submission_id: revile-halustheno-v0-2_v1
developer_uid: Revile
status: inactive
model_repo: Revile/HaluStheno-v0.2
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.7, 'top_p': 1.0, 'min_p': 0.005, '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': "{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}
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: 2024-05-29T07:32:00+00:00
model_name: revile-halustheno-v0-2_v1
model_eval_status: success
model_group: Revile/HaluStheno-v0.2
num_battles: 304603
num_wins: 144524
celo_rating: 1216.02
propriety_score: 0.7065378566614244
propriety_total_count: 44510.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: revile-halustheno-v0-2_v1
ineligible_reason: None
language_model: Revile/HaluStheno-v0.2
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-05-29
win_ratio: 0.47446676493665524
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name revile-halustheno-v0-2-v1-mkmlizer
Waiting for job on revile-halustheno-v0-2-v1-mkmlizer to finish
revile-halustheno-v0-2-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
revile-halustheno-v0-2-v1-mkmlizer: ║ _____ __ __ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
revile-halustheno-v0-2-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
revile-halustheno-v0-2-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ /___/ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ Version: 0.8.14 ║
revile-halustheno-v0-2-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
revile-halustheno-v0-2-v1-mkmlizer: ║ https://mk1.ai ║
revile-halustheno-v0-2-v1-mkmlizer: ║ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ The license key for the current software has been verified as ║
revile-halustheno-v0-2-v1-mkmlizer: ║ belonging to: ║
revile-halustheno-v0-2-v1-mkmlizer: ║ ║
revile-halustheno-v0-2-v1-mkmlizer: ║ Chai Research Corp. ║
revile-halustheno-v0-2-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
revile-halustheno-v0-2-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
revile-halustheno-v0-2-v1-mkmlizer: ║ ║
revile-halustheno-v0-2-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
revile-halustheno-v0-2-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.
revile-halustheno-v0-2-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
revile-halustheno-v0-2-v1-mkmlizer: Downloaded to shared memory in 64.628s
revile-halustheno-v0-2-v1-mkmlizer: quantizing model to /dev/shm/model_cache
revile-halustheno-v0-2-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
revile-halustheno-v0-2-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<11:16, 2.34s/it] Loading 0: 1%| | 3/291 [00:04<06:47, 1.41s/it] Loading 0: 5%|▌ | 16/291 [00:04<00:46, 5.93it/s] Loading 0: 8%|▊ | 23/291 [00:05<00:28, 9.28it/s] Loading 0: 14%|█▎ | 40/291 [00:05<00:12, 20.70it/s] Loading 0: 17%|█▋ | 50/291 [00:05<00:09, 26.60it/s] Loading 0: 22%|██▏ | 63/291 [00:05<00:06, 35.82it/s] Loading 0: 27%|██▋ | 78/291 [00:05<00:04, 50.45it/s] Loading 0: 31%|███ | 89/291 [00:05<00:03, 55.98it/s] Loading 0: 36%|███▌ | 104/291 [00:05<00:02, 63.37it/s] Loading 0: 42%|████▏ | 122/291 [00:05<00:02, 83.55it/s] Loading 0: 46%|████▌ | 134/291 [00:06<00:01, 81.76it/s] Loading 0: 50%|████▉ | 145/291 [00:06<00:01, 79.34it/s] Loading 0: 55%|█████▍ | 160/291 [00:06<00:01, 94.10it/s] Loading 0: 59%|█████▉ | 172/291 [00:06<00:01, 88.23it/s] Loading 0: 64%|██████▎ | 185/291 [00:06<00:01, 82.98it/s] Loading 0: 70%|██████▉ | 203/291 [00:06<00:00, 103.00it/s] Loading 0: 74%|███████▍ | 215/291 [00:06<00:00, 94.40it/s] Loading 0: 78%|███████▊ | 226/291 [00:07<00:00, 85.87it/s] Loading 0: 84%|████████▎ | 243/291 [00:07<00:00, 104.28it/s] Loading 0: 88%|████████▊ | 255/291 [00:07<00:00, 97.13it/s] Loading 0: 91%|█████████▏| 266/291 [00:07<00:00, 85.64it/s] Loading 0: 98%|█████████▊| 284/291 [00:07<00:00, 105.71it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
revile-halustheno-v0-2-v1-mkmlizer: quantized model in 23.572s
revile-halustheno-v0-2-v1-mkmlizer: Processed model Revile/HaluStheno-v0.2 in 90.600s
revile-halustheno-v0-2-v1-mkmlizer: creating bucket guanaco-mkml-models
revile-halustheno-v0-2-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
revile-halustheno-v0-2-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/revile-halustheno-v0-2-v1
revile-halustheno-v0-2-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/revile-halustheno-v0-2-v1/config.json
revile-halustheno-v0-2-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/revile-halustheno-v0-2-v1/tokenizer_config.json
revile-halustheno-v0-2-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/revile-halustheno-v0-2-v1/tokenizer.json
revile-halustheno-v0-2-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/revile-halustheno-v0-2-v1/special_tokens_map.json
revile-halustheno-v0-2-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/revile-halustheno-v0-2-v1/flywheel_model.0.safetensors
revile-halustheno-v0-2-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
revile-halustheno-v0-2-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.
revile-halustheno-v0-2-v1-mkmlizer: warnings.warn(
revile-halustheno-v0-2-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.
revile-halustheno-v0-2-v1-mkmlizer: warnings.warn(
revile-halustheno-v0-2-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()
revile-halustheno-v0-2-v1-mkmlizer: return self.fget.__get__(instance, owner)()
revile-halustheno-v0-2-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
revile-halustheno-v0-2-v1-mkmlizer: Saving duration: 0.398s
revile-halustheno-v0-2-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 4.073s
revile-halustheno-v0-2-v1-mkmlizer: creating bucket guanaco-reward-models
revile-halustheno-v0-2-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
revile-halustheno-v0-2-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/special_tokens_map.json
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/tokenizer_config.json
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/config.json
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/merges.txt
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/vocab.json
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/tokenizer.json
revile-halustheno-v0-2-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/revile-halustheno-v0-2-v1_reward/reward.tensors
Job revile-halustheno-v0-2-v1-mkmlizer completed after 113.74s with status: succeeded
Stopping job with name revile-halustheno-v0-2-v1-mkmlizer
Pipeline stage MKMLizer completed in 117.17s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service revile-halustheno-v0-2-v1
Waiting for inference service revile-halustheno-v0-2-v1 to be ready
Inference service revile-halustheno-v0-2-v1 ready after 30.165025234222412s
Pipeline stage ISVCDeployer completed in 37.43s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1374242305755615s
Received healthy response to inference request in 1.3017966747283936s
Received healthy response to inference request in 1.240870475769043s
Received healthy response to inference request in 1.2835464477539062s
Received healthy response to inference request in 1.2564702033996582s
5 requests
0 failed requests
5th percentile: 1.243990421295166
10th percentile: 1.247110366821289
20th percentile: 1.2533502578735352
30th percentile: 1.2618854522705079
40th percentile: 1.272715950012207
50th percentile: 1.2835464477539062
60th percentile: 1.2908465385437011
70th percentile: 1.298146629333496
80th percentile: 1.4689221858978274
90th percentile: 1.8031732082366945
95th percentile: 1.9702987194061277
99th percentile: 2.1039991283416746
mean time: 1.4440216064453124
Pipeline stage StressChecker completed in 8.00s
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
revile-halustheno-v0-2_v1 status is now deployed due to DeploymentManager action
revile-halustheno-v0-2_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics