developer_uid: shiroe40
submission_id: cycy233-llm2-0_v5
model_name: rolep
model_group: cycy233/llm2.0
status: torndown
timestamp: 2024-05-08T02:53:55+00:00
num_battles: 13712
num_wins: 6026
celo_rating: 1127.18
family_friendly_score: 0.0
submission_type: basic
model_repo: cycy233/llm2.0
model_architecture: LlamaForCausalLM
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
model_num_parameters: 6738415616.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: rolep
is_internal_developer: False
language_model: cycy233/llm2.0
model_size: 7B
ranking_group: single
us_pacific_date: 2024-05-07
win_ratio: 0.4394690781796966
generation_params: {'temperature': 0.8, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 70, 'presence_penalty': 0.9, 'frequency_penalty': 0.9, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': 'system:\n{memory}\n"', 'prompt_template': 'Input:\n# history:{prompt}\n#speaker:{bot_name}\n# Actual conversation:\n"', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}
model_eval_status: success
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 cycy233-llm2-0-v5-mkmlizer
Waiting for job on cycy233-llm2-0-v5-mkmlizer to finish
cycy233-llm2-0-v5-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
cycy233-llm2-0-v5-mkmlizer: ║     _____            __           __                                ║
cycy233-llm2-0-v5-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
cycy233-llm2-0-v5-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
cycy233-llm2-0-v5-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
cycy233-llm2-0-v5-mkmlizer: ║       /___/                                                         ║
cycy233-llm2-0-v5-mkmlizer: ║                                                                     ║
cycy233-llm2-0-v5-mkmlizer: ║  Version: 0.8.10                                                    ║
cycy233-llm2-0-v5-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
cycy233-llm2-0-v5-mkmlizer: ║                                                                     ║
cycy233-llm2-0-v5-mkmlizer: ║  The license key for the current software has been verified as      ║
cycy233-llm2-0-v5-mkmlizer: ║  belonging to:                                                      ║
cycy233-llm2-0-v5-mkmlizer: ║                                                                     ║
cycy233-llm2-0-v5-mkmlizer: ║  Chai Research Corp.                                                ║
cycy233-llm2-0-v5-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
cycy233-llm2-0-v5-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
cycy233-llm2-0-v5-mkmlizer: ║                                                                     ║
cycy233-llm2-0-v5-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
cycy233-llm2-0-v5-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.
cycy233-llm2-0-v5-mkmlizer:   warnings.warn(warning_message, FutureWarning)
cycy233-llm2-0-v5-mkmlizer: Downloaded to shared memory in 10.555s
cycy233-llm2-0-v5-mkmlizer: quantizing model to /dev/shm/model_cache
cycy233-llm2-0-v5-mkmlizer: Saving flywheel model at /dev/shm/model_cache
cycy233-llm2-0-v5-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:  36%|███▋      | 106/291 [00:01<00:01, 101.40it/s]
Loading 0: 100%|██████████| 291/291 [00:02<00:00, 98.84it/s] 
                                                            
You set `add_prefix_space`. The tokenizer needs to be converted from the slow tokenizers
cycy233-llm2-0-v5-mkmlizer: quantized model in 9.715s
cycy233-llm2-0-v5-mkmlizer: Processed model cycy233/llm2.0 in 21.082s
cycy233-llm2-0-v5-mkmlizer: creating bucket guanaco-mkml-models
cycy233-llm2-0-v5-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
cycy233-llm2-0-v5-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/cycy233-llm2-0-v5
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/cycy233-llm2-0-v5/tokenizer.model
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/cycy233-llm2-0-v5/config.json
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/cycy233-llm2-0-v5/tokenizer_config.json
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/cycy233-llm2-0-v5/special_tokens_map.json
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/cycy233-llm2-0-v5/tokenizer.json
cycy233-llm2-0-v5-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/cycy233-llm2-0-v5/flywheel_model.0.safetensors
cycy233-llm2-0-v5-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
cycy233-llm2-0-v5-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.
cycy233-llm2-0-v5-mkmlizer:   warnings.warn(
cycy233-llm2-0-v5-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.
cycy233-llm2-0-v5-mkmlizer:   warnings.warn(
cycy233-llm2-0-v5-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.
cycy233-llm2-0-v5-mkmlizer:   warnings.warn(
cycy233-llm2-0-v5-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()
cycy233-llm2-0-v5-mkmlizer:   return self.fget.__get__(instance, owner)()
cycy233-llm2-0-v5-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
cycy233-llm2-0-v5-mkmlizer: Saving duration: 0.238s
cycy233-llm2-0-v5-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.313s
cycy233-llm2-0-v5-mkmlizer: creating bucket guanaco-reward-models
cycy233-llm2-0-v5-mkmlizer: Bucket 's3://guanaco-reward-models/' created
cycy233-llm2-0-v5-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/cycy233-llm2-0-v5_reward
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/config.json
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/tokenizer_config.json
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/merges.txt
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/special_tokens_map.json
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/vocab.json
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/tokenizer.json
cycy233-llm2-0-v5-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/cycy233-llm2-0-v5_reward/reward.tensors
Job cycy233-llm2-0-v5-mkmlizer completed after 42.23s with status: succeeded
Stopping job with name cycy233-llm2-0-v5-mkmlizer
Pipeline stage MKMLizer completed in 47.56s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service cycy233-llm2-0-v5
Waiting for inference service cycy233-llm2-0-v5 to be ready
Inference service cycy233-llm2-0-v5 ready after 30.189292907714844s
Pipeline stage ISVCDeployer completed in 37.88s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1841654777526855s
Received healthy response to inference request in 1.2185091972351074s
Received healthy response to inference request in 1.1645379066467285s
Received healthy response to inference request in 1.0794153213500977s
Received healthy response to inference request in 1.0391509532928467s
5 requests
0 failed requests
5th percentile: 1.047203826904297
10th percentile: 1.0552567005157472
20th percentile: 1.0713624477386474
30th percentile: 1.0964398384094238
40th percentile: 1.1304888725280762
50th percentile: 1.1645379066467285
60th percentile: 1.18612642288208
70th percentile: 1.2077149391174316
80th percentile: 1.4116404533386233
90th percentile: 1.7979029655456544
95th percentile: 1.9910342216491697
99th percentile: 2.1455392265319824
mean time: 1.3371557712554931
Pipeline stage StressChecker completed in 7.30s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
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 0.07s
cycy233-llm2-0_v5 status is now deployed due to DeploymentManager action
cycy233-llm2-0_v5 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of cycy233-llm2-0_v5
Running pipeline stage ISVCDeleter
Checking if service cycy233-llm2-0-v5 is running
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Tearing down inference service cycy233-llm2-0-v5
Toredown service cycy233-llm2-0-v5
Pipeline stage ISVCDeleter completed in 5.43s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key cycy233-llm2-0-v5/config.json from bucket guanaco-mkml-models
Deleting key cycy233-llm2-0-v5/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key cycy233-llm2-0-v5/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key cycy233-llm2-0-v5/tokenizer.json from bucket guanaco-mkml-models
Deleting key cycy233-llm2-0-v5/tokenizer.model from bucket guanaco-mkml-models
Deleting key cycy233-llm2-0-v5/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key cycy233-llm2-0-v5_reward/config.json from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/merges.txt from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/reward.tensors from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key cycy233-llm2-0-v5_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 2.26s
cycy233-llm2-0_v5 status is now torndown due to DeploymentManager action