hastagaras-l3-cukai-8b-dpo

developer_uid: Hastagaras

submission_id: hastagaras-l3-cukai-8b-dpo_v4

model_name: cool-formatter

model_group: Hastagaras/L3-Cukai-8B-D

status: torndown

timestamp: 2024-05-16T23:33:51+00:00

num_battles: 20074

num_wins: 11126

celo_rating: 1205.89

family_friendly_score: 0.0

submission_type: basic

model_repo: Hastagaras/L3-Cukai-8B-DPO

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: cool-formatter

is_internal_developer: False

language_model: Hastagaras/L3-Cukai-8B-DPO

model_size: 8B

ranking_group: single

us_pacific_date: 2024-05-16

win_ratio: 0.5542492776726113

generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.05, 'top_k': 50, 'presence_penalty': 0.0, 'frequency_penalty': 1.05, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}

formatter: {'memory_template': '### Instruction:\n{memory}\n', 'prompt_template': '### Input:\n{prompt}\n', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '### Response:\n{bot_name}:', 'truncate_by_message': True}

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': True, 'user_template': '{user_name}: {message}\n'}

model_eval_status: success

Resubmit model

Running pipeline stage MKMLizer
Starting job with name hastagaras-l3-cukai-8b-dpo-v4-mkmlizer
Waiting for job on hastagaras-l3-cukai-8b-dpo-v4-mkmlizer to finish
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║     _____            __           __                                ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║       /___/                                                         ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║                                                                     ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  Version: 0.8.14                                                    ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  https://mk1.ai                                                     ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║                                                                     ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  belonging to:                                                      ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║                                                                     ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  Chai Research Corp.                                                ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ║                                                                     ║
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-l3-cukai-8b-dpo-v4-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.
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer:   warnings.warn(warning_message, FutureWarning)
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Downloaded to shared memory in 42.870s
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   7%|▋         | 21/291 [00:00<00:01, 203.06it/s]
Loading 0:  14%|█▍        | 42/291 [00:00<00:01, 204.32it/s]
Loading 0:  23%|██▎       | 67/291 [00:00<00:01, 217.08it/s]
Loading 0:  31%|███       | 89/291 [00:00<00:01, 109.01it/s]
Loading 0:  38%|███▊      | 112/291 [00:00<00:01, 133.79it/s]
Loading 0:  47%|████▋     | 138/291 [00:00<00:00, 160.13it/s]
Loading 0:  54%|█████▍    | 158/291 [00:00<00:00, 169.87it/s]
Loading 0:  63%|██████▎   | 182/291 [00:01<00:00, 187.92it/s]
Loading 0:  70%|███████   | 204/291 [00:01<00:00, 117.35it/s]
Loading 0:  78%|███████▊  | 228/291 [00:01<00:00, 140.28it/s]
Loading 0:  86%|████████▌ | 249/291 [00:01<00:00, 154.90it/s]
Loading 0:  94%|█████████▍| 274/291 [00:01<00:00, 174.41it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: quantized model in 16.897s
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Processed model Hastagaras/L3-Cukai-8B-DPO in 60.717s
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4/special_tokens_map.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4/config.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4/tokenizer_config.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4/tokenizer.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-l3-cukai-8b-dpo-v4/flywheel_model.0.safetensors
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-l3-cukai-8b-dpo-v4-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.
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer:   warnings.warn(
hastagaras-l3-cukai-8b-dpo-v4-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.
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer:   warnings.warn(
hastagaras-l3-cukai-8b-dpo-v4-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.
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer:   warnings.warn(
hastagaras-l3-cukai-8b-dpo-v4-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()
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer:   return self.fget.__get__(instance, owner)()
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Saving duration: 0.218s
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.557s
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: creating bucket guanaco-reward-models
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/special_tokens_map.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/config.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/merges.txt
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/tokenizer_config.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/vocab.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/tokenizer.json
hastagaras-l3-cukai-8b-dpo-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-l3-cukai-8b-dpo-v4_reward/reward.tensors
Job hastagaras-l3-cukai-8b-dpo-v4-mkmlizer completed after 83.65s with status: succeeded
Stopping job with name hastagaras-l3-cukai-8b-dpo-v4-mkmlizer
Pipeline stage MKMLizer completed in 85.81s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-l3-cukai-8b-dpo-v4
Waiting for inference service hastagaras-l3-cukai-8b-dpo-v4 to be ready
Inference service hastagaras-l3-cukai-8b-dpo-v4 ready after 30.19252038002014s
Pipeline stage ISVCDeployer completed in 37.22s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.04913592338562s
Received healthy response to inference request in 1.0141651630401611s
Received healthy response to inference request in 1.0049152374267578s
Received healthy response to inference request in 1.2318851947784424s
Received healthy response to inference request in 1.1949753761291504s
5 requests
0 failed requests
5th percentile: 1.0067652225494386
10th percentile: 1.008615207672119
20th percentile: 1.0123151779174804
30th percentile: 1.050327205657959
40th percentile: 1.1226512908935546
50th percentile: 1.1949753761291504
60th percentile: 1.2097393035888673
70th percentile: 1.224503231048584
80th percentile: 1.395335340499878
90th percentile: 1.7222356319427492
95th percentile: 1.8856857776641844
99th percentile: 2.016445894241333
mean time: 1.2990153789520265
Pipeline stage StressChecker completed in 7.14s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.03s
M-Eval Dataset for topic stay_in_character is loaded
hastagaras-l3-cukai-8b-dpo_v4 status is now deployed due to DeploymentManager action
hastagaras-l3-cukai-8b-dpo_v4 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of hastagaras-l3-cukai-8b-dpo_v4
Running pipeline stage ISVCDeleter
Checking if service hastagaras-l3-cukai-8b-dpo-v4 is running
Tearing down inference service hastagaras-l3-cukai-8b-dpo-v4
Toredown service hastagaras-l3-cukai-8b-dpo-v4
Pipeline stage ISVCDeleter completed in 3.53s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key hastagaras-l3-cukai-8b-dpo-v4/config.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4/tokenizer.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/config.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/merges.txt from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/reward.tensors from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-cukai-8b-dpo-v4_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 1.85s
hastagaras-l3-cukai-8b-dpo_v4 status is now torndown due to DeploymentManager action