submission_id: hastagaras-halu-8b-llama_4167_v2
developer_uid: Hastagaras
status: inactive
model_repo: Hastagaras/Halu-8B-Llama3-CR-v0.45
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.95, 'top_p': 1.0, 'min_p': 0.05, 'top_k': 100, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<|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: {'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}
timestamp: 2024-06-03T08:35:31+00:00
model_name: test
model_eval_status: success
model_group: Hastagaras/Halu-8B-Llama
num_battles: 66360
num_wins: 34995
celo_rating: 1203.83
safety_score: 0.92
propriety_score: 0.6646795827123696
propriety_total_count: 1342.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: test
ineligible_reason: propriety_total_count < 5000
language_model: Hastagaras/Halu-8B-Llama3-CR-v0.45
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-03
win_ratio: 0.5273508137432188
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-halu-8b-llama-4167-v2-mkmlizer
Waiting for job on hastagaras-halu-8b-llama-4167-v2-mkmlizer to finish
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ _____ __ __ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ /___/ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ Version: 0.8.14 ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ https://mk1.ai ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ belonging to: ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ Chai Research Corp. ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4167-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Downloaded to shared memory in 52.759s
hastagaras-halu-8b-llama-4167-v2-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 4%|▍ | 12/291 [00:00<00:02, 111.51it/s] Loading 0: 8%|▊ | 24/291 [00:00<00:02, 110.03it/s] Loading 0: 13%|█▎ | 39/291 [00:00<00:02, 121.45it/s] Loading 0: 18%|█▊ | 52/291 [00:00<00:02, 118.17it/s] Loading 0: 23%|██▎ | 66/291 [00:00<00:01, 123.24it/s] Loading 0: 27%|██▋ | 79/291 [00:00<00:01, 119.23it/s] Loading 0: 31%|███▏ | 91/291 [00:01<00:03, 63.89it/s] Loading 0: 35%|███▌ | 103/291 [00:01<00:02, 73.02it/s] Loading 0: 40%|████ | 117/291 [00:01<00:02, 86.23it/s] Loading 0: 45%|████▍ | 130/291 [00:01<00:01, 93.66it/s] Loading 0: 49%|████▉ | 143/291 [00:01<00:01, 102.38it/s] Loading 0: 54%|█████▍ | 157/291 [00:01<00:01, 107.54it/s] Loading 0: 58%|█████▊ | 170/291 [00:01<00:01, 112.93it/s] Loading 0: 63%|██████▎ | 183/291 [00:01<00:00, 114.20it/s] Loading 0: 67%|██████▋ | 195/291 [00:02<00:01, 55.82it/s] Loading 0: 72%|███████▏ | 210/291 [00:02<00:01, 70.33it/s] Loading 0: 76%|███████▌ | 221/291 [00:02<00:00, 76.40it/s] Loading 0: 81%|████████▏ | 237/291 [00:02<00:00, 92.91it/s] Loading 0: 86%|████████▌ | 250/291 [00:02<00:00, 98.70it/s] Loading 0: 91%|█████████ | 264/291 [00:02<00:00, 107.82it/s] Loading 0: 95%|█████████▌| 277/291 [00:02<00:00, 110.25it/s] Loading 0: 100%|█████████▉| 290/291 [00:08<00:00, 7.65it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-halu-8b-llama-4167-v2-mkmlizer: quantized model in 24.450s
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Processed model Hastagaras/Halu-8B-Llama3-CR-v0.45 in 79.899s
hastagaras-halu-8b-llama-4167-v2-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-halu-8b-llama-4167-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2/special_tokens_map.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2/tokenizer_config.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2/config.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2/tokenizer.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4167-v2/flywheel_model.0.safetensors
hastagaras-halu-8b-llama-4167-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-halu-8b-llama-4167-v2-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.
hastagaras-halu-8b-llama-4167-v2-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4167-v2-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.
hastagaras-halu-8b-llama-4167-v2-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4167-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.
hastagaras-halu-8b-llama-4167-v2-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4167-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()
hastagaras-halu-8b-llama-4167-v2-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Saving duration: 0.444s
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 12.681s
hastagaras-halu-8b-llama-4167-v2-mkmlizer: creating bucket guanaco-reward-models
hastagaras-halu-8b-llama-4167-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-halu-8b-llama-4167-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/config.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/special_tokens_map.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/tokenizer_config.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/vocab.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/merges.txt
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/tokenizer.json
hastagaras-halu-8b-llama-4167-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-halu-8b-llama-4167-v2_reward/reward.tensors
Job hastagaras-halu-8b-llama-4167-v2-mkmlizer completed after 123.71s with status: succeeded
Stopping job with name hastagaras-halu-8b-llama-4167-v2-mkmlizer
Pipeline stage MKMLizer completed in 127.54s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-halu-8b-llama-4167-v2
Waiting for inference service hastagaras-halu-8b-llama-4167-v2 to be ready
Inference service hastagaras-halu-8b-llama-4167-v2 ready after 30.18789291381836s
Pipeline stage ISVCDeployer completed in 37.58s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.186953067779541s
Received healthy response to inference request in 1.375331163406372s
Received healthy response to inference request in 1.3272857666015625s
Received healthy response to inference request in 1.2954344749450684s
Received healthy response to inference request in 1.2339739799499512s
5 requests
0 failed requests
5th percentile: 1.2462660789489746
10th percentile: 1.258558177947998
20th percentile: 1.283142375946045
30th percentile: 1.3018047332763671
40th percentile: 1.314545249938965
50th percentile: 1.3272857666015625
60th percentile: 1.3465039253234863
70th percentile: 1.36572208404541
80th percentile: 1.537655544281006
90th percentile: 1.8623043060302735
95th percentile: 2.024628686904907
99th percentile: 2.154488191604614
mean time: 1.483795690536499
Pipeline stage StressChecker completed in 8.13s
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
hastagaras-halu-8b-llama_4167_v2 status is now deployed due to DeploymentManager action
hastagaras-halu-8b-llama_4167_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics