chaiml-gy-exp78-sft-gy-_32548

developer_uid: bogoconic1

submission_id: chaiml-gy-exp78-sft-gy-_32548_v1

model_name: chaiml-gy-exp78-sft-gy-_32548_v1

model_group: ChaiML/gy-exp78-sft-gy-d

status: torndown

timestamp: 2025-07-04T12:49:32+00:00

num_battles: 10898

num_wins: 5600

celo_rating: 1296.83

family_friendly_score: 0.5502

family_friendly_standard_error: 0.007035338797812086

submission_type: basic

model_repo: ChaiML/gy-exp78-sft-gy-datamix-v1-AR-ep2

model_architecture: MistralForCausalLM

model_num_parameters: 24096691200.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.5202083982895557, 'latency_mean': 1.922203722000122, 'latency_p50': 1.9178752899169922, 'latency_p90': 2.144436407089233}, {'batch_size': 3, 'throughput': 1.0517150882045043, 'latency_mean': 2.847410111427307, 'latency_p50': 2.8590681552886963, 'latency_p90': 3.151772665977478}, {'batch_size': 5, 'throughput': 1.333213899503406, 'latency_mean': 3.722899979352951, 'latency_p50': 3.7236785888671875, 'latency_p90': 4.19783661365509}, {'batch_size': 6, 'throughput': 1.4402771961104635, 'latency_mean': 4.142328170537948, 'latency_p50': 4.150579810142517, 'latency_p90': 4.584323382377624}, {'batch_size': 8, 'throughput': 1.5600151033032663, 'latency_mean': 5.0917341268062595, 'latency_p50': 5.137490391731262, 'latency_p90': 5.691117906570435}, {'batch_size': 10, 'throughput': 1.655829114966321, 'latency_mean': 5.991988573074341, 'latency_p50': 5.985331654548645, 'latency_p90': 6.7636500835418705}]

gpu_counts: {'NVIDIA A100-SXM4-80GB': 1}

display_name: chaiml-gy-exp78-sft-gy-_32548_v1

is_internal_developer: True

language_model: ChaiML/gy-exp78-sft-gy-datamix-v1-AR-ep2

model_size: 24B

ranking_group: single

throughput_3p7s: 1.34

us_pacific_date: 2025-07-04

win_ratio: 0.5138557533492384

generation_params: {'temperature': 0.6, 'top_p': 0.95, 'min_p': 0.025, 'top_k': 60, 'presence_penalty': 0.4, 'frequency_penalty': 0.4, 'stopping_words': ['<|im_start|>', '<|im_end|>', '\n'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '<|system|>Family Friendly{memory}\n', 'prompt_template': '', 'bot_template': '<|im_start|>assistant\n{message}<|im_end|>\n', 'user_template': '<|im_start|>user\nYou:{message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n', 'truncate_by_message': True}

Resubmit model

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
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 chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer
Waiting for job on chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer to finish
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║                                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   ██████   ██████  █████   ████  ████                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  ░░██████ ██████  ░░███   ███░  ░░███                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   ░███░█████░███   ░███  ███     ░███                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   ░███░░███ ░███   ░███████      ░███                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   ░███ ░░░  ░███   ░███░░███     ░███                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   ░███      ░███   ░███ ░░███    ░███                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║   █████     █████  █████ ░░████  █████                              ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  ░░░░░     ░░░░░  ░░░░░   ░░░░  ░░░░░                               ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║                                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Version: 0.29.15                                                   ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Features: FLYWHEEL, CUDA                                           ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Copyright 2023-2025 MK ONE TECHNOLOGIES Inc.                       ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  https://mk1.ai                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║                                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  belonging to:                                                      ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║                                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Chai Research Corp.                                                ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║  Expiration: 2028-03-31 23:59:59                                    ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ║                                                                     ║
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: Downloaded to shared memory in 464.800s
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: Checking if ChaiML/gy-exp78-sft-gy-datamix-v1-AR-ep2 already exists in ChaiML
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp9q0sklkb, device:0
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Unable to record family friendly update due to error: Invalid JSON input: JSON must contain 'User Safety' and 'Response Safety' fields
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: quantized model in 48.353s
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: Processed model ChaiML/gy-exp78-sft-gy-datamix-v1-AR-ep2 in 513.154s
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: creating bucket guanaco-mkml-models
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/config.json
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/special_tokens_map.json
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/tokenizer_config.json
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/tokenizer.json
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/flywheel_model.1.safetensors
chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-gy-exp78-sft-gy-32548-v1/nvidia/flywheel_model.0.safetensors
Job chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer completed after 546.07s with status: succeeded
Stopping job with name chaiml-gy-exp78-sft-gy-32548-v1-mkmlizer
Pipeline stage MKMLizer completed in 546.58s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.17s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service chaiml-gy-exp78-sft-gy-32548-v1
Waiting for inference service chaiml-gy-exp78-sft-gy-32548-v1 to be ready
Inference service chaiml-gy-exp78-sft-gy-32548-v1 ready after 220.9116063117981s
Pipeline stage MKMLDeployer completed in 221.52s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.559305429458618s
Received healthy response to inference request in 2.2621653079986572s
Received healthy response to inference request in 2.016829013824463s
Received healthy response to inference request in 1.8521461486816406s
Received healthy response to inference request in 2.063340187072754s
5 requests
0 failed requests
5th percentile: 1.885082721710205
10th percentile: 1.9180192947387695
20th percentile: 1.9838924407958984
30th percentile: 2.026131248474121
40th percentile: 2.0447357177734373
50th percentile: 2.063340187072754
60th percentile: 2.142870235443115
70th percentile: 2.2224002838134767
80th percentile: 2.3215933322906497
90th percentile: 2.4404493808746337
95th percentile: 2.4998774051666257
99th percentile: 2.5474198246002198
mean time: 2.1507572174072265
Pipeline stage StressChecker completed in 12.19s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 0.84s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 0.72s
Shutdown handler de-registered
chaiml-gy-exp78-sft-gy-_32548_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 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.13s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service chaiml-gy-exp78-sft-gy-32548-v1-profiler
Waiting for inference service chaiml-gy-exp78-sft-gy-32548-v1-profiler to be ready
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyScorer
Evaluating %s Family Friendly Score with %s threads
Pipeline stage OfflineFamilyFriendlyScorer completed in 3291.42s
Shutdown handler de-registered
chaiml-gy-exp78-sft-gy-_32548_v1 status is now inactive due to auto deactivation removed underperforming models
chaiml-gy-exp78-sft-gy-_32548_v1 status is now torndown due to DeploymentManager action