sao10k-mn-12b-lyra-v4a1

developer_uid: chai_backend_admin

submission_id: sao10k-mn-12b-lyra-v4a1_v9

model_name: sao10k-mn-12b-lyra-v4a1_v8

model_group: Sao10K/MN-12B-Lyra-v4a1

status: inactive

timestamp: 2024-09-27T18:54:41+00:00

num_battles: 3786957

num_wins: 1981269

celo_rating: 1268.97

family_friendly_score: 0.6034727330040979

family_friendly_standard_error: 0.004979919881903761

submission_type: basic

model_repo: Sao10K/MN-12B-Lyra-v4a1

model_architecture: MistralForCausalLM

model_num_parameters: 12772070400.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.6166002523448423, 'latency_mean': 1.621728607416153, 'latency_p50': 1.6184449195861816, 'latency_p90': 1.79877769947052}, {'batch_size': 3, 'throughput': 1.0802448732481118, 'latency_mean': 2.770267697572708, 'latency_p50': 2.7811543941497803, 'latency_p90': 3.0714095354080198}, {'batch_size': 5, 'throughput': 1.2389934514001915, 'latency_mean': 4.021822712421417, 'latency_p50': 4.029958128929138, 'latency_p90': 4.60302505493164}, {'batch_size': 6, 'throughput': 1.2536723994603136, 'latency_mean': 4.763927043676376, 'latency_p50': 4.752925872802734, 'latency_p90': 5.446174097061157}, {'batch_size': 8, 'throughput': 1.2320973789524599, 'latency_mean': 6.470800031423568, 'latency_p50': 6.485338807106018, 'latency_p90': 7.305553269386292}, {'batch_size': 10, 'throughput': 1.1973199458876873, 'latency_mean': 8.312187260389328, 'latency_p50': 8.303161978721619, 'latency_p90': 9.445883560180665}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: sao10k-mn-12b-lyra-v4a1_v8

is_internal_developer: True

language_model: Sao10K/MN-12B-Lyra-v4a1

model_size: 13B

ranking_group: single

throughput_3p7s: 1.21

us_pacific_date: 2024-09-27

win_ratio: 0.5231823334672139

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': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '<|im_start|>system\n{memory}<|im_end|>\n', 'prompt_template': '<|im_start|>user\n{prompt}<|im_end|>\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n{bot_name}:', 'truncate_by_message': False}

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 sao10k-mn-12b-lyra-v4a1-v9-mkmlizer
Waiting for job on sao10k-mn-12b-lyra-v4a1-v9-mkmlizer to finish
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║     _____            __           __                                ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║       /___/                                                         ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║                                                                     ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  Version: 0.11.12                                                   ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  https://mk1.ai                                                     ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║                                                                     ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  The license key for the current software has been verified as      ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  belonging to:                                                      ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║                                                                     ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  Chai Research Corp.                                                ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ║                                                                     ║
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: Downloaded to shared memory in 20.194s
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpzzdj1zw_, device:0
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: Saving flywheel model at /dev/shm/model_cache
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: quantized model in 30.517s
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: Processed model Sao10K/MN-12B-Lyra-v4a1 in 50.711s
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: creating bucket guanaco-mkml-models
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9/config.json
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9/special_tokens_map.json
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9/tokenizer_config.json
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9/tokenizer.json
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v9/flywheel_model.0.safetensors
sao10k-mn-12b-lyra-v4a1-v9-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/363 [00:05<15:52,  2.64s/it]
Loading 0:   2%|▏         | 6/363 [00:05<04:13,  1.41it/s]
Loading 0:   4%|▍         | 14/363 [00:05<01:22,  4.22it/s]
Loading 0:   6%|▌         | 20/363 [00:05<00:49,  6.88it/s]
Loading 0:   7%|▋         | 27/363 [00:05<00:30, 10.92it/s]
Loading 0:   9%|▉         | 33/363 [00:05<00:22, 14.54it/s]
Loading 0:  12%|█▏        | 42/363 [00:06<00:15, 21.23it/s]
Loading 0:  14%|█▍        | 51/363 [00:06<00:11, 27.68it/s]
Loading 0:  16%|█▌        | 58/363 [00:06<00:10, 28.44it/s]
Loading 0:  17%|█▋        | 63/363 [00:06<00:09, 31.40it/s]
Loading 0:  19%|█▉        | 69/363 [00:06<00:08, 33.59it/s]
Loading 0:  21%|██▏       | 78/363 [00:06<00:07, 40.06it/s]
Loading 0:  24%|██▍       | 87/363 [00:06<00:06, 45.15it/s]
Loading 0:  26%|██▋       | 96/363 [00:07<00:05, 48.47it/s]
Loading 0:  29%|██▉       | 105/363 [00:07<00:05, 50.89it/s]
Loading 0:  31%|███▏      | 114/363 [00:07<00:04, 52.29it/s]
Loading 0:  34%|███▍      | 123/363 [00:07<00:04, 53.51it/s]
Loading 0:  36%|███▌      | 131/363 [00:07<00:03, 58.96it/s]
Loading 0:  38%|███▊      | 138/363 [00:07<00:03, 57.91it/s]
Loading 0:  40%|███▉      | 145/363 [00:07<00:03, 58.35it/s]
Loading 0:  42%|████▏     | 152/363 [00:08<00:03, 53.69it/s]
Loading 0:  44%|████▎     | 158/363 [00:08<00:04, 41.14it/s]
Loading 0:  45%|████▌     | 164/363 [00:08<00:04, 41.94it/s]
Loading 0:  47%|████▋     | 169/363 [00:08<00:04, 43.53it/s]
Loading 0:  48%|████▊     | 176/363 [00:08<00:03, 49.53it/s]
Loading 0:  50%|█████     | 182/363 [00:08<00:03, 49.58it/s]
Loading 0:  52%|█████▏    | 188/363 [00:08<00:03, 51.28it/s]
Loading 0:  54%|█████▎    | 195/363 [00:09<00:03, 49.24it/s]
Loading 0:  56%|█████▌    | 203/363 [00:09<00:02, 56.77it/s]
Loading 0:  58%|█████▊    | 209/363 [00:09<00:02, 54.42it/s]
Loading 0:  59%|█████▉    | 215/363 [00:09<00:02, 55.14it/s]
Loading 0:  61%|██████    | 222/363 [00:09<00:02, 50.45it/s]
Loading 0:  64%|██████▎   | 231/363 [00:09<00:02, 52.29it/s]
Loading 0:  66%|██████▌   | 240/363 [00:09<00:02, 53.76it/s]
Loading 0:  69%|██████▊   | 249/363 [00:10<00:02, 54.67it/s]
Loading 0:  71%|███████   | 256/363 [00:10<00:02, 44.85it/s]
Loading 0:  72%|███████▏  | 261/363 [00:10<00:02, 45.38it/s]
Loading 0:  74%|███████▎  | 267/363 [00:10<00:02, 43.64it/s]
Loading 0:  76%|███████▌  | 276/363 [00:10<00:01, 47.13it/s]
Loading 0:  79%|███████▊  | 285/363 [00:10<00:01, 49.62it/s]
Loading 0:  81%|████████  | 293/363 [00:10<00:01, 55.94it/s]
Loading 0:  82%|████████▏ | 299/363 [00:11<00:01, 53.65it/s]
Loading 0:  84%|████████▍ | 305/363 [00:11<00:01, 54.26it/s]
Loading 0:  86%|████████▌ | 312/363 [00:11<00:00, 51.03it/s]
Loading 0:  88%|████████▊ | 320/363 [00:11<00:00, 57.81it/s]
Loading 0:  90%|█████████ | 327/363 [00:11<00:00, 57.07it/s]
Loading 0:  92%|█████████▏| 333/363 [00:11<00:00, 55.13it/s]
Loading 0:  93%|█████████▎| 339/363 [00:11<00:00, 49.86it/s]
Loading 0:  96%|█████████▌| 347/363 [00:11<00:00, 56.84it/s]
Loading 0:  97%|█████████▋| 353/363 [00:12<00:00, 54.82it/s]
Loading 0:  99%|█████████▉| 359/363 [00:12<00:00, 38.94it/s]
                                                            
Job sao10k-mn-12b-lyra-v4a1-v9-mkmlizer completed after 69.94s with status: succeeded
Stopping job with name sao10k-mn-12b-lyra-v4a1-v9-mkmlizer
Pipeline stage MKMLizer completed in 71.78s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.65s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service sao10k-mn-12b-lyra-v4a1-v9
Waiting for inference service sao10k-mn-12b-lyra-v4a1-v9 to be ready
Inference service sao10k-mn-12b-lyra-v4a1-v9 ready after 30.820784091949463s
Pipeline stage MKMLDeployer completed in 32.57s
run pipeline stage %s
Running pipeline stage StressChecker
HTTP Request: %s %s "%s %d %s"
Received healthy response to inference request in 3.384308099746704s
HTTP Request: %s %s "%s %d %s"
Received healthy response to inference request in 2.353977918624878s
HTTP Request: %s %s "%s %d %s"
Received healthy response to inference request in 2.1863176822662354s
HTTP Request: %s %s "%s %d %s"
Received healthy response to inference request in 2.2778871059417725s
HTTP Request: %s %s "%s %d %s"
Received healthy response to inference request in 2.9137117862701416s
5 requests
0 failed requests
5th percentile: 2.2046315670013428
10th percentile: 2.22294545173645
20th percentile: 2.259573221206665
30th percentile: 2.2931052684783935
40th percentile: 2.323541593551636
50th percentile: 2.353977918624878
60th percentile: 2.5778714656829833
70th percentile: 2.8017650127410887
80th percentile: 3.0078310489654543
90th percentile: 3.196069574356079
95th percentile: 3.2901888370513914
99th percentile: 3.3654842472076414
mean time: 2.623240518569946
Pipeline stage StressChecker completed in 19.58s
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 2.64s
Shutdown handler de-registered
sao10k-mn-12b-lyra-v4a1_v9 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 0.13s
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 sao10k-mn-12b-lyra-v4a1-v9-profiler
Waiting for inference service sao10k-mn-12b-lyra-v4a1-v9-profiler to be ready
Inference service sao10k-mn-12b-lyra-v4a1-v9-profiler ready after 240.53820180892944s
Pipeline stage MKMLProfilerDeployer completed in 240.90s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/sao10k-mn-12b-lyra-v4a1-v9-profiler-predictor-00001-deployfn9l9:/code/chaiverse_profiler_1727463691 --namespace tenant-chaiml-guanaco
kubectl exec -it sao10k-mn-12b-lyra-v4a1-v9-profiler-predictor-00001-deployfn9l9 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1727463691 && 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_1727463691/summary.json'
kubectl exec -it sao10k-mn-12b-lyra-v4a1-v9-profiler-predictor-00001-deployfn9l9 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1727463691/summary.json'
Pipeline stage MKMLProfilerRunner completed in 1165.04s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service sao10k-mn-12b-lyra-v4a1-v9-profiler is running
Tearing down inference service sao10k-mn-12b-lyra-v4a1-v9-profiler
Service sao10k-mn-12b-lyra-v4a1-v9-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.30s
Shutdown handler de-registered
sao10k-mn-12b-lyra-v4a1_v9 status is now deployed due to admin request
sao10k-mn-12b-lyra-v4a1_v9 status is now torndown due to DeploymentManager action