developer_uid: huohuo12
submission_id: yamatazen-ayla-light-12b-v2_v14
model_name: yamatazen-ayla-light-12b-v2_v2
model_group: yamatazen/Ayla-Light-12B
status: torndown
timestamp: 2025-02-26T09:56:45+00:00
num_battles: 6216
num_wins: 3088
celo_rating: 1264.07
family_friendly_score: 0.5548
family_friendly_standard_error: 0.007028470103799261
submission_type: basic
model_repo: yamatazen/Ayla-Light-12B-v2
model_architecture: MistralForCausalLM
model_num_parameters: 12772070400.0
best_of: 8
max_input_tokens: 1024
max_output_tokens: 64
latencies: [{'batch_size': 1, 'throughput': 0.6030912750116583, 'latency_mean': 1.6580429089069366, 'latency_p50': 1.6618671417236328, 'latency_p90': 1.828090524673462}, {'batch_size': 3, 'throughput': 1.1068433161017877, 'latency_mean': 2.7002204561233523, 'latency_p50': 2.685539722442627, 'latency_p90': 2.9901005983352658}, {'batch_size': 5, 'throughput': 1.320666588026472, 'latency_mean': 3.776322511434555, 'latency_p50': 3.754637122154236, 'latency_p90': 4.203949308395385}, {'batch_size': 6, 'throughput': 1.391086109269058, 'latency_mean': 4.278331043720246, 'latency_p50': 4.289143085479736, 'latency_p90': 4.855329966545105}, {'batch_size': 8, 'throughput': 1.442717531700087, 'latency_mean': 5.500245188474655, 'latency_p50': 5.59487783908844, 'latency_p90': 6.222860288619995}, {'batch_size': 10, 'throughput': 1.4764414183414287, 'latency_mean': 6.730411860942841, 'latency_p50': 6.745451211929321, 'latency_p90': 7.549600553512573}]
gpu_counts: {'NVIDIA RTX A5000': 1}
display_name: yamatazen-ayla-light-12b-v2_v2
is_internal_developer: False
language_model: yamatazen/Ayla-Light-12B-v2
model_size: 13B
ranking_group: single
throughput_3p7s: 1.31
us_pacific_date: 2025-02-26
win_ratio: 0.4967824967824968
generation_params: {'temperature': 1.05, 'top_p': 0.85, 'min_p': 0.24, 'top_k': 55, 'presence_penalty': 0.2, 'frequency_penalty': 0.2, 'stopping_words': ['\n'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}
formatter: {'memory_template': '<|im_start|>system\n{memory}<|im_end|>\n', 'prompt_template': '<|im_start|>user\n{prompt}<|im_end|>\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n{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 yamatazen-ayla-light-12b-v2-v14-mkmlizer
Waiting for job on yamatazen-ayla-light-12b-v2-v14-mkmlizer to finish
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ _____ __ __ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ /___/ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ Version: 0.12.8 ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ https://mk1.ai ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ The license key for the current software has been verified as ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ belonging to: ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ Chai Research Corp. ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ Expiration: 2025-04-15 23:59:59 ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ║ ║
yamatazen-ayla-light-12b-v2-v14-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Failed to get response for submission nischaydnk-exp16-mistra_31861_v2: HTTPConnectionPool(host='nischaydnk-exp16-mistra-31861-v2-predictor.tenant-chaiml-guanaco.k.chaiverse.com', port=80): Read timed out. (read timeout=12.0)
yamatazen-ayla-light-12b-v2-v14-mkmlizer: Downloaded to shared memory in 32.124s
yamatazen-ayla-light-12b-v2-v14-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmphswiq43q, device:0
yamatazen-ayla-light-12b-v2-v14-mkmlizer: Saving flywheel model at /dev/shm/model_cache
yamatazen-ayla-light-12b-v2-v14-mkmlizer: quantized model in 36.890s
yamatazen-ayla-light-12b-v2-v14-mkmlizer: Processed model yamatazen/Ayla-Light-12B-v2 in 69.015s
yamatazen-ayla-light-12b-v2-v14-mkmlizer: creating bucket guanaco-mkml-models
yamatazen-ayla-light-12b-v2-v14-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
yamatazen-ayla-light-12b-v2-v14-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/yamatazen-ayla-light-12b-v2-v14
yamatazen-ayla-light-12b-v2-v14-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/yamatazen-ayla-light-12b-v2-v14/config.json
yamatazen-ayla-light-12b-v2-v14-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/yamatazen-ayla-light-12b-v2-v14/special_tokens_map.json
yamatazen-ayla-light-12b-v2-v14-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/yamatazen-ayla-light-12b-v2-v14/tokenizer_config.json
yamatazen-ayla-light-12b-v2-v14-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/yamatazen-ayla-light-12b-v2-v14/tokenizer.json
yamatazen-ayla-light-12b-v2-v14-mkmlizer: Loading 0: 0%| | 0/363 [00:00<?, ?it/s] Loading 0: 1%| | 2/363 [00:06<18:47, 3.12s/it] Loading 0: 2%|▏ | 6/363 [00:06<04:59, 1.19it/s] Loading 0: 4%|▎ | 13/363 [00:06<01:46, 3.28it/s] Loading 0: 5%|▌ | 19/363 [00:06<01:00, 5.65it/s] Loading 0: 7%|▋ | 24/363 [00:06<00:43, 7.85it/s] Loading 0: 9%|▉ | 32/363 [00:06<00:25, 12.94it/s] Loading 0: 10%|█ | 38/363 [00:07<00:19, 16.50it/s] Loading 0: 12%|█▏ | 43/363 [00:07<00:19, 16.63it/s] Loading 0: 14%|█▍ | 50/363 [00:07<00:13, 22.38it/s] Loading 0: 15%|█▌ | 55/363 [00:07<00:11, 26.15it/s] Loading 0: 17%|█▋ | 60/363 [00:07<00:12, 25.12it/s] Loading 0: 18%|█▊ | 65/363 [00:07<00:10, 28.56it/s] Loading 0: 19%|█▉ | 70/363 [00:08<00:09, 30.25it/s] Loading 0: 21%|██ | 76/363 [00:08<00:07, 36.15it/s] Loading 0: 22%|██▏ | 81/363 [00:08<00:07, 38.27it/s] Loading 0: 24%|██▎ | 86/363 [00:08<00:06, 40.72it/s] Loading 0: 25%|██▌ | 92/363 [00:08<00:06, 39.65it/s] Loading 0: 27%|██▋ | 97/363 [00:08<00:07, 36.98it/s] Loading 0: 28%|██▊ | 101/363 [00:08<00:07, 37.10it/s] Loading 0: 29%|██▉ | 105/363 [00:08<00:07, 34.71it/s] Loading 0: 31%|███ | 112/363 [00:09<00:05, 43.16it/s] Loading 0: 32%|███▏ | 117/363 [00:09<00:05, 43.95it/s] Loading 0: 34%|███▎ | 122/363 [00:09<00:07, 31.32it/s] Loading 0: 35%|███▍ | 126/363 [00:09<00:07, 30.72it/s] Loading 0: 36%|███▌ | 130/363 [00:09<00:07, 31.02it/s] Loading 0: 37%|███▋ | 134/363 [00:09<00:07, 30.68it/s] Loading 0: 38%|███▊ | 139/363 [00:09<00:06, 34.61it/s] Loading 0: 39%|███▉ | 143/363 [00:10<00:06, 35.07it/s] Loading 0: 41%|████ | 149/363 [00:10<00:05, 39.65it/s] Loading 0: 43%|████▎ | 155/363 [00:10<00:05, 39.63it/s] Loading 0: 44%|████▍ | 160/363 [00:10<00:05, 37.28it/s] Loading 0: 45%|████▌ | 164/363 [00:10<00:05, 37.12it/s] Loading 0: 46%|████▋ | 168/363 [00:10<00:05, 33.56it/s] Loading 0: 48%|████▊ | 175/363 [00:10<00:04, 39.87it/s] Loading 0: 50%|████▉ | 180/363 [00:10<00:04, 38.99it/s] Loading 0: 51%|█████ | 184/363 [00:11<00:04, 39.06it/s] Loading 0: 52%|█████▏ | 188/363 [00:11<00:04, 36.74it/s] Loading 0: 53%|█████▎ | 193/363 [00:11<00:04, 37.98it/s] Loading 0: 54%|█████▍ | 197/363 [00:11<00:04, 37.14it/s] Loading 0: 56%|█████▌ | 202/363 [00:11<00:05, 29.60it/s] Loading 0: 57%|█████▋ | 206/363 [00:11<00:05, 30.87it/s] Loading 0: 58%|█████▊ | 212/363 [00:11<00:04, 35.67it/s] Loading 0: 60%|██████ | 218/363 [00:12<00:03, 37.49it/s] Loading 0: 61%|██████ | 222/363 [00:12<00:04, 35.03it/s] Loading 0: 63%|██████▎ | 230/363 [00:12<00:03, 43.68it/s] Loading 0: 65%|██████▌ | 236/363 [00:12<00:03, 42.24it/s] Loading 0: 66%|██████▋ | 241/363 [00:12<00:02, 42.38it/s] Loading 0: 68%|██████▊ | 248/363 [00:12<00:02, 47.40it/s] Loading 0: 70%|██████▉ | 254/363 [00:12<00:02, 46.07it/s] Loading 0: 71%|███████▏ | 259/363 [00:12<00:02, 44.04it/s] Loading 0: 73%|███████▎ | 265/363 [00:13<00:02, 42.58it/s] Loading 0: 74%|███████▍ | 270/363 [00:13<00:02, 41.12it/s] Loading 0: 76%|███████▌ | 275/363 [00:13<00:02, 37.81it/s] Loading 0: 77%|███████▋ | 280/363 [00:13<00:02, 40.54it/s] Loading 0: 79%|███████▊ | 285/363 [00:13<00:02, 26.90it/s] Loading 0: 80%|████████ | 292/363 [00:13<00:02, 34.36it/s] Loading 0: 82%|████████▏ | 297/363 [00:14<00:01, 35.72it/s] Loading 0: 83%|████████▎ | 302/363 [00:14<00:01, 35.39it/s] Loading 0: 85%|████████▍ | 307/363 [00:14<00:01, 36.60it/s] Loading 0: 86%|████████▌ | 312/363 [00:14<00:01, 30.74it/s] Loading 0: 88%|████████▊ | 319/363 [00:14<00:01, 37.46it/s] Loading 0: 89%|████████▉ | 324/363 [00:14<00:00, 39.17it/s] Loading 0: 91%|█████████ | 329/363 [00:14<00:00, 40.51it/s] Loading 0: 92%|█████████▏| 334/363 [00:15<00:00, 41.47it/s] Loading 0: 93%|█████████▎| 339/363 [00:15<00:00, 32.18it/s] Loading 0: 95%|█████████▌| 346/363 [00:15<00:00, 39.47it/s] Loading 0: 97%|█████████▋| 351/363 [00:15<00:00, 41.41it/s] Loading 0: 98%|█████████▊| 356/363 [00:15<00:00, 41.01it/s] Loading 0: 100%|█████████▉| 362/363 [00:15<00:00, 41.67it/s]
Job yamatazen-ayla-light-12b-v2-v14-mkmlizer completed after 94.6s with status: succeeded
Stopping job with name yamatazen-ayla-light-12b-v2-v14-mkmlizer
Pipeline stage MKMLizer completed in 95.10s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.16s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service yamatazen-ayla-light-12b-v2-v14
Waiting for inference service yamatazen-ayla-light-12b-v2-v14 to be ready
Inference service yamatazen-ayla-light-12b-v2-v14 ready after 260.8543450832367s
Pipeline stage MKMLDeployer completed in 261.40s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.3423988819122314s
Received healthy response to inference request in 1.6871986389160156s
Received healthy response to inference request in 1.5144767761230469s
Received healthy response to inference request in 1.5670289993286133s
Received healthy response to inference request in 1.4293644428253174s
5 requests
0 failed requests
5th percentile: 1.4463869094848634
10th percentile: 1.4634093761444091
20th percentile: 1.4974543094635009
30th percentile: 1.5249872207641602
40th percentile: 1.5460081100463867
50th percentile: 1.5670289993286133
60th percentile: 1.6150968551635743
70th percentile: 1.6631647109985352
80th percentile: 1.818238687515259
90th percentile: 2.080318784713745
95th percentile: 2.211358833312988
99th percentile: 2.316190872192383
mean time: 1.7080935478210448
Pipeline stage StressChecker completed in 9.80s
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 0.66s
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 0.69s
Shutdown handler de-registered
yamatazen-ayla-light-12b-v2_v14 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 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service yamatazen-ayla-light-12b-v2-v14-profiler
Waiting for inference service yamatazen-ayla-light-12b-v2-v14-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 2580.19s
Shutdown handler de-registered
yamatazen-ayla-light-12b-v2_v14 status is now inactive due to auto deactivation removed underperforming models
yamatazen-ayla-light-12b-v2_v14 status is now torndown due to DeploymentManager action