thedrummer-cydonia-22b-v1

submission_id: thedrummer-cydonia-22b-v1_v1
developer_uid: Meliodia
best_of: 8
celo_rating: 1246.16
display_name: thedrummer-cydonia-22b-v1_v1
family_friendly_score: 0.0
formatter: {'memory_template': '', 'prompt_template': '', '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': 1024, 'best_of': 8, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A6000': 1}
is_internal_developer: True
language_model: TheDrummer/Cydonia-22B-v1
latencies: [{'batch_size': 1, 'throughput': 0.3800863111184711, 'latency_mean': 2.630906808376312, 'latency_p50': 2.6439945697784424, 'latency_p90': 2.88749680519104}, {'batch_size': 2, 'throughput': 0.6008862680774604, 'latency_mean': 3.3202645444869994, 'latency_p50': 3.3233522176742554, 'latency_p90': 3.620865249633789}, {'batch_size': 3, 'throughput': 0.752498921792413, 'latency_mean': 3.972115000486374, 'latency_p50': 3.9954447746276855, 'latency_p90': 4.407931590080262}, {'batch_size': 4, 'throughput': 0.8922008161248313, 'latency_mean': 4.470612322092056, 'latency_p50': 4.4903724193573, 'latency_p90': 4.992277145385742}, {'batch_size': 5, 'throughput': 0.9663664009401598, 'latency_mean': 5.130267134904861, 'latency_p50': 5.136885166168213, 'latency_p90': 5.785717177391052}]
max_input_tokens: 1024
max_output_tokens: 64
model_architecture: MistralForCausalLM
model_group: TheDrummer/Cydonia-22B-v
model_name: thedrummer-cydonia-22b-v1_v1
model_num_parameters: 22247282688.0
model_repo: TheDrummer/Cydonia-22B-v1
model_size: 22B
num_battles: 15477
num_wins: 7882
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 0.7
timestamp: 2024-09-24T04:30:56+00:00
us_pacific_date: 2024-09-23
win_ratio: 0.5092718227046585
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 thedrummer-cydonia-22b-v1-v1-mkmlizer
Waiting for job on thedrummer-cydonia-22b-v1-v1-mkmlizer to finish
thedrummer-cydonia-22b-v1-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║     _____            __           __                                ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║       /___/                                                         ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║                                                                     ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  https://mk1.ai                                                     ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║                                                                     ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  belonging to:                                                      ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║                                                                     ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  Chai Research Corp.                                                ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ║                                                                     ║
thedrummer-cydonia-22b-v1-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Retrying (%r) after connection broken by '%r': %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
thedrummer-cydonia-22b-v1-v1-mkmlizer: Downloaded to shared memory in 70.571s
thedrummer-cydonia-22b-v1-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpk1l1u7b8, device:0
thedrummer-cydonia-22b-v1-v1-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
thedrummer-cydonia-22b-v1-v1-mkmlizer: quantized model in 42.249s
thedrummer-cydonia-22b-v1-v1-mkmlizer: Processed model TheDrummer/Cydonia-22B-v1 in 112.820s
thedrummer-cydonia-22b-v1-v1-mkmlizer: creating bucket guanaco-mkml-models
thedrummer-cydonia-22b-v1-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
thedrummer-cydonia-22b-v1-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/config.json
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/special_tokens_map.json
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/tokenizer_config.json
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/tokenizer.model
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/tokenizer.json
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
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/flywheel_model.1.safetensors
thedrummer-cydonia-22b-v1-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/thedrummer-cydonia-22b-v1-v1/flywheel_model.0.safetensors
thedrummer-cydonia-22b-v1-v1-mkmlizer: 
Loading 0:   0%|          | 0/507 [00:00<?, ?it/s]
Loading 0:   1%|          | 5/507 [00:00<00:15, 31.57it/s]
Loading 0:   2%|▏         | 12/507 [00:00<00:10, 46.82it/s]
Loading 0:   3%|▎         | 17/507 [00:00<00:11, 43.38it/s]
Loading 0:   4%|▍         | 22/507 [00:00<00:11, 41.29it/s]
Loading 0:   5%|▌         | 27/507 [00:00<00:11, 42.21it/s]
Loading 0:   6%|▋         | 32/507 [00:00<00:14, 32.65it/s]
Loading 0:   8%|▊         | 39/507 [00:00<00:12, 38.62it/s]
Loading 0:   9%|▊         | 44/507 [00:01<00:11, 38.75it/s]
Loading 0:  10%|▉         | 49/507 [00:01<00:13, 34.11it/s]
Loading 0:  10%|█         | 53/507 [00:01<00:16, 27.25it/s]
Loading 0:  11%|█         | 57/507 [00:01<00:16, 27.06it/s]
Loading 0:  12%|█▏        | 63/507 [00:01<00:13, 32.21it/s]
Loading 0:  13%|█▎        | 67/507 [00:01<00:13, 32.61it/s]
Loading 0:  14%|█▍        | 72/507 [00:02<00:11, 36.56it/s]
Loading 0:  15%|█▌        | 78/507 [00:02<00:10, 40.85it/s]
Loading 0:  16%|█▋        | 83/507 [00:02<00:10, 40.82it/s]
Loading 0:  17%|█▋        | 88/507 [00:02<00:10, 40.50it/s]
Loading 0:  18%|█▊        | 93/507 [00:02<00:10, 41.25it/s]
Loading 0:  19%|█▉        | 98/507 [00:02<00:12, 33.90it/s]
Loading 0:  21%|██        | 105/507 [00:02<00:09, 40.22it/s]
Loading 0:  22%|██▏       | 110/507 [00:02<00:10, 39.52it/s]
Loading 0:  23%|██▎       | 115/507 [00:03<00:12, 30.58it/s]
Loading 0:  23%|██▎       | 119/507 [00:03<00:12, 32.20it/s]
Loading 0:  24%|██▍       | 123/507 [00:03<00:12, 30.67it/s]
Loading 0:  25%|██▌       | 129/507 [00:03<00:10, 35.03it/s]
Loading 0:  26%|██▌       | 133/507 [00:03<00:10, 34.79it/s]
Loading 0:  27%|██▋       | 138/507 [00:03<00:10, 36.81it/s]
Loading 0:  28%|██▊       | 142/507 [00:03<00:10, 36.43it/s]
Loading 0:  29%|██▉       | 147/507 [00:04<00:09, 38.80it/s]
Loading 0:  30%|██▉       | 151/507 [00:04<00:09, 37.02it/s]
Loading 0:  31%|███       | 156/507 [00:04<00:09, 38.89it/s]
Loading 0:  32%|███▏      | 160/507 [00:04<00:09, 38.09it/s]
Loading 0:  32%|███▏      | 164/507 [00:04<00:09, 37.71it/s]
Loading 0:  33%|███▎      | 169/507 [00:04<00:11, 30.08it/s]
Loading 0:  34%|███▍      | 173/507 [00:04<00:10, 31.20it/s]
Loading 0:  35%|███▍      | 177/507 [00:05<00:10, 30.35it/s]
Loading 0:  36%|███▌      | 183/507 [00:05<00:09, 35.33it/s]
Loading 0:  37%|███▋      | 187/507 [00:05<00:09, 34.94it/s]
Loading 0:  38%|███▊      | 192/507 [00:05<00:08, 37.79it/s]
Loading 0:  39%|███▊      | 196/507 [00:05<00:08, 37.53it/s]
Loading 0:  40%|███▉      | 201/507 [00:05<00:07, 39.29it/s]
Loading 0:  40%|████      | 205/507 [00:05<00:07, 38.56it/s]
Loading 0:  41%|████▏     | 210/507 [00:05<00:07, 40.47it/s]
Loading 0:  42%|████▏     | 215/507 [00:05<00:07, 40.40it/s]
Loading 0:  43%|████▎     | 220/507 [00:06<00:08, 34.86it/s]
Loading 0:  44%|████▍     | 224/507 [00:06<00:09, 28.47it/s]
Loading 0:  45%|████▌     | 230/507 [00:06<00:08, 30.82it/s]
Loading 0:  47%|████▋     | 237/507 [00:06<00:07, 37.57it/s]
Loading 0:  48%|████▊     | 242/507 [00:06<00:06, 38.17it/s]
Loading 0:  49%|████▊     | 247/507 [00:06<00:06, 38.35it/s]
Loading 0:  50%|████▉     | 251/507 [00:06<00:06, 38.22it/s]
Loading 0:  50%|█████     | 255/507 [00:07<00:06, 37.97it/s]
Loading 0:  51%|█████     | 259/507 [00:07<00:06, 37.19it/s]
Loading 0:  52%|█████▏    | 264/507 [00:07<00:06, 39.00it/s]
Loading 0:  53%|█████▎    | 268/507 [00:07<00:06, 37.93it/s]
Loading 0:  54%|█████▍    | 273/507 [00:07<00:05, 40.06it/s]
Loading 0:  55%|█████▍    | 278/507 [00:07<00:05, 39.99it/s]
Loading 0:  56%|█████▌    | 283/507 [00:07<00:05, 38.34it/s]
Loading 0:  57%|█████▋    | 287/507 [00:08<00:08, 26.34it/s]
Loading 0:  58%|█████▊    | 293/507 [00:08<00:07, 29.70it/s]
Loading 0:  59%|█████▉    | 299/507 [00:22<00:07, 29.70it/s]
Loading 0:  59%|█████▉    | 300/507 [00:22<02:41,  1.28it/s]
Loading 0:  60%|█████▉    | 302/507 [00:22<02:20,  1.46it/s]
Loading 0:  61%|██████    | 307/507 [00:22<01:34,  2.11it/s]
Loading 0:  61%|██████    | 310/507 [00:22<01:15,  2.60it/s]
Loading 0:  62%|██████▏   | 314/507 [00:22<00:54,  3.57it/s]
Loading 0:  63%|██████▎   | 318/507 [00:23<00:39,  4.84it/s]
Loading 0:  64%|██████▎   | 322/507 [00:23<00:28,  6.48it/s]
Loading 0:  64%|██████▍   | 327/507 [00:23<00:19,  9.16it/s]
Loading 0:  65%|██████▌   | 331/507 [00:23<00:15, 11.60it/s]
Loading 0:  66%|██████▌   | 335/507 [00:23<00:12, 14.32it/s]
Loading 0:  67%|██████▋   | 340/507 [00:23<00:10, 16.16it/s]
Loading 0:  68%|██████▊   | 344/507 [00:23<00:08, 19.12it/s]
Loading 0:  69%|██████▊   | 348/507 [00:24<00:07, 21.01it/s]
Loading 0:  70%|██████▉   | 354/507 [00:24<00:05, 26.83it/s]
Loading 0:  71%|███████   | 358/507 [00:24<00:05, 28.43it/s]
Loading 0:  72%|███████▏  | 363/507 [00:24<00:04, 31.46it/s]
Loading 0:  72%|███████▏  | 367/507 [00:24<00:04, 32.38it/s]
Loading 0:  73%|███████▎  | 372/507 [00:24<00:03, 35.64it/s]
Loading 0:  74%|███████▍  | 376/507 [00:24<00:03, 35.38it/s]
Loading 0:  75%|███████▌  | 381/507 [00:24<00:03, 37.42it/s]
Loading 0:  76%|███████▌  | 385/507 [00:24<00:03, 36.75it/s]
Loading 0:  77%|███████▋  | 389/507 [00:25<00:03, 36.66it/s]
Loading 0:  78%|███████▊  | 393/507 [00:25<00:03, 36.81it/s]
Loading 0:  78%|███████▊  | 397/507 [00:25<00:04, 26.79it/s]
Loading 0:  79%|███████▉  | 401/507 [00:25<00:03, 26.88it/s]
Loading 0:  80%|████████  | 408/507 [00:25<00:02, 34.63it/s]
Loading 0:  81%|████████▏ | 412/507 [00:25<00:02, 34.77it/s]
Loading 0:  82%|████████▏ | 417/507 [00:25<00:02, 37.25it/s]
Loading 0:  83%|████████▎ | 421/507 [00:26<00:02, 36.55it/s]
Loading 0:  84%|████████▍ | 426/507 [00:26<00:02, 38.67it/s]
Loading 0:  85%|████████▍ | 430/507 [00:26<00:02, 37.74it/s]
Loading 0:  86%|████████▌ | 435/507 [00:26<00:01, 39.28it/s]
Loading 0:  87%|████████▋ | 439/507 [00:26<00:01, 37.68it/s]
Loading 0:  88%|████████▊ | 444/507 [00:26<00:01, 39.57it/s]
Loading 0:  88%|████████▊ | 448/507 [00:26<00:01, 37.34it/s]
Loading 0:  89%|████████▉ | 453/507 [00:26<00:01, 39.38it/s]
Loading 0:  90%|█████████ | 457/507 [00:29<00:08,  5.87it/s]
Loading 0:  91%|█████████ | 460/507 [00:29<00:06,  7.14it/s]
Loading 0:  92%|█████████▏| 465/507 [00:29<00:04,  9.71it/s]
Loading 0:  93%|█████████▎| 472/507 [00:29<00:02, 14.83it/s]
Loading 0:  94%|█████████▍| 476/507 [00:29<00:01, 17.42it/s]
Loading 0:  95%|█████████▍| 481/507 [00:29<00:01, 21.45it/s]
Loading 0:  96%|█████████▌| 485/507 [00:29<00:00, 23.97it/s]
Loading 0:  97%|█████████▋| 490/507 [00:29<00:00, 28.44it/s]
Loading 0:  98%|█████████▊| 495/507 [00:30<00:00, 31.38it/s]
Loading 0:  99%|█████████▊| 500/507 [00:30<00:00, 33.55it/s]
Loading 0: 100%|█████████▉| 505/507 [00:30<00:00, 36.79it/s]
                                                            
Job thedrummer-cydonia-22b-v1-v1-mkmlizer completed after 137.09s with status: succeeded
Stopping job with name thedrummer-cydonia-22b-v1-v1-mkmlizer
Pipeline stage MKMLizer completed in 139.02s
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 thedrummer-cydonia-22b-v1-v1
Waiting for inference service thedrummer-cydonia-22b-v1-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
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
Failed to get response for submission mistralai-mistral-nemo_9330_v102: ('http://chaiml-llama-8b-pairwis-8189-v27-predictor.tenant-chaiml-guanaco.k2.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'readfrom tcp 127.0.0.1:40542->127.0.0.1:8080: write tcp 127.0.0.1:40542->127.0.0.1:8080: use of closed network connection\n')
Inference service thedrummer-cydonia-22b-v1-v1 ready after 201.8592870235443s
Pipeline stage MKMLDeployer completed in 204.28s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.252293825149536s
Received healthy response to inference request in 2.5037572383880615s
Received healthy response to inference request in 2.8367180824279785s
Received healthy response to inference request in 2.989543914794922s
Received healthy response to inference request in 2.3892245292663574s
5 requests
0 failed requests
5th percentile: 2.2796799659729006
10th percentile: 2.3070661067962646
20th percentile: 2.361838388442993
30th percentile: 2.4121310710906982
40th percentile: 2.45794415473938
50th percentile: 2.5037572383880615
60th percentile: 2.6369415760040282
70th percentile: 2.770125913619995
80th percentile: 2.867283248901367
90th percentile: 2.9284135818481447
95th percentile: 2.9589787483215333
99th percentile: 2.983430881500244
mean time: 2.594307518005371
Pipeline stage StressChecker completed in 14.41s
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 5.78s
Shutdown handler de-registered
thedrummer-cydonia-22b-v1_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 0.15s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.14s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service thedrummer-cydonia-22b-v1-v1-profiler
Waiting for inference service thedrummer-cydonia-22b-v1-v1-profiler to be ready
Inference service thedrummer-cydonia-22b-v1-v1-profiler ready after 200.4417209625244s
Pipeline stage MKMLProfilerDeployer completed in 200.84s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/thedrummer-cydonia-2156bedd95873c0daffe92d1c027914a8-deplogcl87:/code/chaiverse_profiler_1727152873 --namespace tenant-chaiml-guanaco
kubectl exec -it thedrummer-cydonia-2156bedd95873c0daffe92d1c027914a8-deplogcl87 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1727152873 && 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_1727152873/summary.json'
kubectl exec -it thedrummer-cydonia-2156bedd95873c0daffe92d1c027914a8-deplogcl87 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1727152873/summary.json'
Pipeline stage MKMLProfilerRunner completed in 1561.54s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service thedrummer-cydonia-22b-v1-v1-profiler is running
Tearing down inference service thedrummer-cydonia-22b-v1-v1-profiler
Service thedrummer-cydonia-22b-v1-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.18s
Shutdown handler de-registered
thedrummer-cydonia-22b-v1_v1 status is now inactive due to auto deactivation removed underperforming models
thedrummer-cydonia-22b-v1_v1 status is now torndown due to DeploymentManager action