submission_id: riverise-0910-2005-sft_v4
developer_uid: Riverise
best_of: 16
celo_rating: 1253.25
display_name: riverise-0910-2005-sft_v1
family_friendly_score: 0.0
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': 1.15, 'top_p': 0.95, 'min_p': 0.05, 'top_k': 80, '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: Riverise/0910_2005_sft
latencies: [{'batch_size': 1, 'throughput': 0.9146686691850042, 'latency_mean': 1.0931864202022552, 'latency_p50': 1.0899325609207153, 'latency_p90': 1.2333074808120728}, {'batch_size': 4, 'throughput': 1.793014754533477, 'latency_mean': 2.2198899817466735, 'latency_p50': 2.217987298965454, 'latency_p90': 2.4924256086349486}, {'batch_size': 5, 'throughput': 1.8824822981338853, 'latency_mean': 2.6401697981357573, 'latency_p50': 2.6257702112197876, 'latency_p90': 2.966812181472778}, {'batch_size': 8, 'throughput': 2.0218596013171726, 'latency_mean': 3.9257897663116457, 'latency_p50': 3.982595205307007, 'latency_p90': 4.381848287582398}, {'batch_size': 10, 'throughput': 2.038506764575553, 'latency_mean': 4.858266175985336, 'latency_p50': 4.858008146286011, 'latency_p90': 5.559025454521179}, {'batch_size': 12, 'throughput': 2.0482108317018706, 'latency_mean': 5.7822979485988615, 'latency_p50': 5.803446531295776, 'latency_p90': 6.430032086372376}, {'batch_size': 15, 'throughput': 2.0234984375517415, 'latency_mean': 7.257953004837036, 'latency_p50': 7.359744548797607, 'latency_p90': 8.065742015838623}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: Riverise/0910_2005_sft
model_name: riverise-0910-2005-sft_v1
model_num_parameters: 8030261248.0
model_repo: Riverise/0910_2005_sft
model_size: 8B
num_battles: 9555
num_wins: 4892
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 2.02
timestamp: 2024-09-12T07:22:11+00:00
us_pacific_date: 2024-09-12
win_ratio: 0.5119832548403978
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 riverise-0910-2005-sft-v4-mkmlizer
Waiting for job on riverise-0910-2005-sft-v4-mkmlizer to finish
riverise-0910-2005-sft-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
riverise-0910-2005-sft-v4-mkmlizer: ║     _____            __           __                                ║
riverise-0910-2005-sft-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
riverise-0910-2005-sft-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
riverise-0910-2005-sft-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
riverise-0910-2005-sft-v4-mkmlizer: ║       /___/                                                         ║
riverise-0910-2005-sft-v4-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v4-mkmlizer: ║  Version: 0.10.1                                                    ║
riverise-0910-2005-sft-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
riverise-0910-2005-sft-v4-mkmlizer: ║  https://mk1.ai                                                     ║
riverise-0910-2005-sft-v4-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
riverise-0910-2005-sft-v4-mkmlizer: ║  belonging to:                                                      ║
riverise-0910-2005-sft-v4-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v4-mkmlizer: ║  Chai Research Corp.                                                ║
riverise-0910-2005-sft-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
riverise-0910-2005-sft-v4-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
riverise-0910-2005-sft-v4-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Failed to get response for submission zonemercy-virgo-edit-v1-1e5_v12: HTTPConnectionPool(host='zonemercy-virgo-edit-v1-1e5-v12-predictor.tenant-chaiml-guanaco.k2.chaiverse.com', port=80): Max retries exceeded with url: /v1/models/GPT-J-6B-lit-v2:predict (Caused by ConnectTimeoutError(<urllib3.connection.HTTPConnection object at 0x7f2a77d347a0>, 'Connection to zonemercy-virgo-edit-v1-1e5-v12-predictor.tenant-chaiml-guanaco.k2.chaiverse.com timed out. (connect timeout=None)'))
riverise-0910-2005-sft-v4-mkmlizer: Downloaded to shared memory in 21.968s
riverise-0910-2005-sft-v4-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmptn2hzsih, device:0
riverise-0910-2005-sft-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
riverise-0910-2005-sft-v4-mkmlizer: quantized model in 26.539s
riverise-0910-2005-sft-v4-mkmlizer: Processed model Riverise/0910_2005_sft in 48.507s
riverise-0910-2005-sft-v4-mkmlizer: creating bucket guanaco-mkml-models
riverise-0910-2005-sft-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
riverise-0910-2005-sft-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/riverise-0910-2005-sft-v4
riverise-0910-2005-sft-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v4/config.json
riverise-0910-2005-sft-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v4/special_tokens_map.json
riverise-0910-2005-sft-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v4/tokenizer_config.json
riverise-0910-2005-sft-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v4/tokenizer.json
riverise-0910-2005-sft-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/riverise-0910-2005-sft-v4/flywheel_model.0.safetensors
riverise-0910-2005-sft-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:05, 49.74it/s]
Loading 0:   5%|▌         | 16/291 [00:00<00:04, 65.98it/s]
Loading 0:   9%|▊         | 25/291 [00:00<00:03, 71.39it/s]
Loading 0:  12%|█▏        | 34/291 [00:00<00:03, 67.92it/s]
Loading 0:  15%|█▍        | 43/291 [00:00<00:03, 68.14it/s]
Loading 0:  18%|█▊        | 52/291 [00:00<00:03, 71.64it/s]
Loading 0:  21%|██        | 61/291 [00:00<00:03, 69.99it/s]
Loading 0:  24%|██▍       | 70/291 [00:01<00:03, 71.75it/s]
Loading 0:  27%|██▋       | 79/291 [00:01<00:03, 67.41it/s]
Loading 0:  30%|██▉       | 86/291 [00:02<00:10, 18.65it/s]
Loading 0:  32%|███▏      | 94/291 [00:02<00:08, 24.06it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:06, 30.63it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 37.71it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:03, 43.16it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:03, 49.77it/s]
Loading 0:  48%|████▊     | 139/291 [00:03<00:02, 54.96it/s]
Loading 0:  51%|█████     | 148/291 [00:03<00:02, 59.63it/s]
Loading 0:  55%|█████▍    | 159/291 [00:03<00:01, 70.69it/s]
Loading 0:  58%|█████▊    | 169/291 [00:03<00:01, 68.78it/s]
Loading 0:  61%|██████    | 178/291 [00:03<00:01, 66.24it/s]
Loading 0:  64%|██████▍   | 187/291 [00:04<00:04, 20.82it/s]
Loading 0:  67%|██████▋   | 196/291 [00:04<00:03, 26.82it/s]
Loading 0:  70%|███████   | 205/291 [00:04<00:02, 32.99it/s]
Loading 0:  74%|███████▎  | 214/291 [00:05<00:01, 40.36it/s]
Loading 0:  77%|███████▋  | 223/291 [00:05<00:01, 46.57it/s]
Loading 0:  80%|███████▉  | 232/291 [00:05<00:01, 51.44it/s]
Loading 0:  83%|████████▎ | 241/291 [00:05<00:00, 55.47it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 68.89it/s]
Loading 0:  92%|█████████▏| 267/291 [00:05<00:00, 77.52it/s]
Loading 0:  95%|█████████▍| 276/291 [00:05<00:00, 78.03it/s]
Loading 0:  98%|█████████▊| 285/291 [00:05<00:00, 76.27it/s]
                                                            

Job riverise-0910-2005-sft-v4-mkmlizer completed after 73.12s with status: succeeded
Stopping job with name riverise-0910-2005-sft-v4-mkmlizer
Pipeline stage MKMLizer completed in 74.48s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.08s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service riverise-0910-2005-sft-v4
Waiting for inference service riverise-0910-2005-sft-v4 to be ready
Inference service riverise-0910-2005-sft-v4 ready after 170.78907346725464s
Pipeline stage MKMLDeployer completed in 171.17s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.2094058990478516s
Received healthy response to inference request in 3.102872133255005s
Received healthy response to inference request in 11.634801149368286s
Received healthy response to inference request in 1.7501471042633057s
Received healthy response to inference request in 1.742021083831787s
5 requests
0 failed requests
5th percentile: 1.7436462879180907
10th percentile: 1.7452714920043946
20th percentile: 1.748521900177002
30th percentile: 1.8419988632202149
40th percentile: 2.025702381134033
50th percentile: 2.2094058990478516
60th percentile: 2.566792392730713
70th percentile: 2.924178886413574
80th percentile: 4.809257936477662
90th percentile: 8.222029542922975
95th percentile: 9.928415346145629
99th percentile: 11.293523988723754
mean time: 4.087849473953247
Pipeline stage StressChecker completed in 21.59s
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 4.51s
Shutdown handler de-registered
riverise-0910-2005-sft_v4 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 0.11s
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 riverise-0910-2005-sft-v4-profiler
Waiting for inference service riverise-0910-2005-sft-v4-profiler to be ready
Inference service riverise-0910-2005-sft-v4-profiler ready after 170.40673899650574s
Pipeline stage MKMLProfilerDeployer completed in 170.76s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/riverise-0910-2005-sft-v4-profiler-predictor-00001-deploymwbvnr:/code/chaiverse_profiler_1726126220 --namespace tenant-chaiml-guanaco
kubectl exec -it riverise-0910-2005-sft-v4-profiler-predictor-00001-deploymwbvnr --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726126220 && 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_1726126220/summary.json'
kubectl exec -it riverise-0910-2005-sft-v4-profiler-predictor-00001-deploymwbvnr --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726126220/summary.json'
Pipeline stage MKMLProfilerRunner completed in 837.38s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service riverise-0910-2005-sft-v4-profiler is running
Tearing down inference service riverise-0910-2005-sft-v4-profiler
Service riverise-0910-2005-sft-v4-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.77s
Shutdown handler de-registered
riverise-0910-2005-sft_v4 status is now inactive due to auto deactivation removed underperforming models
riverise-0910-2005-sft_v4 status is now torndown due to DeploymentManager action