submission_id: jellywibble-lora-90k-pre_8367_v1
developer_uid: Jellywibble
status: inactive
model_repo: Jellywibble/lora_90k_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: 2024-07-04T03:49:53+00:00
model_name: nitral-ai-hathor-l3-8b-v-01_v1
model_group: Jellywibble/lora_90k_pre
num_battles: 13017
num_wins: 6883
celo_rating: 1212.89
propriety_score: 0.7066963575632257
propriety_total_count: 6287.0
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/lora_90k_pref_data_ep1
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-03
win_ratio: 0.5287700699085811
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jellywibble-lora-90k-pre-8367-v1-mkmlizer
Waiting for job on jellywibble-lora-90k-pre-8367-v1-mkmlizer to finish
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ _____ __ __ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ /___/ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ Version: 0.8.14 ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ https://mk1.ai ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ The license key for the current software has been verified as ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ belonging to: ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ Chai Research Corp. ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ║ ║
jellywibble-lora-90k-pre-8367-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Downloaded to shared memory in 76.292s
jellywibble-lora-90k-pre-8367-v1-mkmlizer: quantizing model to /dev/shm/model_cache
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 2%|▏ | 5/291 [00:00<00:06, 41.95it/s] Loading 0: 4%|▍ | 13/291 [00:00<00:04, 62.55it/s] Loading 0: 7%|▋ | 21/291 [00:00<00:03, 69.20it/s] Loading 0: 10%|▉ | 29/291 [00:00<00:03, 71.80it/s] Loading 0: 13%|█▎ | 37/291 [00:00<00:08, 30.62it/s] Loading 0: 15%|█▍ | 43/291 [00:01<00:07, 35.03it/s] Loading 0: 17%|█▋ | 50/291 [00:01<00:06, 39.32it/s] Loading 0: 20%|█▉ | 58/291 [00:01<00:04, 47.41it/s] Loading 0: 23%|██▎ | 66/291 [00:01<00:04, 54.74it/s] Loading 0: 25%|██▌ | 73/291 [00:01<00:03, 58.31it/s] Loading 0: 27%|██▋ | 80/291 [00:01<00:06, 32.22it/s] Loading 0: 30%|██▉ | 86/291 [00:02<00:05, 35.50it/s] Loading 0: 32%|███▏ | 92/291 [00:02<00:05, 37.81it/s] Loading 0: 34%|███▍ | 100/291 [00:02<00:04, 45.76it/s] Loading 0: 37%|███▋ | 109/291 [00:02<00:03, 51.45it/s] Loading 0: 40%|███▉ | 116/291 [00:02<00:03, 55.05it/s] Loading 0: 43%|████▎ | 125/291 [00:02<00:02, 62.22it/s] Loading 0: 45%|████▌ | 132/291 [00:02<00:02, 64.15it/s] Loading 0: 48%|████▊ | 139/291 [00:03<00:04, 31.79it/s] Loading 0: 51%|█████ | 147/291 [00:03<00:03, 39.13it/s] Loading 0: 54%|█████▎ | 156/291 [00:03<00:02, 45.44it/s] Loading 0: 56%|█████▌ | 163/291 [00:03<00:02, 50.15it/s] Loading 0: 58%|█████▊ | 170/291 [00:03<00:02, 51.32it/s] Loading 0: 61%|██████ | 177/291 [00:03<00:02, 52.94it/s] Loading 0: 64%|██████▍ | 186/291 [00:04<00:03, 34.59it/s] Loading 0: 66%|██████▌ | 191/291 [00:04<00:02, 36.33it/s] Loading 0: 67%|██████▋ | 196/291 [00:04<00:02, 37.97it/s] Loading 0: 70%|██████▉ | 203/291 [00:04<00:02, 41.86it/s] Loading 0: 73%|███████▎ | 212/291 [00:04<00:01, 48.54it/s] Loading 0: 76%|███████▌ | 220/291 [00:04<00:01, 54.79it/s] Loading 0: 78%|███████▊ | 227/291 [00:04<00:01, 57.24it/s] Loading 0: 80%|████████ | 234/291 [00:05<00:01, 32.47it/s] Loading 0: 82%|████████▏ | 239/291 [00:05<00:01, 34.72it/s] Loading 0: 85%|████████▌ | 248/291 [00:05<00:01, 42.27it/s] Loading 0: 88%|████████▊ | 257/291 [00:05<00:00, 48.45it/s] Loading 0: 91%|█████████▏| 266/291 [00:05<00:00, 53.48it/s] Loading 0: 95%|█████████▍| 275/291 [00:06<00:00, 57.09it/s] Loading 0: 98%|█████████▊| 286/291 [00:13<00:01, 4.13it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
jellywibble-lora-90k-pre-8367-v1-mkmlizer: quantized model in 32.017s
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Processed model Jellywibble/lora_90k_pref_data_ep1 in 108.309s
jellywibble-lora-90k-pre-8367-v1-mkmlizer: creating bucket guanaco-mkml-models
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jellywibble-lora-90k-pre-8367-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1/config.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1/special_tokens_map.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1/tokenizer_config.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1/tokenizer.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jellywibble-lora-90k-pre-8367-v1/flywheel_model.0.safetensors
jellywibble-lora-90k-pre-8367-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
jellywibble-lora-90k-pre-8367-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.
jellywibble-lora-90k-pre-8367-v1-mkmlizer: warnings.warn(
jellywibble-lora-90k-pre-8367-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`.
jellywibble-lora-90k-pre-8367-v1-mkmlizer: warnings.warn(
jellywibble-lora-90k-pre-8367-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.
jellywibble-lora-90k-pre-8367-v1-mkmlizer: warnings.warn(
jellywibble-lora-90k-pre-8367-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.
jellywibble-lora-90k-pre-8367-v1-mkmlizer: warnings.warn(
jellywibble-lora-90k-pre-8367-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()
jellywibble-lora-90k-pre-8367-v1-mkmlizer: return self.fget.__get__(instance, owner)()
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Saving duration: 0.534s
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 6.230s
jellywibble-lora-90k-pre-8367-v1-mkmlizer: creating bucket guanaco-reward-models
jellywibble-lora-90k-pre-8367-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jellywibble-lora-90k-pre-8367-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/tokenizer_config.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/special_tokens_map.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/vocab.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/merges.txt
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/config.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/tokenizer.json
jellywibble-lora-90k-pre-8367-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jellywibble-lora-90k-pre-8367-v1_reward/reward.tensors
Job jellywibble-lora-90k-pre-8367-v1-mkmlizer completed after 145.38s with status: succeeded
Stopping job with name jellywibble-lora-90k-pre-8367-v1-mkmlizer
Pipeline stage MKMLizer completed in 146.20s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.12s
Running pipeline stage ISVCDeployer
Creating inference service jellywibble-lora-90k-pre-8367-v1
Waiting for inference service jellywibble-lora-90k-pre-8367-v1 to be ready
Failed to get response for submission trace2333-joint-filtered_3791_v1: ('http://trace2333-joint-filtered-3791-v1-predictor-default.tenant-chaiml-guanaco.knative.ord1.coreweave.cloud/v1/models/GPT-J-6B-lit-v2:predict', 'upstream connect error or disconnect/reset before headers. reset reason: connection failure')
Inference service jellywibble-lora-90k-pre-8367-v1 ready after 40.20548701286316s
Pipeline stage ISVCDeployer completed in 47.04s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.054549217224121s
Received healthy response to inference request in 1.3760406970977783s
Received healthy response to inference request in 1.326026201248169s
Received healthy response to inference request in 1.3096592426300049s
Received healthy response to inference request in 1.3767001628875732s
5 requests
0 failed requests
5th percentile: 1.3129326343536376
10th percentile: 1.3162060260772706
20th percentile: 1.3227528095245362
30th percentile: 1.3360291004180909
40th percentile: 1.3560348987579345
50th percentile: 1.3760406970977783
60th percentile: 1.3763044834136964
70th percentile: 1.3765682697296142
80th percentile: 1.5122699737548828
90th percentile: 1.783409595489502
95th percentile: 1.9189794063568115
99th percentile: 2.0274352550506594
mean time: 1.4885951042175294
Pipeline stage StressChecker completed in 8.18s
jellywibble-lora-90k-pre_8367_v1 status is now deployed due to DeploymentManager action
jellywibble-lora-90k-pre_8367_v1 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics