submission_id: undi95-meta-llama-3-70b-_6209_v3
developer_uid: robert_irvine
status: torndown
model_repo: Undi95/Meta-Llama-3-70B-Instruct-hf
reward_repo: rirv938/gpt2_ties_merge_preference_plus_classic_e2_density_99
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 50, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n', '<|im_end|>', '<|im_start|>', '\n\n'], 'max_input_tokens': 512, 'best_of': 2, 'max_output_tokens': 64}
formatter: {'memory_template': "<|im_start|>system\n{bot_name}'s Persona: {memory}<|im_end|>\n", 'prompt_template': '<|im_start|>system\n{prompt}<|im_end|>\n', 'bot_template': '<|im_start|>assistant\n{bot_name}: {message}<|im_end|>\n', 'user_template': '<|im_start|>user\n{user_name}: {message}<|im_end|>\n', 'response_template': '<|im_start|>assistant\n{bot_name}:', 'truncate_by_message': False}
reward_formatter: {'memory_template': 'Memory: {memory}\n', 'prompt_template': '{prompt}\n', 'bot_template': 'Bot: {message}\n', 'user_template': 'User: {message}\n', 'response_template': 'Bot:', 'truncate_by_message': False}
timestamp: 2024-04-24T17:05:56+00:00
model_name: undi95-meta-llama-3-70b-_6209_v3
model_eval_status: pending
model_group: Undi95/Meta-Llama-3-70B-
num_battles: 100528
num_wins: 50661
celo_rating: 1190.88
propriety_score: 0.7308202174070235
propriety_total_count: 41489.0
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 70553706496.0
best_of: 2
max_input_tokens: 512
max_output_tokens: 64
display_name: undi95-meta-llama-3-70b-_6209_v3
ineligible_reason: None
language_model: Undi95/Meta-Llama-3-70B-Instruct-hf
model_size: 71B
reward_model: rirv938/gpt2_ties_merge_preference_plus_classic_e2_density_99
us_pacific_date: 2024-04-24
win_ratio: 0.5039491484959414
preference_data_url: None
Resubmit model
Running pipeline stage MKMLizer
Starting job with name undi95-meta-llama-3-70b-6209-v3-mkmlizer
Waiting for job on undi95-meta-llama-3-70b-6209-v3-mkmlizer to finish
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ _____ __ __ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ /___/ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ Version: 0.8.10 ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ The license key for the current software has been verified as ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ belonging to: ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ Chai Research Corp. ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ║ ║
undi95-meta-llama-3-70b-6209-v3-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Downloaded to shared memory in 371.989s
undi95-meta-llama-3-70b-6209-v3-mkmlizer: quantizing model to /dev/shm/model_cache
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Saving flywheel model at /dev/shm/model_cache
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Loading 0: 0%| | 0/723 [00:00<?, ?it/s] Loading 0: 2%|▏ | 16/723 [00:01<00:44, 15.79it/s] Loading 0: 6%|▌ | 44/723 [00:02<00:30, 22.40it/s] Loading 0: 10%|▉ | 71/723 [00:03<00:27, 23.90it/s] Loading 0: 14%|█▎ | 99/723 [00:04<00:24, 25.47it/s] Loading 0: 17%|█▋ | 125/723 [00:18<02:13, 4.48it/s] Loading 0: 20%|██ | 147/723 [00:19<01:38, 5.88it/s] Loading 0: 23%|██▎ | 169/723 [00:20<01:13, 7.53it/s] Loading 0: 27%|██▋ | 194/723 [00:21<00:54, 9.65it/s] Loading 0: 30%|███ | 218/723 [00:22<00:43, 11.72it/s] Loading 0: 33%|███▎ | 242/723 [00:23<00:34, 13.90it/s] Loading 0: 37%|███▋ | 264/723 [00:35<01:31, 4.99it/s] Loading 0: 38%|███▊ | 275/723 [00:36<01:21, 5.49it/s] Loading 0: 42%|████▏ | 301/723 [00:37<00:55, 7.67it/s] Loading 0: 45%|████▌ | 327/723 [00:38<00:39, 10.09it/s] Loading 0: 49%|████▊ | 351/723 [00:39<00:30, 12.27it/s] Loading 0: 51%|█████ | 369/723 [00:40<00:26, 13.27it/s] Loading 0: 54%|█████▍ | 394/723 [00:41<00:21, 15.27it/s] Loading 0: 54%|█████▍ | 394/723 [00:52<00:21, 15.27it/s] Loading 0: 55%|█████▌ | 401/723 [00:52<01:16, 4.21it/s] Loading 0: 59%|█████▊ | 423/723 [00:53<00:51, 5.80it/s] Loading 0: 62%|██████▏ | 448/723 [00:54<00:34, 7.95it/s] Loading 0: 65%|██████▌ | 471/723 [00:55<00:25, 10.04it/s] Loading 0: 68%|██████▊ | 495/723 [00:56<00:18, 12.36it/s] Loading 0: 72%|███████▏ | 522/723 [00:57<00:13, 15.00it/s] Loading 0: 75%|███████▍ | 541/723 [01:08<00:36, 5.02it/s] Loading 0: 78%|███████▊ | 562/723 [01:09<00:25, 6.39it/s] Loading 0: 80%|████████ | 581/723 [01:10<00:18, 7.76it/s] Loading 0: 83%|████████▎ | 603/723 [01:11<00:12, 9.69it/s] Loading 0: 86%|████████▋ | 624/723 [01:12<00:08, 11.50it/s] Loading 0: 89%|████████▉ | 647/723 [01:13<00:05, 13.60it/s] Loading 0: 93%|█████████▎| 673/723 [01:14<00:03, 16.18it/s] Loading 0: 93%|█████████▎| 673/723 [01:25<00:03, 16.18it/s] Loading 0: 94%|█████████▍| 679/723 [01:25<00:10, 4.19it/s] Loading 0: 97%|█████████▋| 704/723 [01:26<00:03, 6.08it/s] Loading 0: 97%|█████████▋| 704/723 [01:38<00:03, 6.08it/s] Loading 0: 100%|██████████| 723/723 [01:38<00:00, 3.35it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
undi95-meta-llama-3-70b-6209-v3-mkmlizer: quantized model in 113.632s
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Processed model Undi95/Meta-Llama-3-70B-Instruct-hf in 511.889s
undi95-meta-llama-3-70b-6209-v3-mkmlizer: creating bucket guanaco-mkml-models
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
undi95-meta-llama-3-70b-6209-v3-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/config.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/tokenizer_config.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/tokenizer.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/special_tokens_map.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.5.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.5.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.3.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.3.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.4.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.4.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.2.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.2.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.1.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.1.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/undi95-meta-llama-3-70b-6209-v3/flywheel_model.0.safetensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: loading reward model from rirv938/gpt2_ties_merge_preference_plus_classic_e2_density_99
undi95-meta-llama-3-70b-6209-v3-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.
undi95-meta-llama-3-70b-6209-v3-mkmlizer: warnings.warn(
undi95-meta-llama-3-70b-6209-v3-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.
undi95-meta-llama-3-70b-6209-v3-mkmlizer: warnings.warn(
undi95-meta-llama-3-70b-6209-v3-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.
undi95-meta-llama-3-70b-6209-v3-mkmlizer: warnings.warn(
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Downloading shards: 0%| | 0/1 [00:00<?, ?it/s] Downloading shards: 100%|██████████| 1/1 [00:03<00:00, 3.28s/it] Downloading shards: 100%|██████████| 1/1 [00:03<00:00, 3.28s/it]
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Saving duration: 0.153s
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Processed model rirv938/gpt2_ties_merge_preference_plus_classic_e2_density_99 in 6.296s
undi95-meta-llama-3-70b-6209-v3-mkmlizer: creating bucket guanaco-reward-models
undi95-meta-llama-3-70b-6209-v3-mkmlizer: Bucket 's3://guanaco-reward-models/' created
undi95-meta-llama-3-70b-6209-v3-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/config.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/special_tokens_map.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/tokenizer_config.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/merges.txt
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/vocab.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/tokenizer.json
undi95-meta-llama-3-70b-6209-v3-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/undi95-meta-llama-3-70b-6209-v3_reward/reward.tensors
Job undi95-meta-llama-3-70b-6209-v3-mkmlizer completed after 580.77s with status: succeeded
Stopping job with name undi95-meta-llama-3-70b-6209-v3-mkmlizer
Pipeline stage MKMLizer completed in 582.45s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.30s
Running pipeline stage ISVCDeployer
Creating inference service undi95-meta-llama-3-70b-6209-v3
Waiting for inference service undi95-meta-llama-3-70b-6209-v3 to be ready
Inference service undi95-meta-llama-3-70b-6209-v3 ready after 144.15928769111633s
Pipeline stage ISVCDeployer completed in 150.92s
Running pipeline stage StressChecker
Received healthy response to inference request in 5.363767385482788s
Received healthy response to inference request in 4.2935168743133545s
Received healthy response to inference request in 2.5490381717681885s
Received healthy response to inference request in 4.29552960395813s
Received healthy response to inference request in 4.395754098892212s
5 requests
0 failed requests
5th percentile: 2.897933912277222
10th percentile: 3.246829652786255
20th percentile: 3.9446211338043216
30th percentile: 4.293919420242309
40th percentile: 4.2947245121002195
50th percentile: 4.29552960395813
60th percentile: 4.335619401931763
70th percentile: 4.375709199905396
80th percentile: 4.589356756210328
90th percentile: 4.976562070846557
95th percentile: 5.170164728164672
99th percentile: 5.325046854019165
mean time: 4.179521226882935
Pipeline stage StressChecker completed in 23.28s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.11s
Running M-Eval for topic stay_in_character
Running pipeline stage DaemonicSafetyScorer
M-Eval Dataset for topic stay_in_character is loaded
Pipeline stage DaemonicSafetyScorer completed in 0.29s
%s, retrying in %s seconds...
undi95-meta-llama-3-70b-_6209_v3 status is now deployed due to DeploymentManager action
%s, retrying in %s seconds...
undi95-meta-llama-3-70b-_6209_v3 status is now inactive due to auto deactivation removed underperforming models
admin requested tearing down of undi95-meta-llama-3-70b-_6209_v3
Running pipeline stage ISVCDeleter
Checking if service undi95-meta-llama-3-70b-6209-v3 is running
Tearing down inference service undi95-meta-llama-3-70b-6209-v3
Toredown service undi95-meta-llama-3-70b-6209-v3
Pipeline stage ISVCDeleter completed in 3.41s
Running pipeline stage MKMLModelDeleter
Cleaning model data from S3
Cleaning model data from model cache
Deleting key undi95-meta-llama-3-70b-6209-v3/config.json from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.0.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.1.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.2.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.3.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.4.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/flywheel_model.5.safetensors from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/special_tokens_map.json from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/tokenizer.json from bucket guanaco-mkml-models
Deleting key undi95-meta-llama-3-70b-6209-v3/tokenizer_config.json from bucket guanaco-mkml-models
Cleaning model data from model cache
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/config.json from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/merges.txt from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/reward.tensors from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/special_tokens_map.json from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/tokenizer.json from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/tokenizer_config.json from bucket guanaco-reward-models
Deleting key undi95-meta-llama-3-70b-6209-v3_reward/vocab.json from bucket guanaco-reward-models
Pipeline stage MKMLModelDeleter completed in 7.18s
undi95-meta-llama-3-70b-_6209_v3 status is now torndown due to DeploymentManager action

Usage Metrics

Latency Metrics