submission_id: jieliu-storm-7b_v1
developer_uid: Marjovl
status: inactive
model_repo: jieliu/Storm-7B
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: 2024-07-01T20:51:25+00:00
model_name: jieliu-storm-7b_v1
model_group: jieliu/Storm-7B
num_battles: 12403
num_wins: 4884
celo_rating: 1099.96
propriety_score: 0.7237996902426432
propriety_total_count: 5811.0
submission_type: basic
model_architecture: MistralForCausalLM
model_num_parameters: 7241748480.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: jieliu-storm-7b_v1
ineligible_reason: None
language_model: jieliu/Storm-7B
model_size: 7B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-01
win_ratio: 0.39377569942755786
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jieliu-storm-7b-v1-mkmlizer
Waiting for job on jieliu-storm-7b-v1-mkmlizer to finish
jieliu-storm-7b-v1-mkmlizer: Downloaded to shared memory in 54.609s
jieliu-storm-7b-v1-mkmlizer: quantizing model to /dev/shm/model_cache
jieliu-storm-7b-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jieliu-storm-7b-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   7%|▋         | 20/291 [00:00<00:01, 196.01it/s]
Loading 0:  14%|█▎        | 40/291 [00:00<00:01, 198.06it/s]
Loading 0:  22%|██▏       | 63/291 [00:00<00:01, 207.30it/s]
Loading 0:  29%|██▉       | 84/291 [00:00<00:01, 167.48it/s]
Loading 0:  35%|███▌      | 102/291 [00:00<00:01, 108.02it/s]
Loading 0:  40%|███▉      | 116/291 [00:00<00:01, 114.91it/s]
Loading 0:  47%|████▋     | 138/291 [00:00<00:01, 138.64it/s]
Loading 0:  54%|█████▍    | 157/291 [00:01<00:00, 150.62it/s]
Loading 0:  62%|██████▏   | 179/291 [00:01<00:00, 167.74it/s]
Loading 0:  70%|██████▉   | 203/291 [00:01<00:00, 187.20it/s]
Loading 0:  77%|███████▋  | 224/291 [00:02<00:01, 38.30it/s] 
Loading 0:  83%|████████▎ | 242/291 [00:02<00:01, 48.66it/s]
Loading 0:  89%|████████▊ | 258/291 [00:03<00:00, 58.69it/s]
Loading 0:  95%|█████████▍| 276/291 [00:03<00:00, 72.73it/s]
                                                            
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
jieliu-storm-7b-v1-mkmlizer: quantized model in 9.978s
jieliu-storm-7b-v1-mkmlizer: Processed model jieliu/Storm-7B in 64.587s
jieliu-storm-7b-v1-mkmlizer: creating bucket guanaco-mkml-models
jieliu-storm-7b-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jieliu-storm-7b-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jieliu-storm-7b-v1
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jieliu-storm-7b-v1/config.json
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jieliu-storm-7b-v1/special_tokens_map.json
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/added_tokens.json s3://guanaco-mkml-models/jieliu-storm-7b-v1/added_tokens.json
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/jieliu-storm-7b-v1/tokenizer.model
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jieliu-storm-7b-v1/tokenizer_config.json
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jieliu-storm-7b-v1/tokenizer.json
jieliu-storm-7b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jieliu-storm-7b-v1/flywheel_model.0.safetensors
jieliu-storm-7b-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
jieliu-storm-7b-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:919: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jieliu-storm-7b-v1-mkmlizer:   warnings.warn(
jieliu-storm-7b-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
jieliu-storm-7b-v1-mkmlizer:   warnings.warn(
jieliu-storm-7b-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:769: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jieliu-storm-7b-v1-mkmlizer:   warnings.warn(
jieliu-storm-7b-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.
jieliu-storm-7b-v1-mkmlizer:   warnings.warn(
jieliu-storm-7b-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()
jieliu-storm-7b-v1-mkmlizer:   return self.fget.__get__(instance, owner)()
jieliu-storm-7b-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jieliu-storm-7b-v1-mkmlizer: Saving duration: 1.014s
jieliu-storm-7b-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 5.633s
jieliu-storm-7b-v1-mkmlizer: creating bucket guanaco-reward-models
jieliu-storm-7b-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jieliu-storm-7b-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jieliu-storm-7b-v1_reward
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/config.json
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/special_tokens_map.json
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/tokenizer_config.json
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/merges.txt
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/vocab.json
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/tokenizer.json
jieliu-storm-7b-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jieliu-storm-7b-v1_reward/reward.tensors
Job jieliu-storm-7b-v1-mkmlizer completed after 227.24s with status: succeeded
Stopping job with name jieliu-storm-7b-v1-mkmlizer
Pipeline stage MKMLizer completed in 228.25s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.13s
Running pipeline stage ISVCDeployer
Creating inference service jieliu-storm-7b-v1
Waiting for inference service jieliu-storm-7b-v1 to be ready
Inference service jieliu-storm-7b-v1 ready after 90.6053717136383s
Pipeline stage ISVCDeployer completed in 97.46s
Running pipeline stage StressChecker
%s, retrying in %s seconds...
Received healthy response to inference request in 2.128239154815674s
Received healthy response to inference request in 0.954031229019165s
Received healthy response to inference request in 0.8186659812927246s
Received healthy response to inference request in 0.7578153610229492s
Received healthy response to inference request in 0.7506382465362549s
5 requests
0 failed requests
5th percentile: 0.7520736694335938
10th percentile: 0.7535090923309327
20th percentile: 0.7563799381256103
30th percentile: 0.7699854850769043
40th percentile: 0.7943257331848145
50th percentile: 0.8186659812927246
60th percentile: 0.8728120803833008
70th percentile: 0.9269581794738769
80th percentile: 1.188872814178467
90th percentile: 1.6585559844970703
95th percentile: 1.8933975696563718
99th percentile: 2.0812708377838134
mean time: 1.0818779945373536
Pipeline stage StressChecker completed in 11.35s
jieliu-storm-7b_v1 status is now deployed due to DeploymentManager action
jieliu-storm-7b_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics