submission_id: jellywibble-qlora-30k-pr_4525_v2
developer_uid: Jellywibble
best_of: 16
celo_rating: 1198.24
display_name: nitral-ai-hathor-l3-8b-v-01_v1
family_friendly_score: 0.0
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}
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', '<|eot_id|>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
is_internal_developer: True
language_model: Jellywibble/qlora_30k_pref_data_ep1
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: Jellywibble/qlora_30k_pr
model_name: nitral-ai-hathor-l3-8b-v-01_v1
model_num_parameters: 8030261248.0
model_repo: Jellywibble/qlora_30k_pref_data_ep1
model_size: 8B
num_battles: 14766
num_wins: 7614
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: 2024-07-03T22:21:14+00:00
us_pacific_date: 2024-07-03
win_ratio: 0.5156440471353109
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jellywibble-qlora-30k-pr-4525-v2-mkmlizer
Waiting for job on jellywibble-qlora-30k-pr-4525-v2-mkmlizer to finish
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║     _____            __           __                                ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║       /___/                                                         ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║                                                                     ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  Version: 0.8.14                                                    ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  https://mk1.ai                                                     ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║                                                                     ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  The license key for the current software has been verified as      ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  belonging to:                                                      ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║                                                                     ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  Chai Research Corp.                                                ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ║                                                                     ║
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Downloaded to shared memory in 38.340s
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: quantizing model to /dev/shm/model_cache
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: quantized model in 29.264s
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Processed model Jellywibble/qlora_30k_pref_data_ep1 in 67.605s
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: creating bucket guanaco-mkml-models
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2/special_tokens_map.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2/tokenizer_config.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2/config.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2/tokenizer.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jellywibble-qlora-30k-pr-4525-v2/flywheel_model.0.safetensors
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
jellywibble-qlora-30k-pr-4525-v2-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.
jellywibble-qlora-30k-pr-4525-v2-mkmlizer:   warnings.warn(
jellywibble-qlora-30k-pr-4525-v2-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`.
jellywibble-qlora-30k-pr-4525-v2-mkmlizer:   warnings.warn(
jellywibble-qlora-30k-pr-4525-v2-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.
jellywibble-qlora-30k-pr-4525-v2-mkmlizer:   warnings.warn(
jellywibble-qlora-30k-pr-4525-v2-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.
jellywibble-qlora-30k-pr-4525-v2-mkmlizer:   warnings.warn(
jellywibble-qlora-30k-pr-4525-v2-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()
jellywibble-qlora-30k-pr-4525-v2-mkmlizer:   return self.fget.__get__(instance, owner)()
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Saving duration: 0.547s
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.454s
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: creating bucket guanaco-reward-models
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/config.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/vocab.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/special_tokens_map.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/tokenizer_config.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/merges.txt
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/tokenizer.json
jellywibble-qlora-30k-pr-4525-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jellywibble-qlora-30k-pr-4525-v2_reward/reward.tensors
Job jellywibble-qlora-30k-pr-4525-v2-mkmlizer completed after 94.0s with status: succeeded
Stopping job with name jellywibble-qlora-30k-pr-4525-v2-mkmlizer
Pipeline stage MKMLizer completed in 94.89s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service jellywibble-qlora-30k-pr-4525-v2
Waiting for inference service jellywibble-qlora-30k-pr-4525-v2 to be ready
Inference service jellywibble-qlora-30k-pr-4525-v2 ready after 40.25202679634094s
Pipeline stage ISVCDeployer completed in 47.25s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.105858564376831s
Received healthy response to inference request in 1.3515527248382568s
Received healthy response to inference request in 1.313455581665039s
Received healthy response to inference request in 1.2857475280761719s
Received healthy response to inference request in 1.3707420825958252s
5 requests
0 failed requests
5th percentile: 1.2912891387939454
10th percentile: 1.2968307495117188
20th percentile: 1.3079139709472656
30th percentile: 1.3210750102996827
40th percentile: 1.3363138675689696
50th percentile: 1.3515527248382568
60th percentile: 1.3592284679412843
70th percentile: 1.3669042110443115
80th percentile: 1.5177653789520265
90th percentile: 1.8118119716644288
95th percentile: 1.9588352680206298
99th percentile: 2.076453905105591
mean time: 1.4854712963104248
Pipeline stage StressChecker completed in 8.22s
jellywibble-qlora-30k-pr_4525_v2 status is now deployed due to DeploymentManager action
jellywibble-qlora-30k-pr_4525_v2 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of jellywibble-qlora-30k-pr_4525_v2
Running pipeline stage ISVCDeleter
Checking if service jellywibble-qlora-30k-pr-4525-v2 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.02s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key jellywibble-qlora-30k-pr-4525-v2/config.json from bucket guanaco-mkml-models
Deleting key jellywibble-qlora-30k-pr-4525-v2/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key jellywibble-qlora-30k-pr-4525-v2/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key jellywibble-qlora-30k-pr-4525-v2/tokenizer.json from bucket guanaco-mkml-models
Deleting key jellywibble-qlora-30k-pr-4525-v2/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/config.json from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/merges.txt from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/reward.tensors from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key jellywibble-qlora-30k-pr-4525-v2_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.34s
jellywibble-qlora-30k-pr_4525_v2 status is now torndown due to DeploymentManager action