submission_id: hastagaras-halu-8b-llama_9226_v4
developer_uid: Hastagaras
status: inactive
model_repo: Hastagaras/HALU-8B-LLAMA3-EXP-2
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.85, 'top_p': 0.95, 'min_p': 0.0, '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: 2024-05-31T12:50:05+00:00
model_name: experiment-2-low-temp-test
model_eval_status: success
model_group: Hastagaras/HALU-8B-LLAMA
num_battles: 8430
num_wins: 4465
celo_rating: 1213.65
propriety_score: 0.0
propriety_total_count: 1.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: experiment-2-low-temp-test
ineligible_reason: propriety_total_count < 5000
language_model: Hastagaras/HALU-8B-LLAMA3-EXP-2
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-31
win_ratio: 0.529655990510083
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-halu-8b-llama-9226-v4-mkmlizer
Waiting for job on hastagaras-halu-8b-llama-9226-v4-mkmlizer to finish
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ _____ __ __ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ /___/ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ Version: 0.8.14 ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ https://mk1.ai ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ belonging to: ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ Chai Research Corp. ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ║ ║
hastagaras-halu-8b-llama-9226-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-halu-8b-llama-9226-v4-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/utils/_deprecation.py:131: FutureWarning: 'list_files_info' (from 'huggingface_hub.hf_api') is deprecated and will be removed from version '0.23'. Use `list_repo_tree` and `get_paths_info` instead.
hastagaras-halu-8b-llama-9226-v4-mkmlizer: warnings.warn(warning_message, FutureWarning)
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Downloaded to shared memory in 68.818s
hastagaras-halu-8b-llama-9226-v4-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<11:20, 2.35s/it] Loading 0: 5%|▍ | 14/291 [00:04<01:10, 3.92it/s] Loading 0: 9%|▉ | 27/291 [00:04<00:29, 8.98it/s] Loading 0: 14%|█▍ | 41/291 [00:05<00:15, 15.93it/s] Loading 0: 19%|█▊ | 54/291 [00:05<00:09, 24.04it/s] Loading 0: 23%|██▎ | 66/291 [00:05<00:08, 26.85it/s] Loading 0: 26%|██▋ | 77/291 [00:05<00:06, 34.49it/s] Loading 0: 31%|███ | 90/291 [00:05<00:04, 45.82it/s] Loading 0: 35%|███▌ | 103/291 [00:05<00:03, 58.02it/s] Loading 0: 39%|███▉ | 114/291 [00:05<00:02, 65.43it/s] Loading 0: 45%|████▍ | 130/291 [00:06<00:01, 82.52it/s] Loading 0: 49%|████▉ | 142/291 [00:06<00:01, 87.90it/s] Loading 0: 54%|█████▍ | 157/291 [00:06<00:01, 100.55it/s] Loading 0: 58%|█████▊ | 170/291 [00:06<00:01, 66.67it/s] Loading 0: 63%|██████▎ | 184/291 [00:06<00:01, 78.97it/s] Loading 0: 67%|██████▋ | 195/291 [00:06<00:01, 83.02it/s] Loading 0: 73%|███████▎ | 211/291 [00:06<00:00, 97.90it/s] Loading 0: 77%|███████▋ | 223/291 [00:07<00:00, 99.13it/s] Loading 0: 82%|████████▏ | 238/291 [00:07<00:00, 107.97it/s] Loading 0: 86%|████████▌ | 250/291 [00:07<00:00, 106.25it/s] Loading 0: 91%|█████████ | 265/291 [00:07<00:00, 114.87it/s] Loading 0: 96%|█████████▌| 278/291 [00:07<00:00, 71.17it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-halu-8b-llama-9226-v4-mkmlizer: quantized model in 23.798s
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Processed model Hastagaras/HALU-8B-LLAMA3-EXP-2 in 95.124s
hastagaras-halu-8b-llama-9226-v4-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-halu-8b-llama-9226-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4/config.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4/special_tokens_map.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4/tokenizer_config.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4/tokenizer.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-halu-8b-llama-9226-v4/flywheel_model.0.safetensors
hastagaras-halu-8b-llama-9226-v4-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-halu-8b-llama-9226-v4-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-9226-v4-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-9226-v4-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-9226-v4-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-9226-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.
hastagaras-halu-8b-llama-9226-v4-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-9226-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()
hastagaras-halu-8b-llama-9226-v4-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Saving duration: 0.394s
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 12.221s
hastagaras-halu-8b-llama-9226-v4-mkmlizer: creating bucket guanaco-reward-models
hastagaras-halu-8b-llama-9226-v4-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-halu-8b-llama-9226-v4-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/config.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/special_tokens_map.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/merges.txt
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/vocab.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/tokenizer_config.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/tokenizer.json
hastagaras-halu-8b-llama-9226-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-halu-8b-llama-9226-v4_reward/reward.tensors
Job hastagaras-halu-8b-llama-9226-v4-mkmlizer completed after 134.68s with status: succeeded
Stopping job with name hastagaras-halu-8b-llama-9226-v4-mkmlizer
Pipeline stage MKMLizer completed in 137.10s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-halu-8b-llama-9226-v4
Waiting for inference service hastagaras-halu-8b-llama-9226-v4 to be ready
Inference service hastagaras-halu-8b-llama-9226-v4 ready after 100.50776290893555s
Pipeline stage ISVCDeployer completed in 107.18s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.191030502319336s
Received healthy response to inference request in 1.3670215606689453s
Received healthy response to inference request in 1.3241171836853027s
Received healthy response to inference request in 1.3076541423797607s
Received healthy response to inference request in 1.2520122528076172s
5 requests
0 failed requests
5th percentile: 1.2631406307220459
10th percentile: 1.2742690086364745
20th percentile: 1.296525764465332
30th percentile: 1.3109467506408692
40th percentile: 1.3175319671630858
50th percentile: 1.3241171836853027
60th percentile: 1.3412789344787597
70th percentile: 1.3584406852722168
80th percentile: 1.5318233489990236
90th percentile: 1.8614269256591798
95th percentile: 2.0262287139892576
99th percentile: 2.1580701446533204
mean time: 1.4883671283721924
Pipeline stage StressChecker completed in 8.03s
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 0.03s
M-Eval Dataset for topic stay_in_character is loaded
hastagaras-halu-8b-llama_9226_v4 status is now deployed due to DeploymentManager action
hastagaras-halu-8b-llama_9226_v4 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics