submission_id: thedrummer-llama-3some-8b-v2_v2
developer_uid: Azazelle
best_of: 16
celo_rating: 1204.92
display_name: thedrummer-3some-8b
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': 50, '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: False
language_model: TheDrummer/Llama-3SOME-8B-v2
max_input_tokens: 512
max_output_tokens: 64
model_architecture: LlamaForCausalLM
model_group: TheDrummer/Llama-3SOME-8
model_name: thedrummer-3some-8b
model_num_parameters: 8030261248.0
model_repo: TheDrummer/Llama-3SOME-8B-v2
model_size: 8B
num_battles: 34244
num_wins: 18215
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: 2024-07-06T00:16:59+00:00
us_pacific_date: 2024-07-05
win_ratio: 0.5319180002336176
Resubmit model
Running pipeline stage MKMLizer
Starting job with name thedrummer-llama-3some-8b-v2-v2-mkmlizer
Waiting for job on thedrummer-llama-3some-8b-v2-v2-mkmlizer to finish
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ _____ __ __ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ /___/ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ Version: 0.8.14 ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ https://mk1.ai ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ The license key for the current software has been verified as ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ belonging to: ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ Chai Research Corp. ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ║ ║
thedrummer-llama-3some-8b-v2-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Downloaded to shared memory in 22.636s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: quantizing model to /dev/shm/model_cache
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:05<13:18, 2.76s/it] Loading 0: 4%|▍ | 13/291 [00:05<01:29, 3.12it/s] Loading 0: 8%|▊ | 24/291 [00:05<00:39, 6.80it/s] Loading 0: 14%|█▎ | 40/291 [00:05<00:17, 13.96it/s] Loading 0: 18%|█▊ | 51/291 [00:05<00:12, 19.82it/s] Loading 0: 21%|██▏ | 62/291 [00:06<00:10, 20.87it/s] Loading 0: 26%|██▌ | 76/291 [00:06<00:07, 30.48it/s] Loading 0: 30%|██▉ | 87/291 [00:06<00:05, 38.09it/s] Loading 0: 34%|███▍ | 99/291 [00:06<00:03, 48.48it/s] Loading 0: 38%|███▊ | 110/291 [00:06<00:03, 54.61it/s] Loading 0: 41%|████ | 120/291 [00:07<00:02, 62.10it/s] Loading 0: 45%|████▍ | 130/291 [00:07<00:02, 68.77it/s] Loading 0: 48%|████▊ | 141/291 [00:07<00:01, 75.55it/s] Loading 0: 54%|█████▍ | 157/291 [00:07<00:01, 91.86it/s] Loading 0: 58%|█████▊ | 168/291 [00:07<00:02, 49.92it/s] Loading 0: 63%|██████▎ | 184/291 [00:07<00:01, 65.75it/s] Loading 0: 67%|██████▋ | 195/291 [00:08<00:01, 71.91it/s] Loading 0: 73%|███████▎ | 211/291 [00:08<00:00, 87.37it/s] Loading 0: 77%|███████▋ | 223/291 [00:08<00:00, 90.88it/s] Loading 0: 82%|████████▏ | 238/291 [00:08<00:00, 102.53it/s] Loading 0: 86%|████████▌ | 250/291 [00:08<00:00, 97.52it/s] Loading 0: 91%|█████████ | 265/291 [00:08<00:00, 106.82it/s] Loading 0: 95%|█████████▌| 277/291 [00:09<00:00, 56.53it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
thedrummer-llama-3some-8b-v2-v2-mkmlizer: quantized model in 29.257s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Processed model TheDrummer/Llama-3SOME-8B-v2 in 51.893s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: creating bucket guanaco-mkml-models
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
thedrummer-llama-3some-8b-v2-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2/tokenizer_config.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2/special_tokens_map.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2/config.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2/tokenizer.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/thedrummer-llama-3some-8b-v2-v2/flywheel_model.0.safetensors
thedrummer-llama-3some-8b-v2-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
thedrummer-llama-3some-8b-v2-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.
thedrummer-llama-3some-8b-v2-v2-mkmlizer: warnings.warn(
thedrummer-llama-3some-8b-v2-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`.
thedrummer-llama-3some-8b-v2-v2-mkmlizer: warnings.warn(
thedrummer-llama-3some-8b-v2-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.
thedrummer-llama-3some-8b-v2-v2-mkmlizer: warnings.warn(
thedrummer-llama-3some-8b-v2-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.
thedrummer-llama-3some-8b-v2-v2-mkmlizer: warnings.warn(
thedrummer-llama-3some-8b-v2-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()
thedrummer-llama-3some-8b-v2-v2-mkmlizer: return self.fget.__get__(instance, owner)()
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Saving duration: 0.512s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 7.323s
thedrummer-llama-3some-8b-v2-v2-mkmlizer: creating bucket guanaco-reward-models
thedrummer-llama-3some-8b-v2-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
thedrummer-llama-3some-8b-v2-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/config.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/special_tokens_map.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/tokenizer_config.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/merges.txt
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/vocab.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/tokenizer.json
thedrummer-llama-3some-8b-v2-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/thedrummer-llama-3some-8b-v2-v2_reward/reward.tensors
Job thedrummer-llama-3some-8b-v2-v2-mkmlizer completed after 84.36s with status: succeeded
Stopping job with name thedrummer-llama-3some-8b-v2-v2-mkmlizer
Pipeline stage MKMLizer completed in 85.24s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service thedrummer-llama-3some-8b-v2-v2
Waiting for inference service thedrummer-llama-3some-8b-v2-v2 to be ready
Inference service thedrummer-llama-3some-8b-v2-v2 ready after 50.245522260665894s
Pipeline stage ISVCDeployer completed in 57.26s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.073749542236328s
Received healthy response to inference request in 1.335442304611206s
Received healthy response to inference request in 1.5595972537994385s
Received healthy response to inference request in 1.3017756938934326s
Received healthy response to inference request in 1.3486876487731934s
5 requests
0 failed requests
5th percentile: 1.3085090160369872
10th percentile: 1.315242338180542
20th percentile: 1.3287089824676515
30th percentile: 1.3380913734436035
40th percentile: 1.3433895111083984
50th percentile: 1.3486876487731934
60th percentile: 1.4330514907836913
70th percentile: 1.5174153327941895
80th percentile: 1.6624277114868165
90th percentile: 1.8680886268615724
95th percentile: 1.9709190845489502
99th percentile: 2.0531834506988527
mean time: 1.5238504886627198
Pipeline stage StressChecker completed in 8.36s
thedrummer-llama-3some-8b-v2_v2 status is now deployed due to DeploymentManager action
thedrummer-llama-3some-8b-v2_v2 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of thedrummer-llama-3some-8b-v2_v2
Running pipeline stage ISVCDeleter
Checking if service thedrummer-llama-3some-8b-v2-v2 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.07s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key thedrummer-llama-3some-8b-v2-v2/config.json from bucket guanaco-mkml-models
Deleting key thedrummer-llama-3some-8b-v2-v2/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key thedrummer-llama-3some-8b-v2-v2/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key thedrummer-llama-3some-8b-v2-v2/tokenizer.json from bucket guanaco-mkml-models
Deleting key thedrummer-llama-3some-8b-v2-v2/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/config.json from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/merges.txt from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/reward.tensors from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key thedrummer-llama-3some-8b-v2-v2_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.63s
thedrummer-llama-3some-8b-v2_v2 status is now torndown due to DeploymentManager action