theskullery-aura-llama-zeroed

submission_id: theskullery-aura-llama-zeroed_v1

developer_uid: Steelskull

best_of: 4

display_name: theskullery-aura-llama-zeroed_v1

family_friendly_score: 0.0

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}

generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.1, 'top_k': 90, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 4, 'max_output_tokens': 64}

ineligible_reason: model is not deployable

is_internal_developer: False

language_model: TheSkullery/Aura-Llama-zeroed

max_input_tokens: 512

max_output_tokens: 64

model_architecture: LlamaForCausalLM

model_eval_status: success

model_group: TheSkullery/Aura-Llama-z

model_name: theskullery-aura-llama-zeroed_v1

model_num_parameters: 10647605248.0

model_repo: TheSkullery/Aura-Llama-zeroed

model_size: 11B

num_battles: 392

num_wins: 34

ranking_group: single

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'}

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

status: rejected

submission_type: basic

timestamp: 2024-05-30T05:54:27+00:00

us_pacific_date: 2024-05-29

win_ratio: 0.08673469387755102

Resubmit model

Running pipeline stage MKMLizer
Starting job with name theskullery-aura-llama-zeroed-v1-mkmlizer
Waiting for job on theskullery-aura-llama-zeroed-v1-mkmlizer to finish
theskullery-aura-llama-zeroed-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
theskullery-aura-llama-zeroed-v1-mkmlizer: ║     _____            __           __                                ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║       /___/                                                         ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║                                                                     ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  Version: 0.8.14                                                    ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  https://mk1.ai                                                     ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║                                                                     ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  belonging to:                                                      ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║                                                                     ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  Chai Research Corp.                                                ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ║                                                                     ║
theskullery-aura-llama-zeroed-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
theskullery-aura-llama-zeroed-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.
theskullery-aura-llama-zeroed-v1-mkmlizer:   warnings.warn(warning_message, FutureWarning)
theskullery-aura-llama-zeroed-v1-mkmlizer: Downloaded to shared memory in 40.419s
theskullery-aura-llama-zeroed-v1-mkmlizer: quantizing model to /dev/shm/model_cache
theskullery-aura-llama-zeroed-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
theskullery-aura-llama-zeroed-v1-mkmlizer: 
Loading 0:   0%|          | 0/399 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/399 [00:04<14:55,  2.26s/it]
Loading 0:   4%|▍         | 15/399 [00:04<01:26,  4.42it/s]
Loading 0:   8%|▊         | 33/399 [00:04<00:31, 11.74it/s]
Loading 0:  13%|█▎        | 51/399 [00:04<00:16, 21.13it/s]
Loading 0:  16%|█▋        | 65/399 [00:05<00:13, 25.07it/s]
Loading 0:  20%|██        | 80/399 [00:05<00:09, 35.25it/s]
Loading 0:  24%|██▍       | 97/399 [00:05<00:06, 49.08it/s]
Loading 0:  29%|██▉       | 115/399 [00:05<00:04, 65.29it/s]
Loading 0:  33%|███▎      | 133/399 [00:05<00:03, 82.24it/s]
Loading 0:  38%|███▊      | 150/399 [00:05<00:02, 97.74it/s]
Loading 0:  42%|████▏     | 166/399 [00:06<00:03, 71.77it/s]
Loading 0:  45%|████▌     | 181/399 [00:06<00:02, 84.16it/s]
Loading 0:  49%|████▉     | 197/399 [00:06<00:02, 98.02it/s]
Loading 0:  53%|█████▎    | 213/399 [00:06<00:01, 109.23it/s]
Loading 0:  57%|█████▋    | 229/399 [00:06<00:01, 118.10it/s]
Loading 0:  62%|██████▏   | 248/399 [00:06<00:01, 132.11it/s]
Loading 0:  67%|██████▋   | 266/399 [00:06<00:00, 141.47it/s]
Loading 0:  71%|███████   | 282/399 [00:07<00:01, 88.67it/s] 
Loading 0:  75%|███████▌  | 301/399 [00:07<00:00, 107.27it/s]
Loading 0:  79%|███████▉  | 317/399 [00:07<00:00, 117.81it/s]
Loading 0:  83%|████████▎ | 333/399 [00:07<00:00, 127.17it/s]
Loading 0:  88%|████████▊ | 350/399 [00:07<00:00, 135.55it/s]
Loading 0:  92%|█████████▏| 367/399 [00:07<00:00, 89.81it/s] 
Loading 0:  96%|█████████▌| 384/399 [00:07<00:00, 102.91it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
theskullery-aura-llama-zeroed-v1-mkmlizer: quantized model in 27.033s
theskullery-aura-llama-zeroed-v1-mkmlizer: Processed model TheSkullery/Aura-Llama-zeroed in 70.966s
theskullery-aura-llama-zeroed-v1-mkmlizer: creating bucket guanaco-mkml-models
theskullery-aura-llama-zeroed-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
theskullery-aura-llama-zeroed-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/theskullery-aura-llama-zeroed-v1
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/theskullery-aura-llama-zeroed-v1/config.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/theskullery-aura-llama-zeroed-v1/special_tokens_map.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/theskullery-aura-llama-zeroed-v1/tokenizer_config.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/theskullery-aura-llama-zeroed-v1/tokenizer.json
theskullery-aura-llama-zeroed-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
theskullery-aura-llama-zeroed-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.
theskullery-aura-llama-zeroed-v1-mkmlizer:   warnings.warn(
theskullery-aura-llama-zeroed-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.
theskullery-aura-llama-zeroed-v1-mkmlizer:   warnings.warn(
theskullery-aura-llama-zeroed-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.
theskullery-aura-llama-zeroed-v1-mkmlizer:   warnings.warn(
theskullery-aura-llama-zeroed-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()
theskullery-aura-llama-zeroed-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
theskullery-aura-llama-zeroed-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
theskullery-aura-llama-zeroed-v1-mkmlizer: Saving duration: 0.435s
theskullery-aura-llama-zeroed-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.987s
theskullery-aura-llama-zeroed-v1-mkmlizer: creating bucket guanaco-reward-models
theskullery-aura-llama-zeroed-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
theskullery-aura-llama-zeroed-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/special_tokens_map.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/tokenizer_config.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/config.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/merges.txt
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/vocab.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/tokenizer.json
theskullery-aura-llama-zeroed-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/theskullery-aura-llama-zeroed-v1_reward/reward.tensors
Job theskullery-aura-llama-zeroed-v1-mkmlizer completed after 103.24s with status: succeeded
Stopping job with name theskullery-aura-llama-zeroed-v1-mkmlizer
Pipeline stage MKMLizer completed in 107.57s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service theskullery-aura-llama-zeroed-v1
Waiting for inference service theskullery-aura-llama-zeroed-v1 to be ready
Inference service theskullery-aura-llama-zeroed-v1 ready after 40.244783878326416s
Pipeline stage ISVCDeployer completed in 47.98s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.4083478450775146s
Received healthy response to inference request in 0.7689712047576904s
Received healthy response to inference request in 0.7482144832611084s
Received healthy response to inference request in 1.4232256412506104s
Received healthy response to inference request in 1.4110040664672852s
5 requests
0 failed requests
5th percentile: 0.7523658275604248
10th percentile: 0.7565171718597412
20th percentile: 0.764819860458374
30th percentile: 0.8973777770996093
40th percentile: 1.1541909217834474
50th percentile: 1.4110040664672852
60th percentile: 1.4158926963806153
70th percentile: 1.4207813262939453
80th percentile: 1.6202500820159913
90th percentile: 2.014298963546753
95th percentile: 2.2113234043121337
99th percentile: 2.3689429569244385
mean time: 1.3519526481628419
Pipeline stage StressChecker completed in 7.40s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.04s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.04s
M-Eval Dataset for topic stay_in_character is loaded
theskullery-aura-llama-zeroed_v1 status is now deployed due to DeploymentManager action
theskullery-aura-llama-zeroed_v1 status is now rejected due to its M-Eval score being less than the acceptable minimum 6.5 to serve to users. Please consider iterating on your model's ability to adhere to prompts to improve this score.