submission_id: cgato-borkeddpotest_v1
developer_uid: c.gato
status: inactive
model_repo: cgato/BorkedDPOTest
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.9, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 200, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', 'You:'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<|im_start|>system\n{bot_name}'s Persona: {memory}\n", 'prompt_template': '***\n{prompt}\n***<|im_end|>\n', 'bot_template': '<|im_start|>{bot_name}\n{bot_name}: {message}<|im_end|>\n', 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n', 'response_template': '<|im_start|>{bot_name}\n{bot_name}:', 'truncate_by_message': True}
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': True}
timestamp: 2024-06-03T19:39:35+00:00
model_name: cgato-borkeddpotest_v1
model_eval_status: success
model_group: cgato/BorkedDPOTest
num_battles: 9680
num_wins: 5300
celo_rating: 1201.48
safety_score: 0.95
propriety_score: 0.6575875486381323
propriety_total_count: 257.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030277632.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: cgato-borkeddpotest_v1
ineligible_reason: propriety_total_count < 5000
language_model: cgato/BorkedDPOTest
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-06-03
win_ratio: 0.5475206611570248
Resubmit model
Running pipeline stage MKMLizer
Starting job with name cgato-borkeddpotest-v1-mkmlizer
Waiting for job on cgato-borkeddpotest-v1-mkmlizer to finish
cgato-borkeddpotest-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.
cgato-borkeddpotest-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
cgato-borkeddpotest-v1-mkmlizer: Downloaded to shared memory in 63.392s
cgato-borkeddpotest-v1-mkmlizer: quantizing model to /dev/shm/model_cache
cgato-borkeddpotest-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
cgato-borkeddpotest-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 29%|██▊ | 83/291 [00:01<00:03, 68.82it/s] Loading 0: 64%|██████▍ | 187/291 [00:02<00:01, 77.54it/s] Loading 0: 99%|█████████▊| 287/291 [00:02<00:00, 109.66it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
cgato-borkeddpotest-v1-mkmlizer: quantized model in 24.499s
cgato-borkeddpotest-v1-mkmlizer: Processed model cgato/BorkedDPOTest in 90.434s
cgato-borkeddpotest-v1-mkmlizer: creating bucket guanaco-mkml-models
cgato-borkeddpotest-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
cgato-borkeddpotest-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/cgato-borkeddpotest-v1
cgato-borkeddpotest-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/cgato-borkeddpotest-v1/config.json
cgato-borkeddpotest-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/cgato-borkeddpotest-v1/tokenizer_config.json
cgato-borkeddpotest-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/cgato-borkeddpotest-v1/special_tokens_map.json
cgato-borkeddpotest-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/cgato-borkeddpotest-v1/tokenizer.json
cgato-borkeddpotest-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/cgato-borkeddpotest-v1/flywheel_model.0.safetensors
cgato-borkeddpotest-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
cgato-borkeddpotest-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.
cgato-borkeddpotest-v1-mkmlizer: warnings.warn(
cgato-borkeddpotest-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.
cgato-borkeddpotest-v1-mkmlizer: warnings.warn(
cgato-borkeddpotest-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.
cgato-borkeddpotest-v1-mkmlizer: warnings.warn(
cgato-borkeddpotest-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()
cgato-borkeddpotest-v1-mkmlizer: return self.fget.__get__(instance, owner)()
cgato-borkeddpotest-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
cgato-borkeddpotest-v1-mkmlizer: Saving duration: 0.474s
cgato-borkeddpotest-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 22.625s
cgato-borkeddpotest-v1-mkmlizer: creating bucket guanaco-reward-models
cgato-borkeddpotest-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
cgato-borkeddpotest-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/special_tokens_map.json
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/merges.txt
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/tokenizer_config.json
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/config.json
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/vocab.json
cgato-borkeddpotest-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/cgato-borkeddpotest-v1_reward/tokenizer.json
Job cgato-borkeddpotest-v1-mkmlizer completed after 165.25s with status: succeeded
Stopping job with name cgato-borkeddpotest-v1-mkmlizer
Pipeline stage MKMLizer completed in 165.65s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.08s
Running pipeline stage ISVCDeployer
Creating inference service cgato-borkeddpotest-v1
Waiting for inference service cgato-borkeddpotest-v1 to be ready
Failed to get response for submission blend_ripuf_2024-05-23: ('http://hastagaras-esekembrew-0-3-v7-predictor-default.tenant-chaiml-guanaco.knative.ord1.coreweave.cloud/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:34498->127.0.0.1:8080: read: connection reset by peer\n')
Inference service cgato-borkeddpotest-v1 ready after 100.70156621932983s
Pipeline stage ISVCDeployer completed in 106.29s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.3306496143341064s
Received healthy response to inference request in 1.3395206928253174s
Received healthy response to inference request in 1.2945115566253662s
Received healthy response to inference request in 1.3088083267211914s
Received healthy response to inference request in 1.291520118713379s
5 requests
0 failed requests
5th percentile: 1.2921184062957765
10th percentile: 1.2927166938781738
20th percentile: 1.2939132690429687
30th percentile: 1.2973709106445312
40th percentile: 1.3030896186828613
50th percentile: 1.3088083267211914
60th percentile: 1.3210932731628418
70th percentile: 1.333378219604492
80th percentile: 1.5377464771270755
90th percentile: 1.934198045730591
95th percentile: 2.1324238300323484
99th percentile: 2.291004457473755
mean time: 1.513002061843872
Pipeline stage StressChecker completed in 8.16s
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
cgato-borkeddpotest_v1 status is now deployed due to DeploymentManager action
cgato-borkeddpotest_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics