%|███▍      | 127/363 [00:03<00:06, 34.73it/s]
Loading 0:  37%|███▋      | 134/363 [00:03<00:05, 41.34it/s]
Loading 0:  38%|███▊      | 139/363 [00:03<00:05, 40.22it/s]
Loading 0:  40%|███▉      | 144/363 [00:03<00:08, 26.68it/s]
Loading 0:  41%|████      | 149/363 [00:03<00:07, 29.21it/s]
Loading 0:  43%|████▎     | 156/363 [00:03<00:05, 36.23it/s]
Loading 0:  44%|████▍     | 161/363 [00:04<00:05, 37.63it/s]
Loading 0:  46%|████▌     | 166/363 [00:04<00:04, 39.58it/s]
Loading 0:  47%|████▋     | 172/363 [00:04<00:04, 38.85it/s]
Loading 0:  49%|████▉     | 177/363 [00:04<00:04, 38.49it/s]
Loading 0:  50%|█████     | 183/363 [00:04<00:04, 43.05it/s]
Loading 0:  52%|█████▏    | 188/363 [00:04<00:04, 42.99it/s]
Loading 0:  53%|█████▎    | 193/363 [00:04<00:03, 43.37it/s]
Loading 0:  55%|█████▍    | 198/363 [00:04<00:03, 44.84it/s]
Loading 0:  56%|█████▌    | 203/363 [00:05<00:04, 37.56it/s]
Loading 0:  58%|█████▊    | 210/363 [00:05<00:03, 45.29it/s]
Loading 0:  59%|█████▉    | 215/363 [00:05<00:03, 43.93it/s]
Loading 0:  61%|██████    | 221/363 [00:05<00:02, 47.92it/s]
Loading 0:  63%|██████▎   | 227/363 [00:05<00:04, 31.76it/s]
Loading 0:  64%|██████▍   | 232/363 [00:05<00:03, 33.26it/s]
Loading 0:  66%|██████▌   | 238/363 [00:05<00:03, 37.27it/s]
Loading 0:  67%|██████▋   | 244/363 [00:06<00:03, 37.92it/s]
Loading 0:  69%|██████▊   | 249/363 [00:06<00:03, 37.51it/s]
Loading 0:  70%|███████   | 255/363 [00:06<00:02, 42.17it/s]
Loading 0:  72%|███████▏  | 260/363 [00:06<00:02, 42.13it/s]
Loading 0:  73%|███████▎  | 265/363 [00:06<00:02, 42.73it/s]
Loading 0:  75%|███████▍  | 271/363 [00:06<00:02, 41.55it/s]
Loading 0:  76%|███████▌  | 276/363 [00:06<00:02, 40.49it/s]
Loading 0:  78%|███████▊  | 282/363 [00:06<00:01, 44.20it/s]
Loading 0:  79%|███████▉  | 287/363 [00:07<00:01, 44.09it/s]
Loading 0:  80%|████████  | 292/363 [00:07<00:01, 45.34it/s]
Loading 0:  82%|████████▏ | 298/363 [00:07<00:01, 45.05it/s]
Loading 0:  84%|████████▎ | 304/363 [00:14<00:22,  2.66it/s]
Loading 0:  85%|████████▍ | 308/363 [00:14<00:16,  3.38it/s]
Loading 0:  86%|████████▌ | 312/363 [00:14<00:11,  4.35it/s]
Loading 0:  88%|████████▊ | 320/363 [00:14<00:06,  7.14it/s]
Loading 0:  90%|████████▉ | 326/363 [00:14<00:03,  9.52it/s]
Loading 0:  91%|█████████ | 331/363 [00:14<00:02, 11.98it/s]
Loading 0:  93%|█████████▎| 338/363 [00:14<00:01, 16.73it/s]
Loading 0:  95%|█████████▍| 344/363 [00:15<00:00, 20.29it/s]
Loading 0:  96%|█████████▌| 349/363 [00:15<00:00, 23.40it/s]
Loading 0:  98%|█████████▊| 356/363 [00:15<00:00, 30.02it/s]
Loading 0: 100%|█████████▉| 362/363 [00:15<00:00, 32.93it/s]
                                                            
Job jic062-dpo-v1-2-nemo-c500-v1-mkmlizer completed after 116.39s with status: succeeded
Stopping job with name jic062-dpo-v1-2-nemo-c500-v1-mkmlizer
Pipeline stage MKMLizer completed in 119.92s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service jic062-dpo-v1-2-nemo-c500-v1
Waiting for inference service jic062-dpo-v1-2-nemo-c500-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 jic062-dpo-v1-2-nemo-c500-v1 ready after 160.8551709651947s
Pipeline stage MKMLDeployer completed in 161.52s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.0664753913879395s
Received healthy response to inference request in 1.9238312244415283s
Received healthy response to inference request in 1.9242753982543945s
Received healthy response to inference request in 2.2193262577056885s
Received healthy response to inference request in 1.7061212062835693s
5 requests
0 failed requests
5th percentile: 1.7496632099151612
10th percentile: 1.7932052135467529
20th percentile: 1.8802892208099364
30th percentile: 1.9239200592041015
40th percentile: 1.924097728729248
50th percentile: 1.9242753982543945
60th percentile: 2.042295742034912
70th percentile: 2.1603160858154298
80th percentile: 2.388756084442139
90th percentile: 2.727615737915039
95th percentile: 2.897045564651489
99th percentile: 3.032589426040649
mean time: 2.168005895614624
Pipeline stage StressChecker completed in 11.58s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
Failed to get response for submission trace2333-mistral-dpo-trail2_v1: ('http://trace2333-mistral-dpo-trail2-v1-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:53714->127.0.0.1:8080: read: connection reset by peer\n')
Pipeline stage TriggerMKMLProfilingPipeline completed in 5.22s
Shutdown handler de-registered
jic062-dpo-v1-2-nemo-c500_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 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service jic062-dpo-v1-2-nemo-c500-v1-profiler
Waiting for inference service jic062-dpo-v1-2-nemo-c500-v1-profiler to be ready
Inference service jic062-dpo-v1-2-nemo-c500-v1-profiler ready after 160.4437699317932s
Pipeline stage MKMLProfilerDeployer completed in 160.80s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/jic062-dpo-v1-2-nemoc1e9510ded3fc4fbcc35b90dc82780d1-deplo8x74s:/code/chaiverse_profiler_1726027906 --namespace tenant-chaiml-guanaco
kubectl exec -it jic062-dpo-v1-2-nemoc1e9510ded3fc4fbcc35b90dc82780d1-deplo8x74s --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726027906 && 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 512 --output_tokens 64 --summary /code/chaiverse_profiler_1726027906/summary.json'
kubectl exec -it jic062-dpo-v1-2-nemoc1e9510ded3fc4fbcc35b90dc82780d1-deplo8x74s --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726027906/summary.json'
Pipeline stage MKMLProfilerRunner completed in 946.36s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service jic062-dpo-v1-2-nemo-c500-v1-profiler is running
Tearing down inference service jic062-dpo-v1-2-nemo-c500-v1-profiler
Service jic062-dpo-v1-2-nemo-c500-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.72s
Shutdown handler de-registered
jic062-dpo-v1-2-nemo-c500_v1 status is now inactive due to auto deactivation removed underperforming models
jic062-dpo-v1-2-nemo-c500_v1 status is now torndown due to DeploymentManager action