failspy-meta-llama-3-8b-_2723

submission_id: failspy-meta-llama-3-8b-_2723_v1

developer_uid: Bbbrun0

best_of: 16

celo_rating: 1163.82

display_name: test_failspy_abliterated-v3

family_friendly_score: 0.0

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': 16, 'max_output_tokens': 64}

is_internal_developer: False

language_model: failspy/Meta-Llama-3-8B-Instruct-abliterated-v3

max_input_tokens: 512

max_output_tokens: 64

model_architecture: LlamaForCausalLM

model_group: failspy/Meta-Llama-3-8B-

model_name: test_failspy_abliterated-v3

model_num_parameters: 8030261248.0

model_repo: failspy/Meta-Llama-3-8B-Instruct-abliterated-v3

model_size: 8B

num_battles: 14704

num_wins: 6965

ranking_group: single

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'}

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

status: torndown

submission_type: basic

timestamp: 2024-07-03T02:45:49+00:00

us_pacific_date: 2024-07-02

win_ratio: 0.47368063112078346

Resubmit model

Running pipeline stage MKMLizer
Starting job with name failspy-meta-llama-3-8b-2723-v1-mkmlizer
Waiting for job on failspy-meta-llama-3-8b-2723-v1-mkmlizer to finish
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║     _____            __           __                                ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║       /___/                                                         ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║                                                                     ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  Version: 0.8.14                                                    ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  https://mk1.ai                                                     ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║                                                                     ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  belonging to:                                                      ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║                                                                     ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  Chai Research Corp.                                                ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ║                                                                     ║
failspy-meta-llama-3-8b-2723-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Downloaded to shared memory in 37.525s
failspy-meta-llama-3-8b-2723-v1-mkmlizer: quantizing model to /dev/shm/model_cache
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
failspy-meta-llama-3-8b-2723-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   5%|▍         | 14/291 [00:00<00:02, 130.45it/s]
Loading 0:  11%|█         | 31/291 [00:00<00:01, 152.64it/s]
Loading 0:  16%|█▋        | 48/291 [00:00<00:01, 157.79it/s]
Loading 0:  22%|██▏       | 64/291 [00:00<00:01, 156.68it/s]
Loading 0:  27%|██▋       | 80/291 [00:00<00:01, 151.42it/s]
Loading 0:  33%|███▎      | 96/291 [00:00<00:02, 79.08it/s] 
Loading 0:  38%|███▊      | 112/291 [00:01<00:01, 94.39it/s]
Loading 0:  44%|████▍     | 129/291 [00:01<00:01, 110.22it/s]
Loading 0:  49%|████▉     | 144/291 [00:01<00:01, 118.97it/s]
Loading 0:  55%|█████▍    | 159/291 [00:01<00:01, 123.68it/s]
Loading 0:  62%|██████▏   | 179/291 [00:01<00:00, 138.49it/s]
Loading 0:  67%|██████▋   | 195/291 [00:01<00:01, 83.99it/s] 
Loading 0:  73%|███████▎  | 211/291 [00:01<00:00, 97.45it/s]
Loading 0:  78%|███████▊  | 228/291 [00:02<00:00, 111.96it/s]
Loading 0:  84%|████████▎ | 243/291 [00:02<00:00, 119.86it/s]
Loading 0:  89%|████████▊ | 258/291 [00:02<00:00, 123.18it/s]
Loading 0:  94%|█████████▍| 274/291 [00:02<00:00, 131.83it/s]
Loading 0:  99%|█████████▉| 289/291 [00:07<00:00,  8.80it/s] 
                                                            
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
failspy-meta-llama-3-8b-2723-v1-mkmlizer: quantized model in 23.625s
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Processed model failspy/Meta-Llama-3-8B-Instruct-abliterated-v3 in 61.150s
failspy-meta-llama-3-8b-2723-v1-mkmlizer: creating bucket guanaco-mkml-models
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
failspy-meta-llama-3-8b-2723-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1/special_tokens_map.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1/config.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1/tokenizer_config.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1/tokenizer.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/failspy-meta-llama-3-8b-2723-v1/flywheel_model.0.safetensors
failspy-meta-llama-3-8b-2723-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
failspy-meta-llama-3-8b-2723-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:919: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
failspy-meta-llama-3-8b-2723-v1-mkmlizer:   warnings.warn(
failspy-meta-llama-3-8b-2723-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
failspy-meta-llama-3-8b-2723-v1-mkmlizer:   warnings.warn(
failspy-meta-llama-3-8b-2723-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:769: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
failspy-meta-llama-3-8b-2723-v1-mkmlizer:   warnings.warn(
failspy-meta-llama-3-8b-2723-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.
failspy-meta-llama-3-8b-2723-v1-mkmlizer:   warnings.warn(
failspy-meta-llama-3-8b-2723-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()
failspy-meta-llama-3-8b-2723-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Saving duration: 0.404s
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.691s
failspy-meta-llama-3-8b-2723-v1-mkmlizer: creating bucket guanaco-reward-models
failspy-meta-llama-3-8b-2723-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
failspy-meta-llama-3-8b-2723-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/special_tokens_map.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/tokenizer_config.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/config.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/merges.txt
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/vocab.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/tokenizer.json
failspy-meta-llama-3-8b-2723-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/failspy-meta-llama-3-8b-2723-v1_reward/reward.tensors
Job failspy-meta-llama-3-8b-2723-v1-mkmlizer completed after 83.87s with status: succeeded
Stopping job with name failspy-meta-llama-3-8b-2723-v1-mkmlizer
Pipeline stage MKMLizer completed in 84.85s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service failspy-meta-llama-3-8b-2723-v1
Waiting for inference service failspy-meta-llama-3-8b-2723-v1 to be ready
Inference service failspy-meta-llama-3-8b-2723-v1 ready after 40.23038697242737s
Pipeline stage ISVCDeployer completed in 47.40s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.0715696811676025s
Received healthy response to inference request in 1.2726836204528809s
Received healthy response to inference request in 1.243936538696289s
Received healthy response to inference request in 1.2302806377410889s
Received healthy response to inference request in 1.2220358848571777s
5 requests
0 failed requests
5th percentile: 1.22368483543396
10th percentile: 1.2253337860107423
20th percentile: 1.2286316871643066
30th percentile: 1.233011817932129
40th percentile: 1.238474178314209
50th percentile: 1.243936538696289
60th percentile: 1.2554353713989257
70th percentile: 1.2669342041015625
80th percentile: 1.4324608325958255
90th percentile: 1.752015256881714
95th percentile: 1.911792469024658
99th percentile: 2.0396142387390137
mean time: 1.4081012725830078
Pipeline stage StressChecker completed in 7.72s
failspy-meta-llama-3-8b-_2723_v1 status is now deployed due to DeploymentManager action
failspy-meta-llama-3-8b-_2723_v1 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of failspy-meta-llama-3-8b-_2723_v1
Running pipeline stage ISVCDeleter
Checking if service failspy-meta-llama-3-8b-2723-v1 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.56s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key failspy-meta-llama-3-8b-2723-v1/config.json from bucket guanaco-mkml-models
Deleting key failspy-meta-llama-3-8b-2723-v1/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key failspy-meta-llama-3-8b-2723-v1/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key failspy-meta-llama-3-8b-2723-v1/tokenizer.json from bucket guanaco-mkml-models
Deleting key failspy-meta-llama-3-8b-2723-v1/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/config.json from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/merges.txt from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/reward.tensors from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key failspy-meta-llama-3-8b-2723-v1_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.88s
failspy-meta-llama-3-8b-_2723_v1 status is now torndown due to DeploymentManager action