submission_id: invisietch-etherealrainb_9091_v2
developer_uid: invisietch
status: inactive
model_repo: invisietch/EtherealRainbow-v0.3-8B
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.9, 'top_p': 0.95, 'min_p': 0.025, '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': "<|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: 2024-07-08T03:52:42+00:00
model_name: invisietch-etherealrainbow-v0_3
model_group: invisietch/EtherealRainb
num_battles: 31613
num_wins: 15310
celo_rating: 1187.16
propriety_score: 0.7164234250877517
propriety_total_count: 5413.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 4
max_input_tokens: 512
max_output_tokens: 64
display_name: invisietch-etherealrainbow-v0_3
ineligible_reason: None
language_model: invisietch/EtherealRainbow-v0.3-8B
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
us_pacific_date: 2024-07-07
win_ratio: 0.4842944358333597
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name invisietch-etherealrainb-9091-v2-mkmlizer
Waiting for job on invisietch-etherealrainb-9091-v2-mkmlizer to finish
invisietch-etherealrainb-9091-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
invisietch-etherealrainb-9091-v2-mkmlizer: ║ _____ __ __ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ /___/ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ Version: 0.8.14 ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ https://mk1.ai ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ The license key for the current software has been verified as ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ belonging to: ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ Chai Research Corp. ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
invisietch-etherealrainb-9091-v2-mkmlizer: ║ ║
invisietch-etherealrainb-9091-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
invisietch-etherealrainb-9091-v2-mkmlizer: Downloaded to shared memory in 27.582s
invisietch-etherealrainb-9091-v2-mkmlizer: quantizing model to /dev/shm/model_cache
invisietch-etherealrainb-9091-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
invisietch-etherealrainb-9091-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:05<13:26, 2.79s/it] Loading 0: 5%|▍ | 14/291 [00:05<01:23, 3.33it/s] Loading 0: 8%|▊ | 22/291 [00:06<00:48, 5.50it/s] Loading 0: 11%|█▏ | 33/291 [00:06<00:25, 9.98it/s] Loading 0: 17%|█▋ | 49/291 [00:06<00:13, 18.56it/s] Loading 0: 21%|██ | 60/291 [00:06<00:09, 25.35it/s] Loading 0: 26%|██▌ | 76/291 [00:06<00:05, 38.22it/s] Loading 0: 30%|███ | 88/291 [00:06<00:05, 36.30it/s] Loading 0: 35%|███▌ | 103/291 [00:06<00:03, 49.09it/s] Loading 0: 39%|███▉ | 114/291 [00:07<00:03, 53.59it/s] Loading 0: 45%|████▍ | 130/291 [00:07<00:02, 69.24it/s] Loading 0: 48%|████▊ | 141/291 [00:07<00:02, 51.79it/s] Loading 0: 54%|█████▍ | 157/291 [00:07<00:02, 66.93it/s] Loading 0: 58%|█████▊ | 168/291 [00:07<00:01, 72.82it/s] Loading 0: 63%|██████▎ | 184/291 [00:07<00:01, 88.88it/s] Loading 0: 67%|██████▋ | 196/291 [00:08<00:01, 93.66it/s] Loading 0: 71%|███████▏ | 208/291 [00:08<00:01, 65.25it/s] Loading 0: 76%|███████▌ | 221/291 [00:08<00:00, 74.57it/s] Loading 0: 80%|████████ | 233/291 [00:08<00:00, 83.55it/s] Loading 0: 85%|████████▍ | 247/291 [00:08<00:00, 94.27it/s] Loading 0: 89%|████████▉ | 259/291 [00:08<00:00, 97.35it/s] Loading 0: 93%|█████████▎| 270/291 [00:09<00:00, 64.27it/s] Loading 0: 98%|█████████▊| 284/291 [00:09<00:00, 75.34it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
invisietch-etherealrainb-9091-v2-mkmlizer: quantized model in 29.549s
invisietch-etherealrainb-9091-v2-mkmlizer: Processed model invisietch/EtherealRainbow-v0.3-8B in 57.132s
invisietch-etherealrainb-9091-v2-mkmlizer: creating bucket guanaco-mkml-models
invisietch-etherealrainb-9091-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
invisietch-etherealrainb-9091-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2
invisietch-etherealrainb-9091-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2/special_tokens_map.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2/config.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2/tokenizer_config.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2/tokenizer.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/invisietch-etherealrainb-9091-v2/flywheel_model.0.safetensors
invisietch-etherealrainb-9091-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
invisietch-etherealrainb-9091-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.
invisietch-etherealrainb-9091-v2-mkmlizer: warnings.warn(
invisietch-etherealrainb-9091-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`.
invisietch-etherealrainb-9091-v2-mkmlizer: warnings.warn(
invisietch-etherealrainb-9091-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.
invisietch-etherealrainb-9091-v2-mkmlizer: warnings.warn(
invisietch-etherealrainb-9091-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.
invisietch-etherealrainb-9091-v2-mkmlizer: warnings.warn(
invisietch-etherealrainb-9091-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()
invisietch-etherealrainb-9091-v2-mkmlizer: return self.fget.__get__(instance, owner)()
invisietch-etherealrainb-9091-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
invisietch-etherealrainb-9091-v2-mkmlizer: Saving duration: 0.511s
invisietch-etherealrainb-9091-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 14.746s
invisietch-etherealrainb-9091-v2-mkmlizer: creating bucket guanaco-reward-models
invisietch-etherealrainb-9091-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
invisietch-etherealrainb-9091-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/tokenizer_config.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/special_tokens_map.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/merges.txt
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/vocab.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/tokenizer.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/config.json
invisietch-etherealrainb-9091-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/invisietch-etherealrainb-9091-v2_reward/reward.tensors
Job invisietch-etherealrainb-9091-v2-mkmlizer completed after 93.97s with status: succeeded
Stopping job with name invisietch-etherealrainb-9091-v2-mkmlizer
Pipeline stage MKMLizer completed in 95.10s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.12s
Running pipeline stage ISVCDeployer
Creating inference service invisietch-etherealrainb-9091-v2
Waiting for inference service invisietch-etherealrainb-9091-v2 to be ready
Connection pool is full, discarding connection: %s
Inference service invisietch-etherealrainb-9091-v2 ready after 40.180651903152466s
Pipeline stage ISVCDeployer completed in 47.16s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.9555141925811768s
Received healthy response to inference request in 1.1298036575317383s
Received healthy response to inference request in 1.1434719562530518s
Received healthy response to inference request in 1.1044220924377441s
Received healthy response to inference request in 1.1238105297088623s
5 requests
0 failed requests
5th percentile: 1.1082997798919678
10th percentile: 1.1121774673461915
20th percentile: 1.1199328422546386
30th percentile: 1.1250091552734376
40th percentile: 1.127406406402588
50th percentile: 1.1298036575317383
60th percentile: 1.1352709770202636
70th percentile: 1.140738296508789
80th percentile: 1.3058804035186768
90th percentile: 1.6306972980499268
95th percentile: 1.7931057453155517
99th percentile: 1.9230325031280517
mean time: 1.2914044857025146
Pipeline stage StressChecker completed in 7.22s
invisietch-etherealrainb_9091_v2 status is now deployed due to DeploymentManager action
invisietch-etherealrainb_9091_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics