submission_id: jellywibble-lora-120k-p_2801_v17
developer_uid: chai_backend_admin
best_of: 16
celo_rating: 1293.96
display_name: jellywibble-lora-120k-p_2801_v17
family_friendly_score: 0.0
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}
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': 96}
ineligible_reason: max_output_tokens!=64
is_internal_developer: True
language_model: Jellywibble/lora_120k_pref_data_ep3_stacked_elo_alignment
max_input_tokens: 512
max_output_tokens: 96
model_architecture: LlamaForCausalLM
model_group: Jellywibble/lora_120k_pr
model_name: jellywibble-lora-120k-p_2801_v17
model_num_parameters: 8030261248.0
model_repo: Jellywibble/lora_120k_pref_data_ep3_stacked_elo_alignment
model_size: 8B
num_battles: 217849
num_wins: 134028
ranking_group: single
reward_formatter: {'bot_template': 'Bot: {message}\n', 'memory_template': '', 'prompt_template': '', 'response_template': 'Bot:', 'truncate_by_message': False, 'user_template': 'User: {message}\n'}
reward_repo: ChaiML/gpt2_xl_pairwise_89m_step_347634
status: torndown
submission_type: basic
timestamp: 2024-07-15T17:58:00+00:00
us_pacific_date: 2024-07-15
win_ratio: 0.6152334874155952
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jellywibble-lora-120k-p-2801-v17-mkmlizer
Waiting for job on jellywibble-lora-120k-p-2801-v17-mkmlizer to finish
jellywibble-lora-120k-p-2801-v17-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ _____ __ __ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ /___/ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ Version: 0.9.5.post2 ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ https://mk1.ai ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ The license key for the current software has been verified as ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ belonging to: ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ Chai Research Corp. ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ Expiration: 2024-10-15 23:59:59 ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ║ ║
jellywibble-lora-120k-p-2801-v17-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jellywibble-lora-120k-p-2801-v17-mkmlizer: Downloaded to shared memory in 49.151s
jellywibble-lora-120k-p-2801-v17-mkmlizer: quantizing model to /dev/shm/model_cache
jellywibble-lora-120k-p-2801-v17-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jellywibble-lora-120k-p-2801-v17-mkmlizer: lm_head.weight torch.Size([139542528])
jellywibble-lora-120k-p-2801-v17-mkmlizer: model.layers.31.input_layernorm.weight torch.Size([4096])
jellywibble-lora-120k-p-2801-v17-mkmlizer: model.layers.31.mlp.down_proj.weight torch.Size([11927552])
jellywibble-lora-120k-p-2801-v17-mkmlizer: model.layers.31.mlp.up_gate_proj.weight torch.Size([23855104])
jellywibble-lora-120k-p-2801-v17-mkmlizer: model.layers.31.post_attention_layernorm.weight torch.Size([4096])
jellywibble-lora-120k-p-2801-v17-mkmlizer: model.norm.weight torch.Size([4096])
jellywibble-lora-120k-p-2801-v17-mkmlizer: loading reward model from ChaiML/gpt2_xl_pairwise_89m_step_347634
jellywibble-lora-120k-p-2801-v17-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:950: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jellywibble-lora-120k-p-2801-v17-mkmlizer: warnings.warn(
jellywibble-lora-120k-p-2801-v17-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:778: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
jellywibble-lora-120k-p-2801-v17-mkmlizer: warnings.warn(
jellywibble-lora-120k-p-2801-v17-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.
jellywibble-lora-120k-p-2801-v17-mkmlizer: warnings.warn(
jellywibble-lora-120k-p-2801-v17-mkmlizer: Downloading shards: 0%| | 0/2 [00:00<?, ?it/s] Downloading shards: 50%|█████ | 1/2 [00:07<00:07, 7.03s/it] Downloading shards: 100%|██████████| 2/2 [00:09<00:00, 4.16s/it] Downloading shards: 100%|██████████| 2/2 [00:09<00:00, 4.59s/it]
jellywibble-lora-120k-p-2801-v17-mkmlizer: Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s] Loading checkpoint shards: 50%|█████ | 1/2 [00:00<00:00, 1.46it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 2.38it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 2.17it/s]
jellywibble-lora-120k-p-2801-v17-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jellywibble-lora-120k-p-2801-v17-mkmlizer: Saving duration: 2.371s
jellywibble-lora-120k-p-2801-v17-mkmlizer: Processed model ChaiML/gpt2_xl_pairwise_89m_step_347634 in 14.105s
jellywibble-lora-120k-p-2801-v17-mkmlizer: creating bucket guanaco-reward-models
jellywibble-lora-120k-p-2801-v17-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jellywibble-lora-120k-p-2801-v17-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/tokenizer_config.json
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/merges.txt
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/special_tokens_map.json
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/tokenizer.json
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/vocab.json
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/config.json
jellywibble-lora-120k-p-2801-v17-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jellywibble-lora-120k-p-2801-v17_reward/reward.tensors
Job jellywibble-lora-120k-p-2801-v17-mkmlizer completed after 144.89s with status: succeeded
Stopping job with name jellywibble-lora-120k-p-2801-v17-mkmlizer
Pipeline stage MKMLizer completed in 145.75s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.09s
Running pipeline stage ISVCDeployer
Creating inference service jellywibble-lora-120k-p-2801-v17
Waiting for inference service jellywibble-lora-120k-p-2801-v17 to be ready
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Connection pool is full, discarding connection: %s. Connection pool size: %s
Inference service jellywibble-lora-120k-p-2801-v17 ready after 50.762277126312256s
Pipeline stage ISVCDeployer completed in 57.46s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.898294448852539s
Received healthy response to inference request in 1.7924790382385254s
Received healthy response to inference request in 1.776792049407959s
Received healthy response to inference request in 1.835256814956665s
Received healthy response to inference request in 1.7671949863433838s
5 requests
0 failed requests
5th percentile: 1.7691143989562987
10th percentile: 1.771033811569214
20th percentile: 1.774872636795044
30th percentile: 1.7799294471740723
40th percentile: 1.7862042427062987
50th percentile: 1.7924790382385254
60th percentile: 1.8095901489257813
70th percentile: 1.8267012596130372
80th percentile: 2.04786434173584
90th percentile: 2.4730793952941896
95th percentile: 2.685686922073364
99th percentile: 2.855772943496704
mean time: 2.0140034675598146
Pipeline stage StressChecker completed in 11.04s
jellywibble-lora-120k-p_2801_v17 status is now deployed due to DeploymentManager action
jellywibble-lora-120k-p_2801_v17 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of jellywibble-lora-120k-p_2801_v17
Running pipeline stage ISVCDeleter
Checking if service jellywibble-lora-120k-p-2801-v17 is running
Skipping teardown as no inference service was found
Pipeline stage ISVCDeleter completed in 4.81s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key jellywibble-lora-120k-p-2801-v17/config.json from bucket guanaco-mkml-models
Deleting key jellywibble-lora-120k-p-2801-v17/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key jellywibble-lora-120k-p-2801-v17/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key jellywibble-lora-120k-p-2801-v17/tokenizer.json from bucket guanaco-mkml-models
Deleting key jellywibble-lora-120k-p-2801-v17/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key jellywibble-lora-120k-p-2801-v17_reward/config.json from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/merges.txt from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/reward.tensors from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key jellywibble-lora-120k-p-2801-v17_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 5.65s
jellywibble-lora-120k-p_2801_v17 status is now torndown due to DeploymentManager action