submission_id: chaiml-sao10k-l3-rp-v3-3_v8
developer_uid: robert_irvine
status: inactive
model_repo: ChaiML/sao10k-l3-rp-v3-3
reward_repo: ChaiML/gpt2_xl_pairwise_89m_step_347634
generation_params: {'temperature': 0.95, 'top_p': 0.95, 'min_p': 0.05, 'top_k': 80, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|end_header_id|>,', '<|eot_id|>,', '\n\n{user_name}'], 'max_input_tokens': 1024, 'best_of': 8, '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|>system<|end_header_id|>\n\nrespond with drama<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n{bot_name}:', 'truncate_by_message': False}
reward_formatter: {'memory_template': '""', 'prompt_template': '""', 'bot_template': 'Bot: {message}\n', 'user_template': 'User: {message}\n', 'response_template': 'Bot:', 'truncate_by_message': False}
timestamp: 2024-07-10T18:37:13+00:00
model_name: chaiml-sao10k-l3-rp-v3-3_v8
model_group: ChaiML/sao10k-l3-rp-v3-3
num_battles: 33694
num_wins: 18850
celo_rating: 1213.09
alignment_score: None
alignment_samples: 0
propriety_score: 0.7311980180499027
propriety_total_count: 5651.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 8
max_input_tokens: 1024
max_output_tokens: 64
display_name: chaiml-sao10k-l3-rp-v3-3_v8
ineligible_reason: None
language_model: ChaiML/sao10k-l3-rp-v3-3
model_size: 8B
reward_model: ChaiML/gpt2_xl_pairwise_89m_step_347634
us_pacific_date: 2024-07-10
win_ratio: 0.5594467857778833
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer
Waiting for job on chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer to finish
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ _____ __ __ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ /___/ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ Version: 0.8.14 ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ https://mk1.ai ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ The license key for the current software has been verified as ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ belonging to: ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ Chai Research Corp. ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ║ ║
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Downloaded to shared memory in 23.254s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: quantizing model to /dev/shm/model_cache
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Saving flywheel model at /dev/shm/model_cache
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 1%| | 2/291 [00:04<11:50, 2.46s/it] Loading 0: 5%|▌ | 15/291 [00:05<01:08, 4.05it/s] Loading 0: 11%|█ | 32/291 [00:05<00:24, 10.47it/s] Loading 0: 17%|█▋ | 49/291 [00:05<00:12, 18.77it/s] Loading 0: 22%|██▏ | 63/291 [00:05<00:10, 22.40it/s] Loading 0: 26%|██▋ | 77/291 [00:05<00:06, 31.07it/s] Loading 0: 32%|███▏ | 94/291 [00:05<00:04, 43.82it/s] Loading 0: 37%|███▋ | 108/291 [00:05<00:03, 55.28it/s] Loading 0: 42%|████▏ | 123/291 [00:06<00:02, 68.52it/s] Loading 0: 48%|████▊ | 140/291 [00:06<00:01, 85.86it/s] Loading 0: 54%|█████▍ | 157/291 [00:06<00:01, 101.45it/s] Loading 0: 59%|█████▉ | 172/291 [00:06<00:01, 73.49it/s] Loading 0: 64%|██████▍ | 186/291 [00:06<00:01, 84.31it/s] Loading 0: 70%|██████▉ | 203/291 [00:06<00:00, 99.52it/s] Loading 0: 76%|███████▌ | 220/291 [00:06<00:00, 114.50it/s] Loading 0: 81%|████████ | 235/291 [00:07<00:00, 122.14it/s] Loading 0: 86%|████████▌ | 250/291 [00:07<00:00, 127.65it/s] Loading 0: 91%|█████████▏| 266/291 [00:07<00:00, 80.70it/s] Loading 0: 97%|█████████▋| 283/291 [00:07<00:00, 94.27it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: quantized model in 24.751s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Processed model ChaiML/sao10k-l3-rp-v3-3 in 48.006s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: creating bucket guanaco-mkml-models
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8/config.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8/special_tokens_map.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8/tokenizer_config.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8/tokenizer.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/chaiml-sao10k-l3-rp-v3-3-v8/flywheel_model.0.safetensors
chaiml-sao10k-l3-rp-v3-3-v8-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.
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: warnings.warn(
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Downloading shards: 0%| | 0/2 [00:00<?, ?it/s] Downloading shards: 50%|█████ | 1/2 [00:05<00:05, 5.96s/it] Downloading shards: 100%|██████████| 2/2 [00:07<00:00, 3.40s/it] Downloading shards: 100%|██████████| 2/2 [00:07<00:00, 3.78s/it]
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s] Loading checkpoint shards: 50%|█████ | 1/2 [00:00<00:00, 1.74it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 2.92it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 2.65it/s]
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Saving duration: 1.966s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Processed model ChaiML/gpt2_xl_pairwise_89m_step_347634 in 11.777s
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: creating bucket guanaco-reward-models
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: Bucket 's3://guanaco-reward-models/' created
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/special_tokens_map.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/config.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/tokenizer_config.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/vocab.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/tokenizer.json
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/merges.txt
chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/chaiml-sao10k-l3-rp-v3-3-v8_reward/reward.tensors
Job chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer completed after 104.28s with status: succeeded
Stopping job with name chaiml-sao10k-l3-rp-v3-3-v8-mkmlizer
Pipeline stage MKMLizer completed in 105.21s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.12s
Running pipeline stage ISVCDeployer
Creating inference service chaiml-sao10k-l3-rp-v3-3-v8
Waiting for inference service chaiml-sao10k-l3-rp-v3-3-v8 to be ready
Inference service chaiml-sao10k-l3-rp-v3-3-v8 ready after 50.2649986743927s
Pipeline stage ISVCDeployer completed in 57.73s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.916551113128662s
Received healthy response to inference request in 1.2633264064788818s
Received healthy response to inference request in 1.2668967247009277s
Received healthy response to inference request in 1.2414028644561768s
Received healthy response to inference request in 1.2944557666778564s
5 requests
0 failed requests
5th percentile: 1.2457875728607177
10th percentile: 1.2501722812652587
20th percentile: 1.2589416980743409
30th percentile: 1.264040470123291
40th percentile: 1.2654685974121094
50th percentile: 1.2668967247009277
60th percentile: 1.2779203414916993
70th percentile: 1.2889439582824707
80th percentile: 1.4188748359680177
90th percentile: 1.6677129745483399
95th percentile: 1.7921320438385009
99th percentile: 1.8916672992706298
mean time: 1.396526575088501
Pipeline stage StressChecker completed in 7.74s
chaiml-sao10k-l3-rp-v3-3_v8 status is now deployed due to DeploymentManager action
chaiml-sao10k-l3-rp-v3-3_v8 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics