rinen0721-mistral-dpo-0114

developer_uid: rinen0721

submission_id: rinen0721-mistral-dpo-0114_v2

model_name: rinen0721-mistral-dpo-0114_v2

model_group: rinen0721/Mistral-dpo-01

status: torndown

timestamp: 2025-01-14T05:41:26+00:00

num_battles: 9725

num_wins: 4610

celo_rating: 1244.07

family_friendly_score: 0.5932

family_friendly_standard_error: 0.006947139843129689

submission_type: basic

model_repo: rinen0721/Mistral-dpo-0114

model_architecture: MistralForCausalLM

model_num_parameters: 12772070400.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.6220279126510243, 'latency_mean': 1.6075871324539184, 'latency_p50': 1.6045966148376465, 'latency_p90': 1.7695327520370483}, {'batch_size': 3, 'throughput': 1.1561304150966072, 'latency_mean': 2.5856237971782683, 'latency_p50': 2.5983400344848633, 'latency_p90': 2.865018081665039}, {'batch_size': 5, 'throughput': 1.3988513391359056, 'latency_mean': 3.5519962990283966, 'latency_p50': 3.568021059036255, 'latency_p90': 4.014059901237488}, {'batch_size': 6, 'throughput': 1.4863313784501635, 'latency_mean': 4.015541628599167, 'latency_p50': 4.023710489273071, 'latency_p90': 4.506058835983276}, {'batch_size': 8, 'throughput': 1.5497324913686084, 'latency_mean': 5.129056549072265, 'latency_p50': 5.133368372917175, 'latency_p90': 5.849461221694947}, {'batch_size': 10, 'throughput': 1.5735085415522334, 'latency_mean': 6.294300680160522, 'latency_p50': 6.314275622367859, 'latency_p90': 7.027739953994751}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: rinen0721-mistral-dpo-0114_v2

is_internal_developer: False

language_model: rinen0721/Mistral-dpo-0114

model_size: 13B

ranking_group: single

throughput_3p7s: 1.44

us_pacific_date: 2025-01-13

win_ratio: 0.47403598971722366

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 rinen0721-mistral-dpo-0114-v2-mkmlizer
Waiting for job on rinen0721-mistral-dpo-0114-v2-mkmlizer to finish
rinen0721-mistral-dpo-0114-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║     _____            __           __                                ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║       /___/                                                         ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║                                                                     ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  Version: 0.11.12                                                   ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  https://mk1.ai                                                     ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║                                                                     ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  belonging to:                                                      ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║                                                                     ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  Chai Research Corp.                                                ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║  Expiration: 2025-04-15 23:59:59                                    ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ║                                                                     ║
rinen0721-mistral-dpo-0114-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
rinen0721-mistral-dpo-0114-v2-mkmlizer: Downloaded to shared memory in 30.259s
rinen0721-mistral-dpo-0114-v2-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpl_84k6lh, device:0
rinen0721-mistral-dpo-0114-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
rinen0721-mistral-dpo-0114-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.
rinen0721-mistral-dpo-0114-v2-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
rinen0721-mistral-dpo-0114-v2-mkmlizer: quantized model in 36.947s
rinen0721-mistral-dpo-0114-v2-mkmlizer: Processed model rinen0721/Mistral-dpo-0114 in 67.207s
rinen0721-mistral-dpo-0114-v2-mkmlizer: creating bucket guanaco-mkml-models
rinen0721-mistral-dpo-0114-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
rinen0721-mistral-dpo-0114-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2
rinen0721-mistral-dpo-0114-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2/config.json
rinen0721-mistral-dpo-0114-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2/special_tokens_map.json
rinen0721-mistral-dpo-0114-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2/tokenizer_config.json
rinen0721-mistral-dpo-0114-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2/tokenizer.json
rinen0721-mistral-dpo-0114-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/rinen0721-mistral-dpo-0114-v2/flywheel_model.0.safetensors
rinen0721-mistral-dpo-0114-v2-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 6/363 [00:00<00:05, 59.84it/s]
Loading 0:   4%|▎         | 13/363 [00:00<00:05, 65.10it/s]
Loading 0:   6%|▋         | 23/363 [00:00<00:04, 80.68it/s]
Loading 0:   9%|▉         | 34/363 [00:00<00:04, 78.11it/s]
Loading 0:  12%|█▏        | 44/363 [00:00<00:03, 84.69it/s]
Loading 0:  15%|█▍        | 53/363 [00:00<00:03, 84.68it/s]
Loading 0:  17%|█▋        | 62/363 [00:01<00:15, 19.80it/s]
Loading 0:  21%|██        | 76/363 [00:02<00:09, 29.48it/s]
Loading 0:  23%|██▎       | 85/363 [00:02<00:07, 35.73it/s]
Loading 0:  26%|██▌       | 94/363 [00:02<00:06, 41.35it/s]
Loading 0:  29%|██▉       | 105/363 [00:02<00:04, 52.02it/s]
Loading 0:  32%|███▏      | 115/363 [00:02<00:04, 56.60it/s]
Loading 0:  34%|███▍      | 124/363 [00:02<00:03, 62.78it/s]
Loading 0:  37%|███▋      | 133/363 [00:02<00:03, 66.93it/s]
Loading 0:  39%|███▉      | 142/363 [00:03<00:10, 20.46it/s]
Loading 0:  42%|████▏     | 151/363 [00:03<00:08, 26.41it/s]
Loading 0:  44%|████▍     | 160/363 [00:04<00:06, 32.80it/s]
Loading 0:  47%|████▋     | 169/363 [00:04<00:04, 39.02it/s]
Loading 0:  49%|████▉     | 179/363 [00:04<00:03, 48.25it/s]
Loading 0:  52%|█████▏    | 188/363 [00:04<00:03, 55.64it/s]
Loading 0:  54%|█████▍    | 197/363 [00:04<00:02, 61.75it/s]
Loading 0:  57%|█████▋    | 206/363 [00:04<00:02, 64.97it/s]
Loading 0:  59%|█████▉    | 214/363 [00:04<00:02, 68.26it/s]
Loading 0:  61%|██████▏   | 223/363 [00:05<00:06, 20.46it/s]
Loading 0:  64%|██████▍   | 232/363 [00:06<00:04, 26.72it/s]
Loading 0:  66%|██████▋   | 241/363 [00:06<00:03, 33.40it/s]
Loading 0:  69%|██████▉   | 251/363 [00:06<00:02, 42.43it/s]
Loading 0:  73%|███████▎  | 265/363 [00:06<00:01, 54.73it/s]
Loading 0:  75%|███████▌  | 274/363 [00:06<00:01, 60.78it/s]
Loading 0:  78%|███████▊  | 283/363 [00:06<00:01, 63.22it/s]
Loading 0:  80%|████████  | 292/363 [00:06<00:01, 67.80it/s]
Loading 0:  83%|████████▎ | 302/363 [00:06<00:00, 75.08it/s]
Loading 0:  86%|████████▌ | 311/363 [00:07<00:02, 21.57it/s]
Loading 0:  88%|████████▊ | 319/363 [00:08<00:01, 26.17it/s]
Loading 0:  90%|█████████ | 328/363 [00:08<00:01, 32.31it/s]
Loading 0:  93%|█████████▎| 337/363 [00:08<00:00, 39.55it/s]
Loading 0:  95%|█████████▌| 346/363 [00:08<00:00, 46.50it/s]
Loading 0:  98%|█████████▊| 356/363 [00:08<00:00, 56.18it/s]
                                                            
Job rinen0721-mistral-dpo-0114-v2-mkmlizer completed after 94.42s with status: succeeded
Stopping job with name rinen0721-mistral-dpo-0114-v2-mkmlizer
Pipeline stage MKMLizer completed in 94.87s
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 rinen0721-mistral-dpo-0114-v2
Waiting for inference service rinen0721-mistral-dpo-0114-v2 to be ready
Inference service rinen0721-mistral-dpo-0114-v2 ready after 372.1373176574707s
Pipeline stage MKMLDeployer completed in 372.70s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.8080346584320068s
Received healthy response to inference request in 1.8732950687408447s
Received healthy response to inference request in 1.8551454544067383s
Received healthy response to inference request in 1.5144402980804443s
Received healthy response to inference request in 2.028557777404785s
5 requests
0 failed requests
5th percentile: 1.5731591701507568
10th percentile: 1.6318780422210692
20th percentile: 1.7493157863616944
30th percentile: 1.8174568176269532
40th percentile: 1.8363011360168457
50th percentile: 1.8551454544067383
60th percentile: 1.8624053001403809
70th percentile: 1.8696651458740234
80th percentile: 1.9043476104736328
90th percentile: 1.966452693939209
95th percentile: 1.9975052356719971
99th percentile: 2.0223472690582276
mean time: 1.815894651412964
Pipeline stage StressChecker completed in 10.27s
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 1.06s
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 0.70s
Shutdown handler de-registered
rinen0721-mistral-dpo-0114_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 0.11s
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 rinen0721-mistral-dpo-0114-v2-profiler
Waiting for inference service rinen0721-mistral-dpo-0114-v2-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 2686.28s
Shutdown handler de-registered
rinen0721-mistral-dpo-0114_v2 status is now inactive due to auto deactivation removed underperforming models
rinen0721-mistral-dpo-0114_v2 status is now torndown due to DeploymentManager action