mistralai-ministral-8b-_47735

developer_uid: huohuo12

submission_id: mistralai-ministral-8b-_47735_v6

model_name: mistralai-ministral-8b-_47735_v6

model_group: mistralai/Ministral-8B-I

status: torndown

timestamp: 2025-02-20T07:28:16+00:00

num_battles: 6521

num_wins: 3022

celo_rating: 1216.54

family_friendly_score: 0.6242

family_friendly_standard_error: 0.006849443189048289

submission_type: basic

model_repo: mistralai/Ministral-8B-Instruct-2410

model_architecture: MistralForCausalLM

model_num_parameters: 8019808256.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

reward_model: default

latencies: [{'batch_size': 1, 'throughput': 0.8422857274937068, 'latency_mean': 1.1871831047534942, 'latency_p50': 1.196691632270813, 'latency_p90': 1.3068939924240113}, {'batch_size': 4, 'throughput': 1.7810274279809477, 'latency_mean': 2.241879587173462, 'latency_p50': 2.2184618711471558, 'latency_p90': 2.529356503486633}, {'batch_size': 5, 'throughput': 1.8666569397479513, 'latency_mean': 2.659563421010971, 'latency_p50': 2.668423056602478, 'latency_p90': 3.0088701248168945}, {'batch_size': 8, 'throughput': 2.091040713954194, 'latency_mean': 3.7987653756141664, 'latency_p50': 3.797562599182129, 'latency_p90': 4.270253586769104}, {'batch_size': 10, 'throughput': 2.099424632413149, 'latency_mean': 4.722628809213639, 'latency_p50': 4.7194297313690186, 'latency_p90': 5.295495724678039}, {'batch_size': 12, 'throughput': 2.1866542113916108, 'latency_mean': 5.434813058376312, 'latency_p50': 5.4143335819244385, 'latency_p90': 6.1070517778396605}, {'batch_size': 15, 'throughput': 2.1681072110754234, 'latency_mean': 6.824535669088363, 'latency_p50': 6.87706446647644, 'latency_p90': 7.67938551902771}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: mistralai-ministral-8b-_47735_v6

is_internal_developer: False

language_model: mistralai/Ministral-8B-Instruct-2410

model_size: 8B

ranking_group: single

throughput_3p7s: 2.09

us_pacific_date: 2025-02-19

win_ratio: 0.4634258549302254

generation_params: {'temperature': 1.1, 'top_p': 0.82, 'min_p': 0.035, 'top_k': 35, 'presence_penalty': 0.55, 'frequency_penalty': 0.25, 'stopping_words': ['\n'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': '<|im_start|>system\n{memory}<|im_end|>\n', 'prompt_template': '<|im_start|>user\n{prompt}<|im_end|>\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n{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 mistralai-ministral-8b-47735-v6-mkmlizer
Waiting for job on mistralai-ministral-8b-47735-v6-mkmlizer to finish
mistralai-ministral-8b-47735-v6-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
mistralai-ministral-8b-47735-v6-mkmlizer: ║     _____            __           __                                ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║       /___/                                                         ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║                                                                     ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  Version: 0.12.8                                                    ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  https://mk1.ai                                                     ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║                                                                     ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  The license key for the current software has been verified as      ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  belonging to:                                                      ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║                                                                     ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  Chai Research Corp.                                                ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║  Expiration: 2025-04-15 23:59:59                                    ║
mistralai-ministral-8b-47735-v6-mkmlizer: ║                                                                     ║
mistralai-ministral-8b-47735-v6-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
mistralai-ministral-8b-47735-v6-mkmlizer: Downloaded to shared memory in 40.112s
mistralai-ministral-8b-47735-v6-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpu0ufdlp0, device:0
mistralai-ministral-8b-47735-v6-mkmlizer: Saving flywheel model at /dev/shm/model_cache
mistralai-ministral-8b-47735-v6-mkmlizer: 
Loading 0:   0%|          | 0/327 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 5/327 [00:00<00:10, 30.16it/s]
Loading 0:   3%|▎         | 10/327 [00:00<00:08, 39.25it/s]
Loading 0:   5%|▍         | 15/327 [00:00<00:08, 35.89it/s]
Loading 0:   6%|▌         | 19/327 [00:00<00:08, 37.02it/s]
Loading 0:   7%|▋         | 24/327 [00:00<00:07, 39.93it/s]
Loading 0:   9%|▉         | 29/327 [00:00<00:07, 39.96it/s]
Loading 0:  10%|█         | 34/327 [00:00<00:07, 39.52it/s]
Loading 0:  12%|█▏        | 39/327 [00:01<00:07, 40.56it/s]
Loading 0:  13%|█▎        | 44/327 [00:01<00:08, 33.17it/s]
Loading 0:  15%|█▍        | 49/327 [00:01<00:07, 36.97it/s]
Loading 0:  16%|█▌        | 53/327 [00:01<00:07, 34.69it/s]
Loading 0:  18%|█▊        | 59/327 [00:01<00:06, 40.87it/s]
Loading 0:  20%|█▉        | 64/327 [00:01<00:06, 39.89it/s]
Loading 0:  21%|██        | 69/327 [00:01<00:06, 40.72it/s]
Loading 0:  23%|██▎       | 74/327 [00:01<00:06, 40.09it/s]
Loading 0:  24%|██▍       | 79/327 [00:02<00:06, 39.77it/s]
Loading 0:  26%|██▌       | 84/327 [00:02<00:05, 40.93it/s]
Loading 0:  27%|██▋       | 89/327 [00:02<00:07, 32.23it/s]
Loading 0:  29%|██▉       | 95/327 [00:02<00:08, 26.71it/s]
Loading 0:  30%|███       | 99/327 [00:02<00:08, 27.08it/s]
Loading 0:  32%|███▏      | 104/327 [00:02<00:07, 29.03it/s]
Loading 0:  33%|███▎      | 109/327 [00:03<00:06, 32.96it/s]
Loading 0:  35%|███▍      | 113/327 [00:03<00:06, 31.44it/s]
Loading 0:  36%|███▌      | 118/327 [00:03<00:05, 35.31it/s]
Loading 0:  37%|███▋      | 122/327 [00:03<00:06, 33.72it/s]
Loading 0:  39%|███▉      | 127/327 [00:03<00:05, 36.79it/s]
Loading 0:  40%|████      | 131/327 [00:03<00:05, 34.65it/s]
Loading 0:  42%|████▏     | 136/327 [00:03<00:04, 38.41it/s]
Loading 0:  43%|████▎     | 141/327 [00:03<00:05, 37.06it/s]
Loading 0:  44%|████▍     | 145/327 [00:04<00:04, 37.49it/s]
Loading 0:  46%|████▌     | 149/327 [00:04<00:05, 34.61it/s]
Loading 0:  48%|████▊     | 156/327 [00:04<00:04, 41.84it/s]
Loading 0:  49%|████▉     | 161/327 [00:04<00:03, 41.54it/s]
Loading 0:  51%|█████     | 166/327 [00:04<00:04, 40.19it/s]
Loading 0:  52%|█████▏    | 171/327 [00:04<00:03, 41.14it/s]
Loading 0:  54%|█████▍    | 176/327 [00:04<00:04, 33.19it/s]
Loading 0:  55%|█████▌    | 181/327 [00:05<00:04, 36.06it/s]
Loading 0:  57%|█████▋    | 185/327 [00:05<00:04, 34.42it/s]
Loading 0:  59%|█████▊    | 192/327 [00:05<00:03, 41.06it/s]
Loading 0:  60%|██████    | 197/327 [00:05<00:03, 39.66it/s]
Loading 0:  62%|██████▏   | 202/327 [00:05<00:03, 40.11it/s]
Loading 0:  63%|██████▎   | 207/327 [00:05<00:02, 41.15it/s]
Loading 0:  65%|██████▍   | 212/327 [00:05<00:04, 27.99it/s]
Loading 0:  66%|██████▌   | 216/327 [00:06<00:03, 29.73it/s]
Loading 0:  67%|██████▋   | 220/327 [00:06<00:03, 30.49it/s]
Loading 0:  69%|██████▊   | 224/327 [00:06<00:03, 31.87it/s]
Loading 0:  70%|██████▉   | 228/327 [00:06<00:03, 32.58it/s]
Loading 0:  71%|███████   | 232/327 [00:06<00:02, 32.80it/s]
Loading 0:  72%|███████▏  | 237/327 [00:06<00:02, 35.45it/s]
Loading 0:  74%|███████▎  | 241/327 [00:06<00:02, 34.71it/s]
Loading 0:  75%|███████▌  | 246/327 [00:06<00:02, 36.95it/s]
Loading 0:  76%|███████▋  | 250/327 [00:07<00:02, 36.42it/s]
Loading 0:  78%|███████▊  | 255/327 [00:07<00:01, 38.87it/s]
Loading 0:  79%|███████▉  | 259/327 [00:07<00:01, 36.60it/s]
Loading 0:  81%|████████  | 264/327 [00:07<00:01, 38.64it/s]
Loading 0:  82%|████████▏ | 268/327 [00:07<00:01, 37.23it/s]
Loading 0:  83%|████████▎ | 273/327 [00:07<00:01, 38.78it/s]
Loading 0:  85%|████████▍ | 277/327 [00:07<00:01, 36.89it/s]
Loading 0:  86%|████████▌ | 282/327 [00:07<00:01, 38.72it/s]
Loading 0:  87%|████████▋ | 286/327 [00:07<00:01, 37.21it/s]
Loading 0:  89%|████████▉ | 291/327 [00:08<00:00, 38.81it/s]
Loading 0:  90%|█████████ | 295/327 [00:08<00:00, 36.96it/s]
Loading 0:  92%|█████████▏| 300/327 [00:08<00:00, 38.66it/s]
Loading 0:  93%|█████████▎| 304/327 [00:08<00:00, 37.01it/s]
Loading 0:  94%|█████████▍| 308/327 [00:08<00:00, 37.54it/s]
Loading 0:  95%|█████████▌| 312/327 [00:08<00:00, 34.22it/s]
Loading 0:  97%|█████████▋| 316/327 [00:08<00:00, 35.25it/s]
Loading 0:  98%|█████████▊| 320/327 [00:08<00:00, 32.93it/s]
Loading 0:  99%|█████████▉| 325/327 [00:09<00:00, 35.96it/s]
                                                            
You are using the default legacy behaviour of the <class 'transformers.models.llama.tokenization_llama_fast.LlamaTokenizerFast'>. This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thoroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565 - if you loaded a llama tokenizer from a GGUF file you can ignore this message.
mistralai-ministral-8b-47735-v6-mkmlizer: quantized model in 30.141s
mistralai-ministral-8b-47735-v6-mkmlizer: Processed model mistralai/Ministral-8B-Instruct-2410 in 70.253s
mistralai-ministral-8b-47735-v6-mkmlizer: creating bucket guanaco-mkml-models
mistralai-ministral-8b-47735-v6-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
mistralai-ministral-8b-47735-v6-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6
mistralai-ministral-8b-47735-v6-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6/config.json
mistralai-ministral-8b-47735-v6-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6/special_tokens_map.json
mistralai-ministral-8b-47735-v6-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6/tokenizer_config.json
mistralai-ministral-8b-47735-v6-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6/tokenizer.json
mistralai-ministral-8b-47735-v6-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/mistralai-ministral-8b-47735-v6/flywheel_model.0.safetensors
Job mistralai-ministral-8b-47735-v6-mkmlizer completed after 93.97s with status: succeeded
Stopping job with name mistralai-ministral-8b-47735-v6-mkmlizer
Pipeline stage MKMLizer completed in 94.41s
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 mistralai-ministral-8b-47735-v6
Waiting for inference service mistralai-ministral-8b-47735-v6 to be ready
Inference service mistralai-ministral-8b-47735-v6 ready after 200.74961614608765s
Pipeline stage MKMLDeployer completed in 201.25s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.90643310546875s
Received healthy response to inference request in 1.3080148696899414s
Received healthy response to inference request in 1.190028429031372s
Received healthy response to inference request in 1.152184009552002s
Received healthy response to inference request in 1.0954155921936035s
5 requests
0 failed requests
5th percentile: 1.1067692756652832
10th percentile: 1.1181229591369628
20th percentile: 1.1408303260803223
30th percentile: 1.159752893447876
40th percentile: 1.174890661239624
50th percentile: 1.190028429031372
60th percentile: 1.2372230052948
70th percentile: 1.2844175815582275
80th percentile: 1.4276985168457033
90th percentile: 1.6670658111572267
95th percentile: 1.7867494583129881
99th percentile: 1.8824963760375977
mean time: 1.3304152011871337
Pipeline stage StressChecker completed in 8.04s
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 0.68s
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 0.63s
Shutdown handler de-registered
mistralai-ministral-8b-_47735_v6 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 0.10s
run pipeline stage %s
Running pipeline stage MKMLProfilerTemplater
Pipeline stage MKMLProfilerTemplater completed in 0.09s
run pipeline stage %s
Running pipeline stage MKMLProfilerDeployer
Creating inference service mistralai-ministral-8b-47735-v6-profiler
Waiting for inference service mistralai-ministral-8b-47735-v6-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 2251.64s
Shutdown handler de-registered
mistralai-ministral-8b-_47735_v6 status is now inactive due to auto deactivation removed underperforming models
mistralai-ministral-8b-_47735_v6 status is now torndown due to DeploymentManager action