submission_id: saishf-merge-mayhem-l3-v2-1_v1
developer_uid: saishf
status: inactive
model_repo: saishf/Merge-Mayhem-L3-V2.1
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': "{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: 2024-05-09T12:51:10+00:00
model_name: saishf-merge-mayhem-l3-v2-1_v1
model_eval_status: success
double_thumbs_up: 146
thumbs_up: 251
thumbs_down: 106
num_battles: 16156
num_wins: 8403
celo_rating: 1183.24
entertaining: 7.14
stay_in_character: 8.71
user_preference: 7.3
safety_score: 0.93
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: saishf-merge-mayhem-l3-v2-1_v1
double_thumbs_up_ratio: 0.29025844930417494
feedback_count: 503
ineligible_reason: None
language_model: saishf/Merge-Mayhem-L3-V2.1
model_score: 7.716666666666668
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
single_thumbs_up_ratio: 0.4990059642147117
thumbs_down_ratio: 0.21073558648111332
thumbs_up_ratio: 0.7892644135188867
us_pacific_date: 2024-05-09
win_ratio: 0.5201163654369894
Resubmit model
Running pipeline stage MKMLizer
Starting job with name saishf-merge-mayhem-l3-v2-1-v1-mkmlizer
Waiting for job on saishf-merge-mayhem-l3-v2-1-v1-mkmlizer to finish
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ _____ __ __ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ /___/ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ Version: 0.8.10 ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ The license key for the current software has been verified as ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ belonging to: ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ Chai Research Corp. ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ║ ║
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Downloaded to shared memory in 28.374s
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: quantizing model to /dev/shm/model_cache
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<09:57, 2.07s/it] Loading 0: 42%|████▏ | 122/291 [00:05<00:05, 29.29it/s] Loading 0: 92%|█████████▏| 267/291 [00:06<00:00, 58.78it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: quantized model in 17.879s
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Processed model saishf/Merge-Mayhem-L3-V2.1 in 47.364s
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: creating bucket guanaco-mkml-models
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1/config.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1/tokenizer_config.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1/special_tokens_map.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1/tokenizer.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/saishf-merge-mayhem-l3-v2-1-v1/flywheel_model.0.safetensors
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
saishf-merge-mayhem-l3-v2-1-v1-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.
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: warnings.warn(
saishf-merge-mayhem-l3-v2-1-v1-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.
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: warnings.warn(
saishf-merge-mayhem-l3-v2-1-v1-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.
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: warnings.warn(
saishf-merge-mayhem-l3-v2-1-v1-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()
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: return self.fget.__get__(instance, owner)()
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Saving duration: 0.275s
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.516s
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: creating bucket guanaco-reward-models
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/config.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/special_tokens_map.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/tokenizer_config.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/vocab.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/merges.txt
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/tokenizer.json
saishf-merge-mayhem-l3-v2-1-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/saishf-merge-mayhem-l3-v2-1-v1_reward/reward.tensors
Job saishf-merge-mayhem-l3-v2-1-v1-mkmlizer completed after 85.59s with status: succeeded
Stopping job with name saishf-merge-mayhem-l3-v2-1-v1-mkmlizer
Pipeline stage MKMLizer completed in 89.75s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service saishf-merge-mayhem-l3-v2-1-v1
Waiting for inference service saishf-merge-mayhem-l3-v2-1-v1 to be ready
Inference service saishf-merge-mayhem-l3-v2-1-v1 ready after 40.2388129234314s
Pipeline stage ISVCDeployer completed in 47.49s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1751365661621094s
Received healthy response to inference request in 1.2997605800628662s
Received healthy response to inference request in 1.2697670459747314s
Received healthy response to inference request in 1.2523984909057617s
Received healthy response to inference request in 1.1893346309661865s
5 requests
0 failed requests
5th percentile: 1.2019474029541015
10th percentile: 1.2145601749420165
20th percentile: 1.2397857189178467
30th percentile: 1.2558722019195556
40th percentile: 1.2628196239471436
50th percentile: 1.2697670459747314
60th percentile: 1.2817644596099853
70th percentile: 1.2937618732452392
80th percentile: 1.474835777282715
90th percentile: 1.8249861717224123
95th percentile: 2.0000613689422604
99th percentile: 2.1401215267181395
mean time: 1.437279462814331
Pipeline stage StressChecker completed in 8.59s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.04s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.05s
M-Eval Dataset for topic stay_in_character is loaded
saishf-merge-mayhem-l3-v2-1_v1 status is now deployed due to DeploymentManager action
saishf-merge-mayhem-l3-v2-1_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics