submission_id: jellywibble-qlora-120k-p_350_v2
developer_uid: Jellywibble
status: inactive
model_repo: Jellywibble/qlora_120k_pref_data_ep1
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.95, 'top_p': 1.0, 'min_p': 0.08, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|eot_id|>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<|begin_of_text|><|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: 2024-07-03T22:39:02+00:00
model_name: nitral-ai-hathor-l3-8b-v-01_v1
model_group: Jellywibble/qlora_120k_p
num_battles: 13568
num_wins: 7078
celo_rating: 1203.54
propriety_score: 0.7300565326633166
propriety_total_count: 6368.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: nitral-ai-hathor-l3-8b-v-01_v1
ineligible_reason: None
language_model: Jellywibble/qlora_120k_pref_data_ep1
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-03
win_ratio: 0.5216686320754716
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jellywibble-qlora-120k-p-350-v2-mkmlizer
Waiting for job on jellywibble-qlora-120k-p-350-v2-mkmlizer to finish
jellywibble-qlora-120k-p-350-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ _____ __ __ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ /___/ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ Version: 0.8.14 ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ https://mk1.ai ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ The license key for the current software has been verified as ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ belonging to: ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ Chai Research Corp. ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ║ ║
jellywibble-qlora-120k-p-350-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jellywibble-qlora-120k-p-350-v2-mkmlizer: Downloaded to shared memory in 39.020s
jellywibble-qlora-120k-p-350-v2-mkmlizer: quantizing model to /dev/shm/model_cache
jellywibble-qlora-120k-p-350-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jellywibble-qlora-120k-p-350-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 4%|▍ | 13/291 [00:00<00:02, 112.29it/s] Loading 0: 9%|▉ | 27/291 [00:00<00:02, 124.21it/s] Loading 0: 14%|█▎ | 40/291 [00:00<00:02, 120.08it/s] Loading 0: 18%|█▊ | 53/291 [00:00<00:01, 122.87it/s] Loading 0: 23%|██▎ | 66/291 [00:00<00:01, 124.76it/s] Loading 0: 27%|██▋ | 79/291 [00:00<00:01, 120.61it/s] Loading 0: 32%|███▏ | 92/291 [00:01<00:03, 54.51it/s] Loading 0: 35%|███▌ | 102/291 [00:01<00:03, 60.80it/s] Loading 0: 38%|███▊ | 111/291 [00:01<00:03, 52.31it/s] Loading 0: 42%|████▏ | 121/291 [00:01<00:02, 60.33it/s] Loading 0: 46%|████▌ | 133/291 [00:01<00:02, 72.10it/s] Loading 0: 51%|█████ | 147/291 [00:01<00:01, 85.00it/s] Loading 0: 54%|█████▍ | 158/291 [00:01<00:01, 87.12it/s] Loading 0: 59%|█████▉ | 173/291 [00:02<00:01, 99.70it/s] Loading 0: 64%|██████▍ | 187/291 [00:02<00:01, 54.15it/s] Loading 0: 68%|██████▊ | 197/291 [00:02<00:01, 60.75it/s] Loading 0: 73%|███████▎ | 211/291 [00:02<00:01, 72.58it/s] Loading 0: 77%|███████▋ | 225/291 [00:02<00:00, 85.70it/s] Loading 0: 82%|████████▏ | 238/291 [00:03<00:00, 92.25it/s] Loading 0: 87%|████████▋ | 253/291 [00:03<00:00, 105.54it/s] Loading 0: 91%|█████████▏| 266/291 [00:03<00:00, 105.59it/s] Loading 0: 97%|█████████▋| 281/291 [00:03<00:00, 116.65it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
jellywibble-qlora-120k-p-350-v2-mkmlizer: quantized model in 29.476s
jellywibble-qlora-120k-p-350-v2-mkmlizer: Processed model Jellywibble/qlora_120k_pref_data_ep1 in 68.496s
jellywibble-qlora-120k-p-350-v2-mkmlizer: creating bucket guanaco-mkml-models
jellywibble-qlora-120k-p-350-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jellywibble-qlora-120k-p-350-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2/config.json
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2/tokenizer_config.json
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2/special_tokens_map.json
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2/tokenizer.json
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jellywibble-qlora-120k-p-350-v2/flywheel_model.0.safetensors
jellywibble-qlora-120k-p-350-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
jellywibble-qlora-120k-p-350-v2-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.
jellywibble-qlora-120k-p-350-v2-mkmlizer: warnings.warn(
jellywibble-qlora-120k-p-350-v2-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`.
jellywibble-qlora-120k-p-350-v2-mkmlizer: warnings.warn(
jellywibble-qlora-120k-p-350-v2-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.
jellywibble-qlora-120k-p-350-v2-mkmlizer: warnings.warn(
jellywibble-qlora-120k-p-350-v2-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.
jellywibble-qlora-120k-p-350-v2-mkmlizer: warnings.warn(
jellywibble-qlora-120k-p-350-v2-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()
jellywibble-qlora-120k-p-350-v2-mkmlizer: return self.fget.__get__(instance, owner)()
jellywibble-qlora-120k-p-350-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jellywibble-qlora-120k-p-350-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jellywibble-qlora-120k-p-350-v2_reward/reward.tensors
Job jellywibble-qlora-120k-p-350-v2-mkmlizer completed after 104.45s with status: succeeded
Stopping job with name jellywibble-qlora-120k-p-350-v2-mkmlizer
Pipeline stage MKMLizer completed in 105.36s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service jellywibble-qlora-120k-p-350-v2
Waiting for inference service jellywibble-qlora-120k-p-350-v2 to be ready
Inference service jellywibble-qlora-120k-p-350-v2 ready after 40.556541204452515s
Pipeline stage ISVCDeployer completed in 47.82s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.1371915340423584s
Received healthy response to inference request in 1.336545705795288s
Received healthy response to inference request in 1.3146641254425049s
Received healthy response to inference request in 1.275390863418579s
Received healthy response to inference request in 1.3369436264038086s
5 requests
0 failed requests
5th percentile: 1.2832455158233642
10th percentile: 1.2911001682281493
20th percentile: 1.3068094730377198
30th percentile: 1.3190404415130614
40th percentile: 1.3277930736541748
50th percentile: 1.336545705795288
60th percentile: 1.3367048740386962
70th percentile: 1.3368640422821045
80th percentile: 1.4969932079315187
90th percentile: 1.8170923709869387
95th percentile: 1.9771419525146483
99th percentile: 2.1051816177368163
mean time: 1.4801471710205079
Pipeline stage StressChecker completed in 8.15s
jellywibble-qlora-120k-p_350_v2 status is now deployed due to DeploymentManager action
jellywibble-qlora-120k-p_350_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics