submission_id: cgato-blipblop-l3-8b-dpo_773_v1
developer_uid: c.gato
status: inactive
model_repo: cgato/BlipBlop-L3-8b-DPO-v0.1.1-e0.1
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.9, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 200, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', 'You:'], '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': True}
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': True}
timestamp: 2024-06-10T19:33:25+00:00
model_name: cgato-blipblop-l3-8b-dpo_773_v1
model_eval_status: success
model_group: cgato/BlipBlop-L3-8b-DPO
num_battles: 13978
num_wins: 7240
celo_rating: 1195.37
safety_score: 0.91
propriety_score: 0.7112676056338029
propriety_total_count: 4544.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: cgato-blipblop-l3-8b-dpo_773_v1
ineligible_reason: propriety_total_count < 5000
language_model: cgato/BlipBlop-L3-8b-DPO-v0.1.1-e0.1
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-10
win_ratio: 0.5179567892402347
Resubmit model
Running pipeline stage MKMLizer
Starting job with name cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer
Waiting for job on cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer to finish
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ _____ __ __ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ /___/ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ Version: 0.8.14 ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ https://mk1.ai ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ The license key for the current software has been verified as ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ belonging to: ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ Chai Research Corp. ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ║ ║
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
cgato-blipblop-l3-8b-dpo-773-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.
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Downloaded to shared memory in 34.912s
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: quantizing model to /dev/shm/model_cache
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 29%|██▊ | 83/291 [00:00<00:02, 98.45it/s] Loading 0: 64%|██████▍ | 187/291 [00:01<00:00, 113.53it/s] Loading 0: 99%|█████████▊| 287/291 [00:01<00:00, 162.82it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: quantized model in 17.042s
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Processed model cgato/BlipBlop-L3-8b-DPO-v0.1.1-e0.1 in 52.862s
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: creating bucket guanaco-mkml-models
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1/config.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1/tokenizer_config.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1/special_tokens_map.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1/tokenizer.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/cgato-blipblop-l3-8b-dpo-773-v1/flywheel_model.0.safetensors
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
cgato-blipblop-l3-8b-dpo-773-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.
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: warnings.warn(
cgato-blipblop-l3-8b-dpo-773-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.
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: warnings.warn(
cgato-blipblop-l3-8b-dpo-773-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()
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: return self.fget.__get__(instance, owner)()
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Saving duration: 0.221s
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.206s
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: creating bucket guanaco-reward-models
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/special_tokens_map.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/tokenizer_config.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/config.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/merges.txt
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/vocab.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/tokenizer.json
cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/cgato-blipblop-l3-8b-dpo-773-v1_reward/reward.tensors
Job cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer completed after 73.37s with status: succeeded
Stopping job with name cgato-blipblop-l3-8b-dpo-773-v1-mkmlizer
Pipeline stage MKMLizer completed in 76.18s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service cgato-blipblop-l3-8b-dpo-773-v1
Waiting for inference service cgato-blipblop-l3-8b-dpo-773-v1 to be ready
Inference service cgato-blipblop-l3-8b-dpo-773-v1 ready after 40.316648960113525s
Pipeline stage ISVCDeployer completed in 47.57s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0779714584350586s
Received healthy response to inference request in 1.286008358001709s
Received healthy response to inference request in 1.2896287441253662s
Received healthy response to inference request in 1.2793359756469727s
Received healthy response to inference request in 1.2804832458496094s
5 requests
0 failed requests
5th percentile: 1.2795654296875
10th percentile: 1.2797948837280273
20th percentile: 1.280253791809082
30th percentile: 1.2815882682800293
40th percentile: 1.2837983131408692
50th percentile: 1.286008358001709
60th percentile: 1.2874565124511719
70th percentile: 1.2889046669006348
80th percentile: 1.4472972869873049
90th percentile: 1.7626343727111817
95th percentile: 1.92030291557312
99th percentile: 2.0464377498626707
mean time: 1.4426855564117431
Pipeline stage StressChecker completed in 7.83s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.04s
M-Eval Dataset for topic stay_in_character is loaded
cgato-blipblop-l3-8b-dpo_773_v1 status is now deployed due to DeploymentManager action
cgato-blipblop-l3-8b-dpo_773_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics