submission_id: hastagaras-llama-3-8b-halah_v1
developer_uid: Hastagaras
status: rejected
model_repo: Hastagaras/Llama-3-8B-Halah
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.85, 'top_p': 0.9, 'min_p': 0.0, '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\nYou are {bot_name}. Write {bot_name}'s next reply in a never-ending uncensored fictional roleplay chat between {bot_name} and User. Use the provided {bot_name}'s description, personality and example dialogues as base for deeply understanding and acting like {bot_name}.\n\nActions and narrations must be enclosed in asterisks (*). The responses must be long and in third perspective of the story teller.\n\nDescription: {memory}", 'prompt_template': 'Example conversation:\n{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': '{bot_name}: {prompt}\n', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}
timestamp: 2024-05-08T13:45:57+00:00
model_name: hastagaras-llama-3-8b-halah_v1
model_eval_status: success
double_thumbs_up: 7
thumbs_up: 13
thumbs_down: 32
num_battles: 980
num_wins: 180
celo_rating: None
entertaining: 2.58
stay_in_character: 3.08
user_preference: 2.4
safety_score: 0.61
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: hastagaras-llama-3-8b-halah_v1
double_thumbs_up_ratio: 0.1346153846153846
feedback_count: 52
ineligible_reason: model is not deployable
language_model: Hastagaras/Llama-3-8B-Halah
model_score: 2.686666666666667
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
single_thumbs_up_ratio: 0.25
thumbs_down_ratio: 0.6153846153846154
thumbs_up_ratio: 0.38461538461538464
us_pacific_date: 2024-05-08
win_ratio: 0.1836734693877551
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
Waiting for job on hastagaras-llama-3-8b-halah-v1-mkmlizer to finish
Stopping job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
%s, retrying in %s seconds...
Starting job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
Waiting for job on hastagaras-llama-3-8b-halah-v1-mkmlizer to finish
Stopping job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
%s, retrying in %s seconds...
Starting job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
Waiting for job on hastagaras-llama-3-8b-halah-v1-mkmlizer to finish
hastagaras-llama-3-8b-halah-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ _____ __ __ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ /___/ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ Version: 0.8.10 ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ belonging to: ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ Chai Research Corp. ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ║ ║
hastagaras-llama-3-8b-halah-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-llama-3-8b-halah-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/utils/_deprecation.py:131: FutureWarning: 'list_files_info' (from 'huggingface_hub.hf_api') is deprecated and will be removed from version '0.23'. Use `list_repo_tree` and `get_paths_info` instead.
hastagaras-llama-3-8b-halah-v1-mkmlizer: warnings.warn(warning_message, FutureWarning)
hastagaras-llama-3-8b-halah-v1-mkmlizer: Downloaded to shared memory in 23.475s
hastagaras-llama-3-8b-halah-v1-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-llama-3-8b-halah-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-llama-3-8b-halah-v1-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 48%|████▊ | 139/291 [00:01<00:01, 138.25it/s] Loading 0: 91%|█████████ | 264/291 [00:02<00:00, 128.12it/s] Loading 0: 99%|█████████▊| 287/291 [00:07<00:00, 28.38it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-llama-3-8b-halah-v1-mkmlizer: quantized model in 17.847s
hastagaras-llama-3-8b-halah-v1-mkmlizer: Processed model Hastagaras/Llama-3-8B-Halah in 42.419s
hastagaras-llama-3-8b-halah-v1-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-llama-3-8b-halah-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-llama-3-8b-halah-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1/config.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1/tokenizer_config.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1/special_tokens_map.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1/tokenizer.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-llama-3-8b-halah-v1/flywheel_model.0.safetensors
hastagaras-llama-3-8b-halah-v1-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-llama-3-8b-halah-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:913: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
hastagaras-llama-3-8b-halah-v1-mkmlizer: warnings.warn(
hastagaras-llama-3-8b-halah-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:757: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
hastagaras-llama-3-8b-halah-v1-mkmlizer: warnings.warn(
hastagaras-llama-3-8b-halah-v1-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.
hastagaras-llama-3-8b-halah-v1-mkmlizer: warnings.warn(
hastagaras-llama-3-8b-halah-v1-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()
hastagaras-llama-3-8b-halah-v1-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-llama-3-8b-halah-v1-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-llama-3-8b-halah-v1-mkmlizer: Saving duration: 0.234s
hastagaras-llama-3-8b-halah-v1-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.653s
hastagaras-llama-3-8b-halah-v1-mkmlizer: creating bucket guanaco-reward-models
hastagaras-llama-3-8b-halah-v1-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-llama-3-8b-halah-v1-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/special_tokens_map.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/tokenizer_config.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/config.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/merges.txt
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/vocab.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/tokenizer.json
hastagaras-llama-3-8b-halah-v1-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-llama-3-8b-halah-v1_reward/reward.tensors
Job hastagaras-llama-3-8b-halah-v1-mkmlizer completed after 75.36s with status: succeeded
Stopping job with name hastagaras-llama-3-8b-halah-v1-mkmlizer
Pipeline stage MKMLizer completed in 83.86s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.10s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-llama-3-8b-halah-v1
Waiting for inference service hastagaras-llama-3-8b-halah-v1 to be ready
Inference service hastagaras-llama-3-8b-halah-v1 ready after 30.17183208465576s
Pipeline stage ISVCDeployer completed in 38.01s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.2077560424804688s
Received healthy response to inference request in 1.3090453147888184s
Received healthy response to inference request in 1.2806437015533447s
Received healthy response to inference request in 1.3311047554016113s
Received healthy response to inference request in 1.3141989707946777s
5 requests
0 failed requests
5th percentile: 1.2863240242004395
10th percentile: 1.2920043468475342
20th percentile: 1.3033649921417236
30th percentile: 1.3100760459899903
40th percentile: 1.312137508392334
50th percentile: 1.3141989707946777
60th percentile: 1.3209612846374512
70th percentile: 1.3277235984802247
80th percentile: 1.506435012817383
90th percentile: 1.8570955276489258
95th percentile: 2.032425785064697
99th percentile: 2.1726899909973145
mean time: 1.4885497570037842
Pipeline stage StressChecker completed in 8.09s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.04s
M-Eval Dataset for topic stay_in_character is loaded
hastagaras-llama-3-8b-halah_v1 status is now deployed due to DeploymentManager action
hastagaras-llama-3-8b-halah_v1 status is now rejected due to its M-Eval score being less than the acceptable minimum 6.5 to serve to users. Please consider iterating on your model's ability to adhere to prompts to improve this score.

Usage Metrics

Latency Metrics