submission_id: jellywibble-lora-120k-pr_2801_v2
developer_uid: Jellywibble
status: inactive
model_repo: Jellywibble/lora_120k_pref_data_ep3_stacked_elo_alignment
reward_repo: ChaiML/gpt2_xl_pairwise_89m_step_347634
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': 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: 2024-07-09T07:16:30+00:00
model_name: nitral-ai-hathor-l3-8b-v-01_v1
model_group: Jellywibble/lora_120k_pr
num_battles: 30298
num_wins: 18161
celo_rating: 1267.07
alignment_score: None
alignment_samples: 0
propriety_score: 0.7331581910302121
propriety_total_count: 5329.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: nitral-ai-hathor-l3-8b-v-01_v1
ineligible_reason: None
language_model: Jellywibble/lora_120k_pref_data_ep3_stacked_elo_alignment
model_size: 8B
reward_model: ChaiML/gpt2_xl_pairwise_89m_step_347634
us_pacific_date: 2024-07-09
win_ratio: 0.5994125024754109
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name jellywibble-lora-120k-pr-2801-v2-mkmlizer
Waiting for job on jellywibble-lora-120k-pr-2801-v2-mkmlizer to finish
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ _____ __ __ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ /___/ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ Version: 0.8.14 ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ https://mk1.ai ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ The license key for the current software has been verified as ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ belonging to: ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ Chai Research Corp. ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ║ ║
jellywibble-lora-120k-pr-2801-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Downloaded to shared memory in 55.191s
jellywibble-lora-120k-pr-2801-v2-mkmlizer: quantizing model to /dev/shm/model_cache
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 2%|▏ | 5/291 [00:00<00:06, 44.24it/s] Loading 0: 4%|▍ | 13/291 [00:00<00:04, 64.18it/s] Loading 0: 7%|▋ | 21/291 [00:00<00:03, 71.18it/s] Loading 0: 10%|▉ | 29/291 [00:00<00:03, 73.58it/s] Loading 0: 13%|█▎ | 37/291 [00:00<00:06, 36.66it/s] Loading 0: 15%|█▍ | 43/291 [00:00<00:06, 40.48it/s] Loading 0: 17%|█▋ | 50/291 [00:01<00:05, 44.30it/s] Loading 0: 20%|██ | 59/291 [00:01<00:04, 51.11it/s] Loading 0: 23%|██▎ | 68/291 [00:01<00:03, 56.12it/s] Loading 0: 26%|██▋ | 77/291 [00:01<00:03, 62.80it/s] Loading 0: 29%|██▉ | 84/291 [00:01<00:05, 37.34it/s] Loading 0: 32%|███▏ | 92/291 [00:01<00:04, 42.26it/s] Loading 0: 35%|███▍ | 101/291 [00:02<00:03, 48.17it/s] Loading 0: 38%|███▊ | 110/291 [00:02<00:03, 53.04it/s] Loading 0: 42%|████▏ | 122/291 [00:02<00:02, 64.02it/s] Loading 0: 45%|████▍ | 130/291 [00:02<00:02, 64.35it/s] Loading 0: 47%|████▋ | 138/291 [00:02<00:03, 40.42it/s] Loading 0: 51%|█████ | 147/291 [00:02<00:03, 46.29it/s] Loading 0: 54%|█████▎ | 156/291 [00:03<00:02, 51.32it/s] Loading 0: 57%|█████▋ | 165/291 [00:03<00:02, 55.78it/s] Loading 0: 59%|█████▉ | 172/291 [00:03<00:02, 58.45it/s] Loading 0: 62%|██████▏ | 179/291 [00:03<00:01, 58.12it/s] Loading 0: 64%|██████▍ | 186/291 [00:03<00:02, 40.56it/s] Loading 0: 66%|██████▌ | 192/291 [00:03<00:02, 41.18it/s] Loading 0: 69%|██████▉ | 201/291 [00:04<00:01, 48.14it/s] Loading 0: 72%|███████▏ | 210/291 [00:04<00:01, 53.74it/s] Loading 0: 75%|███████▌ | 219/291 [00:04<00:01, 58.19it/s] Loading 0: 78%|███████▊ | 227/291 [00:04<00:01, 59.52it/s] Loading 0: 80%|████████ | 234/291 [00:04<00:01, 39.25it/s] Loading 0: 82%|████████▏ | 240/291 [00:04<00:01, 42.89it/s] Loading 0: 85%|████████▌ | 248/291 [00:05<00:00, 48.34it/s] Loading 0: 88%|████████▊ | 257/291 [00:05<00:00, 54.66it/s] Loading 0: 91%|█████████▏| 266/291 [00:05<00:00, 59.62it/s] Loading 0: 95%|█████████▍| 275/291 [00:05<00:00, 63.30it/s] Loading 0: 98%|█████████▊| 286/291 [00:11<00:00, 5.10it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
jellywibble-lora-120k-pr-2801-v2-mkmlizer: quantized model in 27.347s
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Processed model Jellywibble/lora_120k_pref_data_ep3_stacked_elo_alignment in 82.538s
jellywibble-lora-120k-pr-2801-v2-mkmlizer: creating bucket guanaco-mkml-models
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
jellywibble-lora-120k-pr-2801-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2/config.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2/tokenizer_config.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2/special_tokens_map.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2/tokenizer.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/jellywibble-lora-120k-pr-2801-v2/flywheel_model.0.safetensors
jellywibble-lora-120k-pr-2801-v2-mkmlizer: loading reward model from ChaiML/gpt2_xl_pairwise_89m_step_347634
jellywibble-lora-120k-pr-2801-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.
jellywibble-lora-120k-pr-2801-v2-mkmlizer: warnings.warn(
jellywibble-lora-120k-pr-2801-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`.
jellywibble-lora-120k-pr-2801-v2-mkmlizer: warnings.warn(
jellywibble-lora-120k-pr-2801-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.
jellywibble-lora-120k-pr-2801-v2-mkmlizer: warnings.warn(
jellywibble-lora-120k-pr-2801-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.
jellywibble-lora-120k-pr-2801-v2-mkmlizer: warnings.warn(
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Downloading shards: 0%| | 0/2 [00:00<?, ?it/s] Downloading shards: 50%|█████ | 1/2 [00:07<00:07, 7.58s/it] Downloading shards: 100%|██████████| 2/2 [00:12<00:00, 5.89s/it] Downloading shards: 100%|██████████| 2/2 [00:12<00:00, 6.14s/it]
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s] Loading checkpoint shards: 50%|█████ | 1/2 [00:00<00:00, 2.03it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 3.28it/s] Loading checkpoint shards: 100%|██████████| 2/2 [00:00<00:00, 3.01it/s]
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
jellywibble-lora-120k-pr-2801-v2-mkmlizer: creating bucket guanaco-reward-models
jellywibble-lora-120k-pr-2801-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
jellywibble-lora-120k-pr-2801-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/config.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/special_tokens_map.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/tokenizer_config.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/merges.txt
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/vocab.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/tokenizer.json
jellywibble-lora-120k-pr-2801-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/jellywibble-lora-120k-pr-2801-v2_reward/reward.tensors
Job jellywibble-lora-120k-pr-2801-v2-mkmlizer completed after 126.1s with status: succeeded
Stopping job with name jellywibble-lora-120k-pr-2801-v2-mkmlizer
Pipeline stage MKMLizer completed in 127.03s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service jellywibble-lora-120k-pr-2801-v2
Waiting for inference service jellywibble-lora-120k-pr-2801-v2 to be ready
Connection pool is full, discarding connection: %s
Inference service jellywibble-lora-120k-pr-2801-v2 ready after 40.176929235458374s
Pipeline stage ISVCDeployer completed in 47.27s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.25789737701416s
Received healthy response to inference request in 1.5677573680877686s
Received healthy response to inference request in 1.5498840808868408s
Received healthy response to inference request in 1.5027661323547363s
Received healthy response to inference request in 1.5635931491851807s
5 requests
0 failed requests
5th percentile: 1.5121897220611573
10th percentile: 1.521613311767578
20th percentile: 1.5404604911804198
30th percentile: 1.5526258945465088
40th percentile: 1.5581095218658447
50th percentile: 1.5635931491851807
60th percentile: 1.565258836746216
70th percentile: 1.566924524307251
80th percentile: 1.705785369873047
90th percentile: 1.9818413734436036
95th percentile: 2.1198693752288817
99th percentile: 2.2302917766571047
mean time: 1.6883796215057374
Pipeline stage StressChecker completed in 9.16s
jellywibble-lora-120k-pr_2801_v2 status is now deployed due to DeploymentManager action
jellywibble-lora-120k-pr_2801_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics