submission_id: rica40325-1300-dpo-v4_v1
developer_uid: rica40325
alignment_samples: 13104
alignment_score: -2.361482741622562
best_of: 16
celo_rating: 1179.25
display_name: rica40325-1300-dpo-v4_v1
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}
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': 512, 'best_of': 16, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: False
language_model: rica40325/1300-dpo-v4
latencies: [{'batch_size': 1, 'throughput': 0.9024170258371432, 'latency_mean': 1.1080404341220855, 'latency_p50': 1.1067068576812744, 'latency_p90': 1.2444321632385253}, {'batch_size': 4, 'throughput': 1.7687489673425418, 'latency_mean': 2.2497894489765167, 'latency_p50': 2.2461944818496704, 'latency_p90': 2.520434927940369}, {'batch_size': 5, 'throughput': 1.8615196924900044, 'latency_mean': 2.677850728034973, 'latency_p50': 2.6747887134552, 'latency_p90': 2.992160749435425}, {'batch_size': 8, 'throughput': 1.9813161119338376, 'latency_mean': 4.004051630496979, 'latency_p50': 3.9918181896209717, 'latency_p90': 4.52749342918396}, {'batch_size': 10, 'throughput': 1.9961259640374858, 'latency_mean': 4.965419933795929, 'latency_p50': 4.967578649520874, 'latency_p90': 5.744495153427124}, {'batch_size': 12, 'throughput': 1.9971006999321088, 'latency_mean': 5.929936105012894, 'latency_p50': 5.956694602966309, 'latency_p90': 6.754742407798767}, {'batch_size': 15, 'throughput': 2.0202607642001986, 'latency_mean': 7.272205016613007, 'latency_p50': 7.406443953514099, 'latency_p90': 8.074227237701416}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: rica40325/1300-dpo-v4
model_name: rica40325-1300-dpo-v4_v1
model_num_parameters: 8030261248.0
model_repo: rica40325/1300-dpo-v4
model_size: 8B
num_battles: 13104
num_wins: 5408
propriety_score: 0.7362924281984334
propriety_total_count: 1149.0
ranking_group: single
status: inactive
submission_type: basic
throughput_3p7s: 1.98
timestamp: 2024-09-05T14:21:34+00:00
us_pacific_date: 2024-09-05
win_ratio: 0.4126984126984127
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 rica40325-1300-dpo-v4-v1-mkmlizer
Waiting for job on rica40325-1300-dpo-v4-v1-mkmlizer to finish
rica40325-1300-dpo-v4-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rica40325-1300-dpo-v4-v1-mkmlizer: ║     _____            __           __                                ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║       /___/                                                         ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║                                                                     ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  https://mk1.ai                                                     ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║                                                                     ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  belonging to:                                                      ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║                                                                     ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  Chai Research Corp.                                                ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
rica40325-1300-dpo-v4-v1-mkmlizer: ║                                                                     ║
rica40325-1300-dpo-v4-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
Connection pool is full, discarding connection: %s. Connection pool size: %s
rica40325-1300-dpo-v4-v1-mkmlizer: Downloaded to shared memory in 62.080s
rica40325-1300-dpo-v4-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp7mlkwl2n, device:0
rica40325-1300-dpo-v4-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Failed to get response for submission mistralai-mixtral-8x7b_3473_v131: ('http://mistralai-mixtral-8x7b-3473-v131-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:35138->127.0.0.1:8080: read: connection reset by peer\n')
rica40325-1300-dpo-v4-v1-mkmlizer: quantized model in 28.640s
rica40325-1300-dpo-v4-v1-mkmlizer: Processed model rica40325/1300-dpo-v4 in 90.720s
rica40325-1300-dpo-v4-v1-mkmlizer: creating bucket guanaco-mkml-models
rica40325-1300-dpo-v4-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rica40325-1300-dpo-v4-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rica40325-1300-dpo-v4-v1
rica40325-1300-dpo-v4-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rica40325-1300-dpo-v4-v1/config.json
rica40325-1300-dpo-v4-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rica40325-1300-dpo-v4-v1/tokenizer_config.json
rica40325-1300-dpo-v4-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rica40325-1300-dpo-v4-v1/special_tokens_map.json
rica40325-1300-dpo-v4-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rica40325-1300-dpo-v4-v1/tokenizer.json
rica40325-1300-dpo-v4-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 5/291 [00:00<00:10, 26.60it/s]
Loading 0:   4%|▍         | 13/291 [00:00<00:06, 43.82it/s]
Loading 0:   6%|▌         | 18/291 [00:00<00:06, 44.80it/s]
Loading 0:   8%|▊         | 23/291 [00:00<00:07, 33.58it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:05, 44.43it/s]
Loading 0:  13%|█▎        | 37/291 [00:01<00:08, 29.46it/s]
Loading 0:  14%|█▍        | 41/291 [00:01<00:08, 28.81it/s]
Loading 0:  16%|█▋        | 48/291 [00:01<00:06, 36.59it/s]
Loading 0:  18%|█▊        | 53/291 [00:01<00:06, 36.92it/s]
Loading 0:  20%|█▉        | 58/291 [00:01<00:06, 37.23it/s]
Loading 0:  22%|██▏       | 63/291 [00:01<00:05, 38.98it/s]
Loading 0:  23%|██▎       | 68/291 [00:01<00:06, 33.51it/s]
Loading 0:  25%|██▌       | 74/291 [00:02<00:05, 38.99it/s]
Loading 0:  27%|██▋       | 79/291 [00:02<00:05, 36.80it/s]
Loading 0:  29%|██▉       | 84/291 [00:02<00:07, 26.87it/s]
Loading 0:  31%|███       | 90/291 [00:02<00:06, 31.63it/s]
Loading 0:  32%|███▏      | 94/291 [00:02<00:06, 31.48it/s]
Loading 0:  34%|███▍      | 99/291 [00:02<00:05, 34.92it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:05, 33.65it/s]
Loading 0:  37%|███▋      | 108/291 [00:03<00:04, 37.47it/s]
Loading 0:  39%|███▉      | 113/291 [00:03<00:04, 35.89it/s]
Loading 0:  40%|████      | 117/291 [00:03<00:04, 36.81it/s]
Loading 0:  42%|████▏     | 122/291 [00:03<00:04, 39.33it/s]
Loading 0:  44%|████▎     | 127/291 [00:03<00:04, 37.13it/s]
Loading 0:  46%|████▌     | 133/291 [00:03<00:04, 33.48it/s]
Loading 0:  47%|████▋     | 137/291 [00:03<00:04, 32.39it/s]
Loading 0:  48%|████▊     | 141/291 [00:04<00:04, 30.11it/s]
Loading 0:  51%|█████     | 147/291 [00:04<00:04, 35.07it/s]
Loading 0:  52%|█████▏    | 151/291 [00:04<00:04, 33.99it/s]
Loading 0:  54%|█████▎    | 156/291 [00:04<00:03, 37.06it/s]
Loading 0:  55%|█████▍    | 160/291 [00:04<00:03, 34.34it/s]
Loading 0:  57%|█████▋    | 165/291 [00:04<00:03, 35.13it/s]
Loading 0:  58%|█████▊    | 169/291 [00:04<00:03, 33.82it/s]
Loading 0:  60%|█████▉    | 174/291 [00:04<00:03, 35.48it/s]
Loading 0:  61%|██████    | 178/291 [00:05<00:03, 33.20it/s]
Loading 0:  63%|██████▎   | 183/291 [00:05<00:02, 36.89it/s]
Loading 0:  64%|██████▍   | 187/291 [00:05<00:03, 27.85it/s]
Loading 0:  66%|██████▌   | 191/291 [00:05<00:03, 28.50it/s]
Loading 0:  67%|██████▋   | 195/291 [00:05<00:03, 27.01it/s]
Loading 0:  69%|██████▉   | 201/291 [00:05<00:02, 31.66it/s]
Loading 0:  70%|███████   | 205/291 [00:06<00:02, 30.39it/s]
Loading 0:  72%|███████▏  | 210/291 [00:06<00:02, 33.65it/s]
Loading 0:  74%|███████▎  | 214/291 [00:06<00:02, 32.23it/s]
Loading 0:  75%|███████▌  | 219/291 [00:06<00:02, 34.92it/s]
Loading 0:  77%|███████▋  | 223/291 [00:06<00:02, 33.33it/s]
Loading 0:  78%|███████▊  | 227/291 [00:06<00:01, 33.80it/s]
Loading 0:  79%|███████▉  | 231/291 [00:06<00:01, 32.77it/s]
Loading 0:  81%|████████  | 235/291 [00:07<00:02, 25.36it/s]
Loading 0:  82%|████████▏ | 239/291 [00:07<00:02, 24.92it/s]
Loading 0:  85%|████████▍ | 246/291 [00:07<00:01, 33.04it/s]
Loading 0:  86%|████████▌ | 250/291 [00:07<00:01, 31.53it/s]
Loading 0:  88%|████████▊ | 255/291 [00:07<00:01, 34.56it/s]
Loading 0:  89%|████████▉ | 259/291 [00:07<00:00, 33.71it/s]
Loading 0:  91%|█████████ | 264/291 [00:07<00:00, 35.38it/s]
Loading 0:  92%|█████████▏| 268/291 [00:07<00:00, 33.25it/s]
Loading 0:  94%|█████████▍| 273/291 [00:08<00:00, 35.25it/s]
Loading 0:  95%|█████████▌| 277/291 [00:08<00:00, 33.27it/s]
Loading 0:  97%|█████████▋| 281/291 [00:08<00:00, 34.41it/s]
Loading 0:  98%|█████████▊| 286/291 [00:13<00:01,  2.61it/s]
Loading 0:  99%|█████████▉| 289/291 [00:13<00:00,  3.24it/s]
                                                            

Job rica40325-1300-dpo-v4-v1-mkmlizer completed after 115.96s with status: succeeded
Stopping job with name rica40325-1300-dpo-v4-v1-mkmlizer
Pipeline stage MKMLizer completed in 116.92s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.10s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rica40325-1300-dpo-v4-v1
Waiting for inference service rica40325-1300-dpo-v4-v1 to be ready
Inference service rica40325-1300-dpo-v4-v1 ready after 140.57843327522278s
Pipeline stage MKMLDeployer completed in 141.14s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.008286952972412s
Received healthy response to inference request in 1.578002691268921s
Received healthy response to inference request in 1.9682393074035645s
Received healthy response to inference request in 1.641249418258667s
Received healthy response to inference request in 1.3667497634887695s
5 requests
0 failed requests
5th percentile: 1.4090003490447998
10th percentile: 1.45125093460083
20th percentile: 1.5357521057128907
30th percentile: 1.5906520366668702
40th percentile: 1.6159507274627685
50th percentile: 1.641249418258667
60th percentile: 1.772045373916626
70th percentile: 1.902841329574585
80th percentile: 1.976248836517334
90th percentile: 1.992267894744873
95th percentile: 2.0002774238586425
99th percentile: 2.006685047149658
mean time: 1.7125056266784668
Pipeline stage StressChecker completed in 10.91s
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 6.45s
Shutdown handler de-registered
rica40325-1300-dpo-v4_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 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service rica40325-1300-dpo-v4-v1-profiler
Waiting for inference service rica40325-1300-dpo-v4-v1-profiler to be ready
Inference service rica40325-1300-dpo-v4-v1-profiler ready after 130.31139993667603s
Pipeline stage MKMLProfilerDeployer completed in 130.69s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/rica40325-1300-dpo-v4-v1-profiler-predictor-00001-deploymeklt7s:/code/chaiverse_profiler_1725546537 --namespace tenant-chaiml-guanaco
kubectl exec -it rica40325-1300-dpo-v4-v1-profiler-predictor-00001-deploymeklt7s --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725546537 && python profiles.py profile --best_of_n 16 --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_1725546537/summary.json'
kubectl exec -it rica40325-1300-dpo-v4-v1-profiler-predictor-00001-deploymeklt7s --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725546537/summary.json'
Pipeline stage MKMLProfilerRunner completed in 848.87s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-1300-dpo-v4-v1-profiler is running
Tearing down inference service rica40325-1300-dpo-v4-v1-profiler
Service rica40325-1300-dpo-v4-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.81s
Shutdown handler de-registered
rica40325-1300-dpo-v4_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics