rirv938-llama-8b-dpo-vs-_5675

developer_uid: robert_irvine

submission_id: rirv938-llama-8b-dpo-vs-_5675_v1

model_name: rirv938-llama-8b-dpo-vs-_5675_v1

model_group: rirv938/llama_8b_dpo_vs_

status: torndown

timestamp: 2024-09-14T20:52:57+00:00

num_battles: 11415

num_wins: 5753

celo_rating: 1257.29

family_friendly_score: 0.0

submission_type: basic

model_repo: rirv938/llama_8b_dpo_vs_dpo_250k_195

model_architecture: LlamaForSequenceClassification

model_num_parameters: 8030261248.0

best_of: 1

max_input_tokens: 256

max_output_tokens: 1

latencies: [{'batch_size': 1, 'throughput': 0.10510586478237989, 'latency_mean': 9.514121218919755, 'latency_p50': 9.517372488975525, 'latency_p90': 9.61135437488556}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: rirv938-llama-8b-dpo-vs-_5675_v1

ineligible_reason: max_output_tokens!=64

is_internal_developer: True

language_model: rirv938/llama_8b_dpo_vs_dpo_250k_195

model_size: 8B

ranking_group: single

us_pacific_date: 2024-09-14

win_ratio: 0.5039859833552344

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': 256, 'best_of': 1, 'max_output_tokens': 1}

formatter: {'memory_template': '', 'prompt_template': '', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '', 'truncate_by_message': False}

Resubmit model

Shutdown handler not registered because Python interpreter is not running in the main thread
Connection pool is full, discarding connection: %s. Connection pool size: %s
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLizer
Starting job with name rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer
Waiting for job on rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer to finish
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║     _____            __           __                                ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║       /___/                                                         ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║                                                                     ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  Version: 0.10.1                                                    ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  https://mk1.ai                                                     ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║                                                                     ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  belonging to:                                                      ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║                                                                     ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  Chai Research Corp.                                                ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ║                                                                     ║
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Failed to get response for submission blend_gusus_2024-09-14: ('http://sao10k-mn-12b-lyra-v4a1-v3-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:43378->127.0.0.1:8080: read: connection reset by peer\n')
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: Downloaded to shared memory in 32.737s
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:t0, folder:/tmp/tmpe5tszx67, device:0
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: quantized model in 84.377s
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: Processed model rirv938/llama_8b_dpo_vs_dpo_250k_195 in 117.114s
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: creating bucket guanaco-mkml-models
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1/special_tokens_map.json
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1/config.json
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1/tokenizer_config.json
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1/tokenizer.json
rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rirv938-llama-8b-dpo-vs-5675-v1/flywheel_model.0.safetensors
Job rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer completed after 146.5s with status: succeeded
Stopping job with name rirv938-llama-8b-dpo-vs-5675-v1-mkmlizer
Pipeline stage MKMLizer completed in 149.62s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.09s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rirv938-llama-8b-dpo-vs-5675-v1
Waiting for inference service rirv938-llama-8b-dpo-vs-5675-v1 to be ready
Failed to get response for submission blend_jitik_2024-08-26: ('http://chaiml-llama-8b-pairwis-8189-v19-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '{"error":"ValueError : [TypeError(\\"\'numpy.int64\' object is not iterable\\"), TypeError(\'vars() argument must have __dict__ attribute\')]"}')
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
Failed to get response for submission blend_hokok_2024-09-09: ('http://neversleep-noromaid-v0-8068-v150-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', '')
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 rirv938-llama-8b-dpo-vs-5675-v1 ready after 170.56834959983826s
Pipeline stage MKMLDeployer completed in 172.04s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.718264579772949s
Received healthy response to inference request in 4.295020341873169s
Received healthy response to inference request in 4.466186285018921s
Received healthy response to inference request in 3.949876546859741s
Received healthy response to inference request in 2.976877450942993s
5 requests
0 failed requests
5th percentile: 3.1251548767089843
10th percentile: 3.2734323024749754
20th percentile: 3.569987154006958
30th percentile: 3.7645869731903074
40th percentile: 3.8572317600250243
50th percentile: 3.949876546859741
60th percentile: 4.087934064865112
70th percentile: 4.225991582870483
80th percentile: 4.329253530502319
90th percentile: 4.39771990776062
95th percentile: 4.431953096389771
99th percentile: 4.459339647293091
mean time: 3.8812450408935546
Pipeline stage StressChecker completed in 22.65s
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 6.07s
Shutdown handler de-registered
rirv938-llama-8b-dpo-vs-_5675_v1 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 0.12s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service rirv938-llama-8b-dpo-vs-5675-v1-profiler
Waiting for inference service rirv938-llama-8b-dpo-vs-5675-v1-profiler to be ready
Inference service rirv938-llama-8b-dpo-vs-5675-v1-profiler ready after 170.42290043830872s
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/rirv938-llama-8b-dpocb09162052dbc1ecbc63406f7e64f5fb-deplozthtz:/code/chaiverse_profiler_1726347748 --namespace tenant-chaiml-guanaco
kubectl exec -it rirv938-llama-8b-dpocb09162052dbc1ecbc63406f7e64f5fb-deplozthtz --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726347748 && 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 256 --output_tokens 1 --summary /code/chaiverse_profiler_1726347748/summary.json'
kubectl exec -it rirv938-llama-8b-dpocb09162052dbc1ecbc63406f7e64f5fb-deplozthtz --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726347748/summary.json'
Pipeline stage MKMLProfilerRunner completed in 1910.36s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rirv938-llama-8b-dpo-vs-5675-v1-profiler is running
Tearing down inference service rirv938-llama-8b-dpo-vs-5675-v1-profiler
Service rirv938-llama-8b-dpo-vs-5675-v1-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.93s
Shutdown handler de-registered
rirv938-llama-8b-dpo-vs-_5675_v1 status is now inactive due to auto deactivation removed underperforming models
rirv938-llama-8b-dpo-vs-_5675_v1 status is now torndown due to DeploymentManager action