submission_id: wespro-omni-llama3-8b_v1
developer_uid: WesPro
status: inactive
model_repo: WesPro/Omni-Llama3-8B
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: 2024-06-14T02:41:53+00:00
model_name: wespro-omni-llama3-8b_v1
model_eval_status: success
model_group: WesPro/Omni-Llama3-8B
num_battles: 18528
num_wins: 9456
celo_rating: 1176.54
propriety_score: 0.6967106044412777
propriety_total_count: 8421.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: wespro-omni-llama3-8b_v1
ineligible_reason: None
language_model: WesPro/Omni-Llama3-8B
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-13
win_ratio: 0.5103626943005182
Resubmit model
Running pipeline stage MKMLizer
Starting job with name wespro-omni-llama3-8b-v1-mkmlizer
Waiting for job on wespro-omni-llama3-8b-v1-mkmlizer to finish
wespro-omni-llama3-8b-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
wespro-omni-llama3-8b-v1-mkmlizer: ║ _____ __ __ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ /___/ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ Version: 0.8.14 ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ https://mk1.ai ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ The license key for the current software has been verified as ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ belonging to: ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ Chai Research Corp. ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
wespro-omni-llama3-8b-v1-mkmlizer: ║ ║
wespro-omni-llama3-8b-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
wespro-omni-llama3-8b-v1-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.
wespro-omni-llama3-8b-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
wespro-omni-llama3-8b-v1-mkmlizer: Downloaded to shared memory in 59.143s
wespro-omni-llama3-8b-v1-mkmlizer: quantizing model to /dev/shm/model_cache
wespro-omni-llama3-8b-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
wespro-omni-llama3-8b-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<10:42, 2.22s/it] Loading 0: 5%|▍ | 14/291 [00:04<01:06, 4.16it/s] Loading 0: 11%|█ | 31/291 [00:04<00:23, 11.16it/s] Loading 0: 15%|█▌ | 45/291 [00:04<00:13, 18.49it/s] Loading 0: 20%|██ | 59/291 [00:04<00:08, 27.57it/s] Loading 0: 25%|██▍ | 72/291 [00:05<00:07, 30.02it/s] Loading 0: 30%|██▉ | 86/291 [00:05<00:05, 40.73it/s] Loading 0: 35%|███▌ | 103/291 [00:05<00:03, 55.89it/s] Loading 0: 40%|███▉ | 116/291 [00:05<00:02, 66.74it/s] Loading 0: 45%|████▌ | 131/291 [00:05<00:02, 79.83it/s] Loading 0: 51%|█████ | 148/291 [00:05<00:01, 95.96it/s] Loading 0: 56%|█████▌ | 162/291 [00:05<00:01, 103.99it/s] Loading 0: 60%|██████ | 176/291 [00:06<00:01, 71.73it/s] Loading 0: 66%|██████▋ | 193/291 [00:06<00:01, 87.48it/s] Loading 0: 71%|███████ | 206/291 [00:06<00:00, 95.72it/s] Loading 0: 76%|███████▌ | 221/291 [00:06<00:00, 105.49it/s] Loading 0: 82%|████████▏ | 238/291 [00:06<00:00, 116.03it/s] Loading 0: 87%|████████▋ | 252/291 [00:06<00:00, 119.97it/s] Loading 0: 91%|█████████▏| 266/291 [00:07<00:00, 75.19it/s] Loading 0: 97%|█████████▋| 283/291 [00:07<00:00, 89.91it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
wespro-omni-llama3-8b-v1-mkmlizer: quantized model in 23.505s
wespro-omni-llama3-8b-v1-mkmlizer: Processed model WesPro/Omni-Llama3-8B in 85.170s
wespro-omni-llama3-8b-v1-mkmlizer: creating bucket guanaco-mkml-models
wespro-omni-llama3-8b-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
wespro-omni-llama3-8b-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/wespro-omni-llama3-8b-v1
wespro-omni-llama3-8b-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/wespro-omni-llama3-8b-v1/special_tokens_map.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/wespro-omni-llama3-8b-v1/tokenizer_config.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/wespro-omni-llama3-8b-v1/config.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/wespro-omni-llama3-8b-v1/tokenizer.json
wespro-omni-llama3-8b-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
wespro-omni-llama3-8b-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.
wespro-omni-llama3-8b-v1-mkmlizer: warnings.warn(
wespro-omni-llama3-8b-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.
wespro-omni-llama3-8b-v1-mkmlizer: warnings.warn(
wespro-omni-llama3-8b-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.
wespro-omni-llama3-8b-v1-mkmlizer: warnings.warn(
wespro-omni-llama3-8b-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()
wespro-omni-llama3-8b-v1-mkmlizer: return self.fget.__get__(instance, owner)()
wespro-omni-llama3-8b-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
wespro-omni-llama3-8b-v1-mkmlizer: Saving duration: 0.402s
wespro-omni-llama3-8b-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 13.006s
wespro-omni-llama3-8b-v1-mkmlizer: creating bucket guanaco-reward-models
wespro-omni-llama3-8b-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
wespro-omni-llama3-8b-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/config.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/special_tokens_map.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/tokenizer_config.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/vocab.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/merges.txt
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/tokenizer.json
wespro-omni-llama3-8b-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/wespro-omni-llama3-8b-v1_reward/reward.tensors
Job wespro-omni-llama3-8b-v1-mkmlizer completed after 113.83s with status: succeeded
Stopping job with name wespro-omni-llama3-8b-v1-mkmlizer
Pipeline stage MKMLizer completed in 117.10s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service wespro-omni-llama3-8b-v1
Waiting for inference service wespro-omni-llama3-8b-v1 to be ready
Inference service wespro-omni-llama3-8b-v1 ready after 40.30631756782532s
Pipeline stage ISVCDeployer completed in 46.84s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0348479747772217s
Received healthy response to inference request in 1.1223058700561523s
Received healthy response to inference request in 1.1165518760681152s
Received healthy response to inference request in 1.1212785243988037s
Received healthy response to inference request in 0.9852571487426758s
5 requests
0 failed requests
5th percentile: 1.0115160942077637
10th percentile: 1.0377750396728516
20th percentile: 1.0902929306030273
30th percentile: 1.117497205734253
40th percentile: 1.1193878650665283
50th percentile: 1.1212785243988037
60th percentile: 1.121689462661743
70th percentile: 1.1221004009246827
80th percentile: 1.3048142910003664
90th percentile: 1.669831132888794
95th percentile: 1.8523395538330076
99th percentile: 1.9983462905883789
mean time: 1.2760482788085938
Pipeline stage StressChecker completed in 7.11s
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
wespro-omni-llama3-8b_v1 status is now deployed due to DeploymentManager action
M-Eval Dataset for topic stay_in_character is loaded
wespro-omni-llama3-8b_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics