submission_id: jebcarter-psyonic-cetace_1919_v4
developer_uid: greg2rod6
status: inactive
model_repo: jebcarter/psyonic-cetacean-20B
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: 2024-06-29T18:01:51+00:00
model_name: jebcarter-psyonic-cetace_1919_v4
model_group: jebcarter/psyonic-cetace
num_battles: 13421
num_wins: 6111
celo_rating: 1141.63
propriety_score: 0.7213013168086755
propriety_total_count: 6455.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 19994362880.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: jebcarter-psyonic-cetace_1919_v4
ineligible_reason: None
language_model: jebcarter/psyonic-cetacean-20B
model_size: 20B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-29
win_ratio: 0.4553311973772446
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jebcarter-psyonic-cetace-1919-v4-mkmlizer
Waiting for job on jebcarter-psyonic-cetace-1919-v4-mkmlizer to finish
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ _____ __ __ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ /___/ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ Version: 0.8.14 ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ https://mk1.ai ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ The license key for the current software has been verified as ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ belonging to: ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ Chai Research Corp. ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ║ ║
jebcarter-psyonic-cetace-1919-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jebcarter-psyonic-cetace-1919-v4-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.
jebcarter-psyonic-cetace-1919-v4-mkmlizer: warnings.warn(warning_message, FutureWarning)
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Downloaded to shared memory in 45.945s
jebcarter-psyonic-cetace-1919-v4-mkmlizer: quantizing model to /dev/shm/model_cache
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jebcarter-psyonic-cetace-1919-v4-mkmlizer: quantized model in 35.831s
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Processed model jebcarter/psyonic-cetacean-20B in 88.972s
jebcarter-psyonic-cetace-1919-v4-mkmlizer: creating bucket guanaco-mkml-models
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jebcarter-psyonic-cetace-1919-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/config.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/special_tokens_map.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/tokenizer_config.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/tokenizer.model
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/tokenizer.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/jebcarter-psyonic-cetace-1919-v4/flywheel_model.1.safetensors
jebcarter-psyonic-cetace-1919-v4-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Loading 0: 0%| | 0/561 [00:00<?, ?it/s] Loading 0: 2%|▏ | 12/561 [00:00<00:05, 103.69it/s] Loading 0: 4%|▍ | 23/561 [00:00<00:05, 100.06it/s] Loading 0: 7%|▋ | 37/561 [00:00<00:04, 115.70it/s] Loading 0: 9%|▊ | 49/561 [00:00<00:04, 112.54it/s] Loading 0: 11%|█ | 61/561 [00:00<00:04, 114.51it/s] Loading 0: 13%|█▎ | 75/561 [00:00<00:03, 122.05it/s] Loading 0: 16%|█▌ | 88/561 [00:00<00:04, 116.82it/s] Loading 0: 18%|█▊ | 100/561 [00:00<00:03, 115.82it/s] Loading 0: 20%|██ | 114/561 [00:00<00:03, 118.92it/s] Loading 0: 23%|██▎ | 127/561 [00:01<00:03, 119.41it/s] Loading 0: 25%|██▍ | 139/561 [00:01<00:03, 117.85it/s] Loading 0: 27%|██▋ | 151/561 [00:01<00:08, 49.95it/s] Loading 0: 29%|██▉ | 164/561 [00:01<00:06, 61.24it/s] Loading 0: 31%|███ | 174/561 [00:02<00:06, 64.32it/s] Loading 0: 33%|███▎ | 187/561 [00:02<00:04, 76.50it/s] Loading 0: 35%|███▌ | 199/561 [00:02<00:04, 85.04it/s] Loading 0: 37%|███▋ | 210/561 [00:02<00:04, 87.11it/s] Loading 0: 40%|███▉ | 224/561 [00:02<00:03, 98.44it/s] Loading 0: 42%|████▏ | 236/561 [00:02<00:03, 103.39it/s] Loading 0: 44%|████▍ | 248/561 [00:02<00:03, 103.41it/s] Loading 0: 47%|████▋ | 261/561 [00:02<00:02, 109.66it/s] Loading 0: 49%|████▊ | 273/561 [00:02<00:02, 109.09it/s] Loading 0: 51%|█████ | 285/561 [00:03<00:05, 50.82it/s] Loading 0: 53%|█████▎ | 299/561 [00:03<00:04, 62.73it/s] Loading 0: 55%|█████▌ | 309/561 [00:03<00:03, 68.87it/s] Loading 0: 57%|█████▋ | 320/561 [00:03<00:03, 76.17it/s] Loading 0: 59%|█████▉ | 332/561 [00:03<00:02, 85.50it/s] Loading 0: 61%|██████▏ | 345/561 [00:03<00:02, 92.60it/s] Loading 0: 63%|██████▎ | 356/561 [00:04<00:02, 91.89it/s] Loading 0: 66%|██████▌ | 369/561 [00:04<00:01, 100.59it/s] Loading 0: 66%|██████▌ | 369/561 [00:20<00:01, 100.59it/s] Loading 0: 66%|██████▌ | 370/561 [00:20<01:47, 1.77it/s] Loading 0: 68%|██████▊ | 382/561 [00:20<01:04, 2.78it/s] Loading 0: 70%|███████ | 393/561 [00:20<00:41, 4.07it/s] Loading 0: 72%|███████▏ | 404/561 [00:20<00:26, 5.83it/s] Loading 0: 74%|███████▍ | 416/561 [00:20<00:17, 8.51it/s] Loading 0: 76%|███████▌ | 427/561 [00:23<00:19, 6.96it/s] Loading 0: 78%|███████▊ | 438/561 [00:23<00:12, 9.70it/s] Loading 0: 80%|███████▉ | 448/561 [00:23<00:08, 12.99it/s] Loading 0: 82%|████████▏ | 459/561 [00:23<00:05, 17.80it/s] Loading 0: 84%|████████▍ | 472/561 [00:23<00:03, 24.92it/s] Loading 0: 86%|████████▌ | 482/561 [00:23<00:02, 31.25it/s] Loading 0: 88%|████████▊ | 492/561 [00:23<00:01, 38.44it/s] Loading 0: 89%|████████▉ | 502/561 [00:23<00:01, 46.43it/s] Loading 0: 91%|█████████▏| 512/561 [00:24<00:00, 53.49it/s] Loading 0: 93%|█████████▎| 522/561 [00:24<00:00, 60.25it/s] Loading 0: 95%|█████████▍| 532/561 [00:24<00:00, 68.01it/s] Loading 0: 97%|█████████▋| 544/561 [00:24<00:00, 77.88it/s] Loading 0: 99%|█████████▉| 554/561 [00:24<00:00, 79.51it/s] /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.
jebcarter-psyonic-cetace-1919-v4-mkmlizer: warnings.warn(
jebcarter-psyonic-cetace-1919-v4-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.
jebcarter-psyonic-cetace-1919-v4-mkmlizer: warnings.warn(
jebcarter-psyonic-cetace-1919-v4-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.
jebcarter-psyonic-cetace-1919-v4-mkmlizer: warnings.warn(
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
jebcarter-psyonic-cetace-1919-v4-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()
jebcarter-psyonic-cetace-1919-v4-mkmlizer: return self.fget.__get__(instance, owner)()
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Saving duration: 0.429s
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 12.213s
jebcarter-psyonic-cetace-1919-v4-mkmlizer: creating bucket guanaco-reward-models
jebcarter-psyonic-cetace-1919-v4-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jebcarter-psyonic-cetace-1919-v4-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/config.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/special_tokens_map.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/tokenizer_config.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/merges.txt
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/vocab.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/tokenizer.json
jebcarter-psyonic-cetace-1919-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jebcarter-psyonic-cetace-1919-v4_reward/reward.tensors
Job jebcarter-psyonic-cetace-1919-v4-mkmlizer completed after 124.74s with status: succeeded
Stopping job with name jebcarter-psyonic-cetace-1919-v4-mkmlizer
Pipeline stage MKMLizer completed in 125.54s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.12s
Running pipeline stage ISVCDeployer
Creating inference service jebcarter-psyonic-cetace-1919-v4
Waiting for inference service jebcarter-psyonic-cetace-1919-v4 to be ready
Inference service jebcarter-psyonic-cetace-1919-v4 ready after 60.287620544433594s
Pipeline stage ISVCDeployer completed in 67.01s
Running pipeline stage StressChecker
Received healthy response to inference request in 3.461137533187866s
Received healthy response to inference request in 2.5971317291259766s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Received healthy response to inference request in 2.4802825450897217s
Received healthy response to inference request in 2.499382734298706s
Received healthy response to inference request in 2.390676975250244s
5 requests
0 failed requests
5th percentile: 2.4085980892181396
10th percentile: 2.426519203186035
20th percentile: 2.4623614311218263
30th percentile: 2.4841025829315186
40th percentile: 2.491742658615112
50th percentile: 2.499382734298706
60th percentile: 2.538482332229614
70th percentile: 2.5775819301605223
80th percentile: 2.769932889938355
90th percentile: 3.1155352115631105
95th percentile: 3.288336372375488
99th percentile: 3.4265773010253904
mean time: 2.685722303390503
Pipeline stage StressChecker completed in 14.22s
jebcarter-psyonic-cetace_1919_v4 status is now deployed due to DeploymentManager action
jebcarter-psyonic-cetace_1919_v4 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics