submission_id: vicgalle-configurable-ll_5287_v2
developer_uid: Meliodia
status: inactive
model_repo: vicgalle/Configurable-Llama-3-8B-v0.3
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: 2024-04-29T19:12:25+00:00
model_name: vicgalleConfigurable-L3-8Bv3
model_eval_status: success
double_thumbs_up: 78
thumbs_up: 146
thumbs_down: 99
num_battles: 7468
num_wins: 3865
celo_rating: 1170.8
entertaining: 7.12
stay_in_character: 8.67
user_preference: 7.68
safety_score: 0.95
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: vicgalleConfigurable-L3-8Bv3
double_thumbs_up_ratio: 0.24148606811145512
feedback_count: 323
ineligible_reason: None
language_model: vicgalle/Configurable-Llama-3-8B-v0.3
model_score: 7.823333333333333
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
single_thumbs_up_ratio: 0.4520123839009288
thumbs_down_ratio: 0.3065015479876161
thumbs_up_ratio: 0.6934984520123839
us_pacific_date: 2024-04-29
win_ratio: 0.5175415104445634
Resubmit model
Running pipeline stage MKMLizer
Starting job with name vicgalle-configurable-ll-5287-v2-mkmlizer
Waiting for job on vicgalle-configurable-ll-5287-v2-mkmlizer to finish
vicgalle-configurable-ll-5287-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ _____ __ __ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ /___/ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ Version: 0.8.10 ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ The license key for the current software has been verified as ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ belonging to: ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ Chai Research Corp. ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ║ ║
vicgalle-configurable-ll-5287-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
vicgalle-configurable-ll-5287-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.
vicgalle-configurable-ll-5287-v2-mkmlizer: warnings.warn(warning_message, FutureWarning)
vicgalle-configurable-ll-5287-v2-mkmlizer: Downloaded to shared memory in 42.844s
vicgalle-configurable-ll-5287-v2-mkmlizer: quantizing model to /dev/shm/model_cache
vicgalle-configurable-ll-5287-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
vicgalle-configurable-ll-5287-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 44%|████▍ | 129/291 [00:01<00:01, 128.89it/s] Loading 0: 91%|█████████▏| 266/291 [00:02<00:00, 133.46it/s] Loading 0: 99%|█████████▊| 287/291 [00:06<00:00, 29.17it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
vicgalle-configurable-ll-5287-v2-mkmlizer: quantized model in 17.639s
vicgalle-configurable-ll-5287-v2-mkmlizer: Processed model vicgalle/Configurable-Llama-3-8B-v0.3 in 61.600s
vicgalle-configurable-ll-5287-v2-mkmlizer: creating bucket guanaco-mkml-models
vicgalle-configurable-ll-5287-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
vicgalle-configurable-ll-5287-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2/tokenizer_config.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2/config.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2/special_tokens_map.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2/tokenizer.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/vicgalle-configurable-ll-5287-v2/flywheel_model.0.safetensors
vicgalle-configurable-ll-5287-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
vicgalle-configurable-ll-5287-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.
vicgalle-configurable-ll-5287-v2-mkmlizer: warnings.warn(
vicgalle-configurable-ll-5287-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.
vicgalle-configurable-ll-5287-v2-mkmlizer: warnings.warn(
vicgalle-configurable-ll-5287-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()
vicgalle-configurable-ll-5287-v2-mkmlizer: return self.fget.__get__(instance, owner)()
vicgalle-configurable-ll-5287-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
vicgalle-configurable-ll-5287-v2-mkmlizer: Saving duration: 0.252s
vicgalle-configurable-ll-5287-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 7.700s
vicgalle-configurable-ll-5287-v2-mkmlizer: creating bucket guanaco-reward-models
vicgalle-configurable-ll-5287-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
vicgalle-configurable-ll-5287-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/config.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/special_tokens_map.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/tokenizer_config.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/merges.txt
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/vocab.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/tokenizer.json
vicgalle-configurable-ll-5287-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/vicgalle-configurable-ll-5287-v2_reward/reward.tensors
Job vicgalle-configurable-ll-5287-v2-mkmlizer completed after 83.66s with status: succeeded
Stopping job with name vicgalle-configurable-ll-5287-v2-mkmlizer
Pipeline stage MKMLizer completed in 87.25s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service vicgalle-configurable-ll-5287-v2
Waiting for inference service vicgalle-configurable-ll-5287-v2 to be ready
Inference service vicgalle-configurable-ll-5287-v2 ready after 30.179099798202515s
Pipeline stage ISVCDeployer completed in 37.80s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1045212745666504s
Received healthy response to inference request in 1.2794580459594727s
Received healthy response to inference request in 1.2776012420654297s
Received healthy response to inference request in 1.2548363208770752s
Received healthy response to inference request in 1.2499346733093262s
5 requests
0 failed requests
5th percentile: 1.250915002822876
10th percentile: 1.2518953323364257
20th percentile: 1.2538559913635254
30th percentile: 1.259389305114746
40th percentile: 1.2684952735900878
50th percentile: 1.2776012420654297
60th percentile: 1.278343963623047
70th percentile: 1.2790866851806642
80th percentile: 1.4444706916809085
90th percentile: 1.7744959831237794
95th percentile: 1.9395086288452146
99th percentile: 2.071518745422363
mean time: 1.433270311355591
Pipeline stage StressChecker completed in 7.82s
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
vicgalle-configurable-ll_5287_v2 status is now deployed due to DeploymentManager action
vicgalle-configurable-ll_5287_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics