meseca-15062024-c1

developer_uid: nguyenzzz

submission_id: meseca-15062024-c1_v3

model_name: meseca-15062024-c1_v3

model_group: meseca/15062024-c1

status: torndown

timestamp: 2024-07-02T03:10:11+00:00

num_battles: 308131

num_wins: 145906

celo_rating: 1213.8

family_friendly_score: 0.0

submission_type: basic

model_repo: meseca/15062024-c1

model_architecture: LlamaForCausalLM

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

model_num_parameters: 8030261248.0

best_of: 8

max_input_tokens: 1024

max_output_tokens: 64

display_name: meseca-15062024-c1_v3

is_internal_developer: False

language_model: meseca/15062024-c1

model_size: 8B

ranking_group: single

us_pacific_date: 2024-07-01

win_ratio: 0.4735193797443295

generation_params: {'temperature': 0.95, 'top_p': 0.95, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|end_header_id|>,', '<|eot_id|>,', '\n\n{user_name}'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}

formatter: {'memory_template': "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\n{bot_name}'s Persona: {memory}\n\n", 'prompt_template': '{prompt}<|eot_id|>', 'bot_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>', 'user_template': '<|start_header_id|>user<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>', 'response_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}:', 'truncate_by_message': False}

reward_formatter: {'bot_template': '{bot_name}: {message}\n', 'memory_template': "{bot_name}'s Persona: {memory}\n####\n", 'prompt_template': '{prompt}\n<START>\n', 'response_template': '{bot_name}:', 'truncate_by_message': False, 'user_template': '{user_name}: {message}\n'}

Resubmit model

Running pipeline stage MKMLizer
Starting job with name meseca-15062024-c1-v3-mkmlizer
Waiting for job on meseca-15062024-c1-v3-mkmlizer to finish
meseca-15062024-c1-v3-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
meseca-15062024-c1-v3-mkmlizer: ║     _____            __           __                                ║
meseca-15062024-c1-v3-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
meseca-15062024-c1-v3-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
meseca-15062024-c1-v3-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
meseca-15062024-c1-v3-mkmlizer: ║       /___/                                                         ║
meseca-15062024-c1-v3-mkmlizer: ║                                                                     ║
meseca-15062024-c1-v3-mkmlizer: ║  Version: 0.8.14                                                    ║
meseca-15062024-c1-v3-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
meseca-15062024-c1-v3-mkmlizer: ║  https://mk1.ai                                                     ║
meseca-15062024-c1-v3-mkmlizer: ║                                                                     ║
meseca-15062024-c1-v3-mkmlizer: ║  The license key for the current software has been verified as      ║
meseca-15062024-c1-v3-mkmlizer: ║  belonging to:                                                      ║
meseca-15062024-c1-v3-mkmlizer: ║                                                                     ║
meseca-15062024-c1-v3-mkmlizer: ║  Chai Research Corp.                                                ║
meseca-15062024-c1-v3-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
meseca-15062024-c1-v3-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
meseca-15062024-c1-v3-mkmlizer: ║                                                                     ║
meseca-15062024-c1-v3-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
meseca-15062024-c1-v3-mkmlizer: Downloaded to shared memory in 18.752s
meseca-15062024-c1-v3-mkmlizer: quantizing model to /dev/shm/model_cache
meseca-15062024-c1-v3-mkmlizer: Saving flywheel model at /dev/shm/model_cache
meseca-15062024-c1-v3-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/291 [00:04<09:57,  2.07s/it]
Loading 0:   8%|▊         | 22/291 [00:04<00:37,  7.15it/s]
Loading 0:  14%|█▍        | 42/291 [00:04<00:15, 15.87it/s]
Loading 0:  21%|██        | 60/291 [00:04<00:09, 23.59it/s]
Loading 0:  29%|██▉       | 85/291 [00:04<00:05, 40.04it/s]
Loading 0:  35%|███▌      | 103/291 [00:04<00:03, 53.20it/s]
Loading 0:  41%|████      | 120/291 [00:04<00:02, 63.82it/s]
Loading 0:  47%|████▋     | 136/291 [00:05<00:02, 74.52it/s]
Loading 0:  53%|█████▎    | 153/291 [00:05<00:01, 89.55it/s]
Loading 0:  58%|█████▊    | 169/291 [00:05<00:01, 80.94it/s]
Loading 0:  66%|██████▋   | 193/291 [00:05<00:00, 107.39it/s]
Loading 0:  73%|███████▎  | 213/291 [00:05<00:00, 123.66it/s]
Loading 0:  82%|████████▏ | 238/291 [00:05<00:00, 150.15it/s]
Loading 0:  89%|████████▊ | 258/291 [00:05<00:00, 159.97it/s]
Loading 0:  95%|█████████▌| 277/291 [00:06<00:00, 126.44it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
meseca-15062024-c1-v3-mkmlizer: quantized model in 17.087s
meseca-15062024-c1-v3-mkmlizer: Processed model meseca/15062024-c1 in 35.839s
meseca-15062024-c1-v3-mkmlizer: creating bucket guanaco-mkml-models
meseca-15062024-c1-v3-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
meseca-15062024-c1-v3-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/meseca-15062024-c1-v3
meseca-15062024-c1-v3-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/meseca-15062024-c1-v3/special_tokens_map.json
meseca-15062024-c1-v3-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/meseca-15062024-c1-v3/config.json
meseca-15062024-c1-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/meseca-15062024-c1-v3/tokenizer_config.json
meseca-15062024-c1-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/meseca-15062024-c1-v3/tokenizer.json
meseca-15062024-c1-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/meseca-15062024-c1-v3/flywheel_model.0.safetensors
meseca-15062024-c1-v3-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
meseca-15062024-c1-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:919: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
meseca-15062024-c1-v3-mkmlizer:   warnings.warn(
meseca-15062024-c1-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
meseca-15062024-c1-v3-mkmlizer:   warnings.warn(
meseca-15062024-c1-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:769: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
meseca-15062024-c1-v3-mkmlizer:   warnings.warn(
meseca-15062024-c1-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py:468: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
meseca-15062024-c1-v3-mkmlizer:   warnings.warn(
meseca-15062024-c1-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly.  To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
meseca-15062024-c1-v3-mkmlizer:   return self.fget.__get__(instance, owner)()
meseca-15062024-c1-v3-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
meseca-15062024-c1-v3-mkmlizer: Saving duration: 1.073s
meseca-15062024-c1-v3-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 8.363s
meseca-15062024-c1-v3-mkmlizer: creating bucket guanaco-reward-models
meseca-15062024-c1-v3-mkmlizer: Bucket 's3://guanaco-reward-models/' created
meseca-15062024-c1-v3-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/meseca-15062024-c1-v3_reward
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/special_tokens_map.json
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/merges.txt
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/config.json
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/tokenizer_config.json
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/vocab.json
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/tokenizer.json
meseca-15062024-c1-v3-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/meseca-15062024-c1-v3_reward/reward.tensors
Job meseca-15062024-c1-v3-mkmlizer completed after 165.0s with status: succeeded
Stopping job with name meseca-15062024-c1-v3-mkmlizer
Pipeline stage MKMLizer completed in 165.87s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.12s
Running pipeline stage ISVCDeployer
Creating inference service meseca-15062024-c1-v3
Waiting for inference service meseca-15062024-c1-v3 to be ready
Inference service meseca-15062024-c1-v3 ready after 40.23662257194519s
Pipeline stage ISVCDeployer completed in 46.82s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.9799790382385254s
Received healthy response to inference request in 1.2230267524719238s
Received healthy response to inference request in 1.2484180927276611s
Received healthy response to inference request in 1.249079942703247s
Received healthy response to inference request in 1.2074847221374512s
5 requests
0 failed requests
5th percentile: 1.2105931282043456
10th percentile: 1.2137015342712403
20th percentile: 1.2199183464050294
30th percentile: 1.2281050205230712
40th percentile: 1.2382615566253663
50th percentile: 1.2484180927276611
60th percentile: 1.2486828327178956
70th percentile: 1.2489475727081298
80th percentile: 1.3952597618103029
90th percentile: 1.6876194000244142
95th percentile: 1.8337992191314696
99th percentile: 1.9507430744171141
mean time: 1.3815977096557617
Pipeline stage StressChecker completed in 7.66s
meseca-15062024-c1_v3 status is now deployed due to DeploymentManager action
meseca-15062024-c1_v3 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of meseca-15062024-c1_v3
Running pipeline stage ISVCDeleter
Checking if service meseca-15062024-c1-v3 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.20s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key meseca-15062024-c1-v3/config.json from bucket guanaco-mkml-models
Deleting key meseca-15062024-c1-v3/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key meseca-15062024-c1-v3/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key meseca-15062024-c1-v3/tokenizer.json from bucket guanaco-mkml-models
Deleting key meseca-15062024-c1-v3/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key meseca-15062024-c1-v3_reward/config.json from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/merges.txt from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/reward.tensors from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key meseca-15062024-c1-v3_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 6.05s
meseca-15062024-c1_v3 status is now torndown due to DeploymentManager action