stark2000s-dpo-fd-mistral

developer_uid: stark2000s

submission_id: stark2000s-dpo-fd-mistral_v3

model_name: stark2000s-dpo-fd-mistral_v3

model_group: stark2000s/dpo_fd_mistra

status: inactive

timestamp: 2024-12-13T00:55:52+00:00

num_battles: 7208

num_wins: 2849

celo_rating: 1182.8

family_friendly_score: 0.5731999999999999

family_friendly_standard_error: 0.00699488041355962

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': 0.8647712000888873, 'latency_mean': 1.15628910779953, 'latency_p50': 1.163068175315857, 'latency_p90': 1.2820693016052247}, {'batch_size': 4, 'throughput': 1.8597203746390107, 'latency_mean': 2.1394544720649717, 'latency_p50': 2.13922917842865, 'latency_p90': 2.373146080970764}, {'batch_size': 5, 'throughput': 2.0000973844260805, 'latency_mean': 2.4849655199050904, 'latency_p50': 2.4714101552963257, 'latency_p90': 2.784777021408081}, {'batch_size': 8, 'throughput': 2.236917001295915, 'latency_mean': 3.554565209150314, 'latency_p50': 3.5487228631973267, 'latency_p90': 4.056117677688599}, {'batch_size': 10, 'throughput': 2.2867110693161448, 'latency_mean': 4.33281857252121, 'latency_p50': 4.337372899055481, 'latency_p90': 4.819346785545349}, {'batch_size': 12, 'throughput': 2.3107790456345096, 'latency_mean': 5.145172080993652, 'latency_p50': 5.192892670631409, 'latency_p90': 5.84217894077301}, {'batch_size': 15, 'throughput': 2.3485938753195175, 'latency_mean': 6.30986151099205, 'latency_p50': 6.36613667011261, 'latency_p90': 7.154149413108826}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: stark2000s-dpo-fd-mistral_v3

is_internal_developer: False

language_model: stark2000s/dpo_fd_mistral

model_size: 8B

ranking_group: single

throughput_3p7s: 2.26

us_pacific_date: 2024-12-12

win_ratio: 0.3952552719200888

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-v3-mkmlizer
Waiting for job on stark2000s-dpo-fd-mistral-v3-mkmlizer to finish
stark2000s-dpo-fd-mistral-v3-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║     _____            __           __                                ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║       /___/                                                         ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  Version: 0.11.12                                                   ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  https://mk1.ai                                                     ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  The license key for the current software has been verified as      ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  belonging to:                                                      ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  Chai Research Corp.                                                ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ║                                                                     ║
stark2000s-dpo-fd-mistral-v3-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
stark2000s-dpo-fd-mistral-v3-mkmlizer: Downloaded to shared memory in 30.267s
stark2000s-dpo-fd-mistral-v3-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp6jhjgi57, device:0
stark2000s-dpo-fd-mistral-v3-mkmlizer: Saving flywheel model at /dev/shm/model_cache
stark2000s-dpo-fd-mistral-v3-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-v3-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
stark2000s-dpo-fd-mistral-v3-mkmlizer: quantized model in 26.970s
stark2000s-dpo-fd-mistral-v3-mkmlizer: Processed model stark2000s/dpo_fd_mistral in 57.238s
stark2000s-dpo-fd-mistral-v3-mkmlizer: creating bucket guanaco-mkml-models
stark2000s-dpo-fd-mistral-v3-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
stark2000s-dpo-fd-mistral-v3-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3
stark2000s-dpo-fd-mistral-v3-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3/config.json
stark2000s-dpo-fd-mistral-v3-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3/special_tokens_map.json
stark2000s-dpo-fd-mistral-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3/tokenizer_config.json
stark2000s-dpo-fd-mistral-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3/tokenizer.json
stark2000s-dpo-fd-mistral-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/stark2000s-dpo-fd-mistral-v3/flywheel_model.0.safetensors
stark2000s-dpo-fd-mistral-v3-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:06, 46.57it/s]
Loading 0:   5%|▌         | 16/291 [00:00<00:04, 66.24it/s]
Loading 0:   9%|▊         | 25/291 [00:00<00:03, 72.69it/s]
Loading 0:  12%|█▏        | 34/291 [00:00<00:03, 73.01it/s]
Loading 0:  15%|█▍        | 43/291 [00:00<00:03, 73.91it/s]
Loading 0:  18%|█▊        | 52/291 [00:00<00:03, 77.46it/s]
Loading 0:  21%|██        | 61/291 [00:00<00:03, 75.70it/s]
Loading 0:  24%|██▍       | 70/291 [00:00<00:02, 75.56it/s]
Loading 0:  27%|██▋       | 79/291 [00:01<00:02, 77.81it/s]
Loading 0:  30%|██▉       | 87/291 [00:02<00:10, 19.79it/s]
Loading 0:  32%|███▏      | 94/291 [00:02<00:08, 24.09it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:06, 31.28it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 37.60it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:03, 45.94it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:02, 53.77it/s]
Loading 0:  48%|████▊     | 139/291 [00:02<00:02, 59.42it/s]
Loading 0:  51%|█████     | 149/291 [00:03<00:02, 68.45it/s]
Loading 0:  54%|█████▍    | 158/291 [00:03<00:01, 73.63it/s]
Loading 0:  57%|█████▋    | 167/291 [00:03<00:01, 77.13it/s]
Loading 0:  60%|██████    | 176/291 [00:03<00:01, 79.14it/s]
Loading 0:  64%|██████▎   | 185/291 [00:03<00:01, 81.02it/s]
Loading 0:  67%|██████▋   | 194/291 [00:04<00:04, 20.88it/s]
Loading 0:  69%|██████▉   | 202/291 [00:04<00:03, 26.05it/s]
Loading 0:  73%|███████▎  | 211/291 [00:04<00:02, 32.61it/s]
Loading 0:  76%|███████▌  | 220/291 [00:04<00:01, 39.93it/s]
Loading 0:  79%|███████▊  | 229/291 [00:05<00:01, 47.18it/s]
Loading 0:  82%|████████▏ | 238/291 [00:05<00:00, 53.83it/s]
Loading 0:  85%|████████▍ | 247/291 [00:05<00:00, 58.83it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 61.43it/s]
Loading 0:  92%|█████████▏| 267/291 [00:05<00:00, 72.21it/s]
Loading 0:  95%|█████████▍| 276/291 [00:05<00:00, 73.63it/s]
Loading 0:  98%|█████████▊| 285/291 [00:05<00:00, 74.31it/s]
                                                            
Job stark2000s-dpo-fd-mistral-v3-mkmlizer completed after 84.15s with status: succeeded
Stopping job with name stark2000s-dpo-fd-mistral-v3-mkmlizer
Pipeline stage MKMLizer completed in 84.67s
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 stark2000s-dpo-fd-mistral-v3
Waiting for inference service stark2000s-dpo-fd-mistral-v3 to be ready
Inference service stark2000s-dpo-fd-mistral-v3 ready after 200.6767122745514s
Pipeline stage MKMLDeployer completed in 201.20s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.6159107685089111s
Received healthy response to inference request in 1.5612859725952148s
Received healthy response to inference request in 1.3356757164001465s
Received healthy response to inference request in 1.2833662033081055s
Received healthy response to inference request in 1.2328977584838867s
5 requests
0 failed requests
5th percentile: 1.2429914474487305
10th percentile: 1.2530851364135742
20th percentile: 1.2732725143432617
30th percentile: 1.2938281059265138
40th percentile: 1.3147519111633301
50th percentile: 1.3356757164001465
60th percentile: 1.4259198188781739
70th percentile: 1.516163921356201
80th percentile: 1.572210931777954
90th percentile: 1.5940608501434326
95th percentile: 1.604985809326172
99th percentile: 1.6137257766723634
mean time: 1.405827283859253
Pipeline stage StressChecker completed in 8.88s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 2.29s
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 2.12s
Shutdown handler de-registered
stark2000s-dpo-fd-mistral_v3 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 2129.78s
Shutdown handler de-registered
stark2000s-dpo-fd-mistral_v3 status is now inactive due to auto deactivation removed underperforming models