submission_id: hastagaras-halu-8b-llama_5820_v1
developer_uid: Hastagaras
status: inactive
model_repo: Hastagaras/HALU-8B-LLAMA3-EXP-3
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.0, 'top_p': 0.95, 'min_p': 0.08, '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|>{bot_name}<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>', 'user_template': '<|start_header_id|>{user_name}<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>', 'response_template': '<|start_header_id|>{bot_name}<|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: 2024-06-01T02:24:18+00:00
model_name: ins-name
model_eval_status: success
model_group: Hastagaras/HALU-8B-LLAMA
num_battles: 7791
num_wins: 4263
celo_rating: 1212.9
safety_score: 0.97
propriety_score: 0.75
propriety_total_count: 4.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: ins-name
ineligible_reason: propriety_total_count < 5000
language_model: Hastagaras/HALU-8B-LLAMA3-EXP-3
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-31
win_ratio: 0.5471698113207547
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-halu-8b-llama-5820-v1-mkmlizer
Waiting for job on hastagaras-halu-8b-llama-5820-v1-mkmlizer to finish
Stopping job with name hastagaras-halu-8b-llama-5820-v1-mkmlizer
%s, retrying in %s seconds...
Starting job with name hastagaras-halu-8b-llama-5820-v1-mkmlizer
Waiting for job on hastagaras-halu-8b-llama-5820-v1-mkmlizer to finish
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ _____ __ __ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ /___/ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ Version: 0.8.14 ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ https://mk1.ai ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ belonging to: ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ Chai Research Corp. ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-5820-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-halu-8b-llama-5820-v1-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-5820-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Downloaded to shared memory in 32.687s
hastagaras-halu-8b-llama-5820-v1-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 5%|▍ | 14/291 [00:00<00:02, 136.81it/s] Loading 0: 11%|█ | 32/291 [00:00<00:01, 153.17it/s] Loading 0: 17%|█▋ | 49/291 [00:00<00:01, 154.94it/s] Loading 0: 22%|██▏ | 65/291 [00:00<00:01, 149.83it/s] Loading 0: 27%|██▋ | 80/291 [00:00<00:01, 116.09it/s] Loading 0: 32%|███▏ | 93/291 [00:00<00:02, 71.65it/s] Loading 0: 38%|███▊ | 111/291 [00:01<00:01, 90.57it/s] Loading 0: 44%|████▍ | 129/291 [00:01<00:01, 106.41it/s] Loading 0: 51%|█████ | 147/291 [00:01<00:01, 120.88it/s] Loading 0: 57%|█████▋ | 165/291 [00:01<00:00, 132.32it/s] Loading 0: 62%|██████▏ | 181/291 [00:01<00:00, 135.99it/s] Loading 0: 67%|██████▋ | 196/291 [00:01<00:01, 86.58it/s] Loading 0: 73%|███████▎ | 212/291 [00:01<00:00, 98.95it/s] Loading 0: 79%|███████▊ | 229/291 [00:02<00:00, 112.61it/s] Loading 0: 85%|████████▍ | 246/291 [00:02<00:00, 123.12it/s] Loading 0: 91%|█████████ | 264/291 [00:02<00:00, 134.62it/s] Loading 0: 97%|█████████▋| 281/291 [00:02<00:00, 140.38it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-halu-8b-llama-5820-v1-mkmlizer: quantized model in 22.753s
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Processed model Hastagaras/HALU-8B-LLAMA3-EXP-3 in 58.042s
hastagaras-halu-8b-llama-5820-v1-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-halu-8b-llama-5820-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-halu-8b-llama-5820-v1
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-5820-v1/config.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-5820-v1/special_tokens_map.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-5820-v1/tokenizer_config.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-5820-v1/tokenizer.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-halu-8b-llama-5820-v1-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-5820-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-5820-v1-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-5820-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-5820-v1-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-5820-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-5820-v1-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-5820-v1-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Saving duration: 0.395s
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.309s
hastagaras-halu-8b-llama-5820-v1-mkmlizer: creating bucket guanaco-reward-models
hastagaras-halu-8b-llama-5820-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-halu-8b-llama-5820-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/special_tokens_map.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/config.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/tokenizer_config.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/merges.txt
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/vocab.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/tokenizer.json
hastagaras-halu-8b-llama-5820-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-halu-8b-llama-5820-v1_reward/reward.tensors
Job hastagaras-halu-8b-llama-5820-v1-mkmlizer completed after 83.25s with status: succeeded
Stopping job with name hastagaras-halu-8b-llama-5820-v1-mkmlizer
Pipeline stage MKMLizer completed in 87.59s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.08s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-halu-8b-llama-5820-v1
Waiting for inference service hastagaras-halu-8b-llama-5820-v1 to be ready
Inference service hastagaras-halu-8b-llama-5820-v1 ready after 30.170434713363647s
Pipeline stage ISVCDeployer completed in 37.50s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1995630264282227s
Received healthy response to inference request in 1.327876091003418s
Received healthy response to inference request in 1.3232412338256836s
Received healthy response to inference request in 1.2902843952178955s
Received healthy response to inference request in 1.2506895065307617s
5 requests
0 failed requests
5th percentile: 1.2586084842681884
10th percentile: 1.2665274620056153
20th percentile: 1.2823654174804688
30th percentile: 1.2968757629394532
40th percentile: 1.3100584983825683
50th percentile: 1.3232412338256836
60th percentile: 1.3250951766967773
70th percentile: 1.326949119567871
80th percentile: 1.5022134780883791
90th percentile: 1.8508882522583008
95th percentile: 2.0252256393432617
99th percentile: 2.1646955490112303
mean time: 1.4783308506011963
Pipeline stage StressChecker completed in 7.96s
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_5820_v1 status is now deployed due to DeploymentManager action
hastagaras-halu-8b-llama_5820_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics