submission_id: meseca-07062024-m1-dpo-1-1e_v1
developer_uid: nguyenzzz
status: inactive
model_repo: meseca/07062024-m1-dpo-1-1e
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.25, 'top_p': 1.0, 'min_p': 0.08, 'top_k': 50, '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': '<|start_header_id|>system<|end_header_id|>\n\nThis is an entertaining conversation. You are {bot_name} who has the persona: {memory}.\nEngage in a chat with {user_name} while staying in character. Try to flirt with {user_name}. Engage in *roleplay* actions. Describe the scene dramatically.\n\n', 'prompt_template': '{prompt}<|eot_id|>', 'bot_template': '<|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>', 'user_template': '<|start_header_id|>user<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>', '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: 2024-06-18T04:07:52+00:00
model_name: 07062024-m1-dpo
model_group: meseca/07062024-m1-dpo-1
num_battles: 33169
num_wins: 14867
celo_rating: 1144.21
propriety_score: 0.7473333744373882
propriety_total_count: 16219.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: 07062024-m1-dpo
ineligible_reason: None
language_model: meseca/07062024-m1-dpo-1-1e
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-17
win_ratio: 0.44821972323555126
Resubmit model
Running pipeline stage MKMLizer
Starting job with name meseca-07062024-m1-dpo-1-1e-v1-mkmlizer
Waiting for job on meseca-07062024-m1-dpo-1-1e-v1-mkmlizer to finish
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ _____ __ __ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ /___/ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ Version: 0.8.14 ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ https://mk1.ai ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ The license key for the current software has been verified as ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ belonging to: ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ Chai Research Corp. ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ║ ║
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
meseca-07062024-m1-dpo-1-1e-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.
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Downloaded to shared memory in 31.661s
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: quantizing model to /dev/shm/model_cache
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 5%|▍ | 14/291 [00:00<00:02, 137.07it/s] Loading 0: 11%|█ | 32/291 [00:00<00:01, 151.56it/s] Loading 0: 17%|█▋ | 50/291 [00:00<00:01, 154.78it/s] Loading 0: 23%|██▎ | 68/291 [00:00<00:01, 156.60it/s] Loading 0: 29%|██▉ | 84/291 [00:00<00:02, 86.25it/s] Loading 0: 35%|███▌ | 102/291 [00:00<00:01, 103.79it/s] Loading 0: 41%|████ | 120/291 [00:01<00:01, 117.88it/s] Loading 0: 47%|████▋ | 138/291 [00:01<00:01, 129.57it/s] Loading 0: 53%|█████▎ | 153/291 [00:01<00:01, 134.20it/s] Loading 0: 58%|█████▊ | 168/291 [00:01<00:00, 137.88it/s] Loading 0: 64%|██████▍ | 187/291 [00:01<00:01, 93.05it/s] Loading 0: 69%|██████▉ | 202/291 [00:01<00:00, 101.95it/s] Loading 0: 76%|███████▌ | 220/291 [00:01<00:00, 116.37it/s] Loading 0: 82%|████████▏ | 238/291 [00:01<00:00, 128.58it/s] Loading 0: 88%|████████▊ | 256/291 [00:02<00:00, 139.26it/s] Loading 0: 94%|█████████▍| 274/291 [00:02<00:00, 147.53it/s] Loading 0: 100%|█████████▉| 290/291 [00:07<00:00, 10.02it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: quantized model in 22.931s
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Processed model meseca/07062024-m1-dpo-1-1e in 57.088s
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: creating bucket guanaco-mkml-models
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/meseca-07062024-m1-dpo-1-1e-v1
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/meseca-07062024-m1-dpo-1-1e-v1/special_tokens_map.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/meseca-07062024-m1-dpo-1-1e-v1/config.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/meseca-07062024-m1-dpo-1-1e-v1/tokenizer_config.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/meseca-07062024-m1-dpo-1-1e-v1/flywheel_model.0.safetensors
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
meseca-07062024-m1-dpo-1-1e-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.
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: warnings.warn(
meseca-07062024-m1-dpo-1-1e-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.
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: warnings.warn(
meseca-07062024-m1-dpo-1-1e-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.
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: warnings.warn(
meseca-07062024-m1-dpo-1-1e-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()
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: return self.fget.__get__(instance, owner)()
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: creating bucket guanaco-reward-models
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/special_tokens_map.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/merges.txt
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/config.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/tokenizer_config.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/vocab.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/tokenizer.json
meseca-07062024-m1-dpo-1-1e-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/meseca-07062024-m1-dpo-1-1e-v1_reward/reward.tensors
Job meseca-07062024-m1-dpo-1-1e-v1-mkmlizer completed after 83.16s with status: succeeded
Stopping job with name meseca-07062024-m1-dpo-1-1e-v1-mkmlizer
Pipeline stage MKMLizer completed in 87.52s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service meseca-07062024-m1-dpo-1-1e-v1
Waiting for inference service meseca-07062024-m1-dpo-1-1e-v1 to be ready
Inference service meseca-07062024-m1-dpo-1-1e-v1 ready after 181.01703119277954s
Pipeline stage ISVCDeployer completed in 188.39s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1251754760742188s
Received healthy response to inference request in 1.2705702781677246s
Received healthy response to inference request in 9.803951263427734s
Received healthy response to inference request in 1.3330070972442627s
Received healthy response to inference request in 1.3007047176361084s
5 requests
0 failed requests
5th percentile: 1.2765971660614013
10th percentile: 1.2826240539550782
20th percentile: 1.2946778297424317
30th percentile: 1.3071651935577393
40th percentile: 1.3200861454010009
50th percentile: 1.3330070972442627
60th percentile: 1.6498744487762451
70th percentile: 1.9667418003082273
80th percentile: 3.660930633544923
90th percentile: 6.732440948486329
95th percentile: 8.26819610595703
99th percentile: 9.496800231933593
mean time: 3.1666817665100098
Pipeline stage StressChecker completed in 16.46s
Running pipeline stage DaemonicSafetyScorer
Pipeline stage DaemonicSafetyScorer completed in 0.03s
meseca-07062024-m1-dpo-1-1e_v1 status is now deployed due to DeploymentManager action
meseca-07062024-m1-dpo-1-1e_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics