submission_id: undi95-bigl-7b_v3
developer_uid: Undi95
status: inactive
model_repo: Undi95/BigL-7B
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '\n###', '</s>'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<s>### Instruction:\nContinue the following conversation in a natural way. Think deeply before replying to {user_name}, keep the conversation entertaining and imaginative. Stay in the role of {bot_name} throughout the entire conversation. Engage in *roleplay* actions.\n{bot_name}'s Persona: {memory}</s>", 'prompt_template': '{prompt}\n<START>\n', 'bot_template': '### Response:\n{bot_name}: {message}\n', 'user_template': '### Input:\n{user_name}: {message}\n', 'response_template': '### Response:\n{bot_name}:'}
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}:'}
timestamp: 2024-03-25T09:59:31+00:00
model_name: undi95-bigl-7b_v3
model_eval_status: success
safety_score: 0.95
entertaining: 7.24
stay_in_character: 8.52
user_preference: 7.46
double_thumbs_up: 478
thumbs_up: 643
thumbs_down: 252
num_battles: 67621
num_wins: 34679
win_ratio: 0.5128436432469203
celo_rating: 1166.39
Resubmit model
Running pipeline stage MKMLizer
Starting job with name undi95-bigl-7b-v3-mkmlizer
Waiting for job on undi95-bigl-7b-v3-mkmlizer to finish
undi95-bigl-7b-v3-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
undi95-bigl-7b-v3-mkmlizer: ║ _____ __ __ ║
undi95-bigl-7b-v3-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
undi95-bigl-7b-v3-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
undi95-bigl-7b-v3-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
undi95-bigl-7b-v3-mkmlizer: ║ /___/ ║
undi95-bigl-7b-v3-mkmlizer: ║ ║
undi95-bigl-7b-v3-mkmlizer: ║ Version: 0.6.11 ║
undi95-bigl-7b-v3-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
undi95-bigl-7b-v3-mkmlizer: ║ ║
undi95-bigl-7b-v3-mkmlizer: ║ The license key for the current software has been verified as ║
undi95-bigl-7b-v3-mkmlizer: ║ belonging to: ║
undi95-bigl-7b-v3-mkmlizer: ║ ║
undi95-bigl-7b-v3-mkmlizer: ║ Chai Research Corp. ║
undi95-bigl-7b-v3-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
undi95-bigl-7b-v3-mkmlizer: ║ Expiration: 2024-04-15 23:59:59 ║
undi95-bigl-7b-v3-mkmlizer: ║ ║
undi95-bigl-7b-v3-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
undi95-bigl-7b-v3-mkmlizer: .gitattributes: 0%| | 0.00/1.52k [00:00<?, ?B/s] .gitattributes: 100%|██████████| 1.52k/1.52k [00:00<00:00, 17.2MB/s]
undi95-bigl-7b-v3-mkmlizer: config.json: 0%| | 0.00/653 [00:00<?, ?B/s] config.json: 100%|██████████| 653/653 [00:00<00:00, 5.12MB/s]
undi95-bigl-7b-v3-mkmlizer: model-00001-of-00002.safetensors: 0%| | 0.00/9.94G [00:00<?, ?B/s] model-00001-of-00002.safetensors: 0%| | 10.5M/9.94G [00:00<14:39, 11.3MB/s] model-00001-of-00002.safetensors: 0%| | 31.5M/9.94G [00:01<04:51, 34.0MB/s] model-00001-of-00002.safetensors: 0%| | 41.9M/9.94G [00:01<05:44, 28.8MB/s] model-00001-of-00002.safetensors: 1%| | 52.4M/9.94G [00:01<04:28, 36.9MB/s] model-00001-of-00002.safetensors: 1%| | 115M/9.94G [00:01<01:24, 117MB/s] model-00001-of-00002.safetensors: 2%|▏ | 189M/9.94G [00:01<00:44, 217MB/s] model-00001-of-00002.safetensors: 5%|▍ | 482M/9.94G [00:02<00:12, 730MB/s] model-00001-of-00002.safetensors: 10%|█ | 1.04G/9.94G [00:02<00:05, 1.72GB/s] model-00001-of-00002.safetensors: 13%|█▎ | 1.29G/9.94G [00:02<00:05, 1.46GB/s] model-00001-of-00002.safetensors: 15%|█▌ | 1.50G/9.94G [00:02<00:08, 987MB/s] model-00001-of-00002.safetensors: 18%|█▊ | 1.74G/9.94G [00:02<00:06, 1.19GB/s] model-00001-of-00002.safetensors: 22%|██▏ | 2.19G/9.94G [00:02<00:04, 1.77GB/s] model-00001-of-00002.safetensors: 25%|██▍ | 2.45G/9.94G [00:03<00:05, 1.26GB/s] model-00001-of-00002.safetensors: 27%|██▋ | 2.66G/9.94G [00:03<00:06, 1.15GB/s] model-00001-of-00002.safetensors: 30%|██▉ | 2.94G/9.94G [00:03<00:05, 1.40GB/s] model-00001-of-00002.safetensors: 33%|███▎ | 3.31G/9.94G [00:03<00:03, 1.83GB/s] model-00001-of-00002.safetensors: 36%|███▌ | 3.57G/9.94G [00:04<00:05, 1.25GB/s] model-00001-of-00002.safetensors: 38%|███▊ | 3.76G/9.94G [00:04<00:04, 1.28GB/s] model-00001-of-00002.safetensors: 41%|████▏ | 4.12G/9.94G [00:04<00:03, 1.66GB/s] model-00001-of-00002.safetensors: 44%|████▍ | 4.38G/9.94G [00:04<00:03, 1.82GB/s] model-00001-of-00002.safetensors: 46%|████▋ | 4.61G/9.94G [00:04<00:04, 1.31GB/s] model-00001-of-00002.safetensors: 48%|████▊ | 4.80G/9.94G [00:04<00:03, 1.30GB/s] model-00001-of-00002.safetensors: 52%|█████▏ | 5.13G/9.94G [00:05<00:02, 1.65GB/s] model-00001-of-00002.safetensors: 54%|█████▍ | 5.39G/9.94G [00:05<00:02, 1.85GB/s] model-00001-of-00002.safetensors: 57%|█████▋ | 5.62G/9.94G [00:05<00:02, 1.63GB/s] model-00001-of-00002.safetensors: 59%|█████▊ | 5.82G/9.94G [00:05<00:02, 1.38GB/s] model-00001-of-00002.safetensors: 62%|██████▏ | 6.14G/9.94G [00:05<00:02, 1.73GB/s] model-00001-of-00002.safetensors: 64%|██████▍ | 6.36G/9.94G [00:05<00:02, 1.78GB/s] model-00001-of-00002.safetensors: 66%|██████▌ | 6.57G/9.94G [00:05<00:01, 1.72GB/s] model-00001-of-00002.safetensors: 68%|██████▊ | 6.77G/9.94G [00:06<00:02, 1.49GB/s] model-00001-of-00002.safetensors: 70%|██████▉ | 6.94G/9.94G [00:06<00:02, 1.39GB/s] model-00001-of-00002.safetensors: 72%|███████▏ | 7.13G/9.94G [00:06<00:01, 1.47GB/s] model-00001-of-00002.safetensors: 73%|███████▎ | 7.29G/9.94G [00:06<00:01, 1.48GB/s] model-00001-of-00002.safetensors: 75%|███████▍ | 7.44G/9.94G [00:06<00:01, 1.48GB/s] model-00001-of-00002.safetensors: 76%|███████▋ | 7.60G/9.94G [00:06<00:01, 1.42GB/s] model-00001-of-00002.safetensors: 78%|███████▊ | 7.80G/9.94G [00:06<00:01, 1.52GB/s] model-00001-of-00002.safetensors: 80%|████████ | 7.96G/9.94G [00:07<00:01, 1.28GB/s] model-00001-of-00002.safetensors: 82%|████████▏ | 8.15G/9.94G [00:07<00:01, 1.42GB/s] model-00001-of-00002.safetensors: 84%|████████▍ | 8.36G/9.94G [00:07<00:01, 1.57GB/s] model-00001-of-00002.safetensors: 86%|████████▌ | 8.57G/9.94G [00:07<00:00, 1.69GB/s] model-00001-of-00002.safetensors: 88%|████████▊ | 8.75G/9.94G [00:07<00:00, 1.66GB/s] model-00001-of-00002.safetensors: 90%|████████▉ | 8.92G/9.94G [00:07<00:00, 1.40GB/s] model-00001-of-00002.safetensors: 92%|█████████▏| 9.14G/9.94G [00:07<00:00, 1.59GB/s] model-00001-of-00002.safetensors: 95%|█████████▍| 9.42G/9.94G [00:07<00:00, 1.88GB/s] model-00001-of-00002.safetensors: 100%|█████████▉| 9.94G/9.94G [00:07<00:00, 1.25GB/s]
undi95-bigl-7b-v3-mkmlizer: model-00002-of-00002.safetensors: 0%| | 0.00/4.54G [00:00<?, ?B/s] model-00002-of-00002.safetensors: 0%| | 10.5M/4.54G [00:00<06:10, 12.2MB/s] model-00002-of-00002.safetensors: 1%| | 31.5M/4.54G [00:01<03:26, 21.8MB/s] model-00002-of-00002.safetensors: 1%| | 41.9M/4.54G [00:01<02:31, 29.7MB/s] model-00002-of-00002.safetensors: 2%|▏ | 94.4M/4.54G [00:01<00:49, 90.4MB/s] model-00002-of-00002.safetensors: 8%|▊ | 346M/4.54G [00:01<00:09, 452MB/s] model-00002-of-00002.safetensors: 22%|██▏ | 1.01G/4.54G [00:01<00:02, 1.53GB/s] model-00002-of-00002.safetensors: 29%|██▊ | 1.30G/4.54G [00:02<00:03, 931MB/s] model-00002-of-00002.safetensors: 33%|███▎ | 1.52G/4.54G [00:02<00:02, 1.02GB/s] model-00002-of-00002.safetensors: 43%|████▎ | 1.95G/4.54G [00:02<00:01, 1.49GB/s] model-00002-of-00002.safetensors: 50%|████▉ | 2.26G/4.54G [00:02<00:01, 1.52GB/s] model-00002-of-00002.safetensors: 55%|█████▍ | 2.50G/4.54G [00:03<00:01, 1.14GB/s] model-00002-of-00002.safetensors: 59%|█████▉ | 2.68G/4.54G [00:03<00:01, 1.20GB/s] model-00002-of-00002.safetensors: 65%|██████▍ | 2.94G/4.54G [00:03<00:01, 1.39GB/s] model-00002-of-00002.safetensors: 72%|███████▏ | 3.25G/4.54G [00:03<00:00, 1.70GB/s] model-00002-of-00002.safetensors: 76%|███████▋ | 3.47G/4.54G [00:04<00:00, 1.14GB/s] model-00002-of-00002.safetensors: 80%|████████ | 3.65G/4.54G [00:04<00:00, 1.19GB/s] model-00002-of-00002.safetensors: 89%|████████▊ | 4.03G/4.54G [00:04<00:00, 1.64GB/s] model-00002-of-00002.safetensors: 100%|█████████▉| 4.54G/4.54G [00:04<00:00, 1.02GB/s]
undi95-bigl-7b-v3-mkmlizer: model.safetensors.index.json: 0%| | 0.00/22.8k [00:00<?, ?B/s] model.safetensors.index.json: 100%|██████████| 22.8k/22.8k [00:00<00:00, 6.30MB/s]
undi95-bigl-7b-v3-mkmlizer: special_tokens_map.json: 0%| | 0.00/414 [00:00<?, ?B/s] special_tokens_map.json: 100%|██████████| 414/414 [00:00<00:00, 3.41MB/s]
undi95-bigl-7b-v3-mkmlizer: tokenizer.json: 0%| | 0.00/1.80M [00:00<?, ?B/s] tokenizer.json: 100%|██████████| 1.80M/1.80M [00:00<00:00, 7.17MB/s] tokenizer.json: 100%|██████████| 1.80M/1.80M [00:00<00:00, 7.14MB/s]
undi95-bigl-7b-v3-mkmlizer: Profiling: 0%| | 0/291 [00:00<?, ?it/s] Profiling: 0%| | 1/291 [00:01<05:45, 1.19s/it] Profiling: 5%|▍ | 14/291 [00:01<00:19, 14.48it/s] Profiling: 11%|█ | 31/291 [00:01<00:07, 34.58it/s] Profiling: 16%|█▋ | 48/291 [00:01<00:04, 55.34it/s] Profiling: 22%|██▏ | 63/291 [00:01<00:03, 72.01it/s] Profiling: 26%|██▋ | 77/291 [00:01<00:02, 84.69it/s] Profiling: 32%|███▏ | 94/291 [00:01<00:01, 102.82it/s] Profiling: 38%|███▊ | 111/291 [00:01<00:01, 117.95it/s] Profiling: 43%|████▎ | 126/291 [00:02<00:01, 124.75it/s] Profiling: 49%|████▉ | 143/291 [00:02<00:01, 135.47it/s] Profiling: 55%|█████▍ | 160/291 [00:02<00:00, 143.15it/s] Profiling: 60%|██████ | 176/291 [00:02<00:00, 146.11it/s] Profiling: 66%|██████▌ | 192/291 [00:02<00:00, 144.96it/s] Profiling: 71%|███████▏ | 208/291 [00:04<00:03, 25.64it/s] Profiling: 76%|███████▋ | 222/291 [00:04<00:02, 32.71it/s] Profiling: 82%|████████▏ | 239/291 [00:04<00:01, 43.90it/s] Profiling: 88%|████████▊ | 256/291 [00:04<00:00, 56.81it/s] Profiling: 93%|█████████▎| 270/291 [00:04<00:00, 67.47it/s] Profiling: 98%|█████████▊| 285/291 [00:04<00:00, 78.49it/s] Profiling: 100%|██████████| 291/291 [00:05<00:00, 57.65it/s]
undi95-bigl-7b-v3-mkmlizer: quantized model in 17.615s
undi95-bigl-7b-v3-mkmlizer: Processed model Undi95/BigL-7B in 35.313s
undi95-bigl-7b-v3-mkmlizer: creating bucket guanaco-mkml-models
undi95-bigl-7b-v3-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
undi95-bigl-7b-v3-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/undi95-bigl-7b-v3
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/undi95-bigl-7b-v3/config.json
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/undi95-bigl-7b-v3/tokenizer_config.json
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/undi95-bigl-7b-v3/special_tokens_map.json
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/undi95-bigl-7b-v3/tokenizer.model
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/undi95-bigl-7b-v3/tokenizer.json
undi95-bigl-7b-v3-mkmlizer: cp /dev/shm/model_cache/mkml_model.tensors s3://guanaco-mkml-models/undi95-bigl-7b-v3/mkml_model.tensors
undi95-bigl-7b-v3-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
undi95-bigl-7b-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:1067: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
undi95-bigl-7b-v3-mkmlizer: warnings.warn(
undi95-bigl-7b-v3-mkmlizer: config.json: 0%| | 0.00/1.05k [00:00<?, ?B/s] config.json: 100%|██████████| 1.05k/1.05k [00:00<00:00, 12.2MB/s]
undi95-bigl-7b-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:690: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
undi95-bigl-7b-v3-mkmlizer: warnings.warn(
undi95-bigl-7b-v3-mkmlizer: tokenizer_config.json: 0%| | 0.00/234 [00:00<?, ?B/s] tokenizer_config.json: 100%|██████████| 234/234 [00:00<00:00, 3.15MB/s]
undi95-bigl-7b-v3-mkmlizer: vocab.json: 0%| | 0.00/1.04M [00:00<?, ?B/s] vocab.json: 100%|██████████| 1.04M/1.04M [00:00<00:00, 4.44MB/s] vocab.json: 100%|██████████| 1.04M/1.04M [00:00<00:00, 4.43MB/s]
undi95-bigl-7b-v3-mkmlizer: tokenizer.json: 0%| | 0.00/2.11M [00:00<?, ?B/s] tokenizer.json: 100%|██████████| 2.11M/2.11M [00:00<00:00, 13.3MB/s] tokenizer.json: 100%|██████████| 2.11M/2.11M [00:00<00:00, 13.2MB/s]
undi95-bigl-7b-v3-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py:472: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
undi95-bigl-7b-v3-mkmlizer: warnings.warn(
undi95-bigl-7b-v3-mkmlizer: pytorch_model.bin: 0%| | 0.00/1.44G [00:00<?, ?B/s] pytorch_model.bin: 1%| | 10.5M/1.44G [00:00<01:56, 12.3MB/s] pytorch_model.bin: 1%|▏ | 21.0M/1.44G [00:00<00:57, 24.9MB/s] pytorch_model.bin: 2%|▏ | 31.5M/1.44G [00:01<01:25, 16.6MB/s] pytorch_model.bin: 7%|▋ | 105M/1.44G [00:01<00:16, 83.5MB/s] pytorch_model.bin: 26%|██▌ | 377M/1.44G [00:02<00:02, 394MB/s] pytorch_model.bin: 64%|██████▍ | 923M/1.44G [00:02<00:00, 1.12GB/s] pytorch_model.bin: 83%|████████▎ | 1.19G/1.44G [00:02<00:00, 1.21GB/s] pytorch_model.bin: 100%|█████████▉| 1.44G/1.44G [00:02<00:00, 606MB/s]
undi95-bigl-7b-v3-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
undi95-bigl-7b-v3-mkmlizer: Saving duration: 0.311s
undi95-bigl-7b-v3-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 8.266s
undi95-bigl-7b-v3-mkmlizer: creating bucket guanaco-reward-models
undi95-bigl-7b-v3-mkmlizer: Bucket 's3://guanaco-reward-models/' created
undi95-bigl-7b-v3-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/undi95-bigl-7b-v3_reward
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/config.json
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/special_tokens_map.json
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/tokenizer_config.json
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/merges.txt
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/vocab.json
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/tokenizer.json
undi95-bigl-7b-v3-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/undi95-bigl-7b-v3_reward/reward.tensors
Job undi95-bigl-7b-v3-mkmlizer completed after 66.62s with status: succeeded
Stopping job with name undi95-bigl-7b-v3-mkmlizer
Pipeline stage MKMLizer completed in 70.45s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.11s
Running pipeline stage ISVCDeployer
Creating inference service undi95-bigl-7b-v3
Waiting for inference service undi95-bigl-7b-v3 to be ready
Inference service undi95-bigl-7b-v3 ready after 40.245747566223145s
Pipeline stage ISVCDeployer completed in 47.22s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.7332961559295654s
Received healthy response to inference request in 1.1988446712493896s
Received healthy response to inference request in 3.422156572341919s
Received healthy response to inference request in 1.1804676055908203s
Received healthy response to inference request in 1.1803743839263916s
5 requests
0 failed requests
5th percentile: 1.1803930282592774
10th percentile: 1.1804116725921632
20th percentile: 1.1804489612579345
30th percentile: 1.1841430187225341
40th percentile: 1.191493844985962
50th percentile: 1.1988446712493896
60th percentile: 1.41262526512146
70th percentile: 1.6264058589935302
80th percentile: 2.0710682392120363
90th percentile: 2.7466124057769776
95th percentile: 3.084384489059448
99th percentile: 3.354602155685425
mean time: 1.7430278778076171
Pipeline stage StressChecker completed in 9.59s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.05s
Running pipeline stage DaemonicSafetyScorer
Pipeline stage DaemonicSafetyScorer completed in 0.04s
Running M-Eval for topic stay_in_character
M-Eval Dataset for topic stay_in_character is loaded
undi95-bigl-7b_v3 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics