submission_id: megumi21-megumi-chat-lla_9784_v2
developer_uid: megumi_10073
status: torndown
model_repo: megumi21/Megumi-Chat-llama3-8b-v0.1
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.1, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': '<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nYou are a creative agent roleplaying as a character called {bot_name}. Stay true to the persona given, reply with short and descriptive sentences.\n\n{memory}\n\n', 'prompt_template': '{prompt}<|eot_id|>', 'bot_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>', 'user_template': '<|start_header_id|>user<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>', 'response_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}:', '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: 2024-04-24T12:41:47+00:00
model_name: megumi-chat-llama3-8b-v1
model_eval_status: success
model_group: megumi21/Megumi-Chat-lla
num_battles: 6029
num_wins: 3035
celo_rating: 1157.59
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: megumi-chat-llama3-8b-v1
ineligible_reason: propriety_total_count < 800
language_model: megumi21/Megumi-Chat-llama3-8b-v0.1
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-04-24
win_ratio: 0.5034002322109803
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name megumi21-megumi-chat-lla-9784-v2-mkmlizer
Waiting for job on megumi21-megumi-chat-lla-9784-v2-mkmlizer to finish
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ _____ __ __ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ /___/ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ Version: 0.8.10 ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ The license key for the current software has been verified as ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ belonging to: ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ Chai Research Corp. ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ║ ║
megumi21-megumi-chat-lla-9784-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
megumi21-megumi-chat-lla-9784-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.
megumi21-megumi-chat-lla-9784-v2-mkmlizer: warnings.warn(warning_message, FutureWarning)
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Downloaded to shared memory in 80.846s
megumi21-megumi-chat-lla-9784-v2-mkmlizer: quantizing model to /dev/shm/model_cache
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 29%|██▊ | 83/291 [00:01<00:02, 75.02it/s] Loading 0: 64%|██████▍ | 187/291 [00:02<00:01, 88.64it/s] Loading 0: 100%|██████████| 291/291 [00:07<00:00, 34.60it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
megumi21-megumi-chat-lla-9784-v2-mkmlizer: quantized model in 18.769s
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Processed model megumi21/Megumi-Chat-llama3-8b-v0.1 in 100.558s
megumi21-megumi-chat-lla-9784-v2-mkmlizer: creating bucket guanaco-mkml-models
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
megumi21-megumi-chat-lla-9784-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/megumi21-megumi-chat-lla-9784-v2
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/megumi21-megumi-chat-lla-9784-v2/config.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/megumi21-megumi-chat-lla-9784-v2/special_tokens_map.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/megumi21-megumi-chat-lla-9784-v2/tokenizer_config.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/megumi21-megumi-chat-lla-9784-v2/flywheel_model.0.safetensors
megumi21-megumi-chat-lla-9784-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
megumi21-megumi-chat-lla-9784-v2-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.
megumi21-megumi-chat-lla-9784-v2-mkmlizer: warnings.warn(
megumi21-megumi-chat-lla-9784-v2-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.
megumi21-megumi-chat-lla-9784-v2-mkmlizer: warnings.warn(
megumi21-megumi-chat-lla-9784-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.
megumi21-megumi-chat-lla-9784-v2-mkmlizer: warnings.warn(
megumi21-megumi-chat-lla-9784-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()
megumi21-megumi-chat-lla-9784-v2-mkmlizer: return self.fget.__get__(instance, owner)()
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Saving duration: 0.226s
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 7.618s
megumi21-megumi-chat-lla-9784-v2-mkmlizer: creating bucket guanaco-reward-models
megumi21-megumi-chat-lla-9784-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
megumi21-megumi-chat-lla-9784-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/config.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/special_tokens_map.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/tokenizer_config.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/vocab.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/merges.txt
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/tokenizer.json
megumi21-megumi-chat-lla-9784-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/megumi21-megumi-chat-lla-9784-v2_reward/reward.tensors
Job megumi21-megumi-chat-lla-9784-v2-mkmlizer completed after 127.11s with status: succeeded
Stopping job with name megumi21-megumi-chat-lla-9784-v2-mkmlizer
Pipeline stage MKMLizer completed in 131.18s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service megumi21-megumi-chat-lla-9784-v2
Waiting for inference service megumi21-megumi-chat-lla-9784-v2 to be ready
Inference service megumi21-megumi-chat-lla-9784-v2 ready after 40.3696973323822s
Pipeline stage ISVCDeployer completed in 47.86s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.2770535945892334s
Received healthy response to inference request in 1.2948017120361328s
Received healthy response to inference request in 1.309119701385498s
Received healthy response to inference request in 3.3776865005493164s
Received healthy response to inference request in 1.3458890914916992s
5 requests
0 failed requests
5th percentile: 1.2976653099060058
10th percentile: 1.3005289077758788
20th percentile: 1.306256103515625
30th percentile: 1.3164735794067384
40th percentile: 1.3311813354492188
50th percentile: 1.3458890914916992
60th percentile: 1.7183548927307128
70th percentile: 2.0908206939697265
80th percentile: 2.49718017578125
90th percentile: 2.9374333381652833
95th percentile: 3.1575599193572996
99th percentile: 3.333661184310913
mean time: 1.920910120010376
Pipeline stage StressChecker completed in 10.22s
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
megumi21-megumi-chat-lla_9784_v2 status is now deployed due to DeploymentManager action
M-Eval Dataset for topic stay_in_character is loaded
megumi21-megumi-chat-lla_9784_v2 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of megumi21-megumi-chat-lla_9784_v2
Running pipeline stage ISVCDeleter
Checking if service megumi21-megumi-chat-lla-9784-v2 is running
Tearing down inference service megumi21-megumi-chat-lla-9784-v2
Toredown service megumi21-megumi-chat-lla-9784-v2
Pipeline stage ISVCDeleter completed in 3.75s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key megumi21-megumi-chat-lla-9784-v2/config.json from bucket guanaco-mkml-models
Deleting key megumi21-megumi-chat-lla-9784-v2/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key megumi21-megumi-chat-lla-9784-v2/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key megumi21-megumi-chat-lla-9784-v2/tokenizer.json from bucket guanaco-mkml-models
Deleting key megumi21-megumi-chat-lla-9784-v2/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/config.json from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/merges.txt from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/reward.tensors from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key megumi21-megumi-chat-lla-9784-v2_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 2.01s
megumi21-megumi-chat-lla_9784_v2 status is now torndown due to DeploymentManager action

Usage Metrics

Latency Metrics