undi95-remm-slerp-l2-13b

developer_uid: fengjiayi
submission_id: undi95-remm-slerp-l2-13b_v1
model_name: undi95-remm-slerp-l2-13b_v1
model_group: Undi95/ReMM-SLERP-L2-13B
status: torndown
timestamp: 2025-01-23T05:30:02+00:00
num_battles: 6232
num_wins: 2875
celo_rating: 1190.27
family_friendly_score: 0.5942000000000001
family_friendly_standard_error: 0.006944441806221721
submission_type: basic
model_repo: Undi95/ReMM-SLERP-L2-13B
model_architecture: LlamaForCausalLM
model_num_parameters: 13015864320.0
best_of: 8
max_input_tokens: 1024
max_output_tokens: 64
display_name: undi95-remm-slerp-l2-13b_v1
is_internal_developer: False
language_model: Undi95/ReMM-SLERP-L2-13B
model_size: 13B
ranking_group: single
us_pacific_date: 2025-01-22
win_ratio: 0.4613286264441592
generation_params: {'temperature': 1.0, 'top_p': 1.0, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['\n'], 'max_input_tokens': 1024, 'best_of': 8, 'max_output_tokens': 64}
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}
Resubmit model
Shutdown handler not registered because Python interpreter is not running in the main thread
run pipeline %s
run pipeline stage %s
Running pipeline stage MKMLizer
Starting job with name undi95-remm-slerp-l2-13b-v1-mkmlizer
Waiting for job on undi95-remm-slerp-l2-13b-v1-mkmlizer to finish
undi95-remm-slerp-l2-13b-v1-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║     _____            __           __                                ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║    / _/ /_ ___    __/ /  ___ ___ / /                                ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║   / _/ / // / |/|/ / _ \/ -_) -_) /                                 ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  /_//_/\_, /|__,__/_//_/\__/\__/_/                                  ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║       /___/                                                         ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║                                                                     ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  Version: 0.11.12                                                   ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  Copyright 2023 MK ONE TECHNOLOGIES Inc.                            ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  https://mk1.ai                                                     ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║                                                                     ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  The license key for the current software has been verified as      ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  belonging to:                                                      ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║                                                                     ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  Chai Research Corp.                                                ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f                   ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║  Expiration: 2025-04-15 23:59:59                                    ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ║                                                                     ║
undi95-remm-slerp-l2-13b-v1-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
undi95-remm-slerp-l2-13b-v1-mkmlizer: Downloaded to shared memory in 59.301s
undi95-remm-slerp-l2-13b-v1-mkmlizer: quantizing model to /dev/shm/model_cache, profile:s0, folder:/tmp/tmpubjlqytc, device:0
undi95-remm-slerp-l2-13b-v1-mkmlizer: Saving flywheel model at /dev/shm/model_cache
undi95-remm-slerp-l2-13b-v1-mkmlizer: /opt/conda/lib/python3.10/site-packages/mk1/flywheel/functional/loader.py:55: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
undi95-remm-slerp-l2-13b-v1-mkmlizer:   tensors = torch.load(model_shard_filename, map_location=torch.device(self.device), mmap=True)
undi95-remm-slerp-l2-13b-v1-mkmlizer: 
Loading 0:   0%|          | 0/363 [00:00<?, ?it/s]
Loading 0:   1%|          | 4/363 [00:00<00:10, 35.76it/s]
Loading 0:   2%|▏         | 8/363 [00:00<00:10, 32.67it/s]
Loading 0:   4%|▍         | 16/363 [00:00<00:08, 40.12it/s]
Loading 0:   7%|▋         | 25/363 [00:00<00:07, 44.83it/s]
Loading 0:   9%|▉         | 34/363 [00:00<00:07, 45.98it/s]
Loading 0:  12%|█▏        | 43/363 [00:00<00:06, 47.70it/s]
Loading 0:  14%|█▍        | 52/363 [00:01<00:06, 48.96it/s]
Loading 0:  17%|█▋        | 61/363 [00:01<00:06, 48.21it/s]
Loading 0:  19%|█▉        | 70/363 [00:01<00:05, 48.88it/s]
Loading 0:  22%|██▏       | 79/363 [00:01<00:05, 49.64it/s]
Loading 0:  24%|██▍       | 88/363 [00:01<00:05, 50.28it/s]
Loading 0:  27%|██▋       | 97/363 [00:02<00:05, 50.44it/s]
Loading 0:  29%|██▉       | 106/363 [00:02<00:05, 49.88it/s]
Loading 0:  32%|███▏      | 115/363 [00:02<00:04, 50.19it/s]
Loading 0:  34%|███▍      | 124/363 [00:02<00:04, 50.27it/s]
Loading 0:  37%|███▋      | 133/363 [00:02<00:04, 49.46it/s]
Loading 0:  39%|███▉      | 142/363 [00:02<00:04, 49.38it/s]
Loading 0:  40%|████      | 147/363 [00:06<00:37,  5.79it/s]
Loading 0:  42%|████▏     | 151/363 [00:07<00:31,  6.69it/s]
Loading 0:  44%|████▍     | 160/363 [00:07<00:20,  9.83it/s]
Loading 0:  47%|████▋     | 169/363 [00:07<00:14, 13.68it/s]
Loading 0:  49%|████▉     | 178/363 [00:07<00:10, 18.10it/s]
Loading 0:  51%|█████     | 185/363 [00:07<00:07, 22.56it/s]
Loading 0:  53%|█████▎    | 193/363 [00:07<00:06, 26.61it/s]
Loading 0:  56%|█████▌    | 202/363 [00:08<00:05, 31.95it/s]
Loading 0:  58%|█████▊    | 211/363 [00:08<00:04, 36.48it/s]
Loading 0:  61%|██████    | 220/363 [00:08<00:03, 40.41it/s]
Loading 0:  63%|██████▎   | 229/363 [00:08<00:03, 44.01it/s]
Loading 0:  66%|██████▌   | 238/363 [00:08<00:02, 45.58it/s]
Loading 0:  68%|██████▊   | 247/363 [00:08<00:02, 47.24it/s]
Loading 0:  71%|███████   | 256/363 [00:09<00:02, 47.04it/s]
Loading 0:  73%|███████▎  | 265/363 [00:09<00:02, 48.25it/s]
Loading 0:  75%|███████▌  | 274/363 [00:09<00:01, 47.92it/s]
Loading 0:  78%|███████▊  | 283/363 [00:09<00:01, 48.05it/s]
Loading 0:  80%|████████  | 292/363 [00:09<00:01, 48.96it/s]
Loading 0:  82%|████████▏ | 297/363 [00:12<00:07,  8.91it/s]
Loading 0:  84%|████████▎ | 304/363 [00:12<00:05, 11.44it/s]
Loading 0:  86%|████████▌ | 313/363 [00:12<00:03, 15.79it/s]
Loading 0:  89%|████████▊ | 322/363 [00:12<00:01, 20.81it/s]
Loading 0:  91%|█████████ | 331/363 [00:12<00:01, 26.61it/s]
Loading 0:  94%|█████████▎| 340/363 [00:13<00:00, 32.31it/s]
Loading 0:  96%|█████████▌| 349/363 [00:13<00:00, 37.73it/s]
Loading 0:  99%|█████████▊| 358/363 [00:13<00:00, 41.23it/s]
                                                            
You are using the default legacy behaviour of the <class 'transformers.models.llama.tokenization_llama.LlamaTokenizer'>. This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thoroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565 - if you loaded a llama tokenizer from a GGUF file you can ignore this message
undi95-remm-slerp-l2-13b-v1-mkmlizer: You are using the default legacy behaviour of the <class 'transformers.models.llama.tokenization_llama_fast.LlamaTokenizerFast'>. This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thoroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565 - if you loaded a llama tokenizer from a GGUF file you can ignore this message.
undi95-remm-slerp-l2-13b-v1-mkmlizer: quantized model in 38.107s
undi95-remm-slerp-l2-13b-v1-mkmlizer: Processed model Undi95/ReMM-SLERP-L2-13B in 97.409s
undi95-remm-slerp-l2-13b-v1-mkmlizer: creating bucket guanaco-mkml-models
undi95-remm-slerp-l2-13b-v1-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
undi95-remm-slerp-l2-13b-v1-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/special_tokens_map.json
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/tokenizer_config.json
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/config.json
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.model s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/tokenizer.model
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/tokenizer.json
undi95-remm-slerp-l2-13b-v1-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/undi95-remm-slerp-l2-13b-v1/flywheel_model.0.safetensors
Job undi95-remm-slerp-l2-13b-v1-mkmlizer completed after 125.23s with status: succeeded
Stopping job with name undi95-remm-slerp-l2-13b-v1-mkmlizer
Pipeline stage MKMLizer completed in 125.71s
run pipeline stage %s
Running pipeline stage MKMLTemplater
Pipeline stage MKMLTemplater completed in 0.17s
run pipeline stage %s
Running pipeline stage MKMLDeployer
Creating inference service undi95-remm-slerp-l2-13b-v1
Waiting for inference service undi95-remm-slerp-l2-13b-v1 to be ready
Inference service undi95-remm-slerp-l2-13b-v1 ready after 140.49409461021423s
Pipeline stage MKMLDeployer completed in 141.00s
run pipeline stage %s
Running pipeline stage StressChecker
Received healthy response to inference request in 1.822190284729004s
Received healthy response to inference request in 1.4925763607025146s
Received healthy response to inference request in 1.741950511932373s
Received healthy response to inference request in 1.5221736431121826s
Received healthy response to inference request in 1.966792345046997s
5 requests
0 failed requests
5th percentile: 1.4984958171844482
10th percentile: 1.5044152736663818
20th percentile: 1.516254186630249
30th percentile: 1.5661290168762207
40th percentile: 1.6540397644042968
50th percentile: 1.741950511932373
60th percentile: 1.7740464210510254
70th percentile: 1.8061423301696777
80th percentile: 1.8511106967926025
90th percentile: 1.9089515209197998
95th percentile: 1.9378719329833984
99th percentile: 1.9610082626342773
mean time: 1.7091366291046142
Pipeline stage StressChecker completed in 9.96s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyTriggerPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage OfflineFamilyFriendlyTriggerPipeline completed in 0.66s
run pipeline stage %s
Running pipeline stage TriggerMKMLProfilingPipeline
run_pipeline:run_in_cloud %s
starting trigger_guanaco_pipeline args=%s
triggered trigger_guanaco_pipeline args=%s
Pipeline stage TriggerMKMLProfilingPipeline completed in 0.66s
Shutdown handler de-registered
undi95-remm-slerp-l2-13b_v1 status is now deployed due to DeploymentManager action
Shutdown handler registered
run pipeline %s
run pipeline stage %s
Running pipeline stage OfflineFamilyFriendlyScorer
Evaluating %s Family Friendly Score with %s threads
Pipeline stage OfflineFamilyFriendlyScorer completed in 3028.99s
Shutdown handler de-registered
undi95-remm-slerp-l2-13b_v1 status is now inactive due to auto deactivation removed underperforming models
undi95-remm-slerp-l2-13b_v1 status is now torndown due to DeploymentManager action