developer_uid: azuruce
submission_id: chaiml-elo-alignment-run-3_v45
model_name: zonemercy-virgo-edit-v1-1e5_v13
model_group: ChaiML/elo_alignment_run
status: torndown
timestamp: 2024-09-20T16:48:01+00:00
num_battles: 11442
num_wins: 6014
celo_rating: 1265.02
family_friendly_score: 0.0
submission_type: basic
model_repo: ChaiML/elo_alignment_run_3
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 8
max_input_tokens: 1024
max_output_tokens: 64
latencies: [{'batch_size': 1, 'throughput': 0.8752742700288311, 'latency_mean': 1.1424320101737977, 'latency_p50': 1.144744634628296, 'latency_p90': 1.2594557285308838}, {'batch_size': 4, 'throughput': 1.8809605034502608, 'latency_mean': 2.1114070665836335, 'latency_p50': 2.1087517738342285, 'latency_p90': 2.328787612915039}, {'batch_size': 5, 'throughput': 2.039316276633964, 'latency_mean': 2.4382073521614074, 'latency_p50': 2.431191325187683, 'latency_p90': 2.727967643737793}, {'batch_size': 8, 'throughput': 2.2938656298241296, 'latency_mean': 3.465954619646072, 'latency_p50': 3.45515775680542, 'latency_p90': 3.9264975786209106}, {'batch_size': 10, 'throughput': 2.329395464889116, 'latency_mean': 4.256330083608628, 'latency_p50': 4.287947297096252, 'latency_p90': 4.793495726585388}, {'batch_size': 12, 'throughput': 2.360046288215602, 'latency_mean': 5.031469687223434, 'latency_p50': 5.037449598312378, 'latency_p90': 5.713597226142883}, {'batch_size': 15, 'throughput': 2.3813255090790952, 'latency_mean': 6.219186326265335, 'latency_p50': 6.294583559036255, 'latency_p90': 6.975516819953919}]
gpu_counts: {'NVIDIA RTX A5000': 1}
display_name: zonemercy-virgo-edit-v1-1e5_v13
is_internal_developer: True
language_model: ChaiML/elo_alignment_run_3
model_size: 8B
ranking_group: single
throughput_3p7s: 2.32
us_pacific_date: 2024-09-20
win_ratio: 0.5256074112917322
generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '</s>', '###', 'Bot:', 'User:', 'You:', '<|im_end|>'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}
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}
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 chaiml-elo-alignment-run-3-v45-mkmlizer
Waiting for job on chaiml-elo-alignment-run-3-v45-mkmlizer to finish
chaiml-elo-alignment-run-3-v45-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ _____ __ __ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ /___/ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ Version: 0.10.1 ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ https://mk1.ai ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ The license key for the current software has been verified as ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ belonging to: ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ Chai Research Corp. ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ║ ║
chaiml-elo-alignment-run-3-v45-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-elo-alignment-run-3-v45-mkmlizer: Downloaded to shared memory in 61.563s
chaiml-elo-alignment-run-3-v45-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpo3_dhspy, device:0
chaiml-elo-alignment-run-3-v45-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Connection pool is full, discarding connection: %s. Connection pool size: %s
chaiml-elo-alignment-run-3-v45-mkmlizer: quantized model in 28.085s
chaiml-elo-alignment-run-3-v45-mkmlizer: Processed model ChaiML/elo_alignment_run_3 in 89.648s
chaiml-elo-alignment-run-3-v45-mkmlizer: creating bucket guanaco-mkml-models
chaiml-elo-alignment-run-3-v45-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-elo-alignment-run-3-v45-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45
chaiml-elo-alignment-run-3-v45-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45/config.json
chaiml-elo-alignment-run-3-v45-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45/special_tokens_map.json
chaiml-elo-alignment-run-3-v45-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45/tokenizer_config.json
chaiml-elo-alignment-run-3-v45-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45/tokenizer.json
chaiml-elo-alignment-run-3-v45-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-elo-alignment-run-3-v45/flywheel_model.0.safetensors
chaiml-elo-alignment-run-3-v45-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 2%|▏ | 5/291 [00:00<00:10, 27.46it/s] Loading 0: 4%|▍ | 12/291 [00:00<00:07, 37.52it/s] Loading 0: 5%|▌ | 16/291 [00:00<00:07, 35.21it/s] Loading 0: 7%|▋ | 21/291 [00:00<00:07, 37.82it/s] Loading 0: 9%|▊ | 25/291 [00:00<00:07, 35.41it/s] Loading 0: 11%|█ | 31/291 [00:00<00:06, 41.72it/s] Loading 0: 12%|█▏ | 36/291 [00:01<00:10, 23.59it/s] Loading 0: 14%|█▍ | 41/291 [00:01<00:09, 25.27it/s] Loading 0: 16%|█▋ | 48/291 [00:01<00:07, 32.65it/s] Loading 0: 18%|█▊ | 53/291 [00:01<00:06, 34.30it/s] Loading 0: 20%|█▉ | 58/291 [00:01<00:06, 35.25it/s] Loading 0: 22%|██▏ | 63/291 [00:01<00:06, 37.82it/s] Loading 0: 23%|██▎ | 68/291 [00:02<00:06, 32.46it/s] Loading 0: 25%|██▌ | 74/291 [00:02<00:05, 36.95it/s] Loading 0: 27%|██▋ | 79/291 [00:02<00:06, 34.46it/s] Loading 0: 29%|██▊ | 83/291 [00:02<00:08, 23.94it/s] Loading 0: 31%|███ | 90/291 [00:02<00:06, 30.46it/s] Loading 0: 32%|███▏ | 94/291 [00:02<00:06, 30.74it/s] Loading 0: 34%|███▍ | 99/291 [00:03<00:05, 33.94it/s] Loading 0: 35%|███▌ | 103/291 [00:03<00:05, 33.25it/s] Loading 0: 37%|███▋ | 108/291 [00:03<00:05, 36.39it/s] Loading 0: 38%|███▊ | 112/291 [00:03<00:05, 34.70it/s] Loading 0: 40%|███▉ | 116/291 [00:03<00:05, 34.78it/s] Loading 0: 42%|████▏ | 122/291 [00:03<00:04, 39.42it/s] Loading 0: 44%|████▎ | 127/291 [00:03<00:04, 36.50it/s] Loading 0: 46%|████▌ | 133/291 [00:04<00:05, 31.30it/s] Loading 0: 47%|████▋ | 137/291 [00:04<00:04, 31.70it/s] Loading 0: 48%|████▊ | 141/291 [00:04<00:04, 30.09it/s] Loading 0: 51%|█████ | 147/291 [00:04<00:04, 34.81it/s] Loading 0: 52%|█████▏ | 151/291 [00:04<00:04, 34.13it/s] Loading 0: 54%|█████▎ | 156/291 [00:04<00:03, 37.16it/s] Loading 0: 55%|█████▍ | 160/291 [00:04<00:03, 35.75it/s] Loading 0: 57%|█████▋ | 165/291 [00:04<00:03, 37.68it/s] Loading 0: 58%|█████▊ | 169/291 [00:05<00:03, 35.51it/s] Loading 0: 60%|█████▉ | 174/291 [00:05<00:03, 38.30it/s] Loading 0: 61%|██████ | 178/291 [00:05<00:03, 36.61it/s] Loading 0: 63%|██████▎ | 184/291 [00:05<00:02, 42.21it/s] Loading 0: 65%|██████▍ | 189/291 [00:05<00:03, 25.85it/s] Loading 0: 67%|██████▋ | 194/291 [00:05<00:03, 27.61it/s] Loading 0: 69%|██████▉ | 201/291 [00:06<00:02, 34.52it/s] Loading 0: 71%|███████ | 206/291 [00:06<00:02, 35.38it/s] Loading 0: 73%|███████▎ | 211/291 [00:06<00:02, 35.59it/s] Loading 0: 74%|███████▍ | 215/291 [00:06<00:02, 36.05it/s] Loading 0: 76%|███████▌ | 220/291 [00:06<00:01, 36.62it/s] Loading 0: 77%|███████▋ | 224/291 [00:06<00:01, 37.29it/s] Loading 0: 78%|███████▊ | 228/291 [00:06<00:02, 31.35it/s] Loading 0: 80%|████████ | 233/291 [00:07<00:02, 27.65it/s] Loading 0: 82%|████████▏ | 238/291 [00:07<00:01, 32.08it/s] Loading 0: 83%|████████▎ | 242/291 [00:07<00:01, 33.65it/s] Loading 0: 85%|████████▍ | 246/291 [00:07<00:01, 34.84it/s] Loading 0: 86%|████████▌ | 250/291 [00:07<00:01, 34.20it/s] Loading 0: 88%|████████▊ | 255/291 [00:07<00:00, 37.32it/s] Loading 0: 89%|████████▉ | 259/291 [00:07<00:00, 35.89it/s] Loading 0: 91%|█████████ | 264/291 [00:07<00:00, 38.83it/s] Loading 0: 92%|█████████▏| 268/291 [00:07<00:00, 37.19it/s] Loading 0: 94%|█████████▍| 273/291 [00:08<00:00, 39.91it/s] Loading 0: 96%|█████████▌| 278/291 [00:08<00:00, 39.55it/s] Loading 0: 97%|█████████▋| 283/291 [00:08<00:00, 41.12it/s] Loading 0: 99%|█████████▉| 288/291 [00:13<00:01, 2.79it/s]
Job chaiml-elo-alignment-run-3-v45-mkmlizer completed after 114.44s with status: succeeded
Stopping job with name chaiml-elo-alignment-run-3-v45-mkmlizer
Pipeline stage MKMLizer completed in 115.57s
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 chaiml-elo-alignment-run-3-v45
Waiting for inference service chaiml-elo-alignment-run-3-v45 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
Inference service chaiml-elo-alignment-run-3-v45 ready after 211.59916996955872s
Pipeline stage MKMLDeployer completed in 212.25s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.9262161254882812s
Received healthy response to inference request in 1.4895684719085693s
Received healthy response to inference request in 1.353630781173706s
Received healthy response to inference request in 1.426417589187622s
Received healthy response to inference request in 1.872208595275879s
5 requests
0 failed requests
5th percentile: 1.3681881427764893
10th percentile: 1.3827455043792725
20th percentile: 1.4118602275848389
30th percentile: 1.4390477657318115
40th percentile: 1.4643081188201905
50th percentile: 1.4895684719085693
60th percentile: 1.642624521255493
70th percentile: 1.7956805706024168
80th percentile: 1.8830101013183593
90th percentile: 1.9046131134033204
95th percentile: 1.9154146194458008
99th percentile: 1.9240558242797852
mean time: 1.6136083126068115
Pipeline stage StressChecker completed in 8.75s
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 5.04s
Shutdown handler de-registered
chaiml-elo-alignment-run-3_v45 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 chaiml-elo-alignment-run-3-v45-profiler
Waiting for inference service chaiml-elo-alignment-run-3-v45-profiler to be ready
Inference service chaiml-elo-alignment-run-3-v45-profiler ready after 210.47120237350464s
Pipeline stage MKMLProfilerDeployer completed in 210.79s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/chaiml-elo-alignment6dcd7395c64e6193164fe5a33ea0328f-deplomrg7n:/code/chaiverse_profiler_1726851480 --namespace tenant-chaiml-guanaco
kubectl exec -it chaiml-elo-alignment6dcd7395c64e6193164fe5a33ea0328f-deplomrg7n --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726851480 && 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_1726851480/summary.json'
kubectl exec -it chaiml-elo-alignment6dcd7395c64e6193164fe5a33ea0328f-deplomrg7n --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726851480/summary.json'
Pipeline stage MKMLProfilerRunner completed in 780.57s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service chaiml-elo-alignment-run-3-v45-profiler is running
Tearing down inference service chaiml-elo-alignment-run-3-v45-profiler
Service chaiml-elo-alignment-run-3-v45-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.19s
Shutdown handler de-registered
chaiml-elo-alignment-run-3_v45 status is now inactive due to auto deactivation removed underperforming models
chaiml-elo-alignment-run-3_v45 status is now torndown due to DeploymentManager action
Shutdown handler not registered because Python interpreter is not running in the main thread
admin requested tearing down of chaiml-test-feed-convo-v1-1e5_v9
Running pipeline stage MKMLModelDeleter
Pipeline stage MKMLDeleter completed in 3.01s
Pipeline stage %s skipped, reason=%s
Running pipeline stage MKMLDeleter
Shutdown handler de-registered
run pipeline stage %s
chaiml-elo-alignment-run-3_v45 status is now torndown due to DeploymentManager action
run pipeline %s
Shutdown handler not registered because Python interpreter is not running in the main thread
Pipeline stage %s skipped, reason=%s
Running pipeline stage MKMLDeleter
run pipeline stage %s
run pipeline %s
Shutdown handler not registered because Python interpreter is not running in the main thread
Shutdown handler de-registered
admin requested tearing down of chaiml-test-feed-convo-v1-1e5_v9
chaiml-elo-alignment-run-3_v45 status is now torndown due to DeploymentManager action
Pipeline stage MKMLModelDeleter completed in 4.01s
Pipeline stage MKMLDeleter completed in 4.01s
Pipeline stage %s skipped, reason=%s
Running pipeline stage MKMLDeleter