submission_id: sao10k-mn-12b-lyra-v4a1_v2
developer_uid: sao10k
best_of: 8
celo_rating: 1262.78
display_name: lyra41
family_friendly_score: 0.0
formatter: {'memory_template': '<|im_start|>system\n{memory}<im_end|>\n', 'prompt_template': '<|im_start|>user\n{prompt}[/INST]\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}[/INST]\n', 'user_template': '<|im_start|>user\n{user_name}: {message}[/INST]\n', 'response_template': '<|im_end|>assistant\n{bot_name}:', 'truncate_by_message': False}
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': 512, 'best_of': 8, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: False
language_model: Sao10K/MN-12B-Lyra-v4a1
latencies: [{'batch_size': 1, 'throughput': 0.6890029277840857, 'latency_mean': 1.4512716162204742, 'latency_p50': 1.4519613981246948, 'latency_p90': 1.6182845830917358}, {'batch_size': 3, 'throughput': 1.3069581944815567, 'latency_mean': 2.286860810518265, 'latency_p50': 2.2844351530075073, 'latency_p90': 2.547550892829895}, {'batch_size': 5, 'throughput': 1.5545807583641187, 'latency_mean': 3.2022209787368774, 'latency_p50': 3.2348639965057373, 'latency_p90': 3.5999825954437257}, {'batch_size': 6, 'throughput': 1.6013016623402776, 'latency_mean': 3.7217323541641236, 'latency_p50': 3.7222198247909546, 'latency_p90': 4.235605788230896}, {'batch_size': 8, 'throughput': 1.5688496823425608, 'latency_mean': 5.067172530889511, 'latency_p50': 5.0726306438446045, 'latency_p90': 5.770679402351379}, {'batch_size': 10, 'throughput': 1.5198218544231639, 'latency_mean': 6.5367644441127775, 'latency_p50': 6.594677925109863, 'latency_p90': 7.445786237716675}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: MistralForCausalLM
model_group: Sao10K/MN-12B-Lyra-v4a1
model_name: lyra41
model_num_parameters: 12772070400.0
model_repo: Sao10K/MN-12B-Lyra-v4a1
model_size: 13B
num_battles: 13133
num_wins: 6958
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 1.61
timestamp: 2024-09-05T15:48:00+00:00
us_pacific_date: 2024-09-05
win_ratio: 0.5298104012792203
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 sao10k-mn-12b-lyra-v4a1-v2-mkmlizer
Waiting for job on sao10k-mn-12b-lyra-v4a1-v2-mkmlizer to finish
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ _____ __ __ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ /___/ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ Version: 0.10.1 ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ https://mk1.ai ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ The license key for the current software has been verified as ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ belonging to: ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ Chai Research Corp. ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ║ ║
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: Downloaded to shared memory in 27.064s
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpof2qiaue, device:0
sao10k-mn-12b-lyra-v4a1-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
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: quantized model in 36.315s
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: Processed model Sao10K/MN-12B-Lyra-v4a1 in 63.379s
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: creating bucket guanaco-mkml-models
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2/config.json
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2/special_tokens_map.json
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2/tokenizer_config.json
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2/tokenizer.json
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/sao10k-mn-12b-lyra-v4a1-v2/flywheel_model.0.safetensors
sao10k-mn-12b-lyra-v4a1-v2-mkmlizer: Loading 0: 0%| | 0/363 [00:00<?, ?it/s] Loading 0: 1%| | 2/363 [00:06<18:09, 3.02s/it] Loading 0: 2%|▏ | 6/363 [00:06<04:49, 1.23it/s] Loading 0: 3%|▎ | 12/363 [00:06<01:53, 3.09it/s] Loading 0: 4%|▍ | 16/363 [00:06<01:14, 4.64it/s] Loading 0: 6%|▋ | 23/363 [00:06<00:40, 8.39it/s] Loading 0: 8%|▊ | 28/363 [00:06<00:29, 11.50it/s] Loading 0: 9%|▉ | 33/363 [00:06<00:23, 14.30it/s] Loading 0: 11%|█▏ | 41/363 [00:06<00:14, 21.65it/s] Loading 0: 13%|█▎ | 47/363 [00:07<00:12, 25.83it/s] Loading 0: 14%|█▍ | 52/363 [00:07<00:10, 29.05it/s] Loading 0: 16%|█▌ | 58/363 [00:07<00:11, 26.46it/s] Loading 0: 17%|█▋ | 62/363 [00:07<00:10, 28.14it/s] Loading 0: 19%|█▊ | 68/363 [00:07<00:08, 33.10it/s] Loading 0: 20%|██ | 73/363 [00:07<00:08, 35.44it/s] Loading 0: 21%|██▏ | 78/363 [00:07<00:08, 32.46it/s] Loading 0: 23%|██▎ | 85/363 [00:08<00:06, 39.94it/s] Loading 0: 25%|██▍ | 90/363 [00:08<00:06, 39.95it/s] Loading 0: 26%|██▌ | 95/363 [00:08<00:06, 41.09it/s] Loading 0: 28%|██▊ | 100/363 [00:08<00:06, 41.64it/s] Loading 0: 29%|██▉ | 105/363 [00:08<00:07, 34.42it/s] Loading 0: 31%|███ | 112/363 [00:08<00:06, 41.65it/s] Loading 0: 32%|███▏ | 117/363 [00:08<00:06, 40.44it/s] Loading 0: 34%|███▎ | 122/363 [00:08<00:05, 40.26it/s] Loading 0: 35%|███▍ | 127/363 [00:09<00:05, 41.34it/s] Loading 0: 36%|███▋ | 132/363 [00:09<00:06, 34.37it/s] Loading 0: 38%|███▊ | 139/363 [00:09<00:05, 40.37it/s] Loading 0: 40%|███▉ | 144/363 [00:09<00:05, 40.89it/s] Loading 0: 41%|████ | 149/363 [00:09<00:05, 41.26it/s] Loading 0: 42%|████▏ | 154/363 [00:09<00:04, 43.33it/s] Loading 0: 44%|████▍ | 159/363 [00:10<00:08, 24.79it/s] Loading 0: 46%|████▌ | 166/363 [00:10<00:06, 32.16it/s] Loading 0: 47%|████▋ | 171/363 [00:10<00:05, 33.18it/s] Loading 0: 48%|████▊ | 176/363 [00:10<00:05, 34.99it/s] Loading 0: 50%|████▉ | 181/363 [00:10<00:04, 36.95it/s] Loading 0: 51%|█████ | 186/363 [00:10<00:05, 32.30it/s] Loading 0: 53%|█████▎ | 193/363 [00:10<00:04, 39.90it/s] Loading 0: 55%|█████▍ | 198/363 [00:11<00:04, 39.41it/s] Loading 0: 56%|█████▌ | 203/363 [00:11<00:04, 39.18it/s] Loading 0: 57%|█████▋ | 208/363 [00:11<00:03, 40.52it/s] Loading 0: 59%|█████▊ | 213/363 [00:11<00:04, 33.24it/s] Loading 0: 61%|██████ | 220/363 [00:11<00:03, 39.60it/s] Loading 0: 62%|██████▏ | 225/363 [00:11<00:03, 39.43it/s] Loading 0: 63%|██████▎ | 230/363 [00:11<00:03, 39.65it/s] Loading 0: 65%|██████▌ | 236/363 [00:12<00:03, 40.39it/s] Loading 0: 66%|██████▋ | 241/363 [00:12<00:03, 38.63it/s] Loading 0: 68%|██████▊ | 247/363 [00:12<00:02, 43.44it/s] Loading 0: 69%|██████▉ | 252/363 [00:12<00:02, 44.51it/s] Loading 0: 71%|███████ | 257/363 [00:12<00:03, 29.12it/s] Loading 0: 72%|███████▏ | 261/363 [00:12<00:03, 29.73it/s] Loading 0: 73%|███████▎ | 266/363 [00:12<00:02, 32.73it/s] Loading 0: 75%|███████▍ | 271/363 [00:13<00:02, 36.34it/s] Loading 0: 76%|███████▌ | 276/363 [00:13<00:02, 30.81it/s] Loading 0: 78%|███████▊ | 283/363 [00:13<00:02, 38.68it/s] Loading 0: 79%|███████▉ | 288/363 [00:13<00:01, 39.18it/s] Loading 0: 81%|████████ | 293/363 [00:13<00:01, 35.96it/s] Loading 0: 82%|████████▏ | 297/363 [00:13<00:01, 36.64it/s] Loading 0: 83%|████████▎ | 302/363 [00:13<00:01, 38.09it/s] Loading 0: 85%|████████▍ | 307/363 [00:14<00:01, 41.00it/s] Loading 0: 86%|████████▌ | 312/363 [00:14<00:01, 35.61it/s] Loading 0: 88%|████████▊ | 320/363 [00:14<00:00, 43.99it/s] Loading 0: 90%|████████▉ | 325/363 [00:14<00:00, 44.18it/s] Loading 0: 91%|█████████ | 330/363 [00:14<00:00, 36.10it/s] Loading 0: 93%|█████████▎| 337/363 [00:14<00:00, 43.47it/s] Loading 0: 94%|█████████▍| 342/363 [00:14<00:00, 44.68it/s] Loading 0: 96%|█████████▌| 347/363 [00:14<00:00, 45.81it/s] Loading 0: 97%|█████████▋| 353/363 [00:15<00:00, 44.22it/s] Loading 0: 99%|█████████▊| 358/363 [00:15<00:00, 29.17it/s]
Job sao10k-mn-12b-lyra-v4a1-v2-mkmlizer completed after 84.66s with status: succeeded
Stopping job with name sao10k-mn-12b-lyra-v4a1-v2-mkmlizer
Pipeline stage MKMLizer completed in 86.69s
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 sao10k-mn-12b-lyra-v4a1-v2
Waiting for inference service sao10k-mn-12b-lyra-v4a1-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
Inference service sao10k-mn-12b-lyra-v4a1-v2 ready after 150.63587713241577s
Pipeline stage MKMLDeployer completed in 150.97s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.682926893234253s
Received healthy response to inference request in 2.469393491744995s
Received healthy response to inference request in 3.3010151386260986s
Received healthy response to inference request in 2.7866439819335938s
Received healthy response to inference request in 1.8369154930114746s
5 requests
0 failed requests
5th percentile: 1.9634110927581787
10th percentile: 2.089906692504883
20th percentile: 2.342897891998291
30th percentile: 2.512100172042847
40th percentile: 2.59751353263855
50th percentile: 2.682926893234253
60th percentile: 2.7244137287139893
70th percentile: 2.7659005641937258
80th percentile: 2.889518213272095
90th percentile: 3.0952666759490968
95th percentile: 3.1981409072875975
99th percentile: 3.2804402923583984
mean time: 2.615378999710083
Pipeline stage StressChecker completed in 14.34s
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 6.57s
Shutdown handler de-registered
sao10k-mn-12b-lyra-v4a1_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 0.12s
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-v2-profiler
Waiting for inference service sao10k-mn-12b-lyra-v4a1-v2-profiler to be ready
Inference service sao10k-mn-12b-lyra-v4a1-v2-profiler ready after 150.3581485748291s
Pipeline stage MKMLProfilerDeployer completed in 150.75s
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-v2-profiler-predictor-00001-deployr8ph7:/code/chaiverse_profiler_1725551729 --namespace tenant-chaiml-guanaco
kubectl exec -it sao10k-mn-12b-lyra-v4a1-v2-profiler-predictor-00001-deployr8ph7 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725551729 && 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_1725551729/summary.json'
kubectl exec -it sao10k-mn-12b-lyra-v4a1-v2-profiler-predictor-00001-deployr8ph7 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725551729/summary.json'
Pipeline stage MKMLProfilerRunner completed in 962.03s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service sao10k-mn-12b-lyra-v4a1-v2-profiler is running
Tearing down inference service sao10k-mn-12b-lyra-v4a1-v2-profiler
Service sao10k-mn-12b-lyra-v4a1-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.74s
Shutdown handler de-registered
sao10k-mn-12b-lyra-v4a1_v2 status is now inactive due to auto deactivation removed underperforming models
Shutdown handler not registered because Python interpreter is not running in the main thread
admin requested tearing down of sao10k-mn-12b-lyra-v4a1_v2
run pipeline %s
Shutdown handler not registered because Python interpreter is not running in the main thread
run pipeline stage %s
run pipeline %s
Running pipeline stage MKMLDeleter
run pipeline stage %s
admin requested tearing down of sao10k-mn-12b-lyra-v4a2_v2
Running pipeline stage MKMLDeleter
Shutdown handler not registered because Python interpreter is not running in the main thread
run pipeline %s
%s, retrying in %s seconds...
run pipeline stage %s
%s, retrying in %s seconds...
Running pipeline stage MKMLDeleter
clean up pipeline due to error=TeardownError("module 'kubernetes.config' has no attribute 'load_kube_config'")
%s, retrying in %s seconds...
Shutdown handler de-registered
%s, retrying in %s seconds...
sao10k-mn-12b-lyra-v4a1_v2 status is now torndown due to DeploymentManager action