kaoeiri-perpicon-l3-expe_1050

developer_uid: helkero

submission_id: kaoeiri-perpicon-l3-expe_1050_v1

model_name: kaoeiri-perpicon-l3-expe_1050_v1

model_group: Kaoeiri/Perpicon-L3-Expe

status: torndown

timestamp: 2024-06-15T11:38:45+00:00

num_battles: 20800

num_wins: 11083

celo_rating: 1195.67

family_friendly_score: 0.0

submission_type: basic

model_repo: Kaoeiri/Perpicon-L3-Experimental-3.5-Rewinded

model_architecture: LlamaForCausalLM

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

display_name: kaoeiri-perpicon-l3-expe_1050_v1

is_internal_developer: False

language_model: Kaoeiri/Perpicon-L3-Experimental-3.5-Rewinded

model_size: 8B

ranking_group: single

us_pacific_date: 2024-06-15

win_ratio: 0.5328365384615384

generation_params: {'temperature': 0.95, 'top_p': 1.0, 'min_p': 0.075, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|eot_id|>'], 'max_input_tokens': 512, 'best_of': 16, '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}

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'}

model_eval_status: success

Resubmit model

Running pipeline stage MKMLizer
Starting job with name kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer
Waiting for job on kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer to finish
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║     _____            __           __                                ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║       /___/                                                         ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║                                                                     ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  Version: 0.8.14                                                    ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  https://mk1.ai                                                     ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║                                                                     ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  belonging to:                                                      ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║                                                                     ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  Chai Research Corp.                                                ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ║                                                                     ║
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/utils/_deprecation.py:131: FutureWarning: 'list_files_info' (from 'huggingface_hub.hf_api') is deprecated and will be removed from version '0.23'. Use `list_repo_tree` and `get_paths_info` instead.
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer:   warnings.warn(warning_message, FutureWarning)
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Downloaded to shared memory in 37.281s
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: quantizing model to /dev/shm/model_cache
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/291 [00:04<11:35,  2.41s/it]
Loading 0:   5%|▍         | 14/291 [00:04<01:12,  3.84it/s]
Loading 0:   9%|▉         | 27/291 [00:05<00:30,  8.78it/s]
Loading 0:  14%|█▎        | 40/291 [00:05<00:16, 15.20it/s]
Loading 0:  18%|█▊        | 51/291 [00:05<00:11, 21.64it/s]
Loading 0:  23%|██▎       | 67/291 [00:05<00:06, 33.84it/s]
Loading 0:  27%|██▋       | 79/291 [00:05<00:04, 42.81it/s]
Loading 0:  32%|███▏      | 94/291 [00:05<00:03, 56.70it/s]
Loading 0:  36%|███▋      | 106/291 [00:05<00:02, 65.09it/s]
Loading 0:  42%|████▏     | 121/291 [00:05<00:02, 78.95it/s]
Loading 0:  46%|████▌     | 134/291 [00:06<00:03, 47.15it/s]
Loading 0:  51%|█████     | 148/291 [00:06<00:02, 58.77it/s]
Loading 0:  55%|█████▍    | 159/291 [00:06<00:02, 65.04it/s]
Loading 0:  60%|██████    | 175/291 [00:06<00:01, 80.21it/s]
Loading 0:  64%|██████▍   | 187/291 [00:06<00:01, 85.21it/s]
Loading 0:  69%|██████▉   | 202/291 [00:06<00:00, 96.94it/s]
Loading 0:  74%|███████▎  | 214/291 [00:07<00:00, 97.85it/s]
Loading 0:  79%|███████▊  | 229/291 [00:07<00:00, 106.73it/s]
Loading 0:  83%|████████▎ | 241/291 [00:07<00:00, 105.17it/s]
Loading 0:  88%|████████▊ | 256/291 [00:07<00:00, 114.22it/s]
Loading 0:  92%|█████████▏| 269/291 [00:07<00:00, 111.58it/s]
Loading 0:  97%|█████████▋| 283/291 [00:07<00:00, 115.33it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: quantized model in 24.121s
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Processed model Kaoeiri/Perpicon-L3-Experimental-3.5-Rewinded in 63.976s
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: creating bucket guanaco-mkml-models
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1/config.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1/special_tokens_map.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1/tokenizer_config.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1/tokenizer.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/kaoeiri-perpicon-l3-expe-1050-v1/flywheel_model.0.safetensors
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:913: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer:   warnings.warn(
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:757: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer:   warnings.warn(
kaoeiri-perpicon-l3-expe-1050-v1-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.
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer:   warnings.warn(
kaoeiri-perpicon-l3-expe-1050-v1-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()
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Saving duration: 0.632s
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 23.324s
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: creating bucket guanaco-reward-models
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/tokenizer_config.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/special_tokens_map.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/config.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/vocab.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/merges.txt
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/tokenizer.json
kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/kaoeiri-perpicon-l3-expe-1050-v1_reward/reward.tensors
Job kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer completed after 113.49s with status: succeeded
Stopping job with name kaoeiri-perpicon-l3-expe-1050-v1-mkmlizer
Pipeline stage MKMLizer completed in 117.15s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.13s
Running pipeline stage ISVCDeployer
Creating inference service kaoeiri-perpicon-l3-expe-1050-v1
Waiting for inference service kaoeiri-perpicon-l3-expe-1050-v1 to be ready
Inference service kaoeiri-perpicon-l3-expe-1050-v1 ready after 40.2944712638855s
Pipeline stage ISVCDeployer completed in 47.68s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1744847297668457s
Received healthy response to inference request in 1.2731108665466309s
Received healthy response to inference request in 1.2717626094818115s
Received healthy response to inference request in 1.2930636405944824s
Received healthy response to inference request in 1.2906060218811035s
5 requests
0 failed requests
5th percentile: 1.2720322608947754
10th percentile: 1.2723019123077393
20th percentile: 1.272841215133667
30th percentile: 1.2766098976135254
40th percentile: 1.2836079597473145
50th percentile: 1.2906060218811035
60th percentile: 1.2915890693664551
70th percentile: 1.2925721168518067
80th percentile: 1.4693478584289552
90th percentile: 1.8219162940979006
95th percentile: 1.998200511932373
99th percentile: 2.139227886199951
mean time: 1.4606055736541748
Pipeline stage StressChecker completed in 7.89s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.04s
Running M-Eval for topic stay_in_character
Running pipeline stage DaemonicSafetyScorer
M-Eval Dataset for topic stay_in_character is loaded
Pipeline stage DaemonicSafetyScorer completed in 0.08s
kaoeiri-perpicon-l3-expe_1050_v1 status is now deployed due to DeploymentManager action
kaoeiri-perpicon-l3-expe_1050_v1 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of kaoeiri-perpicon-l3-expe_1050_v1
Running pipeline stage ISVCDeleter
Checking if service kaoeiri-perpicon-l3-expe-1050-v1 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.21s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key kaoeiri-perpicon-l3-expe-1050-v1/config.json from bucket guanaco-mkml-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1/tokenizer.json from bucket guanaco-mkml-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/config.json from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/merges.txt from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/reward.tensors from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key kaoeiri-perpicon-l3-expe-1050-v1_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 6.28s
kaoeiri-perpicon-l3-expe_1050_v1 status is now torndown due to DeploymentManager action