r136a1-moon-l3-8b

developer_uid: R136a1

submission_id: r136a1-moon-l3-8b_v4

model_name: r136a1-moon-l3-8b_v4

model_group: R136a1/Moon-L3-8B

status: torndown

timestamp: 2024-07-11T10:57:31+00:00

num_battles: 34251

num_wins: 18015

celo_rating: 1207.53

family_friendly_score: 0.0

submission_type: basic

model_repo: R136a1/Moon-L3-8B

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: r136a1-moon-l3-8b_v4

is_internal_developer: False

language_model: R136a1/Moon-L3-8B

model_size: 8B

ranking_group: single

us_pacific_date: 2024-07-11

win_ratio: 0.5259700446702286

generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.1, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|eot_id|>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}

formatter: {'memory_template': "<|begin_of_text|><|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'}

Resubmit model

Running pipeline stage MKMLizer
Starting job with name r136a1-moon-l3-8b-v4-mkmlizer
Waiting for job on r136a1-moon-l3-8b-v4-mkmlizer to finish
r136a1-moon-l3-8b-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
r136a1-moon-l3-8b-v4-mkmlizer: ║     _____            __           __                                ║
r136a1-moon-l3-8b-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
r136a1-moon-l3-8b-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
r136a1-moon-l3-8b-v4-mkmlizer: ║       /___/                                                         ║
r136a1-moon-l3-8b-v4-mkmlizer: ║                                                                     ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  Version: 0.8.14                                                    ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  https://mk1.ai                                                     ║
r136a1-moon-l3-8b-v4-mkmlizer: ║                                                                     ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  belonging to:                                                      ║
r136a1-moon-l3-8b-v4-mkmlizer: ║                                                                     ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  Chai Research Corp.                                                ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
r136a1-moon-l3-8b-v4-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
r136a1-moon-l3-8b-v4-mkmlizer: ║                                                                     ║
r136a1-moon-l3-8b-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
r136a1-moon-l3-8b-v4-mkmlizer: Downloaded to shared memory in 35.501s
r136a1-moon-l3-8b-v4-mkmlizer: quantizing model to /dev/shm/model_cache
r136a1-moon-l3-8b-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Failed to get response for submission blend_pefis_2024-07-04: ('http://mistralai-mixtral-8x7b-3473-v33-predictor-default.tenant-chaiml-guanaco.knative.ord1.coreweave.cloud/v1/models/GPT-J-6B-lit-v2:predict', '{"error":"TypeError : SamplingParameters.__init__() got an unexpected keyword argument \'min_p\'"}')
Failed to get response for submission blend_pefis_2024-07-04: ('http://mistralai-mixtral-8x7b-3473-v33-predictor-default.tenant-chaiml-guanaco.knative.ord1.coreweave.cloud/v1/models/GPT-J-6B-lit-v2:predict', '{"error":"TypeError : SamplingParameters.__init__() got an unexpected keyword argument \'min_p\'"}')
r136a1-moon-l3-8b-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 2/291 [00:05<13:40,  2.84s/it]
Loading 0:   5%|▍         | 14/291 [00:05<01:24,  3.28it/s]
Loading 0:  10%|▉         | 28/291 [00:05<00:33,  7.90it/s]
Loading 0:  14%|█▍        | 41/291 [00:05<00:18, 13.48it/s]
Loading 0:  19%|█▉        | 55/291 [00:06<00:11, 21.26it/s]
Loading 0:  23%|██▎       | 67/291 [00:06<00:09, 22.62it/s]
Loading 0:  27%|██▋       | 78/291 [00:06<00:07, 29.63it/s]
Loading 0:  32%|███▏      | 94/291 [00:06<00:04, 43.21it/s]
Loading 0:  36%|███▋      | 106/291 [00:06<00:03, 52.45it/s]
Loading 0:  42%|████▏     | 121/291 [00:06<00:02, 66.57it/s]
Loading 0:  46%|████▌     | 133/291 [00:07<00:02, 75.02it/s]
Loading 0:  51%|█████     | 148/291 [00:07<00:01, 89.32it/s]
Loading 0:  55%|█████▌    | 161/291 [00:07<00:01, 96.34it/s]
Loading 0:  60%|█████▉    | 174/291 [00:07<00:02, 57.98it/s]
Loading 0:  64%|██████▎   | 185/291 [00:07<00:01, 65.24it/s]
Loading 0:  68%|██████▊   | 199/291 [00:07<00:01, 78.31it/s]
Loading 0:  73%|███████▎  | 212/291 [00:08<00:00, 87.69it/s]
Loading 0:  78%|███████▊  | 226/291 [00:08<00:00, 99.41it/s]
Loading 0:  82%|████████▏ | 239/291 [00:08<00:00, 105.10it/s]
Loading 0:  87%|████████▋ | 252/291 [00:08<00:00, 111.01it/s]
Loading 0:  91%|█████████▏| 266/291 [00:08<00:00, 59.72it/s] 
Loading 0:  96%|█████████▌| 280/291 [00:08<00:00, 72.33it/s]
                                                            
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
r136a1-moon-l3-8b-v4-mkmlizer: quantized model in 29.753s
r136a1-moon-l3-8b-v4-mkmlizer: Processed model R136a1/Moon-L3-8B in 65.254s
r136a1-moon-l3-8b-v4-mkmlizer: creating bucket guanaco-mkml-models
r136a1-moon-l3-8b-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
r136a1-moon-l3-8b-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4
r136a1-moon-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4/config.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4/special_tokens_map.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4/tokenizer_config.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4/tokenizer.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/r136a1-moon-l3-8b-v4/flywheel_model.0.safetensors
r136a1-moon-l3-8b-v4-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
r136a1-moon-l3-8b-v4-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.
r136a1-moon-l3-8b-v4-mkmlizer:   warnings.warn(
r136a1-moon-l3-8b-v4-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`.
r136a1-moon-l3-8b-v4-mkmlizer:   warnings.warn(
r136a1-moon-l3-8b-v4-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.
r136a1-moon-l3-8b-v4-mkmlizer:   warnings.warn(
r136a1-moon-l3-8b-v4-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.
r136a1-moon-l3-8b-v4-mkmlizer:   warnings.warn(
r136a1-moon-l3-8b-v4-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()
r136a1-moon-l3-8b-v4-mkmlizer:   return self.fget.__get__(instance, owner)()
r136a1-moon-l3-8b-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
r136a1-moon-l3-8b-v4-mkmlizer: Saving duration: 0.557s
r136a1-moon-l3-8b-v4-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.563s
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/config.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/special_tokens_map.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/tokenizer_config.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/merges.txt
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/vocab.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/tokenizer.json
r136a1-moon-l3-8b-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/r136a1-moon-l3-8b-v4_reward/reward.tensors
Job r136a1-moon-l3-8b-v4-mkmlizer completed after 93.51s with status: succeeded
Stopping job with name r136a1-moon-l3-8b-v4-mkmlizer
Pipeline stage MKMLizer completed in 94.41s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service r136a1-moon-l3-8b-v4
Waiting for inference service r136a1-moon-l3-8b-v4 to be ready
Inference service r136a1-moon-l3-8b-v4 ready after 40.18396019935608s
Pipeline stage ISVCDeployer completed in 46.99s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.9742891788482666s
Received healthy response to inference request in 1.3352179527282715s
Received healthy response to inference request in 1.3089044094085693s
Received healthy response to inference request in 1.2720320224761963s
Received healthy response to inference request in 1.3665595054626465s
5 requests
0 failed requests
5th percentile: 1.2794064998626709
10th percentile: 1.2867809772491454
20th percentile: 1.3015299320220948
30th percentile: 1.3141671180725099
40th percentile: 1.3246925354003907
50th percentile: 1.3352179527282715
60th percentile: 1.3477545738220216
70th percentile: 1.3602911949157714
80th percentile: 1.4881054401397706
90th percentile: 1.7311973094940187
95th percentile: 1.8527432441711424
99th percentile: 1.9499799919128418
mean time: 1.4514006137847901
Pipeline stage StressChecker completed in 7.94s
r136a1-moon-l3-8b_v4 status is now deployed due to DeploymentManager action
r136a1-moon-l3-8b_v4 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of r136a1-moon-l3-8b_v4
Running pipeline stage ISVCDeleter
Checking if service r136a1-moon-l3-8b-v4 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.75s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key r136a1-moon-l3-8b-v4/config.json from bucket guanaco-mkml-models
Deleting key r136a1-moon-l3-8b-v4/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key r136a1-moon-l3-8b-v4/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key r136a1-moon-l3-8b-v4/tokenizer.json from bucket guanaco-mkml-models
Deleting key r136a1-moon-l3-8b-v4/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key r136a1-moon-l3-8b-v4_reward/config.json from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/merges.txt from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/reward.tensors from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key r136a1-moon-l3-8b-v4_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.69s
r136a1-moon-l3-8b_v4 status is now torndown due to DeploymentManager action