submission_id: cycy233-l3-bdpo-v4-c1_v1
developer_uid: shiroe40
status: inactive
model_repo: cycy233/L3-bdpo-v4-c1
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': 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: 2024-07-12T09:07:25+00:00
model_name: auto
model_group: cycy233/L3-bdpo-v4-c1
num_battles: 20035
num_wins: 10303
celo_rating: 1210.21
alignment_score: None
alignment_samples: 0
propriety_score: 0.7061581744295092
propriety_total_count: 9597.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: auto
ineligible_reason: None
language_model: cycy233/L3-bdpo-v4-c1
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-12
win_ratio: 0.514250062390816
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name cycy233-l3-bdpo-v4-c1-v1-mkmlizer
Waiting for job on cycy233-l3-bdpo-v4-c1-v1-mkmlizer to finish
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ _____ __ __ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ /___/ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ Version: 0.8.14 ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ https://mk1.ai ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ The license key for the current software has been verified as ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ belonging to: ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ Chai Research Corp. ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ║ ║
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Downloaded to shared memory in 37.779s
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: quantizing model to /dev/shm/model_cache
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 4%|▍ | 12/291 [00:00<00:02, 109.93it/s] Loading 0: 8%|▊ | 23/291 [00:00<00:02, 98.76it/s] Loading 0: 13%|█▎ | 39/291 [00:00<00:02, 116.83it/s] Loading 0: 18%|█▊ | 51/291 [00:00<00:02, 110.39it/s] Loading 0: 23%|██▎ | 66/291 [00:00<00:01, 119.73it/s] Loading 0: 27%|██▋ | 79/291 [00:00<00:01, 116.65it/s] Loading 0: 31%|███▏ | 91/291 [00:01<00:03, 50.34it/s] Loading 0: 35%|███▌ | 102/291 [00:01<00:03, 58.91it/s] Loading 0: 39%|███▉ | 113/291 [00:01<00:02, 65.63it/s] Loading 0: 44%|████▍ | 129/291 [00:01<00:01, 82.93it/s] Loading 0: 48%|████▊ | 140/291 [00:01<00:01, 85.64it/s] Loading 0: 54%|█████▎ | 156/291 [00:01<00:01, 99.48it/s] Loading 0: 58%|█████▊ | 168/291 [00:01<00:01, 98.74it/s] Loading 0: 62%|██████▏ | 181/291 [00:02<00:01, 102.91it/s] Loading 0: 66%|██████▋ | 193/291 [00:02<00:01, 54.74it/s] Loading 0: 71%|███████ | 207/291 [00:02<00:01, 67.74it/s] Loading 0: 76%|███████▌ | 220/291 [00:02<00:00, 77.78it/s] Loading 0: 80%|████████ | 233/291 [00:02<00:00, 88.10it/s] Loading 0: 84%|████████▍ | 245/291 [00:02<00:00, 92.61it/s] Loading 0: 88%|████████▊ | 257/291 [00:03<00:00, 91.69it/s] Loading 0: 94%|█████████▍| 273/291 [00:03<00:00, 106.45it/s] Loading 0: 98%|█████████▊| 285/291 [00:03<00:00, 107.24it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: quantized model in 30.884s
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Processed model cycy233/L3-bdpo-v4-c1 in 68.663s
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: creating bucket guanaco-mkml-models
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/cycy233-l3-bdpo-v4-c1-v1
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/cycy233-l3-bdpo-v4-c1-v1/config.json
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/cycy233-l3-bdpo-v4-c1-v1/special_tokens_map.json
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/cycy233-l3-bdpo-v4-c1-v1/tokenizer_config.json
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/cycy233-l3-bdpo-v4-c1-v1/tokenizer.json
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
cycy233-l3-bdpo-v4-c1-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.
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: warnings.warn(
cycy233-l3-bdpo-v4-c1-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`.
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: warnings.warn(
cycy233-l3-bdpo-v4-c1-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.
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: warnings.warn(
cycy233-l3-bdpo-v4-c1-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.
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: warnings.warn(
cycy233-l3-bdpo-v4-c1-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()
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: return self.fget.__get__(instance, owner)()
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Saving duration: 0.518s
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.992s
cycy233-l3-bdpo-v4-c1-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/cycy233-l3-bdpo-v4-c1-v1_reward/reward.tensors
Job cycy233-l3-bdpo-v4-c1-v1-mkmlizer completed after 103.63s with status: succeeded
Stopping job with name cycy233-l3-bdpo-v4-c1-v1-mkmlizer
Pipeline stage MKMLizer completed in 104.47s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service cycy233-l3-bdpo-v4-c1-v1
Waiting for inference service cycy233-l3-bdpo-v4-c1-v1 to be ready
Inference service cycy233-l3-bdpo-v4-c1-v1 ready after 40.25449180603027s
Pipeline stage ISVCDeployer completed in 47.12s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1805059909820557s
Retrying (%r) after connection broken by '%r': %s
Received healthy response to inference request in 1.2467381954193115s
Received healthy response to inference request in 1.2532541751861572s
Received healthy response to inference request in 1.2528786659240723s
Received healthy response to inference request in 1.260793685913086s
5 requests
0 failed requests
5th percentile: 1.2479662895202637
10th percentile: 1.2491943836212158
20th percentile: 1.2516505718231201
30th percentile: 1.2529537677764893
40th percentile: 1.2531039714813232
50th percentile: 1.2532541751861572
60th percentile: 1.2562699794769288
70th percentile: 1.2592857837677003
80th percentile: 1.44473614692688
90th percentile: 1.812621068954468
95th percentile: 1.9965635299682616
99th percentile: 2.1437174987792966
mean time: 1.4388341426849365
Pipeline stage StressChecker completed in 8.00s
cycy233-l3-bdpo-v4-c1_v1 status is now deployed due to DeploymentManager action
cycy233-l3-bdpo-v4-c1_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics