submission_id: meta-llama-meta-llama-gu_1295_v2
developer_uid: chai_backend_admin
alignment_samples: 11839
alignment_score: -0.9468639597444208
best_of: 1
celo_rating: 1063.17
display_name: meta-llama-meta-llama-gu_1295_v2
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': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 1, 'max_output_tokens': 64}
gpu_counts: {'NVIDIA RTX A5000': 1}
is_internal_developer: True
language_model: meta-llama/Meta-Llama-Guard-2-8B
latencies: [{'batch_size': 1, 'throughput': 1.0652574293409152, 'latency_mean': 0.9386442172527313, 'latency_p50': 0.9364891052246094, 'latency_p90': 1.0493179082870483}, {'batch_size': 5, 'throughput': 3.479071577086908, 'latency_mean': 1.4323681235313415, 'latency_p50': 1.4477910995483398, 'latency_p90': 1.6255969762802125}, {'batch_size': 10, 'throughput': 5.1420973662557445, 'latency_mean': 1.9264125680923463, 'latency_p50': 1.907896637916565, 'latency_p90': 2.19656503200531}, {'batch_size': 15, 'throughput': 5.998036893576139, 'latency_mean': 2.4554552137851715, 'latency_p50': 2.469120740890503, 'latency_p90': 2.7524133205413817}, {'batch_size': 20, 'throughput': 6.546976812285759, 'latency_mean': 3.002843954563141, 'latency_p50': 2.9418859481811523, 'latency_p90': 3.5107940673828124}, {'batch_size': 25, 'throughput': 6.811656797314622, 'latency_mean': 3.5813551568984985, 'latency_p50': 3.571694493293762, 'latency_p90': 4.210130596160888}, {'batch_size': 30, 'throughput': 7.023160441454852, 'latency_mean': 4.151380101442337, 'latency_p50': 4.132696151733398, 'latency_p90': 4.839937496185303}, {'batch_size': 35, 'throughput': 7.168742024305675, 'latency_mean': 4.767038826942444, 'latency_p50': 4.676713824272156, 'latency_p90': 5.688713145256042}, {'batch_size': 40, 'throughput': 7.17197024408931, 'latency_mean': 5.380504152774811, 'latency_p50': 5.357659459114075, 'latency_p90': 6.352540612220764}]
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: meta-llama/Meta-Llama-Gu
model_name: meta-llama-meta-llama-gu_1295_v2
model_num_parameters: 8030261248.0
model_repo: meta-llama/Meta-Llama-Guard-2-8B
model_size: 8B
num_battles: 11838
num_wins: 3176
propriety_score: 0.75
propriety_total_count: 1040.0
ranking_group: single
status: inactive
submission_type: basic
throughput_3p7s: 7.04
timestamp: 2024-09-05T19:13:17+00:00
us_pacific_date: 2024-09-05
win_ratio: 0.26828856225713804
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 meta-llama-meta-llama-gu-1295-v2-mkmlizer
Waiting for job on meta-llama-meta-llama-gu-1295-v2-mkmlizer to finish
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ _____ __ __ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ /___/ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ Version: 0.10.1 ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ https://mk1.ai ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ The license key for the current software has been verified as ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ belonging to: ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ Chai Research Corp. ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ║ ║
meta-llama-meta-llama-gu-1295-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
meta-llama-meta-llama-gu-1295-v2-mkmlizer: Downloaded to shared memory in 52.093s
meta-llama-meta-llama-gu-1295-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp74ggrbvl, device:0
meta-llama-meta-llama-gu-1295-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
meta-llama-meta-llama-gu-1295-v2-mkmlizer: quantized model in 26.013s
meta-llama-meta-llama-gu-1295-v2-mkmlizer: Processed model meta-llama/Meta-Llama-Guard-2-8B in 78.106s
meta-llama-meta-llama-gu-1295-v2-mkmlizer: creating bucket guanaco-mkml-models
meta-llama-meta-llama-gu-1295-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
meta-llama-meta-llama-gu-1295-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2
meta-llama-meta-llama-gu-1295-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2/special_tokens_map.json
meta-llama-meta-llama-gu-1295-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2/config.json
meta-llama-meta-llama-gu-1295-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2/tokenizer_config.json
meta-llama-meta-llama-gu-1295-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2/tokenizer.json
meta-llama-meta-llama-gu-1295-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/meta-llama-meta-llama-gu-1295-v2/flywheel_model.0.safetensors
Job meta-llama-meta-llama-gu-1295-v2-mkmlizer completed after 104.66s with status: succeeded
Stopping job with name meta-llama-meta-llama-gu-1295-v2-mkmlizer
Pipeline stage MKMLizer completed in 105.53s
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 meta-llama-meta-llama-gu-1295-v2
Waiting for inference service meta-llama-meta-llama-gu-1295-v2 to be ready
Inference service meta-llama-meta-llama-gu-1295-v2 ready after 151.2551851272583s
Pipeline stage MKMLDeployer completed in 151.78s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.7108867168426514s
Received healthy response to inference request in 0.337430477142334s
Received healthy response to inference request in 0.6747920513153076s
Received healthy response to inference request in 0.6121530532836914s
Received healthy response to inference request in 0.7896416187286377s
5 requests
0 failed requests
5th percentile: 0.39237499237060547
10th percentile: 0.44731950759887695
20th percentile: 0.5572085380554199
30th percentile: 0.6246808528900146
40th percentile: 0.6497364521026612
50th percentile: 0.6747920513153076
60th percentile: 0.7207318782806397
70th percentile: 0.7666717052459716
80th percentile: 0.9738906383514406
90th percentile: 1.342388677597046
95th percentile: 1.5266376972198485
99th percentile: 1.6740369129180908
mean time: 0.8249807834625245
Pipeline stage StressChecker completed in 5.66s
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 4.18s
Shutdown handler de-registered
meta-llama-meta-llama-gu_1295_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 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service meta-llama-meta-llama-gu-1295-v2-profiler
Waiting for inference service meta-llama-meta-llama-gu-1295-v2-profiler to be ready
Inference service meta-llama-meta-llama-gu-1295-v2-profiler ready after 130.312650680542s
Pipeline stage MKMLProfilerDeployer completed in 130.70s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/meta-llama-meta-llam8bddc58073edf1c9d080402d455b1700-deplogc2vr:/code/chaiverse_profiler_1725564052 --namespace tenant-chaiml-guanaco
kubectl exec -it meta-llama-meta-llam8bddc58073edf1c9d080402d455b1700-deplogc2vr --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725564052 && python profiles.py profile --best_of_n 1 --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_1725564052/summary.json'
kubectl exec -it meta-llama-meta-llam8bddc58073edf1c9d080402d455b1700-deplogc2vr --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725564052/summary.json'
Pipeline stage MKMLProfilerRunner completed in 466.88s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service meta-llama-meta-llama-gu-1295-v2-profiler is running
Tearing down inference service meta-llama-meta-llama-gu-1295-v2-profiler
Service meta-llama-meta-llama-gu-1295-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.83s
Shutdown handler de-registered
meta-llama-meta-llama-gu_1295_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics