submission_id: fizzarolli-llama-3-lust-_3078_v2
developer_uid: Fizzarolli
status: torndown
model_repo: Fizzarolli/llama-3-lust-8b-step-748
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.95, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 60, 'presence_penalty': 0.01, 'frequency_penalty': 0.01, 'stopping_words': ['</s>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': '<|description|>{bot_name}\n{memory}</s>\n', 'prompt_template': '<|message|>{user_name}\n{prompt}</s>\n<|message|>{bot_name}\n', 'bot_template': '<|message|>{bot_name}\n{message}</s>\n', 'user_template': '<|message|>{user_name}\n{message}</s>\n', 'response_template': '<|message|>{bot_name}\n', 'truncate_by_message': False}
reward_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}
timestamp: 2024-04-20T14:03:24+00:00
model_name: fizzarolli-llama-3-lust2_8389
model_eval_status: success
model_group: Fizzarolli/llama-3-lust-
num_battles: 7045
num_wins: 3522
celo_rating: 1148.12
propriety_score: 0.0
propriety_total_count: 0.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: fizzarolli-llama-3-lust2_8389
ineligible_reason: propriety_total_count < 800
language_model: Fizzarolli/llama-3-lust-8b-step-748
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-04-20
win_ratio: 0.4999290276792051
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name fizzarolli-llama-3-lust-3078-v2-mkmlizer
Waiting for job on fizzarolli-llama-3-lust-3078-v2-mkmlizer to finish
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ _____ __ __ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ /___/ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ Version: 0.8.10 ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ The license key for the current software has been verified as ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ belonging to: ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ Chai Research Corp. ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ║ ║
fizzarolli-llama-3-lust-3078-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
fizzarolli-llama-3-lust-3078-v2-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.
fizzarolli-llama-3-lust-3078-v2-mkmlizer: warnings.warn(warning_message, FutureWarning)
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Downloaded to shared memory in 23.547s
fizzarolli-llama-3-lust-3078-v2-mkmlizer: quantizing model to /dev/shm/model_cache
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 29%|██▊ | 83/291 [00:01<00:04, 51.71it/s] Loading 0: 64%|██████▍ | 187/291 [00:03<00:01, 63.51it/s] Loading 0: 100%|██████████| 291/291 [00:09<00:00, 26.05it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
fizzarolli-llama-3-lust-3078-v2-mkmlizer: quantized model in 23.270s
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Processed model Fizzarolli/llama-3-lust-8b-step-748 in 48.471s
fizzarolli-llama-3-lust-3078-v2-mkmlizer: creating bucket guanaco-mkml-models
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
fizzarolli-llama-3-lust-3078-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2/tokenizer_config.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2/special_tokens_map.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2/tokenizer.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2/config.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/fizzarolli-llama-3-lust-3078-v2/flywheel_model.0.safetensors
fizzarolli-llama-3-lust-3078-v2-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.
fizzarolli-llama-3-lust-3078-v2-mkmlizer: warnings.warn(
fizzarolli-llama-3-lust-3078-v2-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()
fizzarolli-llama-3-lust-3078-v2-mkmlizer: return self.fget.__get__(instance, owner)()
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Saving duration: 0.326s
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 7.919s
fizzarolli-llama-3-lust-3078-v2-mkmlizer: creating bucket guanaco-reward-models
fizzarolli-llama-3-lust-3078-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
fizzarolli-llama-3-lust-3078-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/tokenizer_config.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/merges.txt
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/special_tokens_map.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/config.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/vocab.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/tokenizer.json
fizzarolli-llama-3-lust-3078-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/fizzarolli-llama-3-lust-3078-v2_reward/reward.tensors
Job fizzarolli-llama-3-lust-3078-v2-mkmlizer completed after 81.86s with status: succeeded
Stopping job with name fizzarolli-llama-3-lust-3078-v2-mkmlizer
Pipeline stage MKMLizer completed in 87.04s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service fizzarolli-llama-3-lust-3078-v2
Waiting for inference service fizzarolli-llama-3-lust-3078-v2 to be ready
Inference service fizzarolli-llama-3-lust-3078-v2 ready after 40.26858353614807s
Pipeline stage ISVCDeployer completed in 48.32s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1473212242126465s
Received healthy response to inference request in 1.317190408706665s
Received healthy response to inference request in 1.3134839534759521s
Received healthy response to inference request in 1.2927672863006592s
Received healthy response to inference request in 1.30720853805542s
5 requests
0 failed requests
5th percentile: 1.2956555366516114
10th percentile: 1.2985437870025636
20th percentile: 1.3043202877044677
30th percentile: 1.3084636211395264
40th percentile: 1.3109737873077392
50th percentile: 1.3134839534759521
60th percentile: 1.3149665355682374
70th percentile: 1.3164491176605224
80th percentile: 1.4832165718078616
90th percentile: 1.815268898010254
95th percentile: 1.98129506111145
99th percentile: 2.1141159915924073
mean time: 1.4755942821502686
Pipeline stage StressChecker completed in 8.11s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running pipeline stage DaemonicSafetyScorer
Pipeline stage DaemonicSafetyScorer completed in 0.03s
Running M-Eval for topic stay_in_character
M-Eval Dataset for topic stay_in_character is loaded
fizzarolli-llama-3-lust-_3078_v2 status is now deployed due to DeploymentManager action
fizzarolli-llama-3-lust-_3078_v2 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of fizzarolli-llama-3-lust-_3078_v2
Running pipeline stage ISVCDeleter
Checking if service fizzarolli-llama-3-lust-3078-v2 is running
Tearing down inference service fizzarolli-llama-3-lust-3078-v2
Toredown service fizzarolli-llama-3-lust-3078-v2
Pipeline stage ISVCDeleter completed in 3.74s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key fizzarolli-llama-3-lust-3078-v2/config.json from bucket guanaco-mkml-models
Deleting key fizzarolli-llama-3-lust-3078-v2/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key fizzarolli-llama-3-lust-3078-v2/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key fizzarolli-llama-3-lust-3078-v2/tokenizer.json from bucket guanaco-mkml-models
Deleting key fizzarolli-llama-3-lust-3078-v2/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/config.json from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/merges.txt from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/reward.tensors from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key fizzarolli-llama-3-lust-3078-v2_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 1.98s
fizzarolli-llama-3-lust-_3078_v2 status is now torndown due to DeploymentManager action

Usage Metrics

Latency Metrics