submission_id: hyperblaze-l3-8b-soliloq_5266_v1

developer_uid: HyperBlaze

status: rejected

model_repo: HyperBlaze/L3-8B-soliloquy-lumimaid-spice-mergetestv1

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

generation_params: {'temperature': 0.9, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 1.0, 'frequency_penalty': 0.2, 'stopping_words': ['\n', 'user:', 'you:'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}

formatter: {'memory_template': "<|start_header_id|>system<|end_header_id|>\n\nYou are `{bot_name}` and engage in a Roleplay with `{user_name}`.\n\n Always stay and respond as {bot_name}. Do not break out. \n\n Your persona as {bot_name}: {memory}\n ## [CoT instruction]\n Before responding to the user, Concisely summarize this scene, considering the main personalities of {bot_name}, recent developments while chatting with {user_name}, scene compositions. actively use {bot_name}'s persona too.", 'prompt_template': 'Your message as {bot_name}: {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: 2024-05-09T16:50:17+00:00

model_name: hyperblaze-l3-8b-soliloq_5266_v1

model_eval_status: error

double_thumbs_up: 24

thumbs_up: 32

thumbs_down: 19

num_battles: 3748

num_wins: 1893

celo_rating: 1178.7

entertaining: None

stay_in_character: None

user_preference: None

safety_score: 0.97

submission_type: basic

model_architecture: LlamaForCausalLM

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

display_name: hyperblaze-l3-8b-soliloq_5266_v1

double_thumbs_up_ratio: 0.32

feedback_count: 75

ineligible_reason: model is not deployable

language_model: HyperBlaze/L3-8B-soliloquy-lumimaid-spice-mergetestv1

model_score: None

model_size: 8B

reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2

single_thumbs_up_ratio: 0.4266666666666667

thumbs_down_ratio: 0.25333333333333335

thumbs_up_ratio: 0.7466666666666667

us_pacific_date: 2024-05-09

win_ratio: 0.5050693703308431