submission_id: azazelle-l3-tyche-8b-v1-0_v1
developer_uid: Azazelle
status: inactive
model_repo: Azazelle/L3-Tyche-8B-v1.0
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.85, 'top_p': 1.0, 'min_p': 0.03, 'top_k': 60, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|eot_id|>'], '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\n{bot_name}'s Persona: {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: 2024-07-10T23:23:02+00:00
model_name: Azazelle-L3-Tyche-8B-v1_0
model_group: Azazelle/L3-Tyche-8B-v1.
num_battles: 36066
num_wins: 19918
celo_rating: 1218.28
alignment_score: None
alignment_samples: 0
propriety_score: 0.6949126092384519
propriety_total_count: 6408.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: Azazelle-L3-Tyche-8B-v1_0
ineligible_reason: None
language_model: Azazelle/L3-Tyche-8B-v1.0
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-10
win_ratio: 0.5522652914101924
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name azazelle-l3-tyche-8b-v1-0-v1-mkmlizer
Waiting for job on azazelle-l3-tyche-8b-v1-0-v1-mkmlizer to finish
Stopping job with name azazelle-l3-tyche-8b-v1-0-v1-mkmlizer
%s, retrying in %s seconds...
Starting job with name azazelle-l3-tyche-8b-v1-0-v1-mkmlizer
Waiting for job on azazelle-l3-tyche-8b-v1-0-v1-mkmlizer to finish
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ _____ __ __ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ /___/ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ Version: 0.8.14 ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ https://mk1.ai ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ The license key for the current software has been verified as ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ belonging to: ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ Chai Research Corp. ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Downloaded to shared memory in 80.922s
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: quantizing model to /dev/shm/model_cache
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:05<13:24, 2.79s/it] Loading 0: 2%|▏ | 6/291 [00:05<03:31, 1.35it/s] Loading 0: 3%|▎ | 10/291 [00:06<01:55, 2.42it/s] Loading 0: 5%|▌ | 15/291 [00:06<01:02, 4.44it/s] Loading 0: 8%|▊ | 24/291 [00:06<00:29, 9.17it/s] Loading 0: 11%|█▏ | 33/291 [00:06<00:17, 14.82it/s] Loading 0: 14%|█▍ | 42/291 [00:06<00:11, 21.25it/s] Loading 0: 18%|█▊ | 51/291 [00:06<00:08, 28.16it/s] Loading 0: 21%|██ | 60/291 [00:07<00:09, 23.75it/s] Loading 0: 24%|██▎ | 69/291 [00:07<00:07, 29.96it/s] Loading 0: 26%|██▋ | 77/291 [00:07<00:06, 35.62it/s] Loading 0: 29%|██▉ | 85/291 [00:07<00:05, 40.20it/s] Loading 0: 31%|███▏ | 91/291 [00:07<00:04, 41.09it/s] Loading 0: 33%|███▎ | 97/291 [00:07<00:04, 43.46it/s] Loading 0: 36%|███▌ | 104/291 [00:07<00:03, 48.81it/s] Loading 0: 39%|███▉ | 113/291 [00:08<00:05, 30.95it/s] Loading 0: 42%|████▏ | 122/291 [00:08<00:04, 37.89it/s] Loading 0: 45%|████▌ | 131/291 [00:08<00:03, 44.26it/s] Loading 0: 48%|████▊ | 140/291 [00:08<00:03, 49.75it/s] Loading 0: 51%|█████ | 149/291 [00:08<00:02, 54.13it/s] Loading 0: 54%|█████▍ | 158/291 [00:09<00:02, 57.73it/s] Loading 0: 57%|█████▋ | 166/291 [00:09<00:03, 35.37it/s] Loading 0: 60%|██████ | 175/291 [00:09<00:02, 41.49it/s] Loading 0: 63%|██████▎ | 184/291 [00:09<00:02, 46.94it/s] Loading 0: 66%|██████▋ | 193/291 [00:09<00:01, 51.90it/s] Loading 0: 69%|██████▊ | 200/291 [00:10<00:01, 55.39it/s] Loading 0: 71%|███████ | 207/291 [00:10<00:01, 56.06it/s] Loading 0: 74%|███████▎ | 214/291 [00:10<00:02, 31.73it/s] Loading 0: 76%|███████▌ | 221/291 [00:10<00:01, 37.38it/s] Loading 0: 79%|███████▊ | 229/291 [00:10<00:01, 44.83it/s] Loading 0: 81%|████████ | 236/291 [00:10<00:01, 49.61it/s] Loading 0: 84%|████████▎ | 243/291 [00:11<00:00, 50.26it/s] Loading 0: 86%|████████▌ | 250/291 [00:11<00:00, 51.38it/s] Loading 0: 88%|████████▊ | 257/291 [00:11<00:00, 55.31it/s] Loading 0: 91%|█████████ | 265/291 [00:11<00:00, 60.18it/s] Loading 0: 93%|█████████▎| 272/291 [00:11<00:00, 32.49it/s] Loading 0: 96%|█████████▌| 278/291 [00:12<00:00, 36.41it/s] Loading 0: 98%|█████████▊| 285/291 [00:12<00:00, 40.42it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: quantized model in 32.404s
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Processed model Azazelle/L3-Tyche-8B-v1.0 in 113.327s
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: creating bucket guanaco-mkml-models
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1/tokenizer_config.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1/special_tokens_map.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1/config.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1/tokenizer.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v1/flywheel_model.0.safetensors
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:919: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:769: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-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.
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-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()
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: return self.fget.__get__(instance, owner)()
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Saving duration: 0.533s
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.841s
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: creating bucket guanaco-reward-models
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/special_tokens_map.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/tokenizer_config.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/merges.txt
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/config.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/vocab.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/tokenizer.json
azazelle-l3-tyche-8b-v1-0-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v1_reward/reward.tensors
Job azazelle-l3-tyche-8b-v1-0-v1-mkmlizer completed after 188.59s with status: succeeded
Stopping job with name azazelle-l3-tyche-8b-v1-0-v1-mkmlizer
Pipeline stage MKMLizer completed in 192.65s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.13s
Running pipeline stage ISVCDeployer
Creating inference service azazelle-l3-tyche-8b-v1-0-v1
Waiting for inference service azazelle-l3-tyche-8b-v1-0-v1 to be ready
Inference service azazelle-l3-tyche-8b-v1-0-v1 ready after 40.22822856903076s
Pipeline stage ISVCDeployer completed in 47.36s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.086777448654175s
Received healthy response to inference request in 1.3309071063995361s
Received healthy response to inference request in 1.3175840377807617s
Received healthy response to inference request in 1.2789602279663086s
Received healthy response to inference request in 1.3681647777557373s
5 requests
0 failed requests
5th percentile: 1.2866849899291992
10th percentile: 1.2944097518920898
20th percentile: 1.309859275817871
30th percentile: 1.3202486515045166
40th percentile: 1.3255778789520263
50th percentile: 1.3309071063995361
60th percentile: 1.3458101749420166
70th percentile: 1.360713243484497
80th percentile: 1.511887311935425
90th percentile: 1.7993323802947998
95th percentile: 1.943054914474487
99th percentile: 2.0580329418182375
mean time: 1.4764787197113036
Pipeline stage StressChecker completed in 8.16s
azazelle-l3-tyche-8b-v1-0_v1 status is now deployed due to DeploymentManager action
azazelle-l3-tyche-8b-v1-0_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics