{
  "schema_version": 2,
  "id": "develop/ai/redisvl/user_guide/threshold_optimization",
  "title": "Threshold Optimization",
  "url": "https://redis.io/docs/latest/develop/ai/redisvl/0.6.0/user_guide/threshold_optimization/",
  "summary": "",
  "content": "\n\nAfter setting up `SemanticRouter` or `SemanticCache` it's best to tune the `distance_threshold` to get the most performance out of your system. RedisVL provides helper classes to make this light weight optimization easy.\n\n**Note:** Threshold optimization relies on `python \u003e 3.9.`\n\n# CacheThresholdOptimizer\n\nLet's say you setup the following semantic cache with a distance_threshold of `X` and store the entries:\n\n- prompt: `what is the capital of france?` response: `paris`\n- prompt: `what is the capital of morocco?` response: `rabat`\n\n\n```python\nfrom redisvl.extensions.cache.llm import SemanticCache\nfrom redisvl.utils.vectorize import HFTextVectorizer\n\nsem_cache = SemanticCache(\n    name=\"sem_cache\",                                       # underlying search index name\n    redis_url=\"redis://localhost:6379\",                     # redis connection url string\n    distance_threshold=0.5,                                 # semantic cache distance threshold\n    vectorizer=HFTextVectorizer(\"redis/langcache-embed-v1\") # embedding model\n)\n\nparis_key = sem_cache.store(prompt=\"what is the capital of france?\", response=\"paris\")\nrabat_key = sem_cache.store(prompt=\"what is the capital of morocco?\", response=\"rabat\")\n\n```\n\n    /Users/justin.cechmanek/.pyenv/versions/3.13/envs/redisvl-dev/lib/python3.13/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n      from .autonotebook import tqdm as notebook_tqdm\n\n\n    16:16:11 sentence_transformers.SentenceTransformer INFO   Use pytorch device_name: mps\n    16:16:11 sentence_transformers.SentenceTransformer INFO   Load pretrained SentenceTransformer: redis/langcache-embed-v1\n\n\n    Batches:   0%|          | 0/1 [00:00\u003c?, ?it/s]Compiling the model with `torch.compile` and using a `torch.mps` device is not supported. Falling back to non-compiled mode.\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  3.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  1.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 25.04it/s]\n\n\nThis works well but we want to make sure the cache only applies for the appropriate questions. If we test the cache with a question we don't want a response to we see that the current distance_threshold is too high. \n\n\n```python\nsem_cache.check(\"what's the capital of britain?\")\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  1.24it/s]\n\n\n\n\n\n    [{'entry_id': 'c990cc06e5e77570e5f03360426d2b7f947cbb5a67daa8af8164bfe0b3e24fe3',\n      'prompt': 'what is the capital of france?',\n      'response': 'paris',\n      'vector_distance': 0.335606634617,\n      'inserted_at': 1746051375.81,\n      'updated_at': 1746051375.81,\n      'key': 'sem_cache:c990cc06e5e77570e5f03360426d2b7f947cbb5a67daa8af8164bfe0b3e24fe3'}]\n\n\n\n### Define test_data and optimize\n\nWith the `CacheThresholdOptimizer` you can quickly tune the distance threshold by providing some test data in the form:\n\n```json\n[\n    {\n        \"query\": \"What's the capital of Britain?\",\n        \"query_match\": \"\"\n    },\n    {\n        \"query\": \"What's the capital of France??\",\n        \"query_match\": paris_key\n    },\n    {\n        \"query\": \"What's the capital city of Morocco?\",\n        \"query_match\": rabat_key\n    },\n]\n```\n\nThe threshold optimizer will then efficiently execute and score different threshold against the what is currently populated in your cache and automatically update the threshold of the cache to the best setting\n\n\n```python\nfrom redisvl.utils.optimize import CacheThresholdOptimizer\n\ntest_data = [\n    {\n        \"query\": \"What's the capital of Britain?\",\n        \"query_match\": \"\"\n    },\n    {\n        \"query\": \"What's the capital of France??\",\n        \"query_match\": paris_key\n    },\n    {\n        \"query\": \"What's the capital city of Morocco?\",\n        \"query_match\": rabat_key\n    },\n]\n\nprint(f\"Distance threshold before: {sem_cache.distance_threshold} \\n\")\noptimizer = CacheThresholdOptimizer(sem_cache, test_data)\noptimizer.optimize()\nprint(f\"Distance threshold after: {sem_cache.distance_threshold} \\n\")\n```\n\n    Distance threshold before: 0.5 \n    \n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  1.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 23.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 24.43it/s]\n    /Users/justin.cechmanek/.pyenv/versions/3.13/envs/redisvl-dev/lib/python3.13/site-packages/ranx/metrics/f1.py:36: NumbaTypeSafetyWarning: unsafe cast from uint64 to int64. Precision may be lost.\n      scores[i] = _f1(qrels[i], run[i], k, rel_lvl)\n\n\n    Distance threshold after: 0.10372881355932204 \n    \n\n\nWe can also see that we no longer match on the incorrect example:\n\n\n```python\nsem_cache.check(\"what's the capital of britain?\")\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 12.39it/s]\n\n\n\n\n\n    []\n\n\n\nBut still match on highly relevant prompts:\n\n\n```python\nsem_cache.check(\"what's the capital city of france?\")\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 25.92it/s]\n\n\n\n\n\n    [{'entry_id': 'c990cc06e5e77570e5f03360426d2b7f947cbb5a67daa8af8164bfe0b3e24fe3',\n      'prompt': 'what is the capital of france?',\n      'response': 'paris',\n      'vector_distance': 0.043138384819,\n      'inserted_at': 1746051375.81,\n      'updated_at': 1746051375.81,\n      'key': 'sem_cache:c990cc06e5e77570e5f03360426d2b7f947cbb5a67daa8af8164bfe0b3e24fe3'}]\n\n\n\n# RouterThresholdOptimizer\n\nVery similar to the caching case, you can optimize your router.\n\n### Define the routes\n\n\n```python\nfrom redisvl.extensions.router import Route\n\nroutes = [\n        Route(\n            name=\"greeting\",\n            references=[\"hello\", \"hi\"],\n            metadata={\"type\": \"greeting\"},\n            distance_threshold=0.5,\n        ),\n        Route(\n            name=\"farewell\",\n            references=[\"bye\", \"goodbye\"],\n            metadata={\"type\": \"farewell\"},\n            distance_threshold=0.5,\n        ),\n    ]\n```\n\n### Initialize the SemanticRouter\n\n\n```python\nimport os\nfrom redisvl.extensions.router import SemanticRouter\nfrom redisvl.utils.vectorize import HFTextVectorizer\n\nos.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n\n# Initialize the SemanticRouter\nrouter = SemanticRouter(\n    name=\"greeting-router\",\n    vectorizer=HFTextVectorizer(),\n    routes=routes,\n    redis_url=\"redis://localhost:6379\",\n    overwrite=True # Blow away any other routing index with this name\n)\n```\n\n    16:16:41 sentence_transformers.SentenceTransformer INFO   Use pytorch device_name: mps\n    16:16:41 sentence_transformers.SentenceTransformer INFO   Load pretrained SentenceTransformer: sentence-transformers/all-mpnet-base-v2\n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  5.90it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  4.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.29it/s]\n\n\n### Provide test_data\n\n\n```python\ntest_data = [\n    # Greetings\n    {\"query\": \"hello\", \"query_match\": \"greeting\"},\n    {\"query\": \"hi\", \"query_match\": \"greeting\"},\n    {\"query\": \"hey\", \"query_match\": \"greeting\"},\n    {\"query\": \"greetings\", \"query_match\": \"greeting\"},\n    {\"query\": \"good morning\", \"query_match\": \"greeting\"},\n    {\"query\": \"good afternoon\", \"query_match\": \"greeting\"},\n    {\"query\": \"good evening\", \"query_match\": \"greeting\"},\n    {\"query\": \"howdy\", \"query_match\": \"greeting\"},\n    {\"query\": \"what's up\", \"query_match\": \"greeting\"},\n    {\"query\": \"yo\", \"query_match\": \"greeting\"},\n    {\"query\": \"hiya\", \"query_match\": \"greeting\"},\n    {\"query\": \"salutations\", \"query_match\": \"greeting\"},\n    {\"query\": \"how's it going\", \"query_match\": \"greeting\"},\n    {\"query\": \"how are you\", \"query_match\": \"greeting\"},\n    {\"query\": \"nice to meet you\", \"query_match\": \"greeting\"},\n    # Farewells\n    {\"query\": \"goodbye\", \"query_match\": \"farewell\"},\n    {\"query\": \"bye\", \"query_match\": \"farewell\"},\n    {\"query\": \"see you later\", \"query_match\": \"farewell\"},\n    {\"query\": \"take care\", \"query_match\": \"farewell\"},\n    {\"query\": \"farewell\", \"query_match\": \"farewell\"},\n    {\"query\": \"have a good day\", \"query_match\": \"farewell\"},\n    {\"query\": \"see you soon\", \"query_match\": \"farewell\"},\n    {\"query\": \"catch you later\", \"query_match\": \"farewell\"},\n    {\"query\": \"so long\", \"query_match\": \"farewell\"},\n    {\"query\": \"peace out\", \"query_match\": \"farewell\"},\n    {\"query\": \"later\", \"query_match\": \"farewell\"},\n    {\"query\": \"all the best\", \"query_match\": \"farewell\"},\n    {\"query\": \"take it easy\", \"query_match\": \"farewell\"},\n    {\"query\": \"have a good one\", \"query_match\": \"farewell\"},\n    {\"query\": \"cheerio\", \"query_match\": \"farewell\"},\n    # Null matches\n    {\"query\": \"what's the capital of britain?\", \"query_match\": \"\"},\n    {\"query\": \"what does laffy taffy taste like?\", \"query_match\": \"\"},\n]\n```\n\n### Optimize\n\nNote: by default route distance threshold optimization will use a random search to find the best threshold since, unlike caching, there are many thresholds to optimize concurrently. \n\n\n```python\nfrom redisvl.utils.optimize import RouterThresholdOptimizer\n\nprint(f\"Route thresholds before: {router.route_thresholds} \\n\")\noptimizer = RouterThresholdOptimizer(router, test_data)\noptimizer.optimize()\n```\n\n    Route thresholds before: {'greeting': 0.5, 'farewell': 0.5} \n    \n\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  7.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.81it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.68it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  9.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 44.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  9.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  9.56it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.82it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.00it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.18it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  7.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00,  9.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.68it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.00it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.86it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 74.82it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.64it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.84it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.29it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.15it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 75.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 73.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.86it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 74.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.81it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 53.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.07it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.29it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.93it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.15it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.62it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.78it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.88it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.76it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.76it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 52.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.46it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.29it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.46it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.76it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 22.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.67it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.04it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.20it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.88it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.29it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.07it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.87it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.90it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.23it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.18it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.82it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.15it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.56it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.76it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.68it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 72.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.74it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.05it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 71.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 20.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.29it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.76it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.89it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.49it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.81it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.82it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.23it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.56it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 52.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.87it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.06it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.49it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.87it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.67it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.84it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.62it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 22.93it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.20it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.99it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.74it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 19.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.93it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.86it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.84it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.81it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.64it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.15it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 48.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.87it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.46it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.88it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.28it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.21it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 24.78it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 51.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.93it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.64it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.04it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.04it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.49it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.78it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.46it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.56it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.11it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 26.60it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 48.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.43it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.19it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.15it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.56it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 50.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.06it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 28.07it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.20it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.04it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 52.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.17it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.80it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.67it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.62it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 31.04it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 54.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.03it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.59it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 54.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 53.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.55it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.89it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.89it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.02it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 38.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 54.75it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.26it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 52.05it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 53.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.93it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.06it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.88it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.79it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.01it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.35it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.24it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.05it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.70it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.73it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 69.52it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.09it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 52.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 54.25it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.34it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.47it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.92it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 28.70it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.89it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.81it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.07it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.85it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.54it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.23it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.84it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.31it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.58it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.87it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.77it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.50it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.51it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.72it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 57.14it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.98it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.41it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 67.63it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 17.08it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.97it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.86it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.74it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.91it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.53it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.64it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 56.45it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.61it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.10it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.22it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 59.16it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.86it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 58.44it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.69it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.12it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.96it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.95it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.36it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 68.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.83it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.57it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 41.90it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 20.01it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.68it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 61.99it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.13it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.30it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.66it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.39it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.27it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 63.94it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.42it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 60.40it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.71it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.32it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 62.48it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 64.38it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.65it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 65.33it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 70.37it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.00it/s]\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 66.63it/s]\n\n\n    Eval metric F1: start 0.438, end 0.812 \n    Ending thresholds: {'greeting': 0.5828282828282831, 'farewell': 0.7545454545454545}\n\n\n### Test it out\n\n\n```python\n# Query the router with a statement\nroute_match = router(\"hi there\")\nroute_match\n```\n\n    Batches: 100%|██████████| 1/1 [00:00\u003c00:00, 55.72it/s]\n\n\n\n\n\n    RouteMatch(name='greeting', distance=0.295984089375)\n\n\n\n## Cleanup\n\n\n```python\nrouter.delete()\nsem_cache.delete()\n```\n",
  "tags": [],
  "last_updated": "0001-01-01T00:00:00Z"
}