submission_id: riverise-0912-1056-sft-9k_v2
developer_uid: Riverise
best_of: 16
celo_rating: 1233.88
display_name: riverise-0912-1056-sft-9k_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/0912_1056_sft_9k
latencies: [{'batch_size': 1, 'throughput': 0.9097967717939018, 'latency_mean': 1.0990527522563935, 'latency_p50': 1.0959552526474, 'latency_p90': 1.22532057762146}, {'batch_size': 4, 'throughput': 1.773047517511903, 'latency_mean': 2.2491184639930726, 'latency_p50': 2.2593854665756226, 'latency_p90': 2.520443296432495}, {'batch_size': 5, 'throughput': 1.835157802210934, 'latency_mean': 2.704769765138626, 'latency_p50': 2.70847487449646, 'latency_p90': 3.033552360534668}, {'batch_size': 8, 'throughput': 1.9589318363035115, 'latency_mean': 4.055997266769409, 'latency_p50': 4.095089793205261, 'latency_p90': 4.561208701133728}, {'batch_size': 10, 'throughput': 1.9942855118557494, 'latency_mean': 4.970189938545227, 'latency_p50': 4.9084084033966064, 'latency_p90': 5.7832465171813965}, {'batch_size': 12, 'throughput': 1.9892410254288437, 'latency_mean': 5.94607985496521, 'latency_p50': 5.9554280042648315, 'latency_p90': 6.808731269836426}, {'batch_size': 15, 'throughput': 2.0131413226572583, 'latency_mean': 7.316072556972504, 'latency_p50': 7.421098947525024, 'latency_p90': 8.20083646774292}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: Riverise/0912_1056_sft_9
model_name: riverise-0912-1056-sft-9k_v1
model_num_parameters: 8030261248.0
model_repo: Riverise/0912_1056_sft_9k
model_size: 8B
num_battles: 11947
num_wins: 5780
ranking_group: single
status: torndown
submission_type: basic
throughput_3p7s: 1.95
timestamp: 2024-09-13T03:01:01+00:00
us_pacific_date: 2024-09-12
win_ratio: 0.4838034653050975
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-0912-1056-sft-9k-v2-mkmlizer
Waiting for job on riverise-0912-1056-sft-9k-v2-mkmlizer to finish
riverise-0912-1056-sft-9k-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ _____ __ __ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ /___/ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ Version: 0.10.1 ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ https://mk1.ai ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ The license key for the current software has been verified as ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ belonging to: ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ Chai Research Corp. ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ║ ║
riverise-0912-1056-sft-9k-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
riverise-0912-1056-sft-9k-v2-mkmlizer: Downloaded to shared memory in 20.807s
riverise-0912-1056-sft-9k-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp59b32mqv, device:0
riverise-0912-1056-sft-9k-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
riverise-0912-1056-sft-9k-v2-mkmlizer: creating bucket guanaco-mkml-models
riverise-0912-1056-sft-9k-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
riverise-0912-1056-sft-9k-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2
riverise-0912-1056-sft-9k-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2/special_tokens_map.json
riverise-0912-1056-sft-9k-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2/config.json
riverise-0912-1056-sft-9k-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2/tokenizer_config.json
riverise-0912-1056-sft-9k-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2/tokenizer.json
riverise-0912-1056-sft-9k-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/riverise-0912-1056-sft-9k-v2/flywheel_model.0.safetensors
riverise-0912-1056-sft-9k-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 2%|▏ | 7/291 [00:00<00:05, 49.02it/s] Loading 0: 8%|▊ | 22/291 [00:00<00:03, 83.33it/s] Loading 0: 12%|█▏ | 34/291 [00:00<00:03, 85.55it/s] Loading 0: 17%|█▋ | 49/291 [00:00<00:02, 91.01it/s] Loading 0: 20%|██ | 59/291 [00:00<00:02, 84.74it/s] Loading 0: 23%|██▎ | 68/291 [00:00<00:02, 82.11it/s] Loading 0: 27%|██▋ | 79/291 [00:00<00:02, 81.57it/s] Loading 0: 30%|███ | 88/291 [00:02<00:09, 22.18it/s] Loading 0: 33%|███▎ | 97/291 [00:02<00:07, 27.44it/s] Loading 0: 36%|███▋ | 106/291 [00:02<00:05, 34.14it/s] Loading 0: 42%|████▏ | 121/291 [00:02<00:03, 47.45it/s] Loading 0: 46%|████▌ | 133/291 [00:02<00:02, 56.15it/s] Loading 0: 51%|█████ | 148/291 [00:02<00:02, 68.90it/s] Loading 0: 54%|█████▍ | 158/291 [00:02<00:01, 74.85it/s] Loading 0: 58%|█████▊ | 168/291 [00:02<00:01, 77.79it/s] Loading 0: 61%|██████ | 178/291 [00:03<00:01, 76.91it/s] Loading 0: 64%|██████▍ | 187/291 [00:04<00:04, 23.82it/s] Loading 0: 69%|██████▉ | 201/291 [00:04<00:02, 34.31it/s] Loading 0: 72%|███████▏ | 210/291 [00:04<00:02, 39.49it/s] Loading 0: 76%|███████▌ | 220/291 [00:04<00:01, 45.39it/s] Loading 0: 80%|███████▉ | 232/291 [00:04<00:01, 53.88it/s] Loading 0: 83%|████████▎ | 241/291 [00:04<00:00, 59.84it/s] Loading 0: 86%|████████▌ | 250/291 [00:04<00:00, 64.50it/s] Loading 0: 91%|█████████ | 265/291 [00:05<00:00, 75.12it/s] Loading 0: 94%|█████████▍| 274/291 [00:05<00:00, 75.43it/s] Loading 0: 98%|█████████▊| 286/291 [00:05<00:00, 80.48it/s]
Job riverise-0912-1056-sft-9k-v2-mkmlizer completed after 64.21s with status: succeeded
Stopping job with name riverise-0912-1056-sft-9k-v2-mkmlizer
Pipeline stage MKMLizer completed in 65.09s
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 riverise-0912-1056-sft-9k-v2
Waiting for inference service riverise-0912-1056-sft-9k-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
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
Inference service riverise-0912-1056-sft-9k-v2 ready after 171.83521795272827s
Pipeline stage MKMLDeployer completed in 172.20s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.5872645378112793s
Received healthy response to inference request in 2.043699026107788s
Received healthy response to inference request in 1.8944156169891357s
Received healthy response to inference request in 1.8539929389953613s
Received healthy response to inference request in 1.378380298614502s
5 requests
0 failed requests
5th percentile: 1.4735028266906738
10th percentile: 1.5686253547668456
20th percentile: 1.7588704109191895
30th percentile: 1.8620774745941162
40th percentile: 1.878246545791626
50th percentile: 1.8944156169891357
60th percentile: 1.9541289806365967
70th percentile: 2.0138423442840576
80th percentile: 2.1524121284484865
90th percentile: 2.369838333129883
95th percentile: 2.478551435470581
99th percentile: 2.5655219173431396
mean time: 1.9515504837036133
Pipeline stage StressChecker completed in 10.58s
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 5.96s
Shutdown handler de-registered
riverise-0912-1056-sft-9k_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 0.11s
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-0912-1056-sft-9k-v2-profiler
Waiting for inference service riverise-0912-1056-sft-9k-v2-profiler to be ready
Inference service riverise-0912-1056-sft-9k-v2-profiler ready after 160.37466478347778s
Pipeline stage MKMLProfilerDeployer completed in 162.49s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/riverise-0912-1056-s936c01e093bb16053212ce1a5f10ac17-deplozx4r7:/code/chaiverse_profiler_1726196925 --namespace tenant-chaiml-guanaco
kubectl exec -it riverise-0912-1056-s936c01e093bb16053212ce1a5f10ac17-deplozx4r7 --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726196925 && 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_1726196925/summary.json'
kubectl exec -it riverise-0912-1056-s936c01e093bb16053212ce1a5f10ac17-deplozx4r7 --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726196925/summary.json'
Pipeline stage MKMLProfilerRunner completed in 849.28s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service riverise-0912-1056-sft-9k-v2-profiler is running
Tearing down inference service riverise-0912-1056-sft-9k-v2-profiler
Service riverise-0912-1056-sft-9k-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.82s
Shutdown handler de-registered
riverise-0912-1056-sft-9k_v2 status is now inactive due to auto deactivation removed underperforming models
riverise-0912-1056-sft-9k_v2 status is now torndown due to DeploymentManager action