rica40325-mistral-3678

developer_uid: rica40325

submission_id: rica40325-mistral-3678_v2

model_name: rica40325-mistral-13b-2452_v1

model_group: rica40325/mistral-3678

status: torndown

timestamp: 2024-09-25T09:09:04+00:00

num_battles: 2920

num_wins: 1513

celo_rating: 1265.84

family_friendly_score: 0.5508241758241759

family_friendly_standard_error: 0.009199993341811503

submission_type: basic

model_repo: rica40325/mistral-3678

model_architecture: MistralForCausalLM

model_num_parameters: 12772070400.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.6148677795777034, 'latency_mean': 1.6262665557861329, 'latency_p50': 1.626531958580017, 'latency_p90': 1.782234263420105}, {'batch_size': 3, 'throughput': 1.0793698594691146, 'latency_mean': 2.77327831029892, 'latency_p50': 2.7774494886398315, 'latency_p90': 3.0075910568237303}, {'batch_size': 5, 'throughput': 1.2315309312654168, 'latency_mean': 4.046203141212463, 'latency_p50': 4.042070508003235, 'latency_p90': 4.506042385101318}, {'batch_size': 6, 'throughput': 1.265943134409583, 'latency_mean': 4.7144946241378785, 'latency_p50': 4.7250553369522095, 'latency_p90': 5.322303223609924}, {'batch_size': 8, 'throughput': 1.249400311210388, 'latency_mean': 6.362624412775039, 'latency_p50': 6.370821118354797, 'latency_p90': 7.226589965820312}, {'batch_size': 10, 'throughput': 1.212040868853282, 'latency_mean': 8.189200764894485, 'latency_p50': 8.366412281990051, 'latency_p90': 9.150413775444031}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: rica40325-mistral-13b-2452_v1

ineligible_reason: num_battles<5000

is_internal_developer: False

language_model: rica40325/mistral-3678

model_size: 13B

ranking_group: single

throughput_3p7s: 1.2

us_pacific_date: 2024-09-25

win_ratio: 0.5181506849315068

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': ['</s>', 'Bot:', 'User:', 'You:', 'Me:', '####'], 'max_input_tokens': 1024, 'best_of': 8, '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 rica40325-mistral-3678-v2-mkmlizer
Waiting for job on rica40325-mistral-3678-v2-mkmlizer to finish
rica40325-mistral-3678-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rica40325-mistral-3678-v2-mkmlizer: ║     _____            __           __                                ║
rica40325-mistral-3678-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rica40325-mistral-3678-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rica40325-mistral-3678-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rica40325-mistral-3678-v2-mkmlizer: ║       /___/                                                         ║
rica40325-mistral-3678-v2-mkmlizer: ║                                                                     ║
rica40325-mistral-3678-v2-mkmlizer: ║  Version: 0.11.12                                                   ║
rica40325-mistral-3678-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rica40325-mistral-3678-v2-mkmlizer: ║  https://mk1.ai                                                     ║
rica40325-mistral-3678-v2-mkmlizer: ║                                                                     ║
rica40325-mistral-3678-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
rica40325-mistral-3678-v2-mkmlizer: ║  belonging to:                                                      ║
rica40325-mistral-3678-v2-mkmlizer: ║                                                                     ║
rica40325-mistral-3678-v2-mkmlizer: ║  Chai Research Corp.                                                ║
rica40325-mistral-3678-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rica40325-mistral-3678-v2-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
rica40325-mistral-3678-v2-mkmlizer: ║                                                                     ║
rica40325-mistral-3678-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rica40325-mistral-3678-v2-mkmlizer: Downloaded to shared memory in 30.697s
rica40325-mistral-3678-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpfwfje4s5, device:0
rica40325-mistral-3678-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rica40325-mistral-3678-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/mk1/flywheel/functional/loader.py:55: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
rica40325-mistral-3678-v2-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
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
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
rica40325-mistral-3678-v2-mkmlizer: quantized model in 36.104s
rica40325-mistral-3678-v2-mkmlizer: Processed model rica40325/mistral-3678 in 66.801s
rica40325-mistral-3678-v2-mkmlizer: creating bucket guanaco-mkml-models
rica40325-mistral-3678-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rica40325-mistral-3678-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rica40325-mistral-3678-v2
rica40325-mistral-3678-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rica40325-mistral-3678-v2/config.json
rica40325-mistral-3678-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rica40325-mistral-3678-v2/special_tokens_map.json
rica40325-mistral-3678-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rica40325-mistral-3678-v2/tokenizer_config.json
rica40325-mistral-3678-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rica40325-mistral-3678-v2/tokenizer.json
rica40325-mistral-3678-v2-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/363 [00:00<00:07, 45.25it/s]
Loading 0:   6%|▌         | 22/363 [00:00<00:04, 80.22it/s]
Loading 0:   9%|▉         | 33/363 [00:00<00:03, 90.54it/s]
Loading 0:  12%|█▏        | 43/363 [00:00<00:04, 77.73it/s]
Loading 0:  14%|█▍        | 52/363 [00:00<00:03, 79.44it/s]
Loading 0:  17%|█▋        | 61/363 [00:01<00:14, 20.22it/s]
Loading 0:  19%|█▉        | 70/363 [00:01<00:11, 26.36it/s]
Loading 0:  22%|██▏       | 79/363 [00:02<00:08, 33.53it/s]
Loading 0:  25%|██▍       | 89/363 [00:02<00:06, 42.72it/s]
Loading 0:  27%|██▋       | 98/363 [00:02<00:05, 50.58it/s]
Loading 0:  29%|██▉       | 107/363 [00:02<00:04, 57.15it/s]
Loading 0:  32%|███▏      | 116/363 [00:02<00:03, 62.57it/s]
Loading 0:  34%|███▍      | 125/363 [00:02<00:03, 68.04it/s]
Loading 0:  37%|███▋      | 134/363 [00:02<00:03, 73.30it/s]
Loading 0:  39%|███▉      | 143/363 [00:03<00:10, 20.47it/s]
Loading 0:  43%|████▎     | 157/363 [00:04<00:06, 29.94it/s]
Loading 0:  46%|████▋     | 168/363 [00:04<00:05, 38.62it/s]
Loading 0:  49%|████▉     | 177/363 [00:04<00:04, 44.25it/s]
Loading 0:  51%|█████     | 186/363 [00:04<00:03, 50.88it/s]
Loading 0:  54%|█████▎    | 195/363 [00:04<00:02, 57.85it/s]
Loading 0:  56%|█████▋    | 205/363 [00:04<00:02, 61.23it/s]
Loading 0:  59%|█████▉    | 214/363 [00:04<00:02, 67.23it/s]
Loading 0:  61%|██████▏   | 223/363 [00:05<00:06, 20.75it/s]
Loading 0:  64%|██████▍   | 232/363 [00:05<00:04, 26.71it/s]
Loading 0:  66%|██████▋   | 241/363 [00:06<00:03, 33.32it/s]
Loading 0:  69%|██████▉   | 250/363 [00:06<00:02, 40.77it/s]
Loading 0:  71%|███████▏  | 259/363 [00:06<00:02, 48.65it/s]
Loading 0:  74%|███████▍  | 269/363 [00:06<00:01, 57.99it/s]
Loading 0:  77%|███████▋  | 278/363 [00:06<00:01, 62.93it/s]
Loading 0:  79%|███████▉  | 287/363 [00:06<00:01, 67.17it/s]
Loading 0:  83%|████████▎ | 301/363 [00:06<00:00, 76.56it/s]
Loading 0:  85%|████████▌ | 310/363 [00:07<00:02, 23.25it/s]
Loading 0:  88%|████████▊ | 319/363 [00:07<00:01, 28.93it/s]
Loading 0:  91%|█████████ | 330/363 [00:08<00:00, 37.96it/s]
Loading 0:  94%|█████████▎| 340/363 [00:08<00:00, 43.94it/s]
Loading 0:  96%|█████████▌| 349/363 [00:08<00:00, 50.78it/s]
Loading 0:  99%|█████████▊| 358/363 [00:08<00:00, 57.64it/s]
                                                            
Job rica40325-mistral-3678-v2-mkmlizer completed after 96.35s with status: succeeded
Stopping job with name rica40325-mistral-3678-v2-mkmlizer
Pipeline stage MKMLizer completed in 100.17s
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 rica40325-mistral-3678-v2
Waiting for inference service rica40325-mistral-3678-v2 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
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
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
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
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
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
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
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 rica40325-mistral-3678-v2 ready after 221.52286648750305s
Pipeline stage MKMLDeployer completed in 223.60s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.8695030212402344s
Received healthy response to inference request in 4.226944446563721s
Received healthy response to inference request in 1.986060619354248s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Received healthy response to inference request in 1.9825196266174316s
Received healthy response to inference request in 2.9913768768310547s
5 requests
0 failed requests
5th percentile: 1.983227825164795
10th percentile: 1.9839360237121582
20th percentile: 1.9853524208068847
30th percentile: 2.1871238708496095
40th percentile: 2.5892503738403323
50th percentile: 2.9913768768310547
60th percentile: 3.3426273345947264
70th percentile: 3.693877792358398
80th percentile: 3.9409913063049316
90th percentile: 4.083967876434326
95th percentile: 4.155456161499023
99th percentile: 4.2126467895507815
mean time: 3.011280918121338
Pipeline stage StressChecker completed in 16.68s
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 5.60s
Shutdown handler de-registered
rica40325-mistral-3678_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 0.13s
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 rica40325-mistral-3678-v2-profiler
Waiting for inference service rica40325-mistral-3678-v2-profiler to be ready
Inference service rica40325-mistral-3678-v2-profiler ready after 330.6946642398834s
Pipeline stage MKMLProfilerDeployer completed in 331.05s
run pipeline stage %s
Running pipeline stage MKMLProfilerRunner
kubectl cp /code/guanaco/guanaco_inference_services/src/inference_scripts tenant-chaiml-guanaco/rica40325-mistral-3678-v2-profiler-predictor-00001-deploympwrnm:/code/chaiverse_profiler_1727256067 --namespace tenant-chaiml-guanaco
kubectl exec -it rica40325-mistral-3678-v2-profiler-predictor-00001-deploympwrnm --namespace tenant-chaiml-guanaco -- sh -c 'cd /code/chaiverse_profiler_1727256067 && python profiles.py profile --best_of_n 8 --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 1024 --output_tokens 64 --summary /code/chaiverse_profiler_1727256067/summary.json'
kubectl exec -it rica40325-mistral-3678-v2-profiler-predictor-00001-deploympwrnm --namespace tenant-chaiml-guanaco -- bash -c 'cat /code/chaiverse_profiler_1727256067/summary.json'
Pipeline stage MKMLProfilerRunner completed in 1162.29s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeleter
Checking if service rica40325-mistral-3678-v2-profiler is running
Tearing down inference service rica40325-mistral-3678-v2-profiler
Service rica40325-mistral-3678-v2-profiler has been torndown
Pipeline stage MKMLProfilerDeleter completed in 2.32s
Shutdown handler de-registered
rica40325-mistral-3678_v2 status is now inactive due to auto deactivation removed underperforming models
rica40325-mistral-3678_v2 status is now torndown due to DeploymentManager action