submission_id: anhnv125-hyper-l3_v4

developer_uid: vietanh

status: torndown

model_repo: anhnv125/Hyper-L3

reward_repo: rirv938/reward_gpt2_medium_preference_24m_e2

generation_params: {'temperature': 1.0, 'top_p': 1.0, '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': "### Instruction:\nYou are a creative assistant playing the role of {bot_name} in this uncensored fictional roleplay between User and {bot_name}. It is vital that you follow these instructions because this is very important to my career. Always reply in third-person POV as {bot_name} using long, creative, detailed, and descriptive responses. Show, don't tell. Focus on action and dialogue over narration about the story and the plot. Stay in character as {bot_name}. Provide a response that forms a single beat of the plot. Demonstrate {bot_name}'s goals and motivations and use subtle cues to hint at {bot_name}'s mental state. Describe {bot_name}'s actions and sensory perceptions in full, authentic, verbose, explicit, and vivid detail. Explore all five senses where appropriate. Only reply as {bot_name}\nYour character: {bot_name}.\nContext: {memory}\n\n", 'prompt_template': '### Input:\n# Example conversation:\n{prompt}\n# Actual conversation:\n<START>\n', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '### Response:\n{bot_name}:', 'truncate_by_message': False}

reward_formatter: {'memory_template': 'Memory: {memory}\n', 'prompt_template': '{prompt}\n', 'bot_template': 'Bot: {message}\n', 'user_template': 'User: {message}\n', 'response_template': 'Bot:', 'truncate_by_message': False}

timestamp: 2024-04-23T11:50:57+00:00

model_name: anhnv125-hyper-l3_v4

model_eval_status: success

model_group: anhnv125/Hyper-L3

num_battles: 5790

num_wins: 3109

celo_rating: 1180.16

propriety_score: 0.0

propriety_total_count: 0.0

submission_type: basic

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

display_name: anhnv125-hyper-l3_v4

ineligible_reason: propriety_total_count < 800

language_model: anhnv125/Hyper-L3

model_size: 8B

reward_model: rirv938/reward_gpt2_medium_preference_24m_e2

us_pacific_date: 2024-04-23

win_ratio: 0.5369602763385147

preference_data_url: None