nitral-ai-pp-1-3-private-test

developer_uid: Nitral-AI

submission_id: nitral-ai-pp-1-3-private-test_v1

model_name: nitral-ai-pp-1-3-private-test_v1

model_group: Nitral-AI/PP_1.3-Private

status: torndown

timestamp: 2024-05-31T01:49:44+00:00

num_battles: 6903

num_wins: 3714

celo_rating: 1212.44

family_friendly_score: 0.0

submission_type: basic

model_repo: Nitral-AI/PP_1.3-Private-Test

model_architecture: LlamaForCausalLM

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

model_num_parameters: 8030261248.0

best_of: 16

max_input_tokens: 512

max_output_tokens: 64

display_name: nitral-ai-pp-1-3-private-test_v1

is_internal_developer: False

language_model: Nitral-AI/PP_1.3-Private-Test

model_size: 8B

ranking_group: single

us_pacific_date: 2024-05-30

win_ratio: 0.5380269448066058

generation_params: {'temperature': 0.95, 'top_p': 1.0, 'min_p': 0.08, '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': "<|start_header_id|>system<|end_header_id|>\n\n{bot_name}'s Persona: {memory}\n\n", 'prompt_template': '{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: {'bot_template': '{bot_name}: {message}\n', 'memory_template': "{bot_name}'s Persona: {memory}\n####\n", 'prompt_template': '{prompt}\n<START>\n', 'response_template': '{bot_name}:', 'truncate_by_message': False, 'user_template': '{user_name}: {message}\n'}

model_eval_status: success

Resubmit model

Running pipeline stage MKMLizer
Starting job with name nitral-ai-pp-1-3-private-test-v1-mkmlizer
Waiting for job on nitral-ai-pp-1-3-private-test-v1-mkmlizer to finish
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║     _____            __           __                                ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║       /___/                                                         ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║                                                                     ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  Version: 0.8.14                                                    ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  https://mk1.ai                                                     ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║                                                                     ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  belonging to:                                                      ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║                                                                     ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  Chai Research Corp.                                                ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ║                                                                     ║
nitral-ai-pp-1-3-private-test-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
nitral-ai-pp-1-3-private-test-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.
nitral-ai-pp-1-3-private-test-v1-mkmlizer:   warnings.warn(warning_message, FutureWarning)
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Downloaded to shared memory in 39.790s
nitral-ai-pp-1-3-private-test-v1-mkmlizer: quantizing model to /dev/shm/model_cache
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
nitral-ai-pp-1-3-private-test-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/291 [00:04<10:51,  2.25s/it]
Loading 0:   6%|▌         | 18/291 [00:04<00:51,  5.34it/s]
Loading 0:  12%|█▏        | 36/291 [00:04<00:20, 12.67it/s]
Loading 0:  19%|█▊        | 54/291 [00:04<00:10, 22.10it/s]
Loading 0:  24%|██▎       | 69/291 [00:05<00:08, 26.64it/s]
Loading 0:  30%|██▉       | 87/291 [00:05<00:05, 39.16it/s]
Loading 0:  36%|███▌      | 105/291 [00:05<00:03, 53.77it/s]
Loading 0:  42%|████▏     | 123/291 [00:05<00:02, 69.60it/s]
Loading 0:  48%|████▊     | 141/291 [00:05<00:01, 86.18it/s]
Loading 0:  55%|█████▍    | 159/291 [00:05<00:01, 102.42it/s]
Loading 0:  60%|██████    | 176/291 [00:06<00:01, 78.70it/s] 
Loading 0:  67%|██████▋   | 194/291 [00:06<00:01, 95.09it/s]
Loading 0:  73%|███████▎  | 212/291 [00:06<00:00, 110.36it/s]
Loading 0:  79%|███████▉  | 230/291 [00:06<00:00, 123.99it/s]
Loading 0:  85%|████████▌ | 248/291 [00:06<00:00, 136.33it/s]
Loading 0:  91%|█████████▏| 266/291 [00:06<00:00, 92.72it/s] 
Loading 0:  98%|█████████▊| 284/291 [00:06<00:00, 107.22it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
nitral-ai-pp-1-3-private-test-v1-mkmlizer: quantized model in 22.861s
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Processed model Nitral-AI/PP_1.3-Private-Test in 65.162s
nitral-ai-pp-1-3-private-test-v1-mkmlizer: creating bucket guanaco-mkml-models
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
nitral-ai-pp-1-3-private-test-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1/config.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1/special_tokens_map.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1/tokenizer_config.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1/tokenizer.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/nitral-ai-pp-1-3-private-test-v1/flywheel_model.0.safetensors
nitral-ai-pp-1-3-private-test-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
nitral-ai-pp-1-3-private-test-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.
nitral-ai-pp-1-3-private-test-v1-mkmlizer:   warnings.warn(
nitral-ai-pp-1-3-private-test-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.
nitral-ai-pp-1-3-private-test-v1-mkmlizer:   warnings.warn(
nitral-ai-pp-1-3-private-test-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.
nitral-ai-pp-1-3-private-test-v1-mkmlizer:   warnings.warn(
nitral-ai-pp-1-3-private-test-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()
nitral-ai-pp-1-3-private-test-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Saving duration: 0.386s
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.453s
nitral-ai-pp-1-3-private-test-v1-mkmlizer: creating bucket guanaco-reward-models
nitral-ai-pp-1-3-private-test-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
nitral-ai-pp-1-3-private-test-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/tokenizer_config.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/config.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/special_tokens_map.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/merges.txt
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/vocab.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/tokenizer.json
nitral-ai-pp-1-3-private-test-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/nitral-ai-pp-1-3-private-test-v1_reward/reward.tensors
Job nitral-ai-pp-1-3-private-test-v1-mkmlizer completed after 93.52s with status: succeeded
Stopping job with name nitral-ai-pp-1-3-private-test-v1-mkmlizer
Pipeline stage MKMLizer completed in 94.40s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service nitral-ai-pp-1-3-private-test-v1
Waiting for inference service nitral-ai-pp-1-3-private-test-v1 to be ready
Inference service nitral-ai-pp-1-3-private-test-v1 ready after 41.110456466674805s
Pipeline stage ISVCDeployer completed in 46.89s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.028916358947754s
Received healthy response to inference request in 1.3532099723815918s
Received healthy response to inference request in 1.3204710483551025s
Received healthy response to inference request in 1.290092945098877s
Received healthy response to inference request in 1.2455673217773438s
5 requests
0 failed requests
5th percentile: 1.2544724464416503
10th percentile: 1.263377571105957
20th percentile: 1.2811878204345704
30th percentile: 1.296168565750122
40th percentile: 1.3083198070526123
50th percentile: 1.3204710483551025
60th percentile: 1.3335666179656982
70th percentile: 1.3466621875762939
80th percentile: 1.4883512496948244
90th percentile: 1.7586338043212892
95th percentile: 1.8937750816345214
99th percentile: 2.0018881034851073
mean time: 1.4476515293121337
Pipeline stage StressChecker completed in 7.84s
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 0.04s
M-Eval Dataset for topic stay_in_character is loaded
nitral-ai-pp-1-3-private-test_v1 status is now deployed due to DeploymentManager action
nitral-ai-pp-1-3-private-test_v1 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of nitral-ai-pp-1-3-private-test_v1
Running pipeline stage ISVCDeleter
Checking if service nitral-ai-pp-1-3-private-test-v1 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 3.89s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key nitral-ai-pp-1-3-private-test-v1/config.json from bucket guanaco-mkml-models
Deleting key nitral-ai-pp-1-3-private-test-v1/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key nitral-ai-pp-1-3-private-test-v1/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key nitral-ai-pp-1-3-private-test-v1/tokenizer.json from bucket guanaco-mkml-models
Deleting key nitral-ai-pp-1-3-private-test-v1/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/config.json from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/merges.txt from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/reward.tensors from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key nitral-ai-pp-1-3-private-test-v1_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.81s
nitral-ai-pp-1-3-private-test_v1 status is now torndown due to DeploymentManager action