submission_id: mylesfriedman30-llama-2-_1165_v2
developer_uid: DDDDDDDDFFFF
status: inactive
model_repo: mylesfriedman30/llama-2-7b-flirtbot
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.0, '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': 4, 'max_output_tokens': 64}
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}
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: 2024-06-15T17:46:52+00:00
model_name: flirtbot
model_eval_status: success
model_group: mylesfriedman30/llama-2-
num_battles: 11911
num_wins: 4364
celo_rating: 1076.99
propriety_score: 0.7160071942446044
propriety_total_count: 5560.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 6738415616.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: flirtbot
ineligible_reason: None
language_model: mylesfriedman30/llama-2-7b-flirtbot
model_size: 7B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-15
win_ratio: 0.36638401477625726
Resubmit model
Running pipeline stage MKMLizer
Starting job with name mylesfriedman30-llama-2-1165-v2-mkmlizer
Waiting for job on mylesfriedman30-llama-2-1165-v2-mkmlizer to finish
mylesfriedman30-llama-2-1165-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ _____ __ __ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ /___/ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ Version: 0.8.14 ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ https://mk1.ai ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ The license key for the current software has been verified as ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ belonging to: ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ Chai Research Corp. ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ║ ║
mylesfriedman30-llama-2-1165-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
mylesfriedman30-llama-2-1165-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.
mylesfriedman30-llama-2-1165-v2-mkmlizer: warnings.warn(warning_message, FutureWarning)
mylesfriedman30-llama-2-1165-v2-mkmlizer: Downloaded to shared memory in 23.354s
mylesfriedman30-llama-2-1165-v2-mkmlizer: quantizing model to /dev/shm/model_cache
mylesfriedman30-llama-2-1165-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
mylesfriedman30-llama-2-1165-v2-mkmlizer: Processed model mylesfriedman30/llama-2-7b-flirtbot in 40.019s
mylesfriedman30-llama-2-1165-v2-mkmlizer: creating bucket guanaco-mkml-models
mylesfriedman30-llama-2-1165-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
mylesfriedman30-llama-2-1165-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/config.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/special_tokens_map.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/added_tokens.json s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/added_tokens.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/tokenizer_config.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/tokenizer.model
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/tokenizer.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/mylesfriedman30-llama-2-1165-v2/flywheel_model.0.safetensors
mylesfriedman30-llama-2-1165-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
mylesfriedman30-llama-2-1165-v2-mkmlizer: Loading 0: 0%| | 0/323 [00:00<?, ?it/s] Loading 0: 75%|███████▍ | 242/323 [00:01<00:00, 207.55it/s] Loading 0: 100%|██████████| 323/323 [00:02<00:00, 116.74it/s] /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.
mylesfriedman30-llama-2-1165-v2-mkmlizer: warnings.warn(
mylesfriedman30-llama-2-1165-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.
mylesfriedman30-llama-2-1165-v2-mkmlizer: warnings.warn(
mylesfriedman30-llama-2-1165-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.
mylesfriedman30-llama-2-1165-v2-mkmlizer: warnings.warn(
mylesfriedman30-llama-2-1165-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()
mylesfriedman30-llama-2-1165-v2-mkmlizer: return self.fget.__get__(instance, owner)()
mylesfriedman30-llama-2-1165-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
mylesfriedman30-llama-2-1165-v2-mkmlizer: Saving duration: 0.515s
mylesfriedman30-llama-2-1165-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 27.570s
mylesfriedman30-llama-2-1165-v2-mkmlizer: creating bucket guanaco-reward-models
mylesfriedman30-llama-2-1165-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
mylesfriedman30-llama-2-1165-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/tokenizer_config.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/special_tokens_map.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/config.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/merges.txt
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/tokenizer.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/vocab.json
mylesfriedman30-llama-2-1165-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/mylesfriedman30-llama-2-1165-v2_reward/reward.tensors
Job mylesfriedman30-llama-2-1165-v2-mkmlizer completed after 93.35s with status: succeeded
Stopping job with name mylesfriedman30-llama-2-1165-v2-mkmlizer
Pipeline stage MKMLizer completed in 96.64s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service mylesfriedman30-llama-2-1165-v2
Waiting for inference service mylesfriedman30-llama-2-1165-v2 to be ready
Inference service mylesfriedman30-llama-2-1165-v2 ready after 50.271082639694214s
Pipeline stage ISVCDeployer completed in 57.16s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.495285987854004s
Received healthy response to inference request in 0.7548122406005859s
Received healthy response to inference request in 1.0012335777282715s
Received healthy response to inference request in 0.6671528816223145s
Received healthy response to inference request in 0.9957993030548096s
5 requests
0 failed requests
5th percentile: 0.6846847534179688
10th percentile: 0.702216625213623
20th percentile: 0.7372803688049316
30th percentile: 0.8030096530914307
40th percentile: 0.8994044780731202
50th percentile: 0.9957993030548096
60th percentile: 0.9979730129241944
70th percentile: 1.0001467227935792
80th percentile: 1.100044059753418
90th percentile: 1.297665023803711
95th percentile: 1.3964755058288574
99th percentile: 1.4755238914489746
mean time: 0.9828567981719971
Pipeline stage StressChecker completed in 5.57s
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
mylesfriedman30-llama-2-_1165_v2 status is now deployed due to DeploymentManager action
mylesfriedman30-llama-2-_1165_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics