rica40325-lora32-third-step

submission_id: rica40325-lora32-third-step_v4

developer_uid: rica40325

best_of: 16

celo_rating: 1260.55

display_name: rica40325-lora32-third-step_v2

family_friendly_score: 0.0

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': 0.95, 'top_p': 0.95, 'min_p': 0.05, '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}

gpu_counts: {'NVIDIA RTX A5000': 1}

is_internal_developer: False

language_model: rica40325/lora32_third_step

latencies: [{'batch_size': 1, 'throughput': 0.916293864151681, 'latency_mean': 1.09129270195961, 'latency_p50': 1.0787122249603271, 'latency_p90': 1.229987382888794}, {'batch_size': 4, 'throughput': 1.752723396425103, 'latency_mean': 2.271913368701935, 'latency_p50': 2.264933466911316, 'latency_p90': 2.571057367324829}, {'batch_size': 5, 'throughput': 1.8515589769155583, 'latency_mean': 2.682383728027344, 'latency_p50': 2.663703680038452, 'latency_p90': 3.038500046730041}, {'batch_size': 8, 'throughput': 1.9440599428617582, 'latency_mean': 4.088932136297226, 'latency_p50': 4.119110703468323, 'latency_p90': 4.641162610054016}, {'batch_size': 10, 'throughput': 1.9986349464515685, 'latency_mean': 4.947938182353973, 'latency_p50': 4.908643126487732, 'latency_p90': 5.692831516265869}, {'batch_size': 12, 'throughput': 1.9919807499320459, 'latency_mean': 5.9406349003314975, 'latency_p50': 5.93007230758667, 'latency_p90': 7.090457344055175}, {'batch_size': 15, 'throughput': 2.0013185605311805, 'latency_mean': 7.340696872472763, 'latency_p50': 7.390771389007568, 'latency_p90': 8.175047373771667}]

max_input_tokens: 512

max_output_tokens: 64

model_architecture: LlamaForCausalLM

model_group: rica40325/lora32_third_s

model_name: rica40325-lora32-third-step_v2

model_num_parameters: 8030261248.0

model_repo: rica40325/lora32_third_step

model_size: 8B

num_battles: 10246

num_wins: 5471

ranking_group: single

status: torndown

submission_type: basic

throughput_3p7s: 1.94

timestamp: 2024-08-31T10:32:04+00:00

us_pacific_date: 2024-08-31

win_ratio: 0.5339644739410502

Resubmit model

run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLizer
Starting job with name rica40325-lora32-third-step-v4-mkmlizer
Waiting for job on rica40325-lora32-third-step-v4-mkmlizer to finish
rica40325-lora32-third-step-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rica40325-lora32-third-step-v4-mkmlizer: ║     _____            __           __                                ║
rica40325-lora32-third-step-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rica40325-lora32-third-step-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rica40325-lora32-third-step-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rica40325-lora32-third-step-v4-mkmlizer: ║       /___/                                                         ║
rica40325-lora32-third-step-v4-mkmlizer: ║                                                                     ║
rica40325-lora32-third-step-v4-mkmlizer: ║  Version: 0.10.1                                                    ║
rica40325-lora32-third-step-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rica40325-lora32-third-step-v4-mkmlizer: ║  https://mk1.ai                                                     ║
rica40325-lora32-third-step-v4-mkmlizer: ║                                                                     ║
rica40325-lora32-third-step-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
rica40325-lora32-third-step-v4-mkmlizer: ║  belonging to:                                                      ║
rica40325-lora32-third-step-v4-mkmlizer: ║                                                                     ║
rica40325-lora32-third-step-v4-mkmlizer: ║  Chai Research Corp.                                                ║
rica40325-lora32-third-step-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rica40325-lora32-third-step-v4-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
rica40325-lora32-third-step-v4-mkmlizer: ║                                                                     ║
rica40325-lora32-third-step-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rica40325-lora32-third-step-v4-mkmlizer: Downloaded to shared memory in 23.531s
rica40325-lora32-third-step-v4-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpo8l26vl7, device:0
rica40325-lora32-third-step-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rica40325-lora32-third-step-v4-mkmlizer: quantized model in 25.790s
rica40325-lora32-third-step-v4-mkmlizer: Processed model rica40325/lora32_third_step in 49.322s
rica40325-lora32-third-step-v4-mkmlizer: creating bucket guanaco-mkml-models
rica40325-lora32-third-step-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rica40325-lora32-third-step-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rica40325-lora32-third-step-v4
rica40325-lora32-third-step-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rica40325-lora32-third-step-v4/config.json
rica40325-lora32-third-step-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rica40325-lora32-third-step-v4/special_tokens_map.json
rica40325-lora32-third-step-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rica40325-lora32-third-step-v4/tokenizer_config.json
rica40325-lora32-third-step-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rica40325-lora32-third-step-v4/tokenizer.json
rica40325-lora32-third-step-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rica40325-lora32-third-step-v4/flywheel_model.0.safetensors
rica40325-lora32-third-step-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|▏         | 4/291 [00:00<00:07, 38.37it/s]
Loading 0:   5%|▌         | 16/291 [00:00<00:03, 71.67it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:02, 91.36it/s]
Loading 0:  14%|█▎        | 40/291 [00:00<00:02, 89.87it/s]
Loading 0:  17%|█▋        | 49/291 [00:00<00:02, 85.09it/s]
Loading 0:  21%|██        | 61/291 [00:00<00:02, 87.41it/s]
Loading 0:  24%|██▍       | 70/291 [00:00<00:02, 86.73it/s]
Loading 0:  27%|██▋       | 79/291 [00:00<00:02, 84.67it/s]
Loading 0:  30%|███       | 88/291 [00:02<00:09, 20.95it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:06, 31.32it/s]
Loading 0:  40%|███▉      | 115/291 [00:02<00:04, 39.52it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:03, 52.02it/s]
Loading 0:  49%|████▉     | 142/291 [00:02<00:02, 59.52it/s]
Loading 0:  54%|█████▍    | 157/291 [00:02<00:01, 68.57it/s]
Loading 0:  57%|█████▋    | 167/291 [00:02<00:01, 73.00it/s]
Loading 0:  61%|██████    | 178/291 [00:03<00:01, 75.67it/s]
Loading 0:  64%|██████▍   | 187/291 [00:04<00:04, 24.41it/s]
Loading 0:  69%|██████▉   | 202/291 [00:04<00:02, 34.21it/s]
Loading 0:  73%|███████▎  | 211/291 [00:04<00:02, 39.45it/s]
Loading 0:  77%|███████▋  | 223/291 [00:04<00:01, 47.41it/s]
Loading 0:  82%|████████▏ | 238/291 [00:04<00:00, 59.28it/s]
Loading 0:  86%|████████▌ | 250/291 [00:04<00:00, 66.23it/s]
Loading 0:  91%|█████████ | 265/291 [00:05<00:00, 77.06it/s]
Loading 0:  95%|█████████▌| 277/291 [00:05<00:00, 77.55it/s]
Loading 0:  99%|█████████▊| 287/291 [00:05<00:00, 48.07it/s]
                                                            
Job rica40325-lora32-third-step-v4-mkmlizer completed after 65.74s with status: succeeded
Stopping job with name rica40325-lora32-third-step-v4-mkmlizer
Pipeline stage MKMLizer completed in 66.73s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.40s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service rica40325-lora32-third-step-v4
Waiting for inference service rica40325-lora32-third-step-v4 to be ready
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Failed to get response for submission blend_lobuf_2024-08-22: ('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\')]"}')
Inference service rica40325-lora32-third-step-v4 ready after 261.4205503463745s
Pipeline stage MKMLDeployer completed in 262.56s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.520103931427002s
Received healthy response to inference request in 1.6391546726226807s
Received healthy response to inference request in 1.4892735481262207s
Received healthy response to inference request in 1.4222381114959717s
Received healthy response to inference request in 1.5793571472167969s
5 requests
0 failed requests
5th percentile: 1.4356451988220216
10th percentile: 1.4490522861480712
20th percentile: 1.4758664608001708
30th percentile: 1.5072902679443358
40th percentile: 1.5433237075805664
50th percentile: 1.5793571472167969
60th percentile: 1.6032761573791503
70th percentile: 1.627195167541504
80th percentile: 1.815344524383545
90th percentile: 2.1677242279052735
95th percentile: 2.3439140796661375
99th percentile: 2.484865961074829
mean time: 1.7300254821777343
Pipeline stage StressChecker completed in 10.45s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
starting trigger_guanaco_pipeline %s
Pipeline stage TriggerMKMLProfilingPipeline completed in 9.16s
rica40325-lora32-third-step_v4 status is now deployed due to DeploymentManager action
run pipeline %s
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 rica40325-lora32-third-step-v4-profiler
Waiting for inference service rica40325-lora32-third-step-v4-profiler to be ready
Inference service rica40325-lora32-third-step-v4-profiler ready after 190.4128134250641s
Pipeline stage MKMLProfilerDeployer completed in 190.79s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
script pods %s
Pipeline stage MKMLProfilerRunner completed in 0.43s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-lora32-third-step-v4-profiler is running
Tearing down inference service rica40325-lora32-third-step-v4-profiler
Service rica40325-lora32-third-step-v4-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.94s
rica40325-lora32-third-step_v4 status is now inactive due to auto deactivation removed underperforming models
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-lora32-third-step-v4-profiler is running
Skipping teardown as no inference service was found
Pipeline stage MKMLProfilerDeleter completed in 1.77s
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 rica40325-lora32-third-step-v4-profiler
Waiting for inference service rica40325-lora32-third-step-v4-profiler to be ready
Inference service rica40325-lora32-third-step-v4-profiler ready after 150.39482164382935s
Pipeline stage MKMLProfilerDeployer completed in 150.88s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/rica40325-lora32-thi01ab30d442342f0578e1ac19b5157733-deplowm7mt:/code/chaiverse_profiler_1725502254 --namespace tenant-chaiml-guanaco
kubectl exec -it rica40325-lora32-thi01ab30d442342f0578e1ac19b5157733-deplowm7mt --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1725502254 && 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_1725502254/summary.json'
kubectl exec -it rica40325-lora32-thi01ab30d442342f0578e1ac19b5157733-deplowm7mt --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1725502254/summary.json'
Pipeline stage MKMLProfilerRunner completed in 849.47s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-lora32-third-step-v4-profiler is running
Tearing down inference service rica40325-lora32-third-step-v4-profiler
Service rica40325-lora32-third-step-v4-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 1.68s
Shutdown handler de-registered
rica40325-lora32-third-step_v4 status is now torndown due to DeploymentManager action