developer_uid: chace9580
submission_id: jic062-instruct-fail_v1
model_name: jic062-instruct-fail_v1
model_group: jic062/instruct_fail
status: torndown
timestamp: 2024-08-07T02:40:40+00:00
num_battles: 12442
num_wins: 6222
celo_rating: 1215.92
family_friendly_score: 0.0
submission_type: basic
model_repo: jic062/instruct_fail
model_architecture: LlamaForCausalLM
reward_repo: ChaiML/gpt2_xl_pairwise_89m_step_347634
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: jic062-instruct-fail_v1
is_internal_developer: False
language_model: jic062/instruct_fail
model_size: 8B
ranking_group: single
us_pacific_date: 2024-08-06
win_ratio: 0.5000803729303971
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|end_of_text|>', '|eot_id|'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64, 'reward_max_token_input': 256}
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: {'bot_template': '{bot_name}: {message}\n', 'memory_template': "{bot_name}'s Persona: {memory}\n####\n", 'prompt_template': '{prompt}\n<START>\n', 'response_template': '{bot_name}:', 'truncate_by_message': False, 'user_template': '{user_name}: {message}\n'}
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jic062-instruct-fail-v1-mkmlizer
Waiting for job on jic062-instruct-fail-v1-mkmlizer to finish
jic062-instruct-fail-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jic062-instruct-fail-v1-mkmlizer: ║     _____            __           __                                ║
jic062-instruct-fail-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
jic062-instruct-fail-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
jic062-instruct-fail-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
jic062-instruct-fail-v1-mkmlizer: ║       /___/                                                         ║
jic062-instruct-fail-v1-mkmlizer: ║                                                                     ║
jic062-instruct-fail-v1-mkmlizer: ║  Version: 0.9.9                                                     ║
jic062-instruct-fail-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
jic062-instruct-fail-v1-mkmlizer: ║  https://mk1.ai                                                     ║
jic062-instruct-fail-v1-mkmlizer: ║                                                                     ║
jic062-instruct-fail-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
jic062-instruct-fail-v1-mkmlizer: ║  belonging to:                                                      ║
jic062-instruct-fail-v1-mkmlizer: ║                                                                     ║
jic062-instruct-fail-v1-mkmlizer: ║  Chai Research Corp.                                                ║
jic062-instruct-fail-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
jic062-instruct-fail-v1-mkmlizer: ║  Expiration: 2024-10-15 23:59:59                                    ║
jic062-instruct-fail-v1-mkmlizer: ║                                                                     ║
jic062-instruct-fail-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jic062-instruct-fail-v1-mkmlizer: Downloaded to shared memory in 31.172s
jic062-instruct-fail-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmp287y2f4g, device:0
jic062-instruct-fail-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jic062-instruct-fail-v1-mkmlizer: quantized model in 25.625s
jic062-instruct-fail-v1-mkmlizer: Processed model jic062/instruct_fail in 56.797s
jic062-instruct-fail-v1-mkmlizer: creating bucket guanaco-mkml-models
jic062-instruct-fail-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jic062-instruct-fail-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jic062-instruct-fail-v1
jic062-instruct-fail-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jic062-instruct-fail-v1/config.json
jic062-instruct-fail-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jic062-instruct-fail-v1/special_tokens_map.json
jic062-instruct-fail-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jic062-instruct-fail-v1/tokenizer_config.json
jic062-instruct-fail-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jic062-instruct-fail-v1/tokenizer.json
jic062-instruct-fail-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jic062-instruct-fail-v1/flywheel_model.0.safetensors
jic062-instruct-fail-v1-mkmlizer: loading reward model from ChaiML/gpt2_xl_pairwise_89m_step_347634
jic062-instruct-fail-v1-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   2%|▏         | 7/291 [00:00<00:05, 52.24it/s]
Loading 0:   5%|▌         | 16/291 [00:00<00:03, 71.62it/s]
Loading 0:   9%|▊         | 25/291 [00:00<00:03, 79.20it/s]
Loading 0:  12%|█▏        | 34/291 [00:00<00:03, 80.25it/s]
Loading 0:  15%|█▍        | 43/291 [00:00<00:02, 82.96it/s]
Loading 0:  18%|█▊        | 52/291 [00:00<00:02, 82.06it/s]
Loading 0:  21%|██        | 61/291 [00:00<00:02, 79.53it/s]
Loading 0:  24%|██▍       | 70/291 [00:00<00:02, 78.94it/s]
Loading 0:  27%|██▋       | 79/291 [00:01<00:02, 77.34it/s]
Loading 0:  30%|██▉       | 87/291 [00:02<00:09, 21.07it/s]
Loading 0:  32%|███▏      | 94/291 [00:02<00:07, 25.62it/s]
Loading 0:  35%|███▌      | 103/291 [00:02<00:05, 32.63it/s]
Loading 0:  38%|███▊      | 112/291 [00:02<00:04, 40.79it/s]
Loading 0:  42%|████▏     | 121/291 [00:02<00:03, 48.74it/s]
Loading 0:  45%|████▍     | 130/291 [00:02<00:02, 55.32it/s]
Loading 0:  48%|████▊     | 139/291 [00:02<00:02, 60.71it/s]
Loading 0:  51%|█████     | 148/291 [00:02<00:02, 67.33it/s]
Loading 0:  54%|█████▍    | 157/291 [00:02<00:01, 71.30it/s]
Loading 0:  57%|█████▋    | 166/291 [00:03<00:01, 74.45it/s]
Loading 0:  60%|██████    | 175/291 [00:03<00:01, 76.89it/s]
Loading 0:  63%|██████▎   | 184/291 [00:03<00:01, 77.23it/s]
Loading 0:  66%|██████▋   | 193/291 [00:04<00:04, 21.96it/s]
Loading 0:  69%|██████▉   | 202/291 [00:04<00:03, 28.27it/s]
Loading 0:  73%|███████▎  | 211/291 [00:04<00:02, 35.59it/s]
Loading 0:  76%|███████▌  | 220/291 [00:04<00:01, 43.26it/s]
Loading 0:  79%|███████▊  | 229/291 [00:04<00:01, 50.88it/s]
Loading 0:  82%|████████▏ | 238/291 [00:04<00:00, 57.45it/s]
Loading 0:  85%|████████▍ | 247/291 [00:05<00:00, 63.87it/s]
Loading 0:  88%|████████▊ | 256/291 [00:05<00:00, 69.92it/s]
Loading 0:  91%|█████████ | 265/291 [00:05<00:00, 73.98it/s]
Loading 0:  94%|█████████▍| 274/291 [00:05<00:00, 77.11it/s]
Loading 0:  97%|█████████▋| 283/291 [00:05<00:00, 78.52it/s]
                                                            
/opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:957: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jic062-instruct-fail-v1-mkmlizer:   warnings.warn(
jic062-instruct-fail-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:785: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jic062-instruct-fail-v1-mkmlizer:   warnings.warn(
jic062-instruct-fail-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py:469: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jic062-instruct-fail-v1-mkmlizer:   warnings.warn(
jic062-instruct-fail-v1-mkmlizer: 
Downloading shards:   0%|          | 0/2 [00:00<?, ?it/s]
Downloading shards:  50%|█████     | 1/2 [00:05<00:05,  5.30s/it]
Downloading shards: 100%|██████████| 2/2 [00:08<00:00,  4.07s/it]
Downloading shards: 100%|██████████| 2/2 [00:08<00:00,  4.26s/it]
jic062-instruct-fail-v1-mkmlizer: 
Loading checkpoint shards:   0%|          | 0/2 [00:00<?, ?it/s]
Loading checkpoint shards:  50%|█████     | 1/2 [00:00<00:00,  2.42it/s]
Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00,  3.98it/s]
Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00,  3.63it/s]
jic062-instruct-fail-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jic062-instruct-fail-v1-mkmlizer: Saving duration: 1.360s
jic062-instruct-fail-v1-mkmlizer: Processed model ChaiML/gpt2_xl_pairwise_89m_step_347634 in 13.406s
jic062-instruct-fail-v1-mkmlizer: creating bucket guanaco-reward-models
jic062-instruct-fail-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jic062-instruct-fail-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jic062-instruct-fail-v1_reward
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/config.json
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/tokenizer_config.json
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/special_tokens_map.json
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/merges.txt
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/vocab.json
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/tokenizer.json
jic062-instruct-fail-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jic062-instruct-fail-v1_reward/reward.tensors
Job jic062-instruct-fail-v1-mkmlizer completed after 104.52s with status: succeeded
Stopping job with name jic062-instruct-fail-v1-mkmlizer
Pipeline stage MKMLizer completed in 105.37s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service jic062-instruct-fail-v1
Waiting for inference service jic062-instruct-fail-v1 to be ready
Failed to get response for submission cycy233-l3-s-v1-c2_v1: ('http://cycy233-l3-s-v1-c2-v1-predictor.tenant-chaiml-guanaco.k.chaiverse.com/v1/models/GPT-J-6B-lit-v2:predict', 'read tcp 127.0.0.1:59114->127.0.0.1:8080: read: connection reset by peer\n')
Inference service jic062-instruct-fail-v1 ready after 191.35445713996887s
Pipeline stage ISVCDeployer completed in 192.96s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.2315430641174316s
Received healthy response to inference request in 1.5044214725494385s
Received healthy response to inference request in 1.4439451694488525s
Received healthy response to inference request in 1.4291324615478516s
Received healthy response to inference request in 1.455031156539917s
5 requests
0 failed requests
5th percentile: 1.4320950031280517
10th percentile: 1.4350575447082519
20th percentile: 1.4409826278686524
30th percentile: 1.4461623668670653
40th percentile: 1.4505967617034912
50th percentile: 1.455031156539917
60th percentile: 1.4747872829437256
70th percentile: 1.4945434093475343
80th percentile: 1.6498457908630373
90th percentile: 1.9406944274902345
95th percentile: 2.086118745803833
99th percentile: 2.202458200454712
mean time: 1.6128146648406982
Pipeline stage StressChecker completed in 8.73s
jic062-instruct-fail_v1 status is now deployed due to DeploymentManager action
jic062-instruct-fail_v1 status is now inactive due to auto deactivation removed underperforming models
jic062-instruct-fail_v1 status is now torndown due to DeploymentManager action