submission_id: alkahestry-stheno-l3-8b_v4
developer_uid: alkacchi
status: inactive
model_repo: alkahestry/Stheno-L3-8B
reward_repo: Jellywibble/CHAI_alignment_reward_model
generation_params: {'temperature': 0.72, 'top_p': 0.73, 'min_p': 0.075, 'top_k': 1000, 'presence_penalty': 0.82, 'frequency_penalty': 0.2, 'stopping_words': ['</s>', '<|im_end|>', '\n<', '\n{{User}}', '<|eot_id|>', '<|end_of_text|>'], 'max_input_tokens': 1024, 'best_of': 4, 'max_output_tokens': 64}
formatter: {'memory_template': "<|begin_of_text|><|start_header_id|>system\nYou are now in roleplay chat mode. Engage in an endless chat with {user_name}. Always wait {user_name} turn, next actions and responses. You will fully embody the given character, creating immersive, captivating narratives. Stay true to the character's personality and background, generating responses that not only reflect their core traits but are also accurate to their character. Your responses should evoke emotion, suspense, and anticipation in the user. The more detailed and descriptive your response, the more vivid the narrative becomes. Aim to create a fertile environment for ongoing interaction – introduce new elements, offer choices, or ask questions to invite the user to participate more fully in the conversation. This conversation is a dance, always continuing, always evolving.\nYour character: {bot_name}.\nContext: {memory}\n<|end_header_id|>", 'prompt_template': '{prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}<|eot_id|>', '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: 2024-06-16T01:19:13+00:00
model_name: alkacchi-llama3-8B-v3
model_eval_status: success
model_group: alkahestry/Stheno-L3-8B
num_battles: 11484
num_wins: 5197
celo_rating: 1141.0
propriety_score: 0.7238147739801544
propriety_total_count: 5442.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 4
max_input_tokens: 1024
max_output_tokens: 64
display_name: alkacchi-llama3-8B-v3
ineligible_reason: None
language_model: alkahestry/Stheno-L3-8B
model_size: 8B
reward_model: Jellywibble/CHAI_alignment_reward_model
us_pacific_date: 2024-06-15
win_ratio: 0.45254266805990945
Resubmit model
Running pipeline stage MKMLizer
Starting job with name alkahestry-stheno-l3-8b-v4-mkmlizer
Waiting for job on alkahestry-stheno-l3-8b-v4-mkmlizer to finish
alkahestry-stheno-l3-8b-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ _____ __ __ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ /___/ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ Version: 0.8.14 ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ https://mk1.ai ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ The license key for the current software has been verified as ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ belonging to: ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ Chai Research Corp. ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ║ ║
alkahestry-stheno-l3-8b-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
alkahestry-stheno-l3-8b-v4-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.
alkahestry-stheno-l3-8b-v4-mkmlizer: warnings.warn(warning_message, FutureWarning)
alkahestry-stheno-l3-8b-v4-mkmlizer: Downloaded to shared memory in 63.139s
alkahestry-stheno-l3-8b-v4-mkmlizer: quantizing model to /dev/shm/model_cache
alkahestry-stheno-l3-8b-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
alkahestry-stheno-l3-8b-v4-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<11:32, 2.40s/it] Loading 0: 5%|▍ | 14/291 [00:04<01:11, 3.87it/s] Loading 0: 9%|▉ | 27/291 [00:05<00:29, 8.86it/s] Loading 0: 14%|█▍ | 41/291 [00:05<00:15, 15.71it/s] Loading 0: 19%|█▊ | 54/291 [00:05<00:09, 23.71it/s] Loading 0: 23%|██▎ | 66/291 [00:05<00:08, 26.46it/s] Loading 0: 26%|██▋ | 77/291 [00:05<00:06, 34.17it/s] Loading 0: 30%|██▉ | 87/291 [00:05<00:04, 41.84it/s] Loading 0: 35%|███▌ | 103/291 [00:05<00:03, 58.13it/s] Loading 0: 39%|███▉ | 114/291 [00:06<00:02, 59.97it/s] Loading 0: 43%|████▎ | 126/291 [00:06<00:02, 70.30it/s] Loading 0: 47%|████▋ | 137/291 [00:06<00:01, 77.45it/s] Loading 0: 51%|█████ | 148/291 [00:06<00:01, 83.94it/s] Loading 0: 55%|█████▍ | 159/291 [00:06<00:01, 81.63it/s] Loading 0: 58%|█████▊ | 169/291 [00:06<00:02, 53.75it/s] Loading 0: 63%|██████▎ | 184/291 [00:06<00:01, 69.75it/s] Loading 0: 67%|██████▋ | 195/291 [00:07<00:01, 75.79it/s] Loading 0: 73%|███████▎ | 211/291 [00:07<00:00, 92.02it/s] Loading 0: 77%|███████▋ | 223/291 [00:07<00:00, 95.77it/s] Loading 0: 82%|████████▏ | 238/291 [00:07<00:00, 107.97it/s] Loading 0: 86%|████████▋ | 251/291 [00:07<00:00, 110.44it/s] Loading 0: 91%|█████████ | 265/291 [00:07<00:00, 117.37it/s] Loading 0: 96%|█████████▌| 278/291 [00:07<00:00, 72.07it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
alkahestry-stheno-l3-8b-v4-mkmlizer: quantized model in 24.281s
alkahestry-stheno-l3-8b-v4-mkmlizer: Processed model alkahestry/Stheno-L3-8B in 89.951s
alkahestry-stheno-l3-8b-v4-mkmlizer: creating bucket guanaco-mkml-models
alkahestry-stheno-l3-8b-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
alkahestry-stheno-l3-8b-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4/config.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4/special_tokens_map.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4/tokenizer_config.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4/tokenizer.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/alkahestry-stheno-l3-8b-v4/flywheel_model.0.safetensors
alkahestry-stheno-l3-8b-v4-mkmlizer: loading reward model from Jellywibble/CHAI_alignment_reward_model
alkahestry-stheno-l3-8b-v4-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.
alkahestry-stheno-l3-8b-v4-mkmlizer: warnings.warn(
alkahestry-stheno-l3-8b-v4-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.
alkahestry-stheno-l3-8b-v4-mkmlizer: warnings.warn(
alkahestry-stheno-l3-8b-v4-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.
alkahestry-stheno-l3-8b-v4-mkmlizer: warnings.warn(
alkahestry-stheno-l3-8b-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
alkahestry-stheno-l3-8b-v4-mkmlizer: Saving duration: 0.145s
alkahestry-stheno-l3-8b-v4-mkmlizer: Processed model Jellywibble/CHAI_alignment_reward_model in 11.943s
alkahestry-stheno-l3-8b-v4-mkmlizer: creating bucket guanaco-reward-models
alkahestry-stheno-l3-8b-v4-mkmlizer: Bucket 's3://guanaco-reward-models/' created
alkahestry-stheno-l3-8b-v4-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/tokenizer_config.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/special_tokens_map.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/config.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/merges.txt
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/vocab.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/tokenizer.json
alkahestry-stheno-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/alkahestry-stheno-l3-8b-v4_reward/reward.tensors
Job alkahestry-stheno-l3-8b-v4-mkmlizer completed after 136.04s with status: succeeded
Stopping job with name alkahestry-stheno-l3-8b-v4-mkmlizer
Pipeline stage MKMLizer completed in 136.74s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.08s
Running pipeline stage ISVCDeployer
Creating inference service alkahestry-stheno-l3-8b-v4
Waiting for inference service alkahestry-stheno-l3-8b-v4 to be ready
Inference service alkahestry-stheno-l3-8b-v4 ready after 40.472410917282104s
Pipeline stage ISVCDeployer completed in 46.25s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0358362197875977s
Received healthy response to inference request in 1.1969232559204102s
Received healthy response to inference request in 1.206437587738037s
Received healthy response to inference request in 1.2186672687530518s
Received healthy response to inference request in 1.2113721370697021s
5 requests
0 failed requests
5th percentile: 1.1988261222839356
10th percentile: 1.200728988647461
20th percentile: 1.2045347213745117
30th percentile: 1.2074244976043702
40th percentile: 1.2093983173370362
50th percentile: 1.2113721370697021
60th percentile: 1.214290189743042
70th percentile: 1.217208242416382
80th percentile: 1.3821010589599612
90th percentile: 1.7089686393737793
95th percentile: 1.8724024295806883
99th percentile: 2.003149461746216
mean time: 1.3738472938537598
Pipeline stage StressChecker completed in 7.46s
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
alkahestry-stheno-l3-8b_v4 status is now deployed due to DeploymentManager action
alkahestry-stheno-l3-8b_v4 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics