jic062-dpo-v1-4-nemo

submission_id: jic062-dpo-v1-4-nemo_v2

developer_uid: chace9580

alignment_samples: 10176

alignment_score: -0.9606863430541009

best_of: 8

celo_rating: 1266.93

display_name: jic062-dpo-v1-4-nemo_v2

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': 0.85, 'top_p': 1.0, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '[/INST]'], '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-v1.4-Nemo

latencies: [{'batch_size': 1, 'throughput': 0.7078953703961188, 'latency_mean': 1.412560646533966, 'latency_p50': 1.4062175750732422, 'latency_p90': 1.5714553356170655}, {'batch_size': 3, 'throughput': 1.3524105028822402, 'latency_mean': 2.2149657344818117, 'latency_p50': 2.207529306411743, 'latency_p90': 2.4440405130386353}, {'batch_size': 5, 'throughput': 1.5819879772347534, 'latency_mean': 3.1412793469429015, 'latency_p50': 3.1524884700775146, 'latency_p90': 3.4978840827941893}, {'batch_size': 6, 'throughput': 1.6420374304116931, 'latency_mean': 3.627544691562653, 'latency_p50': 3.610406756401062, 'latency_p90': 4.091594624519348}, {'batch_size': 8, 'throughput': 1.5979113330502674, 'latency_mean': 4.972033543586731, 'latency_p50': 5.028353691101074, 'latency_p90': 5.651856446266175}, {'batch_size': 10, 'throughput': 1.5655440152020699, 'latency_mean': 6.345912663936615, 'latency_p50': 6.324586987495422, 'latency_p90': 7.236740899085999}]

max_input_tokens: 512

max_output_tokens: 64

model_architecture: MistralForCausalLM

model_group: jic062/dpo-v1.4-Nemo

model_name: jic062-dpo-v1-4-nemo_v2

model_num_parameters: 12772070400.0

model_repo: jic062/dpo-v1.4-Nemo

model_size: 13B

num_battles: 10176

num_wins: 5338

propriety_score: 0.7134894091415831

propriety_total_count: 897.0

ranking_group: single

status: inactive

submission_type: basic

throughput_3p7s: 1.65

timestamp: 2024-09-15T04:47:03+00:00

us_pacific_date: 2024-09-14

win_ratio: 0.5245676100628931

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-v1-4-nemo-v2-mkmlizer
Waiting for job on jic062-dpo-v1-4-nemo-v2-mkmlizer to finish
jic062-dpo-v1-4-nemo-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║     _____            __           __                                ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║       /___/                                                         ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║                                                                     ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  Version: 0.10.1                                                    ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  https://mk1.ai                                                     ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║                                                                     ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  belonging to:                                                      ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║                                                                     ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  Chai Research Corp.                                                ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ║                                                                     ║
jic062-dpo-v1-4-nemo-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jic062-dpo-v1-4-nemo-v2-mkmlizer: Downloaded to shared memory in 44.366s
jic062-dpo-v1-4-nemo-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp9rnwydhs, device:0
jic062-dpo-v1-4-nemo-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jic062-dpo-v1-4-nemo-v2-mkmlizer: quantized model in 35.105s
jic062-dpo-v1-4-nemo-v2-mkmlizer: Processed model jic062/dpo-v1.4-Nemo in 79.472s
jic062-dpo-v1-4-nemo-v2-mkmlizer: creating bucket guanaco-mkml-models
jic062-dpo-v1-4-nemo-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-dpo-v1-4-nemo-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2
jic062-dpo-v1-4-nemo-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2/config.json
jic062-dpo-v1-4-nemo-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2/special_tokens_map.json
jic062-dpo-v1-4-nemo-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2/tokenizer_config.json
jic062-dpo-v1-4-nemo-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2/tokenizer.json
jic062-dpo-v1-4-nemo-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jic062-dpo-v1-4-nemo-v2/flywheel_model.0.safetensors
jic062-dpo-v1-4-nemo-v2-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 5/363 [00:00<00:10, 34.51it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:06, 55.08it/s]
Loading 0:   5%|▌         | 19/363 [00:00<00:06, 51.25it/s]
Loading 0:   7%|▋         | 25/363 [00:00<00:06, 49.87it/s]
Loading 0:   9%|▊         | 31/363 [00:00<00:06, 52.19it/s]
Loading 0:  10%|█         | 37/363 [00:00<00:06, 48.94it/s]
Loading 0:  12%|█▏        | 42/363 [00:00<00:07, 45.19it/s]
Loading 0:  13%|█▎        | 49/363 [00:00<00:06, 50.12it/s]
Loading 0:  15%|█▌        | 55/363 [00:01<00:06, 48.34it/s]
Loading 0:  17%|█▋        | 60/363 [00:01<00:06, 48.48it/s]
Loading 0:  18%|█▊        | 65/363 [00:01<00:09, 32.41it/s]
Loading 0:  20%|█▉        | 72/363 [00:01<00:07, 39.99it/s]
Loading 0:  21%|██▏       | 78/363 [00:01<00:06, 41.87it/s]
Loading 0:  23%|██▎       | 83/363 [00:01<00:06, 42.84it/s]
Loading 0:  25%|██▍       | 90/363 [00:01<00:05, 48.13it/s]
Loading 0:  26%|██▋       | 96/363 [00:02<00:05, 47.39it/s]
Loading 0:  28%|██▊       | 102/363 [00:02<00:05, 48.32it/s]
Loading 0:  30%|███       | 110/363 [00:02<00:05, 50.47it/s]
Loading 0:  32%|███▏      | 116/363 [00:02<00:05, 48.03it/s]
Loading 0:  33%|███▎      | 121/363 [00:02<00:05, 46.77it/s]
Loading 0:  35%|███▍      | 127/363 [00:02<00:05, 40.05it/s]
Loading 0:  37%|███▋      | 134/363 [00:02<00:04, 46.43it/s]
Loading 0:  38%|███▊      | 139/363 [00:03<00:05, 44.03it/s]
Loading 0:  40%|███▉      | 144/363 [00:03<00:07, 29.04it/s]
Loading 0:  41%|████      | 149/363 [00:03<00:06, 31.63it/s]
Loading 0:  43%|████▎     | 157/363 [00:03<00:05, 39.85it/s]
Loading 0:  45%|████▍     | 163/363 [00:03<00:04, 40.93it/s]
Loading 0:  46%|████▋     | 168/363 [00:03<00:04, 41.50it/s]
Loading 0:  48%|████▊     | 175/363 [00:03<00:04, 46.35it/s]
Loading 0:  50%|████▉     | 181/363 [00:04<00:03, 45.59it/s]
Loading 0:  51%|█████     | 186/363 [00:04<00:03, 44.50it/s]
Loading 0:  53%|█████▎    | 193/363 [00:04<00:03, 49.03it/s]
Loading 0:  55%|█████▍    | 199/363 [00:04<00:03, 46.59it/s]
Loading 0:  56%|█████▌    | 204/363 [00:04<00:03, 42.23it/s]
Loading 0:  58%|█████▊    | 211/363 [00:04<00:03, 46.48it/s]
Loading 0:  60%|█████▉    | 217/363 [00:04<00:03, 42.96it/s]
Loading 0:  61%|██████▏   | 223/363 [00:05<00:04, 34.08it/s]
Loading 0:  63%|██████▎   | 227/363 [00:05<00:03, 34.69it/s]
Loading 0:  64%|██████▎   | 231/363 [00:05<00:03, 33.43it/s]
Loading 0:  65%|██████▌   | 237/363 [00:05<00:03, 37.66it/s]
Loading 0:  66%|██████▋   | 241/363 [00:05<00:03, 36.98it/s]
Loading 0:  68%|██████▊   | 247/363 [00:05<00:02, 41.20it/s]
Loading 0:  70%|██████▉   | 253/363 [00:05<00:02, 41.36it/s]
Loading 0:  71%|███████   | 258/363 [00:06<00:02, 42.17it/s]
Loading 0:  73%|███████▎  | 265/363 [00:06<00:02, 47.77it/s]
Loading 0:  75%|███████▍  | 271/363 [00:06<00:01, 47.35it/s]
Loading 0:  76%|███████▌  | 276/363 [00:06<00:01, 45.35it/s]
Loading 0:  78%|███████▊  | 283/363 [00:06<00:01, 50.10it/s]
Loading 0:  80%|███████▉  | 289/363 [00:06<00:01, 47.75it/s]
Loading 0:  81%|████████  | 294/363 [00:06<00:01, 43.95it/s]
Loading 0:  83%|████████▎ | 302/363 [00:06<00:01, 52.36it/s]
Loading 0:  85%|████████▍ | 308/363 [00:13<00:18,  2.94it/s]
Loading 0:  86%|████████▌ | 312/363 [00:13<00:13,  3.64it/s]
Loading 0:  88%|████████▊ | 319/363 [00:13<00:08,  5.44it/s]
Loading 0:  89%|████████▉ | 324/363 [00:14<00:05,  7.10it/s]
Loading 0:  91%|█████████ | 330/363 [00:14<00:03,  9.52it/s]
Loading 0:  93%|█████████▎| 339/363 [00:14<00:01, 14.19it/s]
Loading 0:  96%|█████████▌| 347/363 [00:14<00:00, 19.52it/s]
Loading 0:  97%|█████████▋| 353/363 [00:14<00:00, 23.09it/s]
Loading 0:  99%|█████████▉| 359/363 [00:14<00:00, 26.98it/s]
                                                            
Job jic062-dpo-v1-4-nemo-v2-mkmlizer completed after 108.39s with status: succeeded
Stopping job with name jic062-dpo-v1-4-nemo-v2-mkmlizer
Pipeline stage MKMLizer completed in 110.76s
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 jic062-dpo-v1-4-nemo-v2
Waiting for inference service jic062-dpo-v1-4-nemo-v2 to be ready
Inference service jic062-dpo-v1-4-nemo-v2 ready after 171.97923851013184s
Pipeline stage MKMLDeployer completed in 174.24s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.349841594696045s
Received healthy response to inference request in 1.7979888916015625s
Received healthy response to inference request in 2.2267470359802246s
Received healthy response to inference request in 2.408895254135132s
Received healthy response to inference request in 2.084644079208374s
5 requests
0 failed requests
5th percentile: 1.8553199291229248
10th percentile: 1.912650966644287
20th percentile: 2.0273130416870115
30th percentile: 2.113064670562744
40th percentile: 2.1699058532714846
50th percentile: 2.2267470359802246
60th percentile: 2.275984859466553
70th percentile: 2.325222682952881
80th percentile: 2.3616523265838625
90th percentile: 2.385273790359497
95th percentile: 2.3970845222473143
99th percentile: 2.4065331077575682
mean time: 2.1736233711242674
Pipeline stage StressChecker completed in 12.19s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
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
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 8.23s
Shutdown handler de-registered
jic062-dpo-v1-4-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 0.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service jic062-dpo-v1-4-nemo-v2-profiler
Waiting for inference service jic062-dpo-v1-4-nemo-v2-profiler to be ready
Inference service jic062-dpo-v1-4-nemo-v2-profiler ready after 170.41287231445312s
Pipeline stage MKMLProfilerDeployer completed in 171.40s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/jic062-dpo-v1-4-nemo-v2-profiler-predictor-00001-deploymenszhth:/code/chaiverse_profiler_1726376157 --namespace tenant-chaiml-guanaco
kubectl exec -it jic062-dpo-v1-4-nemo-v2-profiler-predictor-00001-deploymenszhth --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726376157 && 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_1726376157/summary.json'
kubectl exec -it jic062-dpo-v1-4-nemo-v2-profiler-predictor-00001-deploymenszhth --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726376157/summary.json'
Pipeline stage MKMLProfilerRunner completed in 942.11s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service jic062-dpo-v1-4-nemo-v2-profiler is running
Tearing down inference service jic062-dpo-v1-4-nemo-v2-profiler
Service jic062-dpo-v1-4-nemo-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 11.60s
Shutdown handler de-registered
jic062-dpo-v1-4-nemo_v2 status is now inactive due to auto deactivation removed underperforming models