submission_id: azazelle-l3-tyche-8b-v1-0_v2
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}<|eot_id|>\n\n", 'prompt_template': '{prompt}<|eot_id|>\n\n', 'bot_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>\n\n', 'user_template': '<|start_header_id|>user<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>\n\n', '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: 2024-07-11T03:25:13+00:00
model_name: Azazelle-L3-Tyche-8B-v1_0
model_group: Azazelle/L3-Tyche-8B-v1.
num_battles: 340105
num_wins: 165785
celo_rating: 1221.31
alignment_score: None
alignment_samples: 0
propriety_score: 0.7079657171666247
propriety_total_count: 55538.0
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: 0.48745240440452214
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name azazelle-l3-tyche-8b-v1-0-v2-mkmlizer
Waiting for job on azazelle-l3-tyche-8b-v1-0-v2-mkmlizer to finish
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ _____ __ __ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ /___/ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ Version: 0.8.14 ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ https://mk1.ai ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ The license key for the current software has been verified as ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ belonging to: ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ Chai Research Corp. ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ║ ║
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Downloaded to shared memory in 48.477s
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: quantizing model to /dev/shm/model_cache
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:05<13:20, 2.77s/it] Loading 0: 2%|▏ | 6/291 [00:05<03:30, 1.35it/s] Loading 0: 3%|▎ | 10/291 [00:06<01:55, 2.44it/s] Loading 0: 5%|▌ | 15/291 [00:06<01:01, 4.46it/s] Loading 0: 8%|▊ | 24/291 [00:06<00:28, 9.21it/s] Loading 0: 11%|█▏ | 33/291 [00:06<00:17, 14.88it/s] Loading 0: 14%|█▍ | 42/291 [00:06<00:11, 21.32it/s] Loading 0: 18%|█▊ | 51/291 [00:06<00:08, 28.22it/s] Loading 0: 21%|██ | 60/291 [00:07<00:09, 24.30it/s] Loading 0: 24%|██▎ | 69/291 [00:07<00:07, 30.66it/s] Loading 0: 26%|██▋ | 77/291 [00:07<00:05, 37.25it/s] Loading 0: 30%|██▉ | 86/291 [00:07<00:04, 43.43it/s] Loading 0: 32%|███▏ | 94/291 [00:07<00:03, 49.81it/s] Loading 0: 35%|███▌ | 103/291 [00:07<00:03, 54.03it/s] Loading 0: 38%|███▊ | 112/291 [00:07<00:03, 57.46it/s] Loading 0: 41%|████ | 119/291 [00:08<00:05, 33.72it/s] Loading 0: 43%|████▎ | 125/291 [00:08<00:04, 37.17it/s] Loading 0: 45%|████▌ | 132/291 [00:08<00:03, 40.83it/s] Loading 0: 48%|████▊ | 141/291 [00:08<00:03, 47.23it/s] Loading 0: 52%|█████▏ | 150/291 [00:08<00:02, 52.28it/s] Loading 0: 55%|█████▍ | 159/291 [00:08<00:02, 56.35it/s] Loading 0: 57%|█████▋ | 166/291 [00:09<00:03, 34.97it/s] Loading 0: 59%|█████▉ | 172/291 [00:09<00:03, 38.62it/s] Loading 0: 61%|██████ | 178/291 [00:09<00:02, 40.76it/s] Loading 0: 64%|██████▎ | 185/291 [00:09<00:02, 45.66it/s] Loading 0: 66%|██████▋ | 193/291 [00:09<00:01, 51.71it/s] Loading 0: 68%|██████▊ | 199/291 [00:09<00:01, 51.82it/s] Loading 0: 70%|███████ | 205/291 [00:10<00:01, 49.29it/s] Loading 0: 73%|███████▎ | 212/291 [00:10<00:01, 51.59it/s] Loading 0: 75%|███████▍ | 218/291 [00:10<00:02, 29.11it/s] Loading 0: 77%|███████▋ | 223/291 [00:10<00:02, 32.02it/s] Loading 0: 79%|███████▉ | 231/291 [00:10<00:01, 38.43it/s] Loading 0: 82%|████████▏ | 240/291 [00:11<00:01, 45.49it/s] Loading 0: 86%|████████▌ | 249/291 [00:11<00:00, 51.34it/s] Loading 0: 89%|████████▊ | 258/291 [00:11<00:00, 55.66it/s] Loading 0: 91%|█████████▏| 266/291 [00:11<00:00, 34.24it/s] Loading 0: 95%|█████████▍| 275/291 [00:11<00:00, 40.64it/s] Loading 0: 97%|█████████▋| 283/291 [00:12<00:00, 47.38it/s] Loading 0: 100%|█████████▉| 290/291 [00:12<00:00, 51.76it/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-v2-mkmlizer: quantized model in 32.148s
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Processed model Azazelle/L3-Tyche-8B-v1.0 in 80.625s
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: creating bucket guanaco-mkml-models
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2/tokenizer_config.json
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2/special_tokens_map.json
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2/config.json
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2/tokenizer.json
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/azazelle-l3-tyche-8b-v1-0-v2/flywheel_model.0.safetensors
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
azazelle-l3-tyche-8b-v1-0-v2-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-v2-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-v2-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-v2-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-v2-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-v2-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-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.
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: warnings.warn(
azazelle-l3-tyche-8b-v1-0-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()
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: return self.fget.__get__(instance, owner)()
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Saving duration: 0.502s
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.229s
azazelle-l3-tyche-8b-v1-0-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/azazelle-l3-tyche-8b-v1-0-v2_reward/reward.tensors
Job azazelle-l3-tyche-8b-v1-0-v2-mkmlizer completed after 114.45s with status: succeeded
Stopping job with name azazelle-l3-tyche-8b-v1-0-v2-mkmlizer
Pipeline stage MKMLizer completed in 115.41s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service azazelle-l3-tyche-8b-v1-0-v2
Waiting for inference service azazelle-l3-tyche-8b-v1-0-v2 to be ready
Inference service azazelle-l3-tyche-8b-v1-0-v2 ready after 40.2518835067749s
Pipeline stage ISVCDeployer completed in 47.55s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.127350091934204s
Received healthy response to inference request in 1.3970763683319092s
Received healthy response to inference request in 1.3452866077423096s
Received healthy response to inference request in 1.286461353302002s
Received healthy response to inference request in 1.254134178161621s
5 requests
0 failed requests
5th percentile: 1.2605996131896973
10th percentile: 1.2670650482177734
20th percentile: 1.2799959182739258
30th percentile: 1.2982264041900635
40th percentile: 1.3217565059661864
50th percentile: 1.3452866077423096
60th percentile: 1.3660025119781494
70th percentile: 1.3867184162139892
80th percentile: 1.5431311130523684
90th percentile: 1.8352406024932861
95th percentile: 1.981295347213745
99th percentile: 2.0981391429901124
mean time: 1.4820617198944093
Pipeline stage StressChecker completed in 8.17s
azazelle-l3-tyche-8b-v1-0_v2 status is now deployed due to DeploymentManager action
azazelle-l3-tyche-8b-v1-0_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics