developer_uid: Hastagaras
submission_id: hastagaras-l3-8b-dahlah_v2
model_name: hastagaras-l3-8b-dahlah_v1
model_group: Hastagaras/L3-8B-Dahlah
status: torndown
timestamp: 2024-05-08T07:24:19+00:00
num_battles: 17847
num_wins: 8759
celo_rating: 1161.2
family_friendly_score: 0.0
submission_type: basic
model_repo: Hastagaras/L3-8B-Dahlah
model_architecture: LlamaForCausalLM
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: hastagaras-l3-8b-dahlah_v1
is_internal_developer: False
language_model: Hastagaras/L3-8B-Dahlah
model_size: 8B
ranking_group: single
us_pacific_date: 2024-05-08
win_ratio: 0.4907827646102986
generation_params: {'temperature': 0.9, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 1.05, 'stopping_words': ['\n', '<|eot_id|>', '###'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<|begin_of_text|>### Instruction:\nIn this roleplay scenario, you will play the role of `{bot_name}`. Please write vivid and engaging respond as {bot_name} only, without deciding on `{user_name}'s` action or dialogue.\n\nYour persona as {bot_name}: {memory}\n\n", 'prompt_template': 'Your message example: {prompt}\n\n', 'bot_template': '### Response:\n{bot_name}: {message}<|eot_id|>\n\n', 'user_template': '### Input:\n{user_name}: {message}\n\n', 'response_template': '### Response:\n{bot_name}:', 'truncate_by_message': False}
model_eval_status: success
reward_formatter: {'bot_template': '{bot_name}: {message}\n', 'memory_template': "{bot_name}'s Persona: {memory}\n\n", 'prompt_template': '{bot_name}: {prompt}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False, 'user_template': '{user_name}: {message}\n'}
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-l3-8b-dahlah-v2-mkmlizer
Waiting for job on hastagaras-l3-8b-dahlah-v2-mkmlizer to finish
hastagaras-l3-8b-dahlah-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║     _____            __           __                                ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║       /___/                                                         ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║                                                                     ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  Version: 0.8.10                                                    ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║                                                                     ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  belonging to:                                                      ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║                                                                     ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  Chai Research Corp.                                                ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ║                                                                     ║
hastagaras-l3-8b-dahlah-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-l3-8b-dahlah-v2-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.
hastagaras-l3-8b-dahlah-v2-mkmlizer:   warnings.warn(warning_message, FutureWarning)
hastagaras-l3-8b-dahlah-v2-mkmlizer: Downloaded to shared memory in 18.410s
hastagaras-l3-8b-dahlah-v2-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-l3-8b-dahlah-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-l3-8b-dahlah-v2-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:  37%|███▋      | 108/291 [00:01<00:01, 107.77it/s]
Loading 0:  76%|███████▌  | 220/291 [00:02<00:00, 109.26it/s]
Loading 0:  99%|█████████▊| 287/291 [00:07<00:00, 30.24it/s] 
                                                            
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-l3-8b-dahlah-v2-mkmlizer: Processed model Hastagaras/L3-8B-Dahlah in 38.502s
hastagaras-l3-8b-dahlah-v2-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-l3-8b-dahlah-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-l3-8b-dahlah-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2/tokenizer_config.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2/special_tokens_map.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2/config.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2/tokenizer.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-l3-8b-dahlah-v2/flywheel_model.0.safetensors
hastagaras-l3-8b-dahlah-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-l3-8b-dahlah-v2-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.
hastagaras-l3-8b-dahlah-v2-mkmlizer:   warnings.warn(
hastagaras-l3-8b-dahlah-v2-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.
hastagaras-l3-8b-dahlah-v2-mkmlizer:   warnings.warn(
hastagaras-l3-8b-dahlah-v2-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.
hastagaras-l3-8b-dahlah-v2-mkmlizer:   warnings.warn(
hastagaras-l3-8b-dahlah-v2-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()
hastagaras-l3-8b-dahlah-v2-mkmlizer:   return self.fget.__get__(instance, owner)()
hastagaras-l3-8b-dahlah-v2-mkmlizer: Saving duration: 0.240s
hastagaras-l3-8b-dahlah-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.928s
hastagaras-l3-8b-dahlah-v2-mkmlizer: creating bucket guanaco-reward-models
hastagaras-l3-8b-dahlah-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-l3-8b-dahlah-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/config.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/tokenizer_config.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/vocab.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/merges.txt
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/special_tokens_map.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/tokenizer.json
hastagaras-l3-8b-dahlah-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-l3-8b-dahlah-v2_reward/reward.tensors
Job hastagaras-l3-8b-dahlah-v2-mkmlizer completed after 62.49s with status: succeeded
Stopping job with name hastagaras-l3-8b-dahlah-v2-mkmlizer
Pipeline stage MKMLizer completed in 66.94s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-l3-8b-dahlah-v2
Waiting for inference service hastagaras-l3-8b-dahlah-v2 to be ready
Inference service hastagaras-l3-8b-dahlah-v2 ready after 30.19526958465576s
Pipeline stage ISVCDeployer completed in 37.85s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0828857421875s
Received healthy response to inference request in 1.2200093269348145s
Received healthy response to inference request in 1.172429084777832s
Received healthy response to inference request in 1.1277754306793213s
Received healthy response to inference request in 1.15995192527771s
5 requests
0 failed requests
5th percentile: 1.134210729598999
10th percentile: 1.1406460285186768
20th percentile: 1.1535166263580323
30th percentile: 1.1624473571777343
40th percentile: 1.1674382209777832
50th percentile: 1.172429084777832
60th percentile: 1.191461181640625
70th percentile: 1.210493278503418
80th percentile: 1.3925846099853518
90th percentile: 1.737735176086426
95th percentile: 1.9103104591369626
99th percentile: 2.0483706855773924
mean time: 1.3526103019714355
Pipeline stage StressChecker completed in 7.40s
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
hastagaras-l3-8b-dahlah_v2 status is now deployed due to DeploymentManager action
hastagaras-l3-8b-dahlah_v2 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of hastagaras-l3-8b-dahlah_v2
Running pipeline stage ISVCDeleter
Checking if service hastagaras-l3-8b-dahlah-v2 is running
Tearing down inference service hastagaras-l3-8b-dahlah-v2
Toredown service hastagaras-l3-8b-dahlah-v2
Pipeline stage ISVCDeleter completed in 4.12s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key hastagaras-l3-8b-dahlah-v2/config.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-8b-dahlah-v2/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key hastagaras-l3-8b-dahlah-v2/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-8b-dahlah-v2/tokenizer.json from bucket guanaco-mkml-models
Deleting key hastagaras-l3-8b-dahlah-v2/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key hastagaras-l3-8b-dahlah-v2_reward/config.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/merges.txt from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/reward.tensors from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key hastagaras-l3-8b-dahlah-v2_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 2.43s
hastagaras-l3-8b-dahlah_v2 status is now torndown due to DeploymentManager action