developer_uid: Riverise
submission_id: riverise-0910-2005-sft_v2
model_name: riverise-0910-2005-sft_v1
model_group: Riverise/0910_2005_sft
status: torndown
timestamp: 2024-09-11T08:54:20+00:00
num_battles: 11921
num_wins: 6142
celo_rating: 1252.14
family_friendly_score: 0.0
submission_type: basic
model_repo: Riverise/0910_2005_sft
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
latencies: [{'batch_size': 1, 'throughput': 0.9056738907958647, 'latency_mean': 1.1040505838394166, 'latency_p50': 1.0990557670593262, 'latency_p90': 1.2367490768432616}, {'batch_size': 3, 'throughput': 1.6134618949223856, 'latency_mean': 1.848477965593338, 'latency_p50': 1.858961582183838, 'latency_p90': 2.0556514739990233}, {'batch_size': 5, 'throughput': 1.8469144479642576, 'latency_mean': 2.686871590614319, 'latency_p50': 2.693644881248474, 'latency_p90': 3.019568943977356}, {'batch_size': 6, 'throughput': 1.8796971109614469, 'latency_mean': 3.170622926950455, 'latency_p50': 3.2050715684890747, 'latency_p90': 3.5244399309158325}, {'batch_size': 8, 'throughput': 1.9777185507467878, 'latency_mean': 4.008349504470825, 'latency_p50': 4.040689587593079, 'latency_p90': 4.452486371994018}, {'batch_size': 10, 'throughput': 1.9791837426133996, 'latency_mean': 5.004721310138702, 'latency_p50': 5.0085285902023315, 'latency_p90': 5.678121972084045}]
gpu_counts: {'NVIDIA RTX A5000': 1}
display_name: riverise-0910-2005-sft_v1
is_internal_developer: False
language_model: Riverise/0910_2005_sft
model_size: 8B
ranking_group: single
throughput_3p7s: 1.96
us_pacific_date: 2024-09-11
win_ratio: 0.5152252327824847
generation_params: {'temperature': 0.85, 'top_p': 0.95, 'min_p': 0.2, '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}
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}
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-v2-mkmlizer
Waiting for job on riverise-0910-2005-sft-v2-mkmlizer to finish
riverise-0910-2005-sft-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
riverise-0910-2005-sft-v2-mkmlizer: ║     _____            __           __                                ║
riverise-0910-2005-sft-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
riverise-0910-2005-sft-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
riverise-0910-2005-sft-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
riverise-0910-2005-sft-v2-mkmlizer: ║       /___/                                                         ║
riverise-0910-2005-sft-v2-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v2-mkmlizer: ║  Version: 0.10.1                                                    ║
riverise-0910-2005-sft-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
riverise-0910-2005-sft-v2-mkmlizer: ║  https://mk1.ai                                                     ║
riverise-0910-2005-sft-v2-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
riverise-0910-2005-sft-v2-mkmlizer: ║  belonging to:                                                      ║
riverise-0910-2005-sft-v2-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v2-mkmlizer: ║  Chai Research Corp.                                                ║
riverise-0910-2005-sft-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
riverise-0910-2005-sft-v2-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
riverise-0910-2005-sft-v2-mkmlizer: ║                                                                     ║
riverise-0910-2005-sft-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
riverise-0910-2005-sft-v2-mkmlizer: Downloaded to shared memory in 20.767s
riverise-0910-2005-sft-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpvs_xv1px, device:0
riverise-0910-2005-sft-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
riverise-0910-2005-sft-v2-mkmlizer: quantized model in 25.828s
riverise-0910-2005-sft-v2-mkmlizer: Processed model Riverise/0910_2005_sft in 46.595s
riverise-0910-2005-sft-v2-mkmlizer: creating bucket guanaco-mkml-models
riverise-0910-2005-sft-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
riverise-0910-2005-sft-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/riverise-0910-2005-sft-v2
riverise-0910-2005-sft-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v2/special_tokens_map.json
riverise-0910-2005-sft-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v2/config.json
riverise-0910-2005-sft-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v2/tokenizer_config.json
riverise-0910-2005-sft-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/riverise-0910-2005-sft-v2/tokenizer.json
Job riverise-0910-2005-sft-v2-mkmlizer completed after 74.12s with status: succeeded
Stopping job with name riverise-0910-2005-sft-v2-mkmlizer
Pipeline stage MKMLizer completed in 75.90s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.13s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service riverise-0910-2005-sft-v2
Waiting for inference service riverise-0910-2005-sft-v2 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', 'read tcp 127.0.0.1:35298->127.0.0.1:8080: read: connection reset by peer\n')
Failed to get response for submission blend_henub_2024-09-09: ('http://chaiml-llama-8b-pairwis-8189-v19-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:36162->127.0.0.1:8080: read: connection reset by peer\n')
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-0910-2005-sft-v2 ready after 161.1864252090454s
Pipeline stage MKMLDeployer completed in 161.51s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.2514772415161133s
Received healthy response to inference request in 1.8319799900054932s
Received healthy response to inference request in 1.608520746231079s
Received healthy response to inference request in 1.6829147338867188s
Received healthy response to inference request in 1.6401772499084473s
5 requests
0 failed requests
5th percentile: 1.6148520469665528
10th percentile: 1.6211833477020263
20th percentile: 1.6338459491729735
30th percentile: 1.6487247467041015
40th percentile: 1.6658197402954102
50th percentile: 1.6829147338867188
60th percentile: 1.7425408363342285
70th percentile: 1.8021669387817383
80th percentile: 1.9158794403076174
90th percentile: 2.0836783409118653
95th percentile: 2.167577791213989
99th percentile: 2.2346973514556883
mean time: 1.8030139923095703
Pipeline stage StressChecker completed in 9.66s
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 6.44s
Shutdown handler de-registered
riverise-0910-2005-sft_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-0910-2005-sft-v2-profiler
Waiting for inference service riverise-0910-2005-sft-v2-profiler to be ready
Inference service riverise-0910-2005-sft-v2-profiler ready after 160.40316772460938s
Pipeline stage MKMLProfilerDeployer completed in 160.78s
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-v2-profiler-predictor-00001-deploym6k2mw:/code/chaiverse_profiler_1726045308 --namespace tenant-chaiml-guanaco
kubectl exec -it riverise-0910-2005-sft-v2-profiler-predictor-00001-deploym6k2mw --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1726045308 && 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_1726045308/summary.json'
kubectl exec -it riverise-0910-2005-sft-v2-profiler-predictor-00001-deploym6k2mw --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1726045308/summary.json'
Pipeline stage MKMLProfilerRunner completed in 767.21s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service riverise-0910-2005-sft-v2-profiler is running
Tearing down inference service riverise-0910-2005-sft-v2-profiler
Service riverise-0910-2005-sft-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.85s
Shutdown handler de-registered
riverise-0910-2005-sft_v2 status is now inactive due to auto deactivation removed underperforming models
riverise-0910-2005-sft_v2 status is now torndown due to DeploymentManager action