jic062-dpo-v1-nemo-c1000

submission_id: jic062-dpo-v1-nemo-c1000_v1

developer_uid: chace9580

alignment_samples: 10488

alignment_score: 0.416915207143911

best_of: 8

celo_rating: 1235.86

display_name: jic062-dpo-v1-nemo-c1000_v1

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}

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}

gpu_counts: {'NVIDIA RTX A5000': 1}

is_internal_developer: False

language_model: jic062/dpo-v1-Nemo-c1000

latencies: [{'batch_size': 1, 'throughput': 0.6893642555538623, 'latency_mean': 1.4505518293380737, 'latency_p50': 1.445564866065979, 'latency_p90': 1.6106532335281372}, {'batch_size': 3, 'throughput': 1.3435084776407584, 'latency_mean': 2.223885587453842, 'latency_p50': 2.2227851152420044, 'latency_p90': 2.477764439582825}, {'batch_size': 5, 'throughput': 1.582500109222889, 'latency_mean': 3.1440096211433413, 'latency_p50': 3.1697803735733032, 'latency_p90': 3.519953751564026}, {'batch_size': 6, 'throughput': 1.6198450795747412, 'latency_mean': 3.6749135315418244, 'latency_p50': 3.697352886199951, 'latency_p90': 4.1461841583251955}, {'batch_size': 8, 'throughput': 1.613816891199634, 'latency_mean': 4.924161229133606, 'latency_p50': 4.913261532783508, 'latency_p90': 5.566941714286804}, {'batch_size': 10, 'throughput': 1.5606583611464524, 'latency_mean': 6.358254046440124, 'latency_p50': 6.3846518993377686, 'latency_p90': 7.168094205856323}]

max_input_tokens: 512

max_output_tokens: 64

model_architecture: MistralForCausalLM

model_group: jic062/dpo-v1-Nemo-c1000

model_name: jic062-dpo-v1-nemo-c1000_v1

model_num_parameters: 12772070400.0

model_repo: jic062/dpo-v1-Nemo-c1000

model_size: 13B

num_battles: 10488

num_wins: 5132

propriety_score: 0.7269807280513919

propriety_total_count: 934.0

ranking_group: single

status: inactive

submission_type: basic

throughput_3p7s: 1.63

timestamp: 2024-09-10T02:28:51+00:00

us_pacific_date: 2024-09-09

win_ratio: 0.4893211289092296

Download Preference Data

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-nemo-c1000-v1-mkmlizer
Waiting for job on jic062-dpo-v1-nemo-c1000-v1-mkmlizer to finish
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║     _____            __           __                                ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║       /___/                                                         ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  https://mk1.ai                                                     ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  belonging to:                                                      ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  Chai Research Corp.                                                ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ║                                                                     ║
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: Downloaded to shared memory in 45.798s
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp15e78dsz, device:0
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: quantized model in 35.591s
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: Processed model jic062/dpo-v1-Nemo-c1000 in 81.389s
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: creating bucket guanaco-mkml-models
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
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1/config.json
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1/special_tokens_map.json
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1/tokenizer_config.json
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1/tokenizer.json
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jic062-dpo-v1-nemo-c1000-v1/flywheel_model.0.safetensors
jic062-dpo-v1-nemo-c1000-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:10, 33.32it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:06, 54.98it/s]
Loading 0:   5%|▌         | 19/363 [00:00<00:06, 50.38it/s]
Loading 0:   7%|▋         | 25/363 [00:00<00:06, 51.13it/s]
Loading 0:   9%|▊         | 31/363 [00:00<00:06, 50.82it/s]
Loading 0:  10%|█         | 37/363 [00:00<00:06, 47.63it/s]
Loading 0:  12%|█▏        | 42/363 [00:00<00:07, 43.47it/s]
Loading 0:  13%|█▎        | 49/363 [00:01<00:06, 47.79it/s]
Loading 0:  15%|█▍        | 54/363 [00:01<00:06, 46.90it/s]
Loading 0:  17%|█▋        | 60/363 [00:01<00:06, 44.22it/s]
Loading 0:  18%|█▊        | 65/363 [00:01<00:09, 31.46it/s]
Loading 0:  20%|█▉        | 72/363 [00:01<00:07, 38.69it/s]
Loading 0:  21%|██▏       | 78/363 [00:01<00:07, 38.28it/s]
Loading 0:  23%|██▎       | 83/363 [00:01<00:07, 37.17it/s]
Loading 0:  25%|██▍       | 89/363 [00:02<00:06, 41.08it/s]
Loading 0:  26%|██▌       | 94/363 [00:02<00:06, 40.35it/s]
Loading 0:  27%|██▋       | 99/363 [00:02<00:06, 40.61it/s]
Loading 0:  29%|██▊       | 104/363 [00:02<00:06, 42.75it/s]
Loading 0:  30%|███       | 110/363 [00:02<00:05, 43.10it/s]
Loading 0:  32%|███▏      | 115/363 [00:02<00:05, 44.79it/s]
Loading 0:  33%|███▎      | 120/363 [00:02<00:05, 41.51it/s]
Loading 0:  34%|███▍      | 125/363 [00:02<00:05, 43.23it/s]
Loading 0:  36%|███▌      | 130/363 [00:03<00:05, 44.51it/s]
Loading 0:  37%|███▋      | 136/363 [00:03<00:05, 40.86it/s]
Loading 0:  39%|███▉      | 142/363 [00:03<00:06, 36.03it/s]
Loading 0:  40%|████      | 146/363 [00:03<00:05, 36.41it/s]
Loading 0:  41%|████▏     | 150/363 [00:03<00:06, 34.13it/s]
Loading 0:  43%|████▎     | 156/363 [00:03<00:05, 38.10it/s]
Loading 0:  44%|████▍     | 160/363 [00:03<00:05, 38.24it/s]
Loading 0:  46%|████▌     | 166/363 [00:03<00:04, 42.85it/s]
Loading 0:  47%|████▋     | 172/363 [00:04<00:04, 43.34it/s]
Loading 0:  49%|████▉     | 177/363 [00:04<00:04, 40.61it/s]
Loading 0:  51%|█████     | 184/363 [00:04<00:03, 46.40it/s]
Loading 0:  52%|█████▏    | 190/363 [00:04<00:03, 45.47it/s]
Loading 0:  54%|█████▎    | 195/363 [00:04<00:03, 44.25it/s]
Loading 0:  55%|█████▌    | 201/363 [00:04<00:03, 46.54it/s]
Loading 0:  57%|█████▋    | 206/363 [00:04<00:03, 45.54it/s]
Loading 0:  58%|█████▊    | 211/363 [00:04<00:03, 46.00it/s]
Loading 0:  60%|█████▉    | 217/363 [00:05<00:03, 42.89it/s]
Loading 0:  61%|██████    | 222/363 [00:05<00:03, 44.19it/s]
Loading 0:  63%|██████▎   | 227/363 [00:05<00:04, 32.20it/s]
Loading 0:  64%|██████▎   | 231/363 [00:05<00:04, 32.22it/s]
Loading 0:  65%|██████▌   | 237/363 [00:05<00:03, 36.81it/s]
Loading 0:  67%|██████▋   | 242/363 [00:05<00:03, 37.37it/s]
Loading 0:  68%|██████▊   | 246/363 [00:05<00:03, 37.58it/s]
Loading 0:  69%|██████▉   | 250/363 [00:06<00:03, 36.96it/s]
Loading 0:  71%|███████   | 256/363 [00:06<00:02, 40.94it/s]
Loading 0:  72%|███████▏  | 262/363 [00:06<00:02, 39.46it/s]
Loading 0:  74%|███████▎  | 267/363 [00:06<00:02, 39.28it/s]
Loading 0:  75%|███████▌  | 273/363 [00:06<00:02, 44.12it/s]
Loading 0:  77%|███████▋  | 278/363 [00:06<00:01, 44.99it/s]
Loading 0:  78%|███████▊  | 283/363 [00:06<00:01, 45.37it/s]
Loading 0:  80%|███████▉  | 289/363 [00:06<00:01, 44.48it/s]
Loading 0:  81%|████████  | 294/363 [00:07<00:01, 41.73it/s]
Loading 0:  82%|████████▏ | 299/363 [00:07<00:01, 43.28it/s]
Loading 0:  84%|████████▎ | 304/363 [00:13<00:24,  2.45it/s]
Loading 0:  85%|████████▍ | 308/363 [00:14<00:17,  3.19it/s]
Loading 0:  86%|████████▌ | 312/363 [00:14<00:12,  4.18it/s]
Loading 0:  88%|████████▊ | 320/363 [00:14<00:06,  7.00it/s]
Loading 0:  90%|████████▉ | 326/363 [00:14<00:03,  9.56it/s]
Loading 0:  91%|█████████ | 331/363 [00:14<00:02, 12.09it/s]
Loading 0:  93%|█████████▎| 338/363 [00:14<00:01, 17.07it/s]
Loading 0:  95%|█████████▍| 344/363 [00:14<00:00, 21.08it/s]
Loading 0:  96%|█████████▌| 349/363 [00:14<00:00, 24.49it/s]
Loading 0:  98%|█████████▊| 356/363 [00:14<00:00, 31.44it/s]
Loading 0: 100%|█████████▉| 362/363 [00:15<00:00, 34.48it/s]
                                                            
Job jic062-dpo-v1-nemo-c1000-v1-mkmlizer completed after 107.58s with status: succeeded
Stopping job with name jic062-dpo-v1-nemo-c1000-v1-mkmlizer
Pipeline stage MKMLizer completed in 108.91s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.09s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service jic062-dpo-v1-nemo-c1000-v1
Waiting for inference service jic062-dpo-v1-nemo-c1000-v1 to be ready
Failed to get response for submission blend_fedek_2024-08-24: ('http://zonemercy-lexical-nemo-1518-v18-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:40876->127.0.0.1:8080: read: connection reset by peer\n')
Failed to get response for submission blend_siken_2024-09-09: ('http://chaiml-llama-8b-pairwis-8189-v19-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:35884->127.0.0.1:8080: read: connection reset by peer\n')
Inference service jic062-dpo-v1-nemo-c1000-v1 ready after 151.37575387954712s
Pipeline stage MKMLDeployer completed in 151.79s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.4045560359954834s
Received healthy response to inference request in 2.2352495193481445s
Received healthy response to inference request in 2.3153674602508545s
Received healthy response to inference request in 2.6802732944488525s
Received healthy response to inference request in 2.3530852794647217s
5 requests
0 failed requests
5th percentile: 2.2512731075286867
10th percentile: 2.2672966957092284
20th percentile: 2.2993438720703123
30th percentile: 2.322911024093628
40th percentile: 2.337998151779175
50th percentile: 2.3530852794647217
60th percentile: 2.483960485458374
70th percentile: 2.614835691452026
80th percentile: 2.825129842758179
90th percentile: 3.114842939376831
95th percentile: 3.259699487686157
99th percentile: 3.375584726333618
mean time: 2.5977063179016113
Pipeline stage StressChecker completed in 13.93s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 7.41s
Shutdown handler de-registered
jic062-dpo-v1-nemo-c1000_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 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service jic062-dpo-v1-nemo-c1000-v1-profiler
Waiting for inference service jic062-dpo-v1-nemo-c1000-v1-profiler to be ready
Inference service jic062-dpo-v1-nemo-c1000-v1-profiler ready after 160.38475584983826s
Pipeline stage MKMLProfilerDeployer completed in 160.75s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/jic062-dpo-v1-nemo-c1000-v1-profiler-predictor-00001-deplokxx8z:/code/chaiverse_profiler_1725935823 --namespace tenant-chaiml-guanaco
kubectl exec -it jic062-dpo-v1-nemo-c1000-v1-profiler-predictor-00001-deplokxx8z --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725935823 && 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_1725935823/summary.json'
kubectl exec -it jic062-dpo-v1-nemo-c1000-v1-profiler-predictor-00001-deplokxx8z --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725935823/summary.json'
Pipeline stage MKMLProfilerRunner completed in 945.60s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service jic062-dpo-v1-nemo-c1000-v1-profiler is running
Tearing down inference service jic062-dpo-v1-nemo-c1000-v1-profiler
Service jic062-dpo-v1-nemo-c1000-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.68s
Shutdown handler de-registered
jic062-dpo-v1-nemo-c1000_v1 status is now inactive due to auto deactivation removed underperforming models