developer_uid: stark2000s
submission_id: stark2000s-dpo-fd-mistral_v2
model_name: stark2000s-dpo-fd-mistral_v2
model_group: stark2000s/dpo_fd_mistra
status: inactive
timestamp: 2024-12-08T13:40:58+00:00
num_battles: 20761
num_wins: 7829
celo_rating: 1176.17
family_friendly_score: 0.5754
family_friendly_standard_error: 0.0069902051472041935
submission_type: basic
model_repo: stark2000s/dpo_fd_mistral
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 8
max_input_tokens: 1024
max_output_tokens: 64
latencies: [{'batch_size': 1, 'throughput': 0.8616476732885439, 'latency_mean': 1.1604672062397003, 'latency_p50': 1.1531461477279663, 'latency_p90': 1.2793917179107666}, {'batch_size': 4, 'throughput': 1.8405117986137018, 'latency_mean': 2.1688475000858305, 'latency_p50': 2.1568284034729004, 'latency_p90': 2.420129084587097}, {'batch_size': 5, 'throughput': 1.9746536267086527, 'latency_mean': 2.522358611822128, 'latency_p50': 2.512492299079895, 'latency_p90': 2.827845811843872}, {'batch_size': 8, 'throughput': 2.194510561302872, 'latency_mean': 3.6189764344692232, 'latency_p50': 3.5984413623809814, 'latency_p90': 4.110013961791992}, {'batch_size': 10, 'throughput': 2.247952601394079, 'latency_mean': 4.4135514569282535, 'latency_p50': 4.371548891067505, 'latency_p90': 4.984894967079162}, {'batch_size': 12, 'throughput': 2.2989582130307977, 'latency_mean': 5.170410799980163, 'latency_p50': 5.175815582275391, 'latency_p90': 5.735412168502807}, {'batch_size': 15, 'throughput': 2.2948654250216194, 'latency_mean': 6.444146897792816, 'latency_p50': 6.450862526893616, 'latency_p90': 7.164945602416992}]
gpu_counts: {'NVIDIA RTX A5000': 1}
display_name: stark2000s-dpo-fd-mistral_v2
is_internal_developer: False
language_model: stark2000s/dpo_fd_mistral
model_size: 8B
ranking_group: single
throughput_3p7s: 2.22
us_pacific_date: 2024-12-08
win_ratio: 0.3771012956986658
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], '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 stark2000s-dpo-fd-mistral-v2-mkmlizer
Waiting for job on stark2000s-dpo-fd-mistral-v2-mkmlizer to finish
stark2000s-dpo-fd-mistral-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║     _____            __           __                                ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║       /___/                                                         ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  Version: 0.11.12                                                   ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  https://mk1.ai                                                     ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  belonging to:                                                      ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  Chai Research Corp.                                                ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
stark2000s-dpo-fd-mistral-v2-mkmlizer: Downloaded to shared memory in 33.790s
stark2000s-dpo-fd-mistral-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmps182w4bj, device:0
stark2000s-dpo-fd-mistral-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
stark2000s-dpo-fd-mistral-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.
stark2000s-dpo-fd-mistral-v2-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
stark2000s-dpo-fd-mistral-v2-mkmlizer: quantized model in 26.201s
stark2000s-dpo-fd-mistral-v2-mkmlizer: Processed model stark2000s/dpo_fd_mistral in 59.991s
stark2000s-dpo-fd-mistral-v2-mkmlizer: creating bucket guanaco-mkml-models
stark2000s-dpo-fd-mistral-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
stark2000s-dpo-fd-mistral-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2
stark2000s-dpo-fd-mistral-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2/config.json
stark2000s-dpo-fd-mistral-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2/special_tokens_map.json
stark2000s-dpo-fd-mistral-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2/tokenizer_config.json
stark2000s-dpo-fd-mistral-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2/tokenizer.json
stark2000s-dpo-fd-mistral-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v2/flywheel_model.0.safetensors
Job stark2000s-dpo-fd-mistral-v2-mkmlizer completed after 84.93s with status: succeeded
Stopping job with name stark2000s-dpo-fd-mistral-v2-mkmlizer
Pipeline stage MKMLizer completed in 85.43s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.15s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service stark2000s-dpo-fd-mistral-v2
Waiting for inference service stark2000s-dpo-fd-mistral-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
Inference service stark2000s-dpo-fd-mistral-v2 ready after 170.94679236412048s
Pipeline stage MKMLDeployer completed in 171.46s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.8431506156921387s
Received healthy response to inference request in 1.4703266620635986s
Received healthy response to inference request in 1.3108067512512207s
Received healthy response to inference request in 1.3568236827850342s
Received healthy response to inference request in 1.2125205993652344s
5 requests
0 failed requests
5th percentile: 1.2321778297424317
10th percentile: 1.2518350601196289
20th percentile: 1.2911495208740233
30th percentile: 1.3200101375579834
40th percentile: 1.3384169101715089
50th percentile: 1.3568236827850342
60th percentile: 1.40222487449646
70th percentile: 1.4476260662078857
80th percentile: 1.5448914527893067
90th percentile: 1.6940210342407227
95th percentile: 1.7685858249664306
99th percentile: 1.828237657546997
mean time: 1.4387256622314453
Pipeline stage StressChecker completed in 8.60s
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 2.33s
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 2.29s
Shutdown handler de-registered
stark2000s-dpo-fd-mistral_v2 status is now deployed due to DeploymentManager action
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 2328.27s
Shutdown handler de-registered
stark2000s-dpo-fd-mistral_v2 status is now inactive due to auto deactivation removed underperforming models