dans-discountmodels-mist_8327

developer_uid: PocketDoc

submission_id: dans-discountmodels-mist_8327_v1

model_name: Dans-Discount-Mistral-Instruct

model_group: Dans-DiscountModels/Mist

status: inactive

timestamp: 2024-11-20T04:35:29+00:00

num_battles: 6873

num_wins: 3238

celo_rating: 1228.72

family_friendly_score: 0.5642

family_friendly_standard_error: 0.0070125367735221185

submission_type: basic

model_repo: Dans-DiscountModels/Mistral-7b-v0.3-Test-E0.7

model_architecture: MistralForCausalLM

model_num_parameters: 7248023552.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

latencies: [{'batch_size': 1, 'throughput': 0.9040542288640595, 'latency_mean': 1.1060654497146607, 'latency_p50': 1.1088320016860962, 'latency_p90': 1.2205981016159058}, {'batch_size': 4, 'throughput': 1.9650816341348976, 'latency_mean': 2.029604251384735, 'latency_p50': 2.0222259759902954, 'latency_p90': 2.2351365804672243}, {'batch_size': 5, 'throughput': 2.1314634351361303, 'latency_mean': 2.3427968716621397, 'latency_p50': 2.330844759941101, 'latency_p90': 2.6293666839599608}, {'batch_size': 8, 'throughput': 2.3355138940865867, 'latency_mean': 3.4025909399986265, 'latency_p50': 3.3925697803497314, 'latency_p90': 3.858400654792786}, {'batch_size': 10, 'throughput': 2.388132173592213, 'latency_mean': 4.160032922029496, 'latency_p50': 4.170439720153809, 'latency_p90': 4.695250606536865}, {'batch_size': 12, 'throughput': 2.432900420395488, 'latency_mean': 4.888100597858429, 'latency_p50': 4.8737486600875854, 'latency_p90': 5.619600439071655}, {'batch_size': 15, 'throughput': 2.4396534399483496, 'latency_mean': 6.060987627506256, 'latency_p50': 6.106120347976685, 'latency_p90': 6.802632641792297}]

gpu_counts: {'NVIDIA RTX A5000': 1}

display_name: Dans-Discount-Mistral-Instruct

is_internal_developer: False

language_model: Dans-DiscountModels/Mistral-7b-v0.3-Test-E0.7

model_size: 7B

ranking_group: single

throughput_3p7s: 2.37

us_pacific_date: 2024-11-19

win_ratio: 0.4711188709442747

generation_params: {'temperature': 1.0, 'top_p': 0.9, 'min_p': 0.02, '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': '<|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': True}

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 dans-discountmodels-mist-8327-v1-mkmlizer
Waiting for job on dans-discountmodels-mist-8327-v1-mkmlizer to finish
dans-discountmodels-mist-8327-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
dans-discountmodels-mist-8327-v1-mkmlizer: ║     _____            __           __                                ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║       /___/                                                         ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║                                                                     ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  Version: 0.11.12                                                   ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  https://mk1.ai                                                     ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║                                                                     ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  belonging to:                                                      ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║                                                                     ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  Chai Research Corp.                                                ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║  Expiration: 2025-01-15 23:59:59                                    ║
dans-discountmodels-mist-8327-v1-mkmlizer: ║                                                                     ║
dans-discountmodels-mist-8327-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
dans-discountmodels-mist-8327-v1-mkmlizer: Downloaded to shared memory in 52.663s
dans-discountmodels-mist-8327-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpohiljz0y, device:0
dans-discountmodels-mist-8327-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
dans-discountmodels-mist-8327-v1-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.
dans-discountmodels-mist-8327-v1-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
dans-discountmodels-mist-8327-v1-mkmlizer: quantized model in 17.414s
dans-discountmodels-mist-8327-v1-mkmlizer: Processed model Dans-DiscountModels/Mistral-7b-v0.3-Test-E0.7 in 70.077s
dans-discountmodels-mist-8327-v1-mkmlizer: creating bucket guanaco-mkml-models
dans-discountmodels-mist-8327-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
dans-discountmodels-mist-8327-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/dans-discountmodels-mist-8327-v1
dans-discountmodels-mist-8327-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/dans-discountmodels-mist-8327-v1/config.json
dans-discountmodels-mist-8327-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/dans-discountmodels-mist-8327-v1/special_tokens_map.json
dans-discountmodels-mist-8327-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/dans-discountmodels-mist-8327-v1/tokenizer_config.json
dans-discountmodels-mist-8327-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/dans-discountmodels-mist-8327-v1/tokenizer.json
dans-discountmodels-mist-8327-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:05, 49.36it/s]
Loading 0:   8%|▊         | 22/291 [00:00<00:03, 79.96it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:03, 81.24it/s]
Loading 0:  14%|█▎        | 40/291 [00:00<00:02, 84.08it/s]
Loading 0:  17%|█▋        | 49/291 [00:00<00:02, 84.65it/s]
Loading 0:  20%|█▉        | 58/291 [00:00<00:02, 85.61it/s]
Loading 0:  23%|██▎       | 67/291 [00:00<00:02, 84.82it/s]
Loading 0:  26%|██▌       | 76/291 [00:00<00:02, 86.09it/s]
Loading 0:  29%|██▉       | 85/291 [00:01<00:02, 87.11it/s]
Loading 0:  32%|███▏      | 94/291 [00:01<00:02, 87.08it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:09, 20.53it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:06, 26.00it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:05, 31.94it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:04, 38.63it/s]
Loading 0:  48%|████▊     | 139/291 [00:02<00:03, 44.14it/s]
Loading 0:  51%|█████     | 148/291 [00:02<00:02, 48.75it/s]
Loading 0:  54%|█████▍    | 157/291 [00:03<00:02, 53.83it/s]
Loading 0:  57%|█████▋    | 166/291 [00:03<00:02, 58.60it/s]
Loading 0:  60%|██████    | 175/291 [00:03<00:01, 61.83it/s]
Loading 0:  63%|██████▎   | 184/291 [00:03<00:01, 65.45it/s]
Loading 0:  66%|██████▋   | 193/291 [00:03<00:01, 69.31it/s]
Loading 0:  69%|██████▉   | 202/291 [00:03<00:01, 70.00it/s]
Loading 0:  72%|███████▏  | 210/291 [00:04<00:03, 22.51it/s]
Loading 0:  74%|███████▍  | 216/291 [00:04<00:02, 26.11it/s]
Loading 0:  77%|███████▋  | 223/291 [00:04<00:02, 31.53it/s]
Loading 0:  80%|███████▉  | 232/291 [00:05<00:01, 39.19it/s]
Loading 0:  83%|████████▎ | 241/291 [00:05<00:01, 45.51it/s]
Loading 0:  86%|████████▌ | 250/291 [00:05<00:00, 52.24it/s]
Loading 0:  89%|████████▉ | 259/291 [00:05<00:00, 58.52it/s]
Loading 0:  92%|█████████▏| 268/291 [00:05<00:00, 63.60it/s]
Loading 0:  95%|█████████▌| 277/291 [00:05<00:00, 69.33it/s]
Loading 0:  98%|█████████▊| 286/291 [00:05<00:00, 73.65it/s]
                                                            
Job dans-discountmodels-mist-8327-v1-mkmlizer completed after 94.99s with status: succeeded
Stopping job with name dans-discountmodels-mist-8327-v1-mkmlizer
Pipeline stage MKMLizer completed in 95.54s
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 dans-discountmodels-mist-8327-v1
Waiting for inference service dans-discountmodels-mist-8327-v1 to be ready
Connection pool is full, discarding connection: %s. Connection pool size: %s
Inference service dans-discountmodels-mist-8327-v1 ready after 201.19192743301392s
Pipeline stage MKMLDeployer completed in 201.75s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.6748838424682617s
Received healthy response to inference request in 1.452359676361084s
Received healthy response to inference request in 1.3704960346221924s
Received healthy response to inference request in 1.2230479717254639s
Received healthy response to inference request in 1.5530812740325928s
5 requests
0 failed requests
5th percentile: 1.2525375843048097
10th percentile: 1.2820271968841552
20th percentile: 1.3410064220428466
30th percentile: 1.3868687629699707
40th percentile: 1.4196142196655273
50th percentile: 1.452359676361084
60th percentile: 1.4926483154296875
70th percentile: 1.532936954498291
80th percentile: 1.5774417877197267
90th percentile: 1.6261628150939942
95th percentile: 1.6505233287811278
99th percentile: 1.670011739730835
mean time: 1.454773759841919
Pipeline stage StressChecker completed in 8.68s
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 2.31s
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 2.68s
Shutdown handler de-registered
dans-discountmodels-mist_8327_v1 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 dans-discountmodels-mist-8327-v1-profiler
Waiting for inference service dans-discountmodels-mist-8327-v1-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 2084.24s
Shutdown handler de-registered
dans-discountmodels-mist_8327_v1 status is now inactive due to auto deactivation removed underperforming models