submission_id: hastagaras-halu-8b-llama_4766_v1
developer_uid: Hastagaras
status: inactive
model_repo: Hastagaras/HALU-8B-LLAMA3-BRSLURP
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.1, 'top_p': 1.0, '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|>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: 2024-05-31T22:13:14+00:00
model_name: renamed-from-exp2
model_eval_status: success
model_group: Hastagaras/HALU-8B-LLAMA
num_battles: 12658
num_wins: 7092
celo_rating: 1211.25
safety_score: 0.96
propriety_score: 1.0
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: renamed-from-exp2
ineligible_reason: propriety_total_count < 5000
language_model: Hastagaras/HALU-8B-LLAMA3-BRSLURP
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-31
win_ratio: 0.5602780850055301
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-halu-8b-llama-4766-v1-mkmlizer
Waiting for job on hastagaras-halu-8b-llama-4766-v1-mkmlizer to finish
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ _____ __ __ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ /___/ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ Version: 0.8.14 ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ https://mk1.ai ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ belonging to: ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ Chai Research Corp. ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ║ ║
hastagaras-halu-8b-llama-4766-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-halu-8b-llama-4766-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-4766-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Downloaded to shared memory in 17.307s
hastagaras-halu-8b-llama-4766-v1-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:03<09:37, 2.00s/it] Loading 0: 8%|▊ | 23/291 [00:04<00:34, 7.70it/s] Loading 0: 15%|█▌ | 45/291 [00:04<00:13, 17.63it/s] Loading 0: 21%|██▏ | 62/291 [00:04<00:09, 23.56it/s] Loading 0: 29%|██▉ | 85/291 [00:04<00:05, 38.26it/s] Loading 0: 36%|███▌ | 105/291 [00:04<00:03, 52.83it/s] Loading 0: 45%|████▍ | 130/291 [00:04<00:02, 75.17it/s] Loading 0: 52%|█████▏ | 150/291 [00:04<00:01, 91.89it/s] Loading 0: 58%|█████▊ | 169/291 [00:05<00:01, 76.92it/s] Loading 0: 66%|██████▋ | 193/291 [00:05<00:00, 99.42it/s] Loading 0: 73%|███████▎ | 213/291 [00:05<00:00, 114.43it/s] Loading 0: 82%|████████▏ | 238/291 [00:05<00:00, 138.49it/s] Loading 0: 89%|████████▊ | 258/291 [00:05<00:00, 148.76it/s] Loading 0: 95%|█████████▌| 277/291 [00:06<00:00, 101.31it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-halu-8b-llama-4766-v1-mkmlizer: quantized model in 16.850s
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Processed model Hastagaras/HALU-8B-LLAMA3-BRSLURP in 35.079s
hastagaras-halu-8b-llama-4766-v1-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-halu-8b-llama-4766-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4766-v1
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4766-v1/config.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4766-v1/special_tokens_map.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4766-v1/tokenizer_config.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-halu-8b-llama-4766-v1/tokenizer.json
hastagaras-halu-8b-llama-4766-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-4766-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4766-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-halu-8b-llama-4766-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-4766-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4766-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-4766-v1-mkmlizer: warnings.warn(
hastagaras-halu-8b-llama-4766-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-4766-v1-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Saving duration: 0.217s
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.597s
hastagaras-halu-8b-llama-4766-v1-mkmlizer: creating bucket guanaco-reward-models
hastagaras-halu-8b-llama-4766-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-halu-8b-llama-4766-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/config.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/special_tokens_map.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/vocab.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/tokenizer_config.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/tokenizer.json
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/merges.txt
hastagaras-halu-8b-llama-4766-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-halu-8b-llama-4766-v1_reward/reward.tensors
Job hastagaras-halu-8b-llama-4766-v1-mkmlizer completed after 63.47s with status: succeeded
Stopping job with name hastagaras-halu-8b-llama-4766-v1-mkmlizer
Pipeline stage MKMLizer completed in 63.81s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.08s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-halu-8b-llama-4766-v1
Waiting for inference service hastagaras-halu-8b-llama-4766-v1 to be ready
Retrying (%r) after connection broken by '%r': %s
Inference service hastagaras-halu-8b-llama-4766-v1 ready after 40.40028667449951s
Pipeline stage ISVCDeployer completed in 46.48s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.304567813873291s
Received healthy response to inference request in 1.330390453338623s
Received healthy response to inference request in 1.3100149631500244s
Received healthy response to inference request in 1.2884602546691895s
Received healthy response to inference request in 1.2499706745147705s
5 requests
0 failed requests
5th percentile: 1.2576685905456544
10th percentile: 1.265366506576538
20th percentile: 1.2807623386383056
30th percentile: 1.2927711963653565
40th percentile: 1.3013930797576905
50th percentile: 1.3100149631500244
60th percentile: 1.318165159225464
70th percentile: 1.3263153553009033
80th percentile: 1.525225925445557
90th percentile: 1.914896869659424
95th percentile: 2.109732341766357
99th percentile: 2.2656007194519043
mean time: 1.4966808319091798
Pipeline stage StressChecker completed in 8.09s
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_4766_v1 status is now deployed due to DeploymentManager action
hastagaras-halu-8b-llama_4766_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics