submission_id: neversleep-lumi-8b-v0-2-_7154_v1
developer_uid: IkariDev
status: inactive
model_repo: NeverSleep/Lumi-8B-v0.2-WIP-Train2
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.074, 'top_k': 70, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['<|eot_id|>'], '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: 2024-06-29T17:13:34+00:00
model_name: neversleep-lumi-8b-v0-2-_7154_v1
model_group: NeverSleep/Lumi-8B-v0.2-
num_battles: 25936
num_wins: 13140
celo_rating: 1182.52
propriety_score: 0.7093302425697492
propriety_total_count: 12079.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: neversleep-lumi-8b-v0-2-_7154_v1
ineligible_reason: None
language_model: NeverSleep/Lumi-8B-v0.2-WIP-Train2
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-29
win_ratio: 0.506631708821715
Resubmit model
Running pipeline stage MKMLizer
Starting job with name neversleep-lumi-8b-v0-2-7154-v1-mkmlizer
Waiting for job on neversleep-lumi-8b-v0-2-7154-v1-mkmlizer to finish
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ _____ __ __ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ /___/ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ Version: 0.8.14 ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ https://mk1.ai ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ The license key for the current software has been verified as ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ belonging to: ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ Chai Research Corp. ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ║ ║
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
neversleep-lumi-8b-v0-2-7154-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.
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Downloaded to shared memory in 37.415s
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: quantizing model to /dev/shm/model_cache
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 5%|▍ | 14/291 [00:00<00:02, 133.49it/s] Loading 0: 11%|█ | 31/291 [00:00<00:01, 152.72it/s] Loading 0: 16%|█▋ | 48/291 [00:00<00:01, 157.59it/s] Loading 0: 22%|██▏ | 64/291 [00:00<00:01, 156.04it/s] Loading 0: 27%|██▋ | 80/291 [00:00<00:01, 149.92it/s] Loading 0: 33%|███▎ | 96/291 [00:00<00:02, 74.77it/s] Loading 0: 38%|███▊ | 111/291 [00:01<00:02, 84.45it/s] Loading 0: 43%|████▎ | 124/291 [00:01<00:01, 93.51it/s] Loading 0: 48%|████▊ | 139/291 [00:01<00:01, 105.48it/s] Loading 0: 54%|█████▎ | 156/291 [00:01<00:01, 120.15it/s] Loading 0: 58%|█████▊ | 170/291 [00:01<00:00, 121.83it/s] Loading 0: 64%|██████▍ | 187/291 [00:01<00:01, 77.72it/s] Loading 0: 69%|██████▉ | 201/291 [00:01<00:01, 87.88it/s] Loading 0: 74%|███████▍ | 216/291 [00:02<00:00, 100.23it/s] Loading 0: 79%|███████▉ | 230/291 [00:02<00:00, 107.55it/s] Loading 0: 85%|████████▍ | 247/291 [00:02<00:00, 122.19it/s] Loading 0: 91%|█████████ | 264/291 [00:02<00:00, 133.03it/s] Loading 0: 96%|█████████▌| 279/291 [00:02<00:00, 130.44it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: quantized model in 25.079s
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Processed model NeverSleep/Lumi-8B-v0.2-WIP-Train2 in 65.243s
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: creating bucket guanaco-mkml-models
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1/config.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1/special_tokens_map.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1/tokenizer.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1/tokenizer_config.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/neversleep-lumi-8b-v0-2-7154-v1/flywheel_model.0.safetensors
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
neversleep-lumi-8b-v0-2-7154-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.
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: warnings.warn(
neversleep-lumi-8b-v0-2-7154-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.
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: warnings.warn(
neversleep-lumi-8b-v0-2-7154-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.
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: warnings.warn(
neversleep-lumi-8b-v0-2-7154-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()
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: return self.fget.__get__(instance, owner)()
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Saving duration: 0.437s
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.187s
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: creating bucket guanaco-reward-models
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/config.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/merges.txt
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/tokenizer_config.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/vocab.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/special_tokens_map.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/tokenizer.json
neversleep-lumi-8b-v0-2-7154-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/neversleep-lumi-8b-v0-2-7154-v1_reward/reward.tensors
Job neversleep-lumi-8b-v0-2-7154-v1-mkmlizer completed after 93.99s with status: succeeded
Stopping job with name neversleep-lumi-8b-v0-2-7154-v1-mkmlizer
Pipeline stage MKMLizer completed in 94.92s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service neversleep-lumi-8b-v0-2-7154-v1
Waiting for inference service neversleep-lumi-8b-v0-2-7154-v1 to be ready
Inference service neversleep-lumi-8b-v0-2-7154-v1 ready after 40.197736501693726s
Pipeline stage ISVCDeployer completed in 47.43s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.19785213470459s
Received healthy response to inference request in 1.2893624305725098s
Received healthy response to inference request in 1.306769847869873s
Received healthy response to inference request in 1.2967603206634521s
Received healthy response to inference request in 1.2931814193725586s
5 requests
0 failed requests
5th percentile: 1.2901262283325194
10th percentile: 1.2908900260925293
20th percentile: 1.292417621612549
30th percentile: 1.2938971996307373
40th percentile: 1.2953287601470946
50th percentile: 1.2967603206634521
60th percentile: 1.3007641315460206
70th percentile: 1.304767942428589
80th percentile: 1.4849863052368166
90th percentile: 1.8414192199707031
95th percentile: 2.0196356773376465
99th percentile: 2.162208843231201
mean time: 1.4767852306365967
Pipeline stage StressChecker completed in 8.02s
neversleep-lumi-8b-v0-2-_7154_v1 status is now deployed due to DeploymentManager action
neversleep-lumi-8b-v0-2-_7154_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics