submission_id: function_dikes_2024-08-09
developer_uid: chai_backend_admin
alignment_samples: 137
alignment_score: 6.391515445387325
celo_rating: 1095.6
display_name: gpt-4o
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}
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': 1, 'max_output_tokens': 64, 'reward_max_token_input': 256}
is_internal_developer: True
model_group:
model_name: gpt-4o
num_battles: 1442
num_wins: 480
propriety_score: 0.7099236641221374
propriety_total_count: 131.0
ranking_group: single
reward_repo: ChaiML/gpt2_xl_pairwise_89m_step_347634
status: torndown
submission_type: function
timestamp: 2024-08-09T18:04:23+00:00
us_pacific_date: 2024-08-09
win_ratio: 0.332871012482663
Download Preference Data
Resubmit model
Running pipeline stage StressChecker
Received healthy response to inference request in 0.6179914474487305s
Received healthy response to inference request in 0.5799520015716553s
Received healthy response to inference request in 0.8843789100646973s
Received healthy response to inference request in 0.7411012649536133s
Received healthy response to inference request in 1.571326494216919s
5 requests
0 failed requests
5th percentile: 0.5875598907470703
10th percentile: 0.5951677799224854
20th percentile: 0.6103835582733155
30th percentile: 0.642613410949707
40th percentile: 0.6918573379516602
50th percentile: 0.7411012649536133
60th percentile: 0.7984123229980469
70th percentile: 0.8557233810424805
80th percentile: 1.0217684268951417
90th percentile: 1.2965474605560303
95th percentile: 1.4339369773864745
99th percentile: 1.5438485908508301
mean time: 0.8789500236511231
Pipeline stage StressChecker completed in 4.94s
function_dikes_2024-08-09 status is now deployed due to DeploymentManager action
function_dikes_2024-08-09 status is now deployed due to admin request
function_dikes_2024-08-09 status is now torndown due to DeploymentManager action

Usage Metrics

Latency Metrics