submission_id: nousresearch-nous-herme_4824_v10
developer_uid: Meliodia
status: inactive
model_repo: NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 512, 'best_of': 4, '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: 2024-05-06T22:12:47+00:00
model_name: nousresearch-nous-herme_4824_v10
model_eval_status: success
double_thumbs_up: 414
thumbs_up: 625
thumbs_down: 308
num_battles: 31656
num_wins: 14914
celo_rating: 1162.74
entertaining: 6.8
stay_in_character: 8.79
user_preference: 7.14
safety_score: 0.96
submission_type: basic
model_architecture: MixtralForCausalLM
model_num_parameters: 46702809088.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: nousresearch-nous-herme_4824_v10
double_thumbs_up_ratio: 0.30734966592427615
feedback_count: 1347
ineligible_reason: None
language_model: NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO
model_score: 7.576666666666667
model_size: 47B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
single_thumbs_up_ratio: 0.4639940608760208
thumbs_down_ratio: 0.22865627319970305
thumbs_up_ratio: 0.771343726800297
us_pacific_date: 2024-05-06
win_ratio: 0.4711271165024008
Resubmit model
Running pipeline stage MKMLizer
Starting job with name nousresearch-nous-herme-4824-v10-mkmlizer
Waiting for job on nousresearch-nous-herme-4824-v10-mkmlizer to finish
nousresearch-nous-herme-4824-v10-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
nousresearch-nous-herme-4824-v10-mkmlizer: ║ _____ __ __ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ /___/ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ Version: 0.8.10 ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ The license key for the current software has been verified as ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ belonging to: ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ Chai Research Corp. ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
nousresearch-nous-herme-4824-v10-mkmlizer: ║ ║
nousresearch-nous-herme-4824-v10-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
nousresearch-nous-herme-4824-v10-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.
nousresearch-nous-herme-4824-v10-mkmlizer: warnings.warn(warning_message, FutureWarning)
Failed to get response for submission neversleep-noromaidxopen_2041_v3: ('http://neversleep-noromaidxopen-2041-v3-predictor-default.tenant-chaiml-guanaco.knative.ord1.coreweave.cloud/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:40104->127.0.0.1:8080: read: connection reset by peer\n')
nousresearch-nous-herme-4824-v10-mkmlizer: Downloaded to shared memory in 82.036s
nousresearch-nous-herme-4824-v10-mkmlizer: quantizing model to /dev/shm/model_cache
nousresearch-nous-herme-4824-v10-mkmlizer: Saving flywheel model at /dev/shm/model_cache
nousresearch-nous-herme-4824-v10-mkmlizer: Loading 0: 0%| | 0/995 [00:00<?, ?it/s] Loading 0: 6%|▌ | 57/995 [00:01<00:16, 56.91it/s] Loading 0: 11%|█ | 107/995 [00:02<00:17, 51.75it/s] Loading 0: 16%|█▋ | 162/995 [00:03<00:15, 52.15it/s] Loading 0: 22%|██▏ | 215/995 [00:04<00:14, 52.45it/s] Loading 0: 27%|██▋ | 272/995 [00:05<00:13, 54.04it/s] Loading 0: 28%|██▊ | 282/995 [00:17<01:25, 8.29it/s] Loading 0: 33%|███▎ | 324/995 [00:18<00:58, 11.39it/s] Loading 0: 37%|███▋ | 368/995 [00:19<00:41, 15.13it/s] Loading 0: 43%|████▎ | 424/995 [00:20<00:27, 20.90it/s] Loading 0: 48%|████▊ | 482/995 [00:21<00:18, 27.12it/s] Loading 0: 53%|█████▎ | 532/995 [00:22<00:14, 31.52it/s] Loading 0: 56%|█████▋ | 561/995 [00:33<00:45, 9.62it/s] Loading 0: 59%|█████▉ | 592/995 [00:34<00:35, 11.49it/s] Loading 0: 64%|██████▍ | 636/995 [00:35<00:24, 14.86it/s] Loading 0: 69%|██████▉ | 691/995 [00:36<00:15, 20.06it/s] Loading 0: 74%|███████▍ | 740/995 [00:37<00:10, 24.69it/s] Loading 0: 80%|███████▉ | 795/995 [00:38<00:06, 30.29it/s] Loading 0: 80%|███████▉ | 795/995 [00:50<00:06, 30.29it/s] Loading 0: 85%|████████▍ | 842/995 [00:50<00:14, 10.26it/s] Loading 0: 88%|████████▊ | 874/995 [00:51<00:10, 12.07it/s] Loading 0: 92%|█████████▏| 914/995 [00:52<00:05, 15.04it/s] Loading 0: 96%|█████████▌| 952/995 [00:54<00:02, 15.81it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
nousresearch-nous-herme-4824-v10-mkmlizer: quantized model in 63.098s
nousresearch-nous-herme-4824-v10-mkmlizer: Processed model NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO in 151.116s
nousresearch-nous-herme-4824-v10-mkmlizer: creating bucket guanaco-mkml-models
nousresearch-nous-herme-4824-v10-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
nousresearch-nous-herme-4824-v10-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/config.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/tokenizer_config.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/tokenizer.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/added_tokens.json s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/added_tokens.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/tokenizer.model
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/special_tokens_map.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/flywheel_model.3.safetensors s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/flywheel_model.3.safetensors
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/flywheel_model.0.safetensors
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/flywheel_model.1.safetensors
nousresearch-nous-herme-4824-v10-mkmlizer: cp /dev/shm/model_cache/flywheel_model.2.safetensors s3://guanaco-mkml-models/nousresearch-nous-herme-4824-v10/flywheel_model.2.safetensors
nousresearch-nous-herme-4824-v10-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
nousresearch-nous-herme-4824-v10-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.
nousresearch-nous-herme-4824-v10-mkmlizer: warnings.warn(
nousresearch-nous-herme-4824-v10-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.
nousresearch-nous-herme-4824-v10-mkmlizer: warnings.warn(
nousresearch-nous-herme-4824-v10-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.
nousresearch-nous-herme-4824-v10-mkmlizer: warnings.warn(
nousresearch-nous-herme-4824-v10-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()
nousresearch-nous-herme-4824-v10-mkmlizer: return self.fget.__get__(instance, owner)()
nousresearch-nous-herme-4824-v10-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
nousresearch-nous-herme-4824-v10-mkmlizer: Saving duration: 0.279s
nousresearch-nous-herme-4824-v10-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.948s
nousresearch-nous-herme-4824-v10-mkmlizer: creating bucket guanaco-reward-models
nousresearch-nous-herme-4824-v10-mkmlizer: Bucket 's3://guanaco-reward-models/' created
nousresearch-nous-herme-4824-v10-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/config.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/special_tokens_map.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/tokenizer_config.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/vocab.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/merges.txt
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/tokenizer.json
nousresearch-nous-herme-4824-v10-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/nousresearch-nous-herme-4824-v10_reward/reward.tensors
Job nousresearch-nous-herme-4824-v10-mkmlizer completed after 329.95s with status: succeeded
Stopping job with name nousresearch-nous-herme-4824-v10-mkmlizer
Pipeline stage MKMLizer completed in 334.37s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service nousresearch-nous-herme-4824-v10
Waiting for inference service nousresearch-nous-herme-4824-v10 to be ready
Inference service nousresearch-nous-herme-4824-v10 ready after 60.38810896873474s
Pipeline stage ISVCDeployer completed in 67.82s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.3693525791168213s
Received healthy response to inference request in 0.6937441825866699s
Received healthy response to inference request in 0.7465949058532715s
Received healthy response to inference request in 1.3469414710998535s
Received healthy response to inference request in 1.3580844402313232s
5 requests
0 failed requests
5th percentile: 0.7043143272399902
10th percentile: 0.7148844718933105
20th percentile: 0.7360247611999512
30th percentile: 0.8666642189025878
40th percentile: 1.1068028450012208
50th percentile: 1.3469414710998535
60th percentile: 1.3513986587524414
70th percentile: 1.3558558464050292
80th percentile: 1.5603380680084231
90th percentile: 1.9648453235626222
95th percentile: 2.1670989513397214
99th percentile: 2.3289018535614012
mean time: 1.302943515777588
Pipeline stage StressChecker completed in 7.21s
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
nousresearch-nous-herme_4824_v10 status is now deployed due to DeploymentManager action
nousresearch-nous-herme_4824_v10 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics