submission_id: cgato-l3-thespice-8b-dpo_6068_v1
developer_uid: c.gato
status: torndown
model_repo: cgato/L3-TheSpice-8b-DPO-v3.2-e6.1-HighLR
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.8, 'top_p': 0.9, '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: 2024-05-14T21:20:35+00:00
model_name: cgato-l3-thespice-8b-dpo_6068_v1
model_eval_status: success
model_group: cgato/L3-TheSpice-8b-DPO
num_battles: 17325
num_wins: 9687
celo_rating: 1200.94
propriety_score: 0.0
propriety_total_count: 0.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030277632.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: cgato-l3-thespice-8b-dpo_6068_v1
ineligible_reason: propriety_total_count < 800
language_model: cgato/L3-TheSpice-8b-DPO-v3.2-e6.1-HighLR
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-14
win_ratio: 0.5591341991341992
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer
Waiting for job on cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer to finish
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ _____ __ __ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ /___/ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ Version: 0.8.10 ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ The license key for the current software has been verified as ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ belonging to: ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ Chai Research Corp. ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ║ ║
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
cgato-l3-thespice-8b-dpo-6068-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-l3-thespice-8b-dpo-6068-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Downloaded to shared memory in 30.099s
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: quantizing model to /dev/shm/model_cache
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 64%|██████▍ | 187/291 [00:01<00:00, 108.37it/s] Loading 0: 100%|██████████| 291/291 [00:06<00:00, 38.36it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: quantized model in 17.995s
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Processed model cgato/L3-TheSpice-8b-DPO-v3.2-e6.1-HighLR in 49.209s
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: creating bucket guanaco-mkml-models
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1/tokenizer_config.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1/special_tokens_map.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1/config.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1/tokenizer.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/cgato-l3-thespice-8b-dpo-6068-v1/flywheel_model.0.safetensors
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
cgato-l3-thespice-8b-dpo-6068-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-l3-thespice-8b-dpo-6068-v1-mkmlizer: warnings.warn(
cgato-l3-thespice-8b-dpo-6068-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-l3-thespice-8b-dpo-6068-v1-mkmlizer: warnings.warn(
cgato-l3-thespice-8b-dpo-6068-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-l3-thespice-8b-dpo-6068-v1-mkmlizer: return self.fget.__get__(instance, owner)()
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Saving duration: 0.238s
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.491s
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: creating bucket guanaco-reward-models
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/special_tokens_map.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/config.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/tokenizer_config.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/merges.txt
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/vocab.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/tokenizer.json
cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/cgato-l3-thespice-8b-dpo-6068-v1_reward/reward.tensors
Job cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer completed after 73.05s with status: succeeded
Stopping job with name cgato-l3-thespice-8b-dpo-6068-v1-mkmlizer
Pipeline stage MKMLizer completed in 77.43s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service cgato-l3-thespice-8b-dpo-6068-v1
Waiting for inference service cgato-l3-thespice-8b-dpo-6068-v1 to be ready
Inference service cgato-l3-thespice-8b-dpo-6068-v1 ready after 40.25491762161255s
Pipeline stage ISVCDeployer completed in 50.62s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.3143086433410645s
Received healthy response to inference request in 1.286229133605957s
Received healthy response to inference request in 1.312291145324707s
Received healthy response to inference request in 1.2955152988433838s
Received healthy response to inference request in 1.289609432220459s
5 requests
0 failed requests
5th percentile: 1.2869051933288573
10th percentile: 1.2875812530517579
20th percentile: 1.2889333724975587
30th percentile: 1.290790605545044
40th percentile: 1.2931529521942138
50th percentile: 1.2955152988433838
60th percentile: 1.3022256374359131
70th percentile: 1.3089359760284425
80th percentile: 1.5126946449279788
90th percentile: 1.9135016441345216
95th percentile: 2.1139051437377927
99th percentile: 2.27422794342041
mean time: 1.4995907306671143
Pipeline stage StressChecker completed in 8.11s
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 0.03s
M-Eval Dataset for topic stay_in_character is loaded
cgato-l3-thespice-8b-dpo_6068_v1 status is now deployed due to DeploymentManager action
cgato-l3-thespice-8b-dpo_6068_v1 status is now inactive due to auto deactivation removed underperforming models
cgato-l3-thespice-8b-dpo_3603_v1 status is now torndown due to DeploymentManager action
admin requested tearing down of cgato-l3-thespice-8b-dpo_6068_v1
Running pipeline stage ISVCDeleter
Checking if service cgato-l3-thespice-8b-dpo-6068-v1 is running
Tearing down inference service cgato-l3-thespice-8b-dpo-6068-v1
Toredown service cgato-l3-thespice-8b-dpo-6068-v1
Pipeline stage ISVCDeleter completed in 4.46s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key cgato-l3-thespice-8b-dpo-6068-v1/config.json from bucket guanaco-mkml-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1/tokenizer.json from bucket guanaco-mkml-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/config.json from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/merges.txt from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/reward.tensors from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key cgato-l3-thespice-8b-dpo-6068-v1_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 3.17s
cgato-l3-thespice-8b-dpo_6068_v1 status is now torndown due to DeploymentManager action

Usage Metrics

Latency Metrics