submission_id: meseca-sophon-1-5_v3
developer_uid: nguyenzzz
status: inactive
model_repo: meseca/sophon-1.5
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.95, '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': 4, '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': True}
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: 2024-05-19T14:26:28+00:00
model_name: meseca-sophon-1_v1
model_eval_status: success
model_group: meseca/sophon-1.5
num_battles: 22230
num_wins: 11322
celo_rating: 1176.68
safety_score: 0.97
propriety_score: 0.0
propriety_total_count: 0.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 11520053248.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: meseca-sophon-1_v1
ineligible_reason: propriety_total_count < 5000
language_model: meseca/sophon-1.5
model_size: 12B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-19
win_ratio: 0.5093117408906883
Resubmit model
Running pipeline stage MKMLizer
Starting job with name meseca-sophon-1-5-v3-mkmlizer
Waiting for job on meseca-sophon-1-5-v3-mkmlizer to finish
meseca-sophon-1-5-v3-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
meseca-sophon-1-5-v3-mkmlizer: ║ _____ __ __ ║
meseca-sophon-1-5-v3-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
meseca-sophon-1-5-v3-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
meseca-sophon-1-5-v3-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
meseca-sophon-1-5-v3-mkmlizer: ║ /___/ ║
meseca-sophon-1-5-v3-mkmlizer: ║ ║
meseca-sophon-1-5-v3-mkmlizer: ║ Version: 0.8.14 ║
meseca-sophon-1-5-v3-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
meseca-sophon-1-5-v3-mkmlizer: ║ https://mk1.ai ║
meseca-sophon-1-5-v3-mkmlizer: ║ ║
meseca-sophon-1-5-v3-mkmlizer: ║ The license key for the current software has been verified as ║
meseca-sophon-1-5-v3-mkmlizer: ║ belonging to: ║
meseca-sophon-1-5-v3-mkmlizer: ║ ║
meseca-sophon-1-5-v3-mkmlizer: ║ Chai Research Corp. ║
meseca-sophon-1-5-v3-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
meseca-sophon-1-5-v3-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
meseca-sophon-1-5-v3-mkmlizer: ║ ║
meseca-sophon-1-5-v3-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
meseca-sophon-1-5-v3-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.
meseca-sophon-1-5-v3-mkmlizer: warnings.warn(warning_message, FutureWarning)
meseca-sophon-1-5-v3-mkmlizer: Downloaded to shared memory in 41.747s
meseca-sophon-1-5-v3-mkmlizer: quantizing model to /dev/shm/model_cache
meseca-sophon-1-5-v3-mkmlizer: Saving flywheel model at /dev/shm/model_cache
meseca-sophon-1-5-v3-mkmlizer: Loading 0: 0%| | 0/435 [00:00<?, ?it/s] Loading 0: 0%| | 2/435 [00:04<15:26, 2.14s/it] Loading 0: 5%|▌ | 22/435 [00:04<00:59, 6.91it/s] Loading 0: 10%|▉ | 43/435 [00:04<00:24, 15.79it/s] Loading 0: 14%|█▍ | 61/435 [00:04<00:17, 21.88it/s] Loading 0: 18%|█▊ | 78/435 [00:04<00:11, 31.50it/s] Loading 0: 24%|██▎ | 103/435 [00:05<00:06, 50.03it/s] Loading 0: 28%|██▊ | 123/435 [00:05<00:04, 65.70it/s] Loading 0: 33%|███▎ | 143/435 [00:05<00:03, 83.10it/s] Loading 0: 37%|███▋ | 161/435 [00:05<00:03, 72.86it/s] Loading 0: 41%|████▏ | 180/435 [00:05<00:02, 89.58it/s] Loading 0: 45%|████▌ | 196/435 [00:05<00:02, 95.30it/s] Loading 0: 49%|████▉ | 213/435 [00:05<00:02, 106.94it/s] Loading 0: 54%|█████▍ | 234/435 [00:06<00:01, 127.33it/s] Loading 0: 59%|█████▉ | 256/435 [00:06<00:01, 144.88it/s] Loading 0: 63%|██████▎ | 274/435 [00:06<00:01, 95.85it/s] Loading 0: 68%|██████▊ | 294/435 [00:06<00:01, 112.82it/s] Loading 0: 73%|███████▎ | 319/435 [00:06<00:00, 138.59it/s] Loading 0: 78%|███████▊ | 338/435 [00:06<00:00, 147.67it/s] Loading 0: 82%|████████▏ | 356/435 [00:06<00:00, 152.43it/s] Loading 0: 86%|████████▌ | 374/435 [00:07<00:00, 98.63it/s] Loading 0: 91%|█████████ | 395/435 [00:07<00:00, 118.25it/s] Loading 0: 96%|█████████▌| 416/435 [00:07<00:00, 136.90it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
meseca-sophon-1-5-v3-mkmlizer: quantized model in 21.165s
meseca-sophon-1-5-v3-mkmlizer: Processed model meseca/sophon-1.5 in 64.248s
meseca-sophon-1-5-v3-mkmlizer: creating bucket guanaco-mkml-models
meseca-sophon-1-5-v3-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
meseca-sophon-1-5-v3-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/meseca-sophon-1-5-v3
meseca-sophon-1-5-v3-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/meseca-sophon-1-5-v3/config.json
meseca-sophon-1-5-v3-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/meseca-sophon-1-5-v3/special_tokens_map.json
meseca-sophon-1-5-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/meseca-sophon-1-5-v3/tokenizer_config.json
meseca-sophon-1-5-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/meseca-sophon-1-5-v3/tokenizer.json
meseca-sophon-1-5-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/meseca-sophon-1-5-v3/flywheel_model.0.safetensors
meseca-sophon-1-5-v3-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
meseca-sophon-1-5-v3-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.
meseca-sophon-1-5-v3-mkmlizer: warnings.warn(
meseca-sophon-1-5-v3-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.
meseca-sophon-1-5-v3-mkmlizer: warnings.warn(
meseca-sophon-1-5-v3-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.
meseca-sophon-1-5-v3-mkmlizer: warnings.warn(
meseca-sophon-1-5-v3-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()
meseca-sophon-1-5-v3-mkmlizer: return self.fget.__get__(instance, owner)()
meseca-sophon-1-5-v3-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
meseca-sophon-1-5-v3-mkmlizer: Saving duration: 0.234s
meseca-sophon-1-5-v3-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.553s
meseca-sophon-1-5-v3-mkmlizer: creating bucket guanaco-reward-models
meseca-sophon-1-5-v3-mkmlizer: Bucket 's3://guanaco-reward-models/' created
meseca-sophon-1-5-v3-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/config.json
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/special_tokens_map.json
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/merges.txt
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/vocab.json
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/tokenizer_config.json
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/tokenizer.json
meseca-sophon-1-5-v3-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/meseca-sophon-1-5-v3_reward/reward.tensors
Job meseca-sophon-1-5-v3-mkmlizer completed after 93.93s with status: succeeded
Stopping job with name meseca-sophon-1-5-v3-mkmlizer
Pipeline stage MKMLizer completed in 96.53s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service meseca-sophon-1-5-v3
Waiting for inference service meseca-sophon-1-5-v3 to be ready
Inference service meseca-sophon-1-5-v3 ready after 100.6625988483429s
Pipeline stage ISVCDeployer completed in 107.99s
Running pipeline stage StressChecker
%s, retrying in %s seconds...
%s, retrying in %s seconds...
Received healthy response to inference request in 19.082892894744873s
Received healthy response to inference request in 1.5435459613800049s
Received healthy response to inference request in 1.5239152908325195s
Received healthy response to inference request in 1.5689446926116943s
Received healthy response to inference request in 1.5515336990356445s
5 requests
0 failed requests
5th percentile: 1.5278414249420167
10th percentile: 1.5317675590515136
20th percentile: 1.5396198272705077
30th percentile: 1.5451435089111327
40th percentile: 1.5483386039733886
50th percentile: 1.5515336990356445
60th percentile: 1.5584980964660644
70th percentile: 1.5654624938964843
80th percentile: 5.071734333038333
90th percentile: 12.077313613891604
95th percentile: 15.580103254318235
99th percentile: 18.382334966659545
mean time: 5.054166507720947
Pipeline stage StressChecker completed in 66.57s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running M-Eval for topic stay_in_character
Running pipeline stage DaemonicSafetyScorer
M-Eval Dataset for topic stay_in_character is loaded
Pipeline stage DaemonicSafetyScorer completed in 0.06s
meseca-sophon-1-5_v3 status is now deployed due to DeploymentManager action
meseca-sophon-1-5_v3 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics