submission_id: steelskull-l3-ms-astoria-70b_v1
developer_uid: Steelskull
status: inactive
model_repo: Steelskull/L3-MS-Astoria-70b
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.85, 'top_p': 1.0, 'min_p': 0.1, 'top_k': 90, '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': 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: 2024-05-22T06:23:31+00:00
model_name: steelskull-l3-ms-astoria-70b_v1
model_eval_status: success
model_group: Steelskull/L3-MS-Astoria
num_battles: 97957
num_wins: 50029
celo_rating: 1196.78
safety_score: 0.95
propriety_score: 0.7158384720589294
propriety_total_count: 41677.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 70553706496.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: steelskull-l3-ms-astoria-70b_v1
ineligible_reason: None
language_model: Steelskull/L3-MS-Astoria-70b
model_size: 71B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-21
win_ratio: 0.5107240932245781
Resubmit model
Running pipeline stage MKMLizer
Starting job with name steelskull-l3-ms-astoria-70b-v1-mkmlizer
Waiting for job on steelskull-l3-ms-astoria-70b-v1-mkmlizer to finish
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ _____ __ __ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ /___/ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ Version: 0.8.14 ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ https://mk1.ai ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ The license key for the current software has been verified as ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ belonging to: ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ Chai Research Corp. ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ║ ║
steelskull-l3-ms-astoria-70b-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
steelskull-l3-ms-astoria-70b-v1-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.
steelskull-l3-ms-astoria-70b-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Downloaded to shared memory in 154.413s
steelskull-l3-ms-astoria-70b-v1-mkmlizer: quantizing model to /dev/shm/model_cache
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Loading 0: 0%| | 0/723 [00:00<?, ?it/s] Loading 0: 0%| | 2/723 [00:08<53:35, 4.46s/it] Loading 0: 1%| | 6/723 [00:09<14:07, 1.18s/it] Loading 0: 2%|▏ | 14/723 [00:09<04:36, 2.57it/s] Loading 0: 3%|▎ | 22/723 [00:09<02:23, 4.87it/s] Loading 0: 4%|▎ | 27/723 [00:09<01:43, 6.70it/s] Loading 0: 4%|▍ | 32/723 [00:10<01:40, 6.88it/s] Loading 0: 6%|▌ | 40/723 [00:10<01:02, 10.87it/s] Loading 0: 6%|▌ | 45/723 [00:10<00:49, 13.70it/s] Loading 0: 7%|▋ | 50/723 [00:10<00:39, 17.12it/s] Loading 0: 8%|▊ | 58/723 [00:10<00:27, 24.06it/s] Loading 0: 9%|▉ | 64/723 [00:10<00:23, 28.39it/s] Loading 0: 10%|▉ | 70/723 [00:10<00:20, 31.62it/s] Loading 0: 11%|█ | 77/723 [00:10<00:17, 37.41it/s] Loading 0: 12%|█▏ | 85/723 [00:11<00:28, 22.41it/s] Loading 0: 12%|█▏ | 90/723 [00:11<00:24, 25.52it/s] Loading 0: 13%|█▎ | 96/723 [00:11<00:21, 28.76it/s] Loading 0: 14%|█▍ | 104/723 [00:11<00:17, 36.23it/s] Loading 0: 15%|█▌ | 112/723 [00:24<05:32, 1.84it/s] Loading 0: 16%|█▌ | 117/723 [00:24<04:16, 2.37it/s] Loading 0: 17%|█▋ | 123/723 [00:24<03:05, 3.23it/s] Loading 0: 18%|█▊ | 131/723 [00:24<02:01, 4.86it/s] Loading 0: 19%|█▉ | 137/723 [00:25<01:43, 5.67it/s] Loading 0: 20%|█▉ | 142/723 [00:25<01:21, 7.16it/s] Loading 0: 21%|██ | 149/723 [00:25<00:56, 10.11it/s] Loading 0: 22%|██▏ | 157/723 [00:25<00:39, 14.39it/s] Loading 0: 23%|██▎ | 163/723 [00:25<00:31, 17.94it/s] Loading 0: 23%|██▎ | 169/723 [00:25<00:25, 21.62it/s] Loading 0: 24%|██▍ | 176/723 [00:26<00:19, 27.39it/s] Loading 0: 25%|██▌ | 184/723 [00:26<00:15, 34.62it/s] Loading 0: 26%|██▋ | 190/723 [00:26<00:25, 20.88it/s] Loading 0: 27%|██▋ | 195/723 [00:26<00:22, 23.51it/s] Loading 0: 28%|██▊ | 203/723 [00:27<00:16, 30.77it/s] Loading 0: 29%|██▉ | 211/723 [00:27<00:13, 37.87it/s] Loading 0: 30%|███ | 217/723 [00:27<00:12, 41.00it/s] Loading 0: 31%|███ | 223/723 [00:27<00:11, 42.31it/s] Loading 0: 32%|███▏ | 230/723 [00:27<00:10, 47.40it/s] Loading 0: 33%|███▎ | 238/723 [00:27<00:09, 52.89it/s] Loading 0: 34%|███▎ | 244/723 [00:28<00:19, 24.77it/s] Loading 0: 34%|███▍ | 247/723 [00:40<00:19, 24.77it/s] Loading 0: 34%|███▍ | 248/723 [00:40<05:22, 1.47it/s] Loading 0: 35%|███▌ | 256/723 [00:41<03:23, 2.30it/s] Loading 0: 36%|███▌ | 262/723 [00:41<02:26, 3.14it/s] Loading 0: 37%|███▋ | 268/723 [00:41<01:46, 4.29it/s] Loading 0: 38%|███▊ | 275/723 [00:41<01:12, 6.14it/s] Loading 0: 39%|███▉ | 283/723 [00:41<00:49, 8.98it/s] Loading 0: 40%|███▉ | 289/723 [00:41<00:37, 11.56it/s] Loading 0: 41%|████ | 295/723 [00:42<00:40, 10.69it/s] Loading 0: 42%|████▏ | 302/723 [00:42<00:29, 14.46it/s] Loading 0: 43%|████▎ | 310/723 [00:42<00:20, 19.85it/s] Loading 0: 44%|████▎ | 316/723 [00:42<00:17, 23.83it/s] Loading 0: 45%|████▍ | 322/723 [00:42<00:14, 27.46it/s] Loading 0: 46%|████▌ | 329/723 [00:42<00:11, 33.52it/s] Loading 0: 47%|████▋ | 337/723 [00:43<00:09, 40.79it/s] Loading 0: 47%|████▋ | 343/723 [00:43<00:16, 22.97it/s] Loading 0: 48%|████▊ | 348/723 [00:43<00:14, 25.36it/s] Loading 0: 49%|████▉ | 356/723 [00:43<00:11, 32.86it/s] Loading 0: 50%|█████ | 364/723 [00:43<00:08, 40.01it/s] Loading 0: 51%|█████ | 370/723 [00:44<00:08, 42.93it/s] Loading 0: 52%|█████▏ | 376/723 [00:44<00:07, 43.88it/s] Loading 0: 53%|█████▎ | 383/723 [00:44<00:06, 48.79it/s] Loading 0: 54%|█████▍ | 389/723 [00:57<03:32, 1.57it/s] Loading 0: 54%|█████▍ | 393/723 [00:58<02:56, 1.87it/s] Loading 0: 55%|█████▌ | 401/723 [00:58<01:49, 2.95it/s] Loading 0: 57%|█████▋ | 409/723 [00:58<01:11, 4.40it/s] Loading 0: 57%|█████▋ | 415/723 [00:58<00:52, 5.87it/s] Loading 0: 58%|█████▊ | 421/723 [00:58<00:38, 7.75it/s] Loading 0: 59%|█████▉ | 428/723 [00:58<00:27, 10.76it/s] Loading 0: 60%|██████ | 436/723 [00:59<00:18, 15.15it/s] Loading 0: 61%|██████ | 442/723 [00:59<00:14, 18.74it/s] Loading 0: 62%|██████▏ | 448/723 [00:59<00:17, 15.39it/s] Loading 0: 63%|██████▎ | 455/723 [00:59<00:13, 20.22it/s] Loading 0: 64%|██████▍ | 463/723 [00:59<00:09, 26.65it/s] Loading 0: 65%|██████▍ | 469/723 [01:00<00:08, 30.76it/s] Loading 0: 66%|██████▌ | 475/723 [01:00<00:07, 33.88it/s] Loading 0: 67%|██████▋ | 482/723 [01:00<00:06, 39.70it/s] Loading 0: 68%|██████▊ | 490/723 [01:00<00:04, 46.89it/s] Loading 0: 69%|██████▊ | 497/723 [01:00<00:04, 50.45it/s] Loading 0: 70%|██████▉ | 504/723 [01:01<00:08, 25.14it/s] Loading 0: 71%|███████ | 510/723 [01:01<00:07, 28.34it/s] Loading 0: 72%|███████▏ | 518/723 [01:01<00:05, 35.38it/s] Loading 0: 72%|███████▏ | 523/723 [01:14<00:05, 35.38it/s] Loading 0: 72%|███████▏ | 524/723 [01:14<01:58, 1.68it/s] Loading 0: 73%|███████▎ | 528/723 [01:14<01:33, 2.08it/s] Loading 0: 74%|███████▍ | 536/723 [01:14<00:57, 3.23it/s] Loading 0: 75%|███████▌ | 544/723 [01:14<00:37, 4.80it/s] Loading 0: 76%|███████▌ | 550/723 [01:15<00:31, 5.57it/s] Loading 0: 77%|███████▋ | 555/723 [01:15<00:24, 7.00it/s] Loading 0: 78%|███████▊ | 563/723 [01:15<00:15, 10.25it/s] Loading 0: 79%|███████▉ | 571/723 [01:15<00:10, 14.31it/s] Loading 0: 80%|███████▉ | 577/723 [01:15<00:08, 17.58it/s] Loading 0: 81%|████████ | 583/723 [01:15<00:06, 20.86it/s] Loading 0: 82%|████████▏ | 590/723 [01:16<00:05, 26.10it/s] Loading 0: 83%|████████▎ | 598/723 [01:16<00:03, 32.71it/s] Loading 0: 84%|████████▎ | 604/723 [01:16<00:06, 18.65it/s] Loading 0: 84%|████████▍ | 609/723 [01:17<00:05, 20.96it/s] Loading 0: 85%|████████▌ | 617/723 [01:17<00:03, 27.60it/s] Loading 0: 86%|████████▋ | 625/723 [01:17<00:02, 34.16it/s] Loading 0: 87%|████████▋ | 631/723 [01:17<00:02, 37.16it/s] Loading 0: 88%|████████▊ | 637/723 [01:17<00:02, 38.42it/s] Loading 0: 89%|████████▉ | 644/723 [01:17<00:01, 43.13it/s] Loading 0: 90%|█████████ | 652/723 [01:17<00:01, 48.70it/s] Loading 0: 91%|█████████ | 658/723 [01:18<00:02, 21.87it/s] Loading 0: 92%|█████████▏| 663/723 [01:30<00:36, 1.66it/s] Loading 0: 93%|█████████▎| 671/723 [01:30<00:20, 2.54it/s] Loading 0: 94%|█████████▍| 679/723 [01:30<00:11, 3.76it/s] Loading 0: 95%|█████████▍| 685/723 [01:30<00:07, 4.98it/s] Loading 0: 96%|█████████▌| 691/723 [01:30<00:04, 6.57it/s] Loading 0: 97%|█████████▋| 698/723 [01:30<00:02, 9.11it/s] Loading 0: 98%|█████████▊| 706/723 [01:31<00:01, 9.70it/s] Loading 0: 98%|█████████▊| 711/723 [01:31<00:01, 11.86it/s] Loading 0: 99%|█████████▉| 716/723 [01:31<00:00, 14.57it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
steelskull-l3-ms-astoria-70b-v1-mkmlizer: quantized model in 106.966s
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Processed model Steelskull/L3-MS-Astoria-70b in 270.320s
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
steelskull-l3-ms-astoria-70b-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/config.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/tokenizer_config.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/tokenizer.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/special_tokens_map.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.5.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.5.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.3.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.3.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.0.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.4.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.4.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.1.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.2.safetensors s3://guanaco-mkml-models/steelskull-l3-ms-astoria-70b-v1/flywheel_model.2.safetensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
steelskull-l3-ms-astoria-70b-v1-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.
steelskull-l3-ms-astoria-70b-v1-mkmlizer: warnings.warn(
steelskull-l3-ms-astoria-70b-v1-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.
steelskull-l3-ms-astoria-70b-v1-mkmlizer: warnings.warn(
steelskull-l3-ms-astoria-70b-v1-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.
steelskull-l3-ms-astoria-70b-v1-mkmlizer: warnings.warn(
steelskull-l3-ms-astoria-70b-v1-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()
steelskull-l3-ms-astoria-70b-v1-mkmlizer: return self.fget.__get__(instance, owner)()
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Saving duration: 0.283s
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.043s
steelskull-l3-ms-astoria-70b-v1-mkmlizer: creating bucket guanaco-reward-models
steelskull-l3-ms-astoria-70b-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
steelskull-l3-ms-astoria-70b-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/config.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/special_tokens_map.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/tokenizer_config.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/vocab.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/merges.txt
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/tokenizer.json
steelskull-l3-ms-astoria-70b-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/steelskull-l3-ms-astoria-70b-v1_reward/reward.tensors
Job steelskull-l3-ms-astoria-70b-v1-mkmlizer completed after 321.67s with status: succeeded
Stopping job with name steelskull-l3-ms-astoria-70b-v1-mkmlizer
Pipeline stage MKMLizer completed in 325.22s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service steelskull-l3-ms-astoria-70b-v1
Waiting for inference service steelskull-l3-ms-astoria-70b-v1 to be ready
Inference service steelskull-l3-ms-astoria-70b-v1 ready after 130.60918426513672s
Pipeline stage ISVCDeployer completed in 137.73s
Running pipeline stage StressChecker
Received healthy response to inference request in 5.560669422149658s
Received healthy response to inference request in 4.248526334762573s
Received healthy response to inference request in 4.26397705078125s
Received healthy response to inference request in 4.2216269969940186s
Received healthy response to inference request in 4.200321435928345s
5 requests
0 failed requests
5th percentile: 4.204582548141479
10th percentile: 4.208843660354614
20th percentile: 4.217365884780884
30th percentile: 4.22700686454773
40th percentile: 4.237766599655151
50th percentile: 4.248526334762573
60th percentile: 4.254706621170044
70th percentile: 4.260886907577515
80th percentile: 4.523315525054932
90th percentile: 5.041992473602295
95th percentile: 5.301330947875976
99th percentile: 5.508801727294922
mean time: 4.499024248123169
Pipeline stage StressChecker completed in 23.26s
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 0.03s
M-Eval Dataset for topic stay_in_character is loaded
steelskull-l3-ms-astoria-70b_v1 status is now deployed due to DeploymentManager action
steelskull-l3-ms-astoria-70b_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics