submission_id: jic062-dpo-v1-6-nemo_v2
developer_uid: chace9580
best_of: 8
celo_rating: 1251.29
display_name: jic062-dpo-v1-6-nemo_v2
family_friendly_score: 0.0
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': 1.0, 'top_p': 0.9, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '/s', '[/INST]'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: False
language_model: jic062/dpo-v1.6-Nemo
latencies: [{'batch_size': 1, 'throughput': 0.6125217585540602, 'latency_mean': 1.632495746612549, 'latency_p50': 1.6309312582015991, 'latency_p90': 1.7907677173614502}, {'batch_size': 3, 'throughput': 1.0817938304541401, 'latency_mean': 2.760161876678467, 'latency_p50': 2.760886788368225, 'latency_p90': 2.999288535118103}, {'batch_size': 5, 'throughput': 1.2309812398526432, 'latency_mean': 4.046174170970917, 'latency_p50': 4.069713830947876, 'latency_p90': 4.472532343864441}, {'batch_size': 6, 'throughput': 1.2690452327040453, 'latency_mean': 4.702583719491958, 'latency_p50': 4.726541042327881, 'latency_p90': 5.3465454339981076}, {'batch_size': 8, 'throughput': 1.2526353815203415, 'latency_mean': 6.354067895412445, 'latency_p50': 6.353281378746033, 'latency_p90': 7.102592658996582}, {'batch_size': 10, 'throughput': 1.223552959222131, 'latency_mean': 8.121063417196273, 'latency_p50': 8.167364239692688, 'latency_p90': 9.319572615623475}]
max_input_tokens: 1024
max_output_tokens: 64
model_architecture: MistralForCausalLM
model_group: jic062/dpo-v1.6-Nemo
model_name: jic062-dpo-v1-6-nemo_v2
model_num_parameters: 12772070400.0
model_repo: jic062/dpo-v1.6-Nemo
model_size: 13B
num_battles: 22714
num_wins: 11752
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 1.21
timestamp: 2024-09-24T03:26:19+00:00
us_pacific_date: 2024-09-23
win_ratio: 0.517390155851017
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-v1-6-nemo-v2-mkmlizer
Waiting for job on jic062-dpo-v1-6-nemo-v2-mkmlizer to finish
jic062-dpo-v1-6-nemo-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ _____ __ __ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ /___/ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ Version: 0.10.1 ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ https://mk1.ai ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ The license key for the current software has been verified as ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ belonging to: ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ Chai Research Corp. ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ║ ║
jic062-dpo-v1-6-nemo-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
jic062-dpo-v1-6-nemo-v2-mkmlizer: Downloaded to shared memory in 31.142s
jic062-dpo-v1-6-nemo-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp1yuz_tyk, device:0
jic062-dpo-v1-6-nemo-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
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
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
jic062-dpo-v1-6-nemo-v2-mkmlizer: quantized model in 36.557s
jic062-dpo-v1-6-nemo-v2-mkmlizer: Processed model jic062/dpo-v1.6-Nemo in 67.699s
jic062-dpo-v1-6-nemo-v2-mkmlizer: creating bucket guanaco-mkml-models
jic062-dpo-v1-6-nemo-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-dpo-v1-6-nemo-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2
jic062-dpo-v1-6-nemo-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2/config.json
jic062-dpo-v1-6-nemo-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2/special_tokens_map.json
jic062-dpo-v1-6-nemo-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2/tokenizer_config.json
jic062-dpo-v1-6-nemo-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2/tokenizer.json
jic062-dpo-v1-6-nemo-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jic062-dpo-v1-6-nemo-v2/flywheel_model.0.safetensors
jic062-dpo-v1-6-nemo-v2-mkmlizer: Loading 0: 0%| | 0/363 [00:00<?, ?it/s] Loading 0: 1%|▏ | 5/363 [00:00<00:11, 32.19it/s] Loading 0: 4%|▎ | 13/363 [00:00<00:06, 52.75it/s] Loading 0: 5%|▌ | 19/363 [00:00<00:06, 49.58it/s] Loading 0: 7%|▋ | 25/363 [00:00<00:06, 49.54it/s] Loading 0: 9%|▊ | 31/363 [00:00<00:06, 51.91it/s] Loading 0: 10%|█ | 37/363 [00:00<00:06, 50.27it/s] Loading 0: 12%|█▏ | 43/363 [00:00<00:06, 48.17it/s] Loading 0: 13%|█▎ | 49/363 [00:00<00:06, 50.62it/s] Loading 0: 15%|█▌ | 55/363 [00:01<00:06, 49.29it/s] Loading 0: 17%|█▋ | 61/363 [00:01<00:09, 33.17it/s] Loading 0: 18%|█▊ | 66/363 [00:01<00:08, 34.46it/s] Loading 0: 20%|█▉ | 72/363 [00:01<00:07, 39.34it/s] Loading 0: 21%|██▏ | 78/363 [00:01<00:07, 40.66it/s] Loading 0: 23%|██▎ | 83/363 [00:01<00:06, 41.13it/s] Loading 0: 25%|██▍ | 90/363 [00:02<00:05, 46.97it/s] Loading 0: 26%|██▋ | 96/363 [00:02<00:05, 46.18it/s] Loading 0: 28%|██▊ | 101/363 [00:02<00:05, 44.06it/s] Loading 0: 30%|███ | 109/363 [00:02<00:04, 52.47it/s] Loading 0: 32%|███▏ | 115/363 [00:02<00:05, 47.12it/s] Loading 0: 33%|███▎ | 120/363 [00:02<00:05, 44.76it/s] Loading 0: 35%|███▍ | 126/363 [00:02<00:05, 46.66it/s] Loading 0: 36%|███▋ | 132/363 [00:02<00:04, 46.21it/s] Loading 0: 38%|███▊ | 137/363 [00:03<00:05, 44.81it/s] Loading 0: 39%|███▉ | 142/363 [00:03<00:06, 33.44it/s] Loading 0: 40%|████ | 146/363 [00:03<00:06, 33.87it/s] Loading 0: 41%|████▏ | 150/363 [00:03<00:06, 33.08it/s] Loading 0: 43%|████▎ | 156/363 [00:03<00:05, 39.10it/s] Loading 0: 44%|████▍ | 161/363 [00:03<00:05, 39.04it/s] Loading 0: 46%|████▌ | 166/363 [00:03<00:04, 39.78it/s] Loading 0: 47%|████▋ | 171/363 [00:04<00:04, 41.93it/s] Loading 0: 48%|████▊ | 176/363 [00:04<00:05, 36.82it/s] Loading 0: 50%|█████ | 183/363 [00:04<00:04, 44.58it/s] Loading 0: 52%|█████▏ | 188/363 [00:04<00:03, 44.46it/s] Loading 0: 53%|█████▎ | 193/363 [00:04<00:03, 45.38it/s] Loading 0: 55%|█████▍ | 199/363 [00:04<00:03, 45.16it/s] Loading 0: 56%|█████▌ | 204/363 [00:04<00:03, 43.49it/s] Loading 0: 58%|█████▊ | 210/363 [00:04<00:03, 47.50it/s] Loading 0: 59%|█████▉ | 215/363 [00:04<00:03, 47.08it/s] Loading 0: 61%|██████ | 222/363 [00:05<00:02, 47.62it/s] Loading 0: 63%|██████▎ | 227/363 [00:05<00:04, 31.59it/s] Loading 0: 64%|██████▎ | 231/363 [00:05<00:04, 31.35it/s] Loading 0: 66%|██████▌ | 238/363 [00:05<00:03, 37.69it/s] Loading 0: 67%|██████▋ | 244/363 [00:05<00:03, 38.51it/s] Loading 0: 69%|██████▊ | 249/363 [00:05<00:02, 38.96it/s] Loading 0: 71%|███████ | 256/363 [00:06<00:02, 44.44it/s] Loading 0: 72%|███████▏ | 262/363 [00:06<00:02, 45.07it/s] Loading 0: 74%|███████▎ | 267/363 [00:06<00:02, 43.64it/s] Loading 0: 75%|███████▍ | 272/363 [00:06<00:02, 44.53it/s] Loading 0: 76%|███████▋ | 277/363 [00:06<00:02, 36.88it/s] Loading 0: 77%|███████▋ | 281/363 [00:06<00:02, 37.55it/s] Loading 0: 79%|███████▊ | 285/363 [00:06<00:02, 35.61it/s] Loading 0: 80%|████████ | 292/363 [00:06<00:01, 42.14it/s] Loading 0: 82%|████████▏ | 298/363 [00:07<00:01, 42.97it/s] Loading 0: 83%|████████▎ | 303/363 [00:07<00:01, 44.56it/s] Loading 0: 85%|████████▍ | 308/363 [00:14<00:22, 2.42it/s] Loading 0: 86%|████████▌ | 312/363 [00:14<00:16, 3.13it/s] Loading 0: 88%|████████▊ | 320/363 [00:14<00:08, 5.18it/s] Loading 0: 90%|████████▉ | 326/363 [00:14<00:05, 7.13it/s] Loading 0: 91%|█████████ | 331/363 [00:14<00:03, 9.19it/s] Loading 0: 93%|█████████▎| 338/363 [00:14<00:01, 13.13it/s] Loading 0: 95%|█████████▍| 344/363 [00:14<00:01, 16.45it/s] Loading 0: 96%|█████████▌| 349/363 [00:15<00:00, 19.44it/s] Loading 0: 98%|█████████▊| 356/363 [00:15<00:00, 25.56it/s] Loading 0: 100%|█████████▉| 362/363 [00:15<00:00, 29.36it/s]
Job jic062-dpo-v1-6-nemo-v2-mkmlizer completed after 104.56s with status: succeeded
Stopping job with name jic062-dpo-v1-6-nemo-v2-mkmlizer
Pipeline stage MKMLizer completed in 105.56s
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 jic062-dpo-v1-6-nemo-v2
Waiting for inference service jic062-dpo-v1-6-nemo-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
Failed to get response for submission blend_negek_2024-09-20: ('http://chaiml-llama-8b-pairwis-8189-v27-predictor.tenant-chaiml-guanaco.k2.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'readfrom tcp 127.0.0.1:55092->127.0.0.1:8080: write tcp 127.0.0.1:55092->127.0.0.1:8080: use of closed network connection\n')
Inference service jic062-dpo-v1-6-nemo-v2 ready after 231.86827540397644s
Pipeline stage MKMLDeployer completed in 232.60s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.3749678134918213s
Received healthy response to inference request in 2.108210325241089s
Received healthy response to inference request in 2.757500171661377s
Received healthy response to inference request in 2.5550007820129395s
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 1.8289728164672852s
5 requests
0 failed requests
5th percentile: 1.884820318222046
10th percentile: 1.9406678199768066
20th percentile: 2.052362823486328
30th percentile: 2.1615618228912354
40th percentile: 2.268264818191528
50th percentile: 2.3749678134918213
60th percentile: 2.4469810009002684
70th percentile: 2.518994188308716
80th percentile: 2.595500659942627
90th percentile: 2.676500415802002
95th percentile: 2.7170002937316893
99th percentile: 2.7494001960754395
mean time: 2.3249303817749025
Pipeline stage StressChecker completed in 14.53s
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 9.51s
Shutdown handler de-registered
jic062-dpo-v1-6-nemo_v2 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 0.20s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.18s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service jic062-dpo-v1-6-nemo-v2-profiler
Waiting for inference service jic062-dpo-v1-6-nemo-v2-profiler to be ready
Inference service jic062-dpo-v1-6-nemo-v2-profiler ready after 200.53335189819336s
Pipeline stage MKMLProfilerDeployer completed in 201.05s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/jic062-dpo-v1-6-nemo-v2-profiler-predictor-00001-deploymenzggx5:/code/chaiverse_profiler_1727149008 --namespace tenant-chaiml-guanaco
kubectl exec -it jic062-dpo-v1-6-nemo-v2-profiler-predictor-00001-deploymenzggx5 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1727149008 && 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_1727149008/summary.json'
kubectl exec -it jic062-dpo-v1-6-nemo-v2-profiler-predictor-00001-deploymenzggx5 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1727149008/summary.json'
Pipeline stage MKMLProfilerRunner completed in 1160.44s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service jic062-dpo-v1-6-nemo-v2-profiler is running
Tearing down inference service jic062-dpo-v1-6-nemo-v2-profiler
Service jic062-dpo-v1-6-nemo-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.25s
Shutdown handler de-registered
jic062-dpo-v1-6-nemo_v2 status is now inactive due to auto deactivation removed underperforming models
jic062-dpo-v1-6-nemo_v2 status is now torndown due to DeploymentManager action