trace2333-joint-filtered_3791

submission_id: trace2333-joint-filtered_3791_v4

developer_uid: Trace2333

best_of: 16

celo_rating: 1148.16

display_name: trace2333-joint-filtered_3791_v4

family_friendly_score: 0.0

formatter: {'memory_template': "{bot_name}'s Persona####: {memory}", 'prompt_template': '<|im_start|>user\n{prompt}<|im_end|>\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'user_template': 'You: {message}\\\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}

generation_params: {'temperature': 1.25, 'top_p': 1.0, 'min_p': 0.12, 'top_k': 200, 'presence_penalty': 0.2, 'frequency_penalty': 0.2, 'stopping_words': ['</s>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}

is_internal_developer: False

language_model: Trace2333/joint_filtered_llama2_v0

max_input_tokens: 512

max_output_tokens: 64

model_architecture: LlamaForCausalLM

model_group: Trace2333/joint_filtered

model_name: trace2333-joint-filtered_3791_v4

model_num_parameters: 6738415616.0

model_repo: Trace2333/joint_filtered_llama2_v0

model_size: 7B

num_battles: 28013

num_wins: 12556

ranking_group: single

reward_formatter: {'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'memory_template': '<|im_start|>system\n{memory}<|im_end|>\n', 'prompt_template': '<|im_start|>user\n{prompt}<|im_end|>\n', 'response_template': '{bot_name}:', 'truncate_by_message': True, 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n'}

reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2

status: torndown

submission_type: basic

timestamp: 2024-07-04T12:50:41+00:00

us_pacific_date: 2024-07-04

win_ratio: 0.4482204690679327

Resubmit model

Running pipeline stage MKMLizer
Starting job with name trace2333-joint-filtered-3791-v4-mkmlizer
Waiting for job on trace2333-joint-filtered-3791-v4-mkmlizer to finish
trace2333-joint-filtered-3791-v4-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
trace2333-joint-filtered-3791-v4-mkmlizer: ║     _____            __           __                                ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║       /___/                                                         ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║                                                                     ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  Version: 0.8.14                                                    ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  https://mk1.ai                                                     ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║                                                                     ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  The license key for the current software has been verified as      ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  belonging to:                                                      ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║                                                                     ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  Chai Research Corp.                                                ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║  Expiration: 2024-07-15 23:59:59                                    ║
trace2333-joint-filtered-3791-v4-mkmlizer: ║                                                                     ║
trace2333-joint-filtered-3791-v4-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
Connection pool is full, discarding connection: %s
trace2333-joint-filtered-3791-v4-mkmlizer: Downloaded to shared memory in 74.661s
trace2333-joint-filtered-3791-v4-mkmlizer: quantizing model to /dev/shm/model_cache
trace2333-joint-filtered-3791-v4-mkmlizer: Saving flywheel model at /dev/shm/model_cache
trace2333-joint-filtered-3791-v4-mkmlizer: quantized model in 27.338s
trace2333-joint-filtered-3791-v4-mkmlizer: Processed model Trace2333/joint_filtered_llama2_v0 in 101.999s
trace2333-joint-filtered-3791-v4-mkmlizer: creating bucket guanaco-mkml-models
trace2333-joint-filtered-3791-v4-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
trace2333-joint-filtered-3791-v4-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/config.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/special_tokens_map.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/tokenizer.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/tokenizer.model
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/tokenizer_config.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/trace2333-joint-filtered-3791-v4/flywheel_model.0.safetensors
trace2333-joint-filtered-3791-v4-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
trace2333-joint-filtered-3791-v4-mkmlizer: 
Loading 0:   0%|          | 0/291 [00:00<?, ?it/s]
Loading 0:   1%|          | 3/291 [00:00<00:12, 22.59it/s]
Loading 0:   2%|▏         | 6/291 [00:00<00:19, 14.31it/s]
Loading 0:   3%|▎         | 8/291 [00:00<00:26, 10.58it/s]
Loading 0:   3%|▎         | 10/291 [00:00<00:32,  8.64it/s]
Loading 0:   4%|▍         | 12/291 [00:01<00:42,  6.64it/s]
Loading 0:   4%|▍         | 13/291 [00:01<00:55,  5.01it/s]
Loading 0:   5%|▍         | 14/291 [00:02<01:03,  4.36it/s]
Loading 0:   5%|▌         | 16/291 [00:02<00:48,  5.70it/s]
Loading 0:   6%|▌         | 17/291 [00:02<00:45,  6.04it/s]
Loading 0:   6%|▌         | 18/291 [00:02<00:43,  6.29it/s]
Loading 0:   7%|▋         | 19/291 [00:02<00:42,  6.37it/s]
Loading 0:   7%|▋         | 21/291 [00:03<00:43,  6.27it/s]
Loading 0:   9%|▊         | 25/291 [00:03<00:23, 11.41it/s]
Loading 0:  11%|█         | 32/291 [00:03<00:11, 21.80it/s]
Loading 0:  14%|█▍        | 41/291 [00:03<00:07, 34.59it/s]
Loading 0:  18%|█▊        | 51/291 [00:03<00:07, 31.05it/s]
Loading 0:  20%|█▉        | 57/291 [00:03<00:06, 35.41it/s]
Loading 0:  23%|██▎       | 66/291 [00:04<00:05, 44.80it/s]
Loading 0:  25%|██▍       | 72/291 [00:04<00:04, 47.68it/s]
Loading 0:  28%|██▊       | 81/291 [00:04<00:03, 55.58it/s]
Loading 0:  31%|███       | 90/291 [00:04<00:03, 61.52it/s]
Loading 0:  34%|███▍      | 99/291 [00:04<00:02, 65.18it/s]
Loading 0:  36%|███▋      | 106/291 [00:04<00:04, 41.75it/s]
Loading 0:  38%|███▊      | 112/291 [00:04<00:04, 44.56it/s]
Loading 0:  42%|████▏     | 121/291 [00:05<00:03, 52.35it/s]
Loading 0:  45%|████▍     | 130/291 [00:05<00:02, 58.96it/s]
Loading 0:  48%|████▊     | 139/291 [00:05<00:02, 64.07it/s]
Loading 0:  51%|█████     | 148/291 [00:05<00:02, 67.50it/s]
Loading 0:  54%|█████▍    | 158/291 [00:05<00:01, 74.47it/s]
Loading 0:  57%|█████▋    | 166/291 [00:05<00:02, 43.08it/s]
Loading 0:  60%|██████    | 175/291 [00:05<00:02, 49.80it/s]
Loading 0:  63%|██████▎   | 184/291 [00:06<00:01, 55.64it/s]
Loading 0:  66%|██████▋   | 193/291 [00:06<00:01, 60.98it/s]
Loading 0:  69%|██████▉   | 202/291 [00:06<00:01, 64.93it/s]
Loading 0:  73%|███████▎  | 212/291 [00:06<00:01, 72.70it/s]
Loading 0:  76%|███████▌  | 221/291 [00:06<00:01, 42.80it/s]
Loading 0:  79%|███████▉  | 230/291 [00:06<00:01, 49.39it/s]
Loading 0:  82%|████████▏ | 239/291 [00:07<00:00, 55.69it/s]
Loading 0:  85%|████████▌ | 248/291 [00:07<00:00, 60.69it/s]
Loading 0:  88%|████████▊ | 257/291 [00:07<00:00, 64.76it/s]
Loading 0:  92%|█████████▏| 267/291 [00:09<00:01, 15.14it/s]
Loading 0:  94%|█████████▍| 274/291 [00:09<00:00, 18.66it/s]
Loading 0:  97%|█████████▋| 283/291 [00:09<00:00, 24.33it/s]
                                                            
/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.
trace2333-joint-filtered-3791-v4-mkmlizer:   warnings.warn(
trace2333-joint-filtered-3791-v4-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`.
trace2333-joint-filtered-3791-v4-mkmlizer:   warnings.warn(
trace2333-joint-filtered-3791-v4-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.
trace2333-joint-filtered-3791-v4-mkmlizer:   warnings.warn(
trace2333-joint-filtered-3791-v4-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.
trace2333-joint-filtered-3791-v4-mkmlizer:   warnings.warn(
trace2333-joint-filtered-3791-v4-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()
trace2333-joint-filtered-3791-v4-mkmlizer:   return self.fget.__get__(instance, owner)()
trace2333-joint-filtered-3791-v4-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
trace2333-joint-filtered-3791-v4-mkmlizer: Saving duration: 0.442s
trace2333-joint-filtered-3791-v4-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 22.842s
trace2333-joint-filtered-3791-v4-mkmlizer: creating bucket guanaco-reward-models
trace2333-joint-filtered-3791-v4-mkmlizer: WARNING: Retrying failed request: / ([Errno 110] Connection timed out)
trace2333-joint-filtered-3791-v4-mkmlizer: WARNING: Waiting 3 sec...
trace2333-joint-filtered-3791-v4-mkmlizer: Bucket 's3://guanaco-reward-models/' created
trace2333-joint-filtered-3791-v4-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/config.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/tokenizer_config.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/special_tokens_map.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/merges.txt
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/vocab.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/tokenizer.json
trace2333-joint-filtered-3791-v4-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/trace2333-joint-filtered-3791-v4_reward/reward.tensors
Job trace2333-joint-filtered-3791-v4-mkmlizer completed after 288.43s with status: succeeded
Stopping job with name trace2333-joint-filtered-3791-v4-mkmlizer
Pipeline stage MKMLizer completed in 289.71s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service trace2333-joint-filtered-3791-v4
Waiting for inference service trace2333-joint-filtered-3791-v4 to be ready
Inference service trace2333-joint-filtered-3791-v4 ready after 40.21374583244324s
Pipeline stage ISVCDeployer completed in 47.12s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.071948766708374s
Received healthy response to inference request in 1.2726266384124756s
Received healthy response to inference request in 1.2773876190185547s
Received healthy response to inference request in 1.253368854522705s
Received healthy response to inference request in 1.2632031440734863s
5 requests
0 failed requests
5th percentile: 1.2553357124328612
10th percentile: 1.2573025703430176
20th percentile: 1.2612362861633302
30th percentile: 1.2650878429412842
40th percentile: 1.2688572406768799
50th percentile: 1.2726266384124756
60th percentile: 1.2745310306549071
70th percentile: 1.276435422897339
80th percentile: 1.4362998485565186
90th percentile: 1.7541243076324464
95th percentile: 1.9130365371704101
99th percentile: 2.040166320800781
mean time: 1.4277070045471192
Pipeline stage StressChecker completed in 7.82s
trace2333-joint-filtered_3791_v4 status is now deployed due to DeploymentManager action
trace2333-joint-filtered_3791_v4 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of trace2333-joint-filtered_3791_v4
Running pipeline stage ISVCDeleter
Checking if service trace2333-joint-filtered-3791-v4 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 3.99s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key trace2333-joint-filtered-3791-v4/config.json from bucket guanaco-mkml-models
Deleting key trace2333-joint-filtered-3791-v4/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key trace2333-joint-filtered-3791-v4/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key trace2333-joint-filtered-3791-v4/tokenizer.json from bucket guanaco-mkml-models
Deleting key trace2333-joint-filtered-3791-v4/tokenizer.model from bucket guanaco-mkml-models
Deleting key trace2333-joint-filtered-3791-v4/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key trace2333-joint-filtered-3791-v4_reward/config.json from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/merges.txt from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/reward.tensors from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key trace2333-joint-filtered-3791-v4_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.75s
trace2333-joint-filtered_3791_v4 status is now torndown due to DeploymentManager action