v000000-l3-8b-blueserpentine

developer_uid: v000000

submission_id: v000000-l3-8b-blueserpentine_v1

model_name: v000000-l3-8b-blueserpentine_v1

model_group: v000000/L3-8B-BlueSerpen

status: torndown

timestamp: 2024-06-13T00:53:17+00:00

num_battles: 41181

num_wins: 22193

celo_rating: 1206.09

family_friendly_score: 0.0

submission_type: basic

model_repo: v000000/L3-8B-BlueSerpentine

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: v000000-l3-8b-blueserpentine_v1

is_internal_developer: False

language_model: v000000/L3-8B-BlueSerpentine

model_size: 8B

ranking_group: single

us_pacific_date: 2024-06-12

win_ratio: 0.5389135766494257

generation_params: {'temperature': 1.22, 'top_p': 0.95, 'min_p': 0.08, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|end_of_text|>', '<|eot_id|>'], 'max_input_tokens': 512, 'best_of': 16, '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: {'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 v000000-l3-8b-blueserpentine-v1-mkmlizer
Waiting for job on v000000-l3-8b-blueserpentine-v1-mkmlizer to finish
v000000-l3-8b-blueserpentine-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║     _____            __           __                                ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║       /___/                                                         ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║                                                                     ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  Version: 0.8.14                                                    ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  https://mk1.ai                                                     ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║                                                                     ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  belonging to:                                                      ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║                                                                     ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  Chai Research Corp.                                                ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ║                                                                     ║
v000000-l3-8b-blueserpentine-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
v000000-l3-8b-blueserpentine-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.
v000000-l3-8b-blueserpentine-v1-mkmlizer:   warnings.warn(warning_message, FutureWarning)
v000000-l3-8b-blueserpentine-v1-mkmlizer: Downloaded to shared memory in 37.362s
v000000-l3-8b-blueserpentine-v1-mkmlizer: quantizing model to /dev/shm/model_cache
v000000-l3-8b-blueserpentine-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Failed to get response for submission blend_junes_2024-06-13: name 'requests' is not defined
v000000-l3-8b-blueserpentine-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/291 [00:04<11:03,  2.30s/it]
Loading 0:   5%|▌         | 15/291 [00:04<01:03,  4.34it/s]
Loading 0:  11%|█▏        | 33/291 [00:04<00:22, 11.54it/s]
Loading 0:  18%|█▊        | 51/291 [00:04<00:11, 20.78it/s]
Loading 0:  22%|██▏       | 65/291 [00:05<00:08, 25.21it/s]
Loading 0:  28%|██▊       | 81/291 [00:05<00:05, 36.18it/s]
Loading 0:  33%|███▎      | 96/291 [00:05<00:04, 47.82it/s]
Loading 0:  39%|███▉      | 114/291 [00:05<00:02, 64.07it/s]
Loading 0:  45%|████▌     | 132/291 [00:05<00:01, 80.47it/s]
Loading 0:  52%|█████▏    | 150/291 [00:05<00:01, 96.34it/s]
Loading 0:  57%|█████▋    | 166/291 [00:06<00:01, 74.52it/s]
Loading 0:  63%|██████▎   | 184/291 [00:06<00:01, 90.67it/s]
Loading 0:  69%|██████▉   | 202/291 [00:06<00:00, 105.89it/s]
Loading 0:  76%|███████▌  | 220/291 [00:06<00:00, 118.32it/s]
Loading 0:  81%|████████  | 235/291 [00:06<00:00, 125.10it/s]
Loading 0:  86%|████████▌ | 250/291 [00:06<00:00, 130.78it/s]
Loading 0:  91%|█████████▏| 266/291 [00:06<00:00, 85.04it/s] 
Loading 0:  98%|█████████▊| 284/291 [00:07<00:00, 101.06it/s]
                                                             
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
v000000-l3-8b-blueserpentine-v1-mkmlizer: quantized model in 23.695s
v000000-l3-8b-blueserpentine-v1-mkmlizer: Processed model v000000/L3-8B-BlueSerpentine in 63.684s
v000000-l3-8b-blueserpentine-v1-mkmlizer: creating bucket guanaco-mkml-models
v000000-l3-8b-blueserpentine-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
v000000-l3-8b-blueserpentine-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1/special_tokens_map.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1/config.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1/tokenizer_config.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1/tokenizer.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/v000000-l3-8b-blueserpentine-v1/flywheel_model.0.safetensors
v000000-l3-8b-blueserpentine-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
v000000-l3-8b-blueserpentine-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.
v000000-l3-8b-blueserpentine-v1-mkmlizer:   warnings.warn(
v000000-l3-8b-blueserpentine-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.
v000000-l3-8b-blueserpentine-v1-mkmlizer:   warnings.warn(
v000000-l3-8b-blueserpentine-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.
v000000-l3-8b-blueserpentine-v1-mkmlizer:   warnings.warn(
v000000-l3-8b-blueserpentine-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()
v000000-l3-8b-blueserpentine-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
v000000-l3-8b-blueserpentine-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
v000000-l3-8b-blueserpentine-v1-mkmlizer: Saving duration: 0.416s
v000000-l3-8b-blueserpentine-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 7.124s
v000000-l3-8b-blueserpentine-v1-mkmlizer: creating bucket guanaco-reward-models
v000000-l3-8b-blueserpentine-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
v000000-l3-8b-blueserpentine-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/config.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/tokenizer_config.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/special_tokens_map.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/merges.txt
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/vocab.json
v000000-l3-8b-blueserpentine-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/v000000-l3-8b-blueserpentine-v1_reward/tokenizer.json
Failed to get response for submission blend_junes_2024-06-13: name 'requests' is not defined
Failed to get response for submission blend_junes_2024-06-13: name 'requests' is not defined
Job v000000-l3-8b-blueserpentine-v1-mkmlizer completed after 93.37s with status: succeeded
Stopping job with name v000000-l3-8b-blueserpentine-v1-mkmlizer
Pipeline stage MKMLizer completed in 96.60s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service v000000-l3-8b-blueserpentine-v1
Waiting for inference service v000000-l3-8b-blueserpentine-v1 to be ready
Inference service v000000-l3-8b-blueserpentine-v1 ready after 42.18260383605957s
Pipeline stage ISVCDeployer completed in 49.49s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1985435485839844s
Received healthy response to inference request in 1.2653264999389648s
Received healthy response to inference request in 1.2528889179229736s
Received healthy response to inference request in 1.2837388515472412s
Received healthy response to inference request in 1.2781012058258057s
5 requests
0 failed requests
5th percentile: 1.255376434326172
10th percentile: 1.25786395072937
20th percentile: 1.2628389835357665
30th percentile: 1.267881441116333
40th percentile: 1.2729913234710692
50th percentile: 1.2781012058258057
60th percentile: 1.28035626411438
70th percentile: 1.282611322402954
80th percentile: 1.46669979095459
90th percentile: 1.8326216697692872
95th percentile: 2.0155826091766356
99th percentile: 2.1619513607025147
mean time: 1.455719804763794
Pipeline stage StressChecker completed in 7.89s
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
v000000-l3-8b-blueserpentine_v1 status is now deployed due to DeploymentManager action
v000000-l3-8b-blueserpentine_v1 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of v000000-l3-8b-blueserpentine_v1
Running pipeline stage ISVCDeleter
Checking if service v000000-l3-8b-blueserpentine-v1 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.18s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key v000000-l3-8b-blueserpentine-v1/config.json from bucket guanaco-mkml-models
Deleting key v000000-l3-8b-blueserpentine-v1/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key v000000-l3-8b-blueserpentine-v1/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key v000000-l3-8b-blueserpentine-v1/tokenizer.json from bucket guanaco-mkml-models
Deleting key v000000-l3-8b-blueserpentine-v1/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key v000000-l3-8b-blueserpentine-v1_reward/config.json from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/merges.txt from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/reward.tensors from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key v000000-l3-8b-blueserpentine-v1_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 6.11s
v000000-l3-8b-blueserpentine_v1 status is now torndown due to DeploymentManager action