colbert-distilroberta-base

PyLate model based on distilbert/distilroberta-base

This is a PyLate model finetuned from distilbert/distilroberta-base on the msmarco-train dataset. It maps sentences & paragraphs to sequences of 128-dimensional dense vectors and can be used for semantic textual similarity using the MaxSim operator.

Model Details

Model Description

• Model Type: PyLate model
• Base model: distilbert/distilroberta-base <!-- at revision fb53ab8802853c8e4fbdbcd0529f21fc6f459b2b -->
• Document Length: 256 tokens
• Query Length: 32 tokens
• Output Dimensionality: 128 tokens
• Similarity Function: MaxSim
• Training Dataset:
  • msmarco-train

<!-- - **Language:** Unknown --> <!-- - **License:** Unknown -->

Model Sources

• Documentation: PyLate Documentation
• Repository: PyLate on GitHub
• Hugging Face: PyLate models on Hugging Face

Full Model Architecture

ColBERT(
  (0): Transformer({'max_seq_length': 255, 'do_lower_case': False}) with Transformer model: RobertaModel 
  (1): Dense({'in_features': 768, 'out_features': 128, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
)

Usage

First install the PyLate library:

pip install -U pylate

Retrieval

PyLate provides a streamlined interface to index and retrieve documents using ColBERT models. The index leverages the Voyager HNSW index to efficiently handle document embeddings and enable fast retrieval.

Indexing documents

First, load the ColBERT model and initialize the Voyager index, then encode and index your documents:

from pylate import indexes, models, retrieve

# Step 1: Load the ColBERT model
model = models.ColBERT(
    model_name_or_path=yosefw/colbert-distilroberta-base,
)

# Step 2: Initialize the Voyager index
index = indexes.Voyager(
    index_folder="pylate-index",
    index_name="index",
    override=True,  # This overwrites the existing index if any
)

# Step 3: Encode the documents
documents_ids = ["1", "2", "3"]
documents = ["document 1 text", "document 2 text", "document 3 text"]

documents_embeddings = model.encode(
    documents,
    batch_size=32,
    is_query=False,  # Ensure that it is set to False to indicate that these are documents, not queries
    show_progress_bar=True,
)

# Step 4: Add document embeddings to the index by providing embeddings and corresponding ids
index.add_documents(
    documents_ids=documents_ids,
    documents_embeddings=documents_embeddings,
)

Note that you do not have to recreate the index and encode the documents every time. Once you have created an index and added the documents, you can re-use the index later by loading it:

# To load an index, simply instantiate it with the correct folder/name and without overriding it
index = indexes.Voyager(
    index_folder="pylate-index",
    index_name="index",
)

Retrieving top-k documents for queries

Once the documents are indexed, you can retrieve the top-k most relevant documents for a given set of queries. To do so, initialize the ColBERT retriever with the index you want to search in, encode the queries and then retrieve the top-k documents to get the top matches ids and relevance scores:

# Step 1: Initialize the ColBERT retriever
retriever = retrieve.ColBERT(index=index)

# Step 2: Encode the queries
queries_embeddings = model.encode(
    ["query for document 3", "query for document 1"],
    batch_size=32,
    is_query=True,  #  # Ensure that it is set to False to indicate that these are queries
    show_progress_bar=True,
)

# Step 3: Retrieve top-k documents
scores = retriever.retrieve(
    queries_embeddings=queries_embeddings,
    k=10,  # Retrieve the top 10 matches for each query
)

Reranking

If you only want to use the ColBERT model to perform reranking on top of your first-stage retrieval pipeline without building an index, you can simply use rank function and pass the queries and documents to rerank:

from pylate import rank, models

queries = [
    "query A",
    "query B",
]

documents = [
    ["document A", "document B"],
    ["document 1", "document C", "document B"],
]

documents_ids = [
    [1, 2],
    [1, 3, 2],
]

model = models.ColBERT(
    model_name_or_path=yosefw/colbert-distilroberta-base,
)

queries_embeddings = model.encode(
    queries,
    is_query=True,
)

documents_embeddings = model.encode(
    documents,
    is_query=False,
)

reranked_documents = rank.rerank(
    documents_ids=documents_ids,
    queries_embeddings=queries_embeddings,
    documents_embeddings=documents_embeddings,
)

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* -->

Evaluation

Metrics

Col BERTTriplet

• Evaluated with <code>pylate.evaluation.colbert_triplet.ColBERTTripletEvaluator</code>

<!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Dataset

msmarco-train

• Dataset: msmarco-train at 6853021
• Size: 500,000 training samples
• Columns: <code>query_id</code>, <code>query</code>, <code>positive</code>, <code>negative_1</code>, <code>negative_2</code>, <code>negative_3</code>, and <code>negative_4</code>
• Approximate statistics based on the first 1000 samples:

  	query_id	query	positive	negative_1	negative_2	negative_3	negative_4
  type	int	string	string	string	string	string	string
  details	<ul><li>13501: ~0.20%</li><li>13502: ~0.20%</li><li>13503: ~0.20%</li><li>13506: ~0.20%</li><li>13508: ~0.20%</li><li>13510: ~0.20%</li><li>13515: ~0.20%</li><li>13517: ~0.20%</li><li>13520: ~0.20%</li><li>13525: ~0.20%</li><li>13528: ~0.20%</li><li>13529: ~0.20%</li><li>13531: ~0.20%</li><li>13533: ~0.20%</li><li>13535: ~0.20%</li><li>13536: ~0.20%</li><li>13537: ~0.20%</li><li>13539: ~0.20%</li><li>13540: ~0.20%</li><li>13541: ~0.20%</li><li>13542: ~0.20%</li><li>13543: ~0.20%</li><li>13545: ~0.20%</li><li>13550: ~0.20%</li><li>13551: ~0.20%</li><li>13562: ~0.20%</li><li>13565: ~0.20%</li><li>13566: ~0.20%</li><li>13567: ~0.20%</li><li>13568: ~0.20%</li><li>13569: ~0.20%</li><li>13570: ~0.20%</li><li>13571: ~0.20%</li><li>13573: ~0.20%</li><li>13574: ~0.20%</li><li>13575: ~0.20%</li><li>13576: ~0.20%</li><li>13580: ~0.20%</li><li>13581: ~0.20%</li><li>13583: ~0.20%</li><li>13585: ~0.20%</li><li>13589: ~0.20%</li><li>13590: ~0.20%</li><li>13593: ~0.20%</li><li>13595: ~0.20%</li><li>13596: ~0.20%</li><li>13597: ~0.20%</li><li>13599: ~0.20%</li><li>13604: ~0.20%</li><li>13609: ~0.20%</li><li>13611: ~0.20%</li><li>13613: ~0.20%</li><li>13618: ~0.20%</li><li>13619: ~0.20%</li><li>13620: ~0.20%</li><li>13622: ~0.20%</li><li>13623: ~0.20%</li><li>13624: ~0.20%</li><li>13626: ~0.20%</li><li>13627: ~0.20%</li><li>13628: ~0.20%</li><li>13629: ~0.20%</li><li>13632: ~0.20%</li><li>13636: ~0.20%</li><li>13637: ~0.20%</li><li>13638: ~0.20%</li><li>13642: ~0.20%</li><li>13644: ~0.20%</li><li>13649: ~0.20%</li><li>13653: ~0.20%</li><li>13656: ~0.20%</li><li>13664: ~0.20%</li><li>13675: ~0.20%</li><li>13678: ~0.20%</li><li>13679: ~0.20%</li><li>13680: ~0.20%</li><li>13682: ~0.20%</li><li>13687: ~0.20%</li><li>13688: ~0.20%</li><li>13690: ~0.20%</li><li>13692: ~0.20%</li><li>13693: ~0.20%</li><li>13695: ~0.20%</li><li>13696: ~0.20%</li><li>13698: ~0.20%</li><li>13700: ~0.20%</li><li>13701: ~0.20%</li><li>13702: ~0.20%</li><li>13704: ~0.20%</li><li>13705: ~0.20%</li><li>13706: ~0.20%</li><li>13715: ~0.20%</li><li>13725: ~0.20%</li><li>13726: ~0.20%</li><li>13732: ~0.20%</li><li>13736: ~0.20%</li><li>13737: ~0.20%</li><li>13738: ~0.20%</li><li>13740: ~0.20%</li><li>13741: ~0.20%</li><li>13743: ~0.20%</li><li>13745: ~0.20%</li><li>13746: ~0.20%</li><li>13748: ~0.20%</li><li>13750: ~0.20%</li><li>13751: ~0.20%</li><li>13752: ~0.20%</li><li>13753: ~0.20%</li><li>13754: ~0.20%</li><li>13755: ~0.20%</li><li>13757: ~0.20%</li><li>13758: ~0.20%</li><li>13759: ~0.20%</li><li>13762: ~0.20%</li><li>13764: ~0.20%</li><li>13767: ~0.20%</li><li>13768: ~0.20%</li><li>13769: ~0.20%</li><li>13771: ~0.20%</li><li>13772: ~0.20%</li><li>13774: ~0.20%</li><li>13775: ~0.20%</li><li>13778: ~0.20%</li><li>13782: ~0.20%</li><li>13783: ~0.20%</li><li>13785: ~0.20%</li><li>13787: ~0.20%</li><li>13789: ~0.20%</li><li>13790: ~0.20%</li><li>13791: ~0.20%</li><li>13795: ~0.20%</li><li>13798: ~0.20%</li><li>13799: ~0.20%</li><li>13802: ~0.20%</li><li>13803: ~0.20%</li><li>13805: ~0.20%</li><li>13806: ~0.20%</li><li>13810: ~0.20%</li><li>13812: ~0.20%</li><li>13814: ~0.20%</li><li>13822: ~0.20%</li><li>13826: ~0.20%</li><li>13829: ~0.20%</li><li>13832: ~0.20%</li><li>13835: ~0.20%</li><li>13837: ~0.20%</li><li>13841: ~0.20%</li><li>13842: ~0.20%</li><li>13844: ~0.20%</li><li>13850: ~0.20%</li><li>13851: ~0.20%</li><li>13852: ~0.20%</li><li>13857: ~0.20%</li><li>13859: ~0.20%</li><li>13860: ~0.20%</li><li>13862: ~0.20%</li><li>13863: ~0.20%</li><li>13867: ~0.20%</li><li>13873: ~0.20%</li><li>13874: ~0.20%</li><li>13876: ~0.20%</li><li>13879: ~0.20%</li><li>13880: ~0.20%</li><li>13882: ~0.20%</li><li>13883: ~0.20%</li><li>13887: ~0.20%</li><li>13890: ~0.20%</li><li>13894: ~0.20%</li><li>13896: ~0.20%</li><li>13897: ~0.20%</li><li>13898: ~0.20%</li><li>13899: ~0.20%</li><li>13903: ~0.20%</li><li>13911: ~0.20%</li><li>13914: ~0.20%</li><li>13915: ~0.20%</li><li>13923: ~0.20%</li><li>13924: ~0.20%</li><li>13927: ~0.20%</li><li>13930: ~0.20%</li><li>13931: ~0.20%</li><li>13933: ~0.20%</li><li>13936: ~0.20%</li><li>13940: ~0.20%</li><li>13944: ~0.20%</li><li>13947: ~0.20%</li><li>13948: ~0.20%</li><li>13949: ~0.20%</li><li>13955: ~0.20%</li><li>13956: ~0.20%</li><li>13958: ~0.20%</li><li>13959: ~0.20%</li><li>13964: ~0.20%</li><li>13966: ~0.20%</li><li>13968: ~0.20%</li><li>13971: ~0.20%</li><li>13974: ~0.20%</li><li>13975: ~0.20%</li><li>13976: ~0.20%</li><li>13983: ~0.20%</li><li>13984: ~0.20%</li><li>13986: ~0.20%</li><li>13990: ~0.20%</li><li>13996: ~0.20%</li><li>14000: ~0.20%</li><li>14001: ~0.20%</li><li>14002: ~0.20%</li><li>14003: ~0.20%</li><li>14004: ~0.20%</li><li>14006: ~0.20%</li><li>14007: ~0.20%</li><li>14010: ~0.20%</li><li>14012: ~0.20%</li><li>14015: ~0.20%</li><li>14016: ~0.20%</li><li>14017: ~0.20%</li><li>14019: ~0.20%</li><li>14020: ~0.20%</li><li>14022: ~0.20%</li><li>14025: ~0.20%</li><li>14029: ~0.20%</li><li>14035: ~0.20%</li><li>14036: ~0.20%</li><li>14043: ~0.20%</li><li>14045: ~0.20%</li><li>14047: ~0.20%</li><li>14049: ~0.20%</li><li>14050: ~0.20%</li><li>14053: ~0.20%</li><li>14055: ~0.20%</li><li>14061: ~0.20%</li><li>14062: ~0.20%</li><li>14063: ~0.20%</li><li>14068: ~0.20%</li><li>14072: ~0.20%</li><li>14075: ~0.20%</li><li>14076: ~0.20%</li><li>14077: ~0.20%</li><li>14078: ~0.20%</li><li>14079: ~0.20%</li><li>14080: ~0.20%</li><li>14081: ~0.20%</li><li>14082: ~0.20%</li><li>14083: ~0.20%</li><li>14086: ~0.20%</li><li>14087: ~0.20%</li><li>14088: ~0.20%</li><li>14091: ~0.20%</li><li>14092: ~0.20%</li><li>14101: ~0.20%</li><li>14102: ~0.20%</li><li>14103: ~0.20%</li><li>14104: ~0.20%</li><li>14105: ~0.20%</li><li>14111: ~0.20%</li><li>14112: ~0.20%</li><li>14114: ~0.20%</li><li>14115: ~0.20%</li><li>14117: ~0.20%</li><li>14124: ~0.20%</li><li>14125: ~0.20%</li><li>14133: ~0.20%</li><li>14135: ~0.20%</li><li>14137: ~0.20%</li><li>14138: ~0.20%</li><li>14139: ~0.20%</li><li>14140: ~0.20%</li><li>14141: ~0.20%</li><li>14142: ~0.20%</li><li>14144: ~0.20%</li><li>14145: ~0.20%</li><li>14149: ~0.20%</li><li>14150: ~0.20%</li><li>14152: ~0.20%</li><li>14154: ~0.20%</li><li>14157: ~0.20%</li><li>14161: ~0.20%</li><li>14163: ~0.20%</li><li>14164: ~0.20%</li><li>14171: ~0.20%</li><li>14174: ~0.20%</li><li>14175: ~0.20%</li><li>14180: ~0.20%</li><li>14181: ~0.20%</li><li>14183: ~0.20%</li><li>14185: ~0.20%</li><li>14186: ~0.20%</li><li>14187: ~0.20%</li><li>14190: ~0.20%</li><li>14191: ~0.20%</li><li>14192: ~0.20%</li><li>14193: ~0.20%</li><li>14195: ~0.20%</li><li>14196: ~0.20%</li><li>14197: ~0.20%</li><li>14198: ~0.20%</li><li>14199: ~0.20%</li><li>14200: ~0.20%</li><li>14202: ~0.20%</li><li>14204: ~0.20%</li><li>14205: ~0.20%</li><li>14206: ~0.20%</li><li>14208: ~0.20%</li><li>14209: ~0.20%</li><li>14215: ~0.20%</li><li>14216: ~0.20%</li><li>14218: ~0.20%</li><li>14219: ~0.20%</li><li>14221: ~0.20%</li><li>14223: ~0.20%</li><li>14225: ~0.20%</li><li>14226: ~0.20%</li><li>14228: ~0.20%</li><li>14229: ~0.20%</li><li>14230: ~0.20%</li><li>14231: ~0.20%</li><li>14232: ~0.20%</li><li>14233: ~0.20%</li><li>14234: ~0.20%</li><li>14235: ~0.20%</li><li>14237: ~0.20%</li><li>14241: ~0.20%</li><li>14242: ~0.20%</li><li>14243: ~0.20%</li><li>14246: ~0.20%</li><li>14248: ~0.20%</li><li>14249: ~0.20%</li><li>14250: ~0.20%</li><li>14252: ~0.20%</li><li>14253: ~0.20%</li><li>14256: ~0.20%</li><li>14258: ~0.20%</li><li>14260: ~0.20%</li><li>14261: ~0.20%</li><li>14262: ~0.20%</li><li>14263: ~0.20%</li><li>14264: ~0.20%</li><li>14268: ~0.20%</li><li>14269: ~0.20%</li><li>14270: ~0.20%</li><li>14271: ~0.20%</li><li>14272: ~0.20%</li><li>14273: ~0.20%</li><li>14279: ~0.20%</li><li>14280: ~0.20%</li><li>14281: ~0.20%</li><li>14283: ~0.20%</li><li>14284: ~0.20%</li><li>14285: ~0.20%</li><li>14287: ~0.20%</li><li>14289: ~0.20%</li><li>14292: ~0.20%</li><li>14293: ~0.20%</li><li>14295: ~0.20%</li><li>14296: ~0.20%</li><li>14297: ~0.20%</li><li>14299: ~0.20%</li><li>14300: ~0.20%</li><li>14303: ~0.20%</li><li>14305: ~0.20%</li><li>14306: ~0.20%</li><li>14307: ~0.20%</li><li>14308: ~0.20%</li><li>14309: ~0.20%</li><li>14310: ~0.20%</li><li>14311: ~0.20%</li><li>14313: ~0.20%</li><li>14316: ~0.20%</li><li>14320: ~0.20%</li><li>14321: ~0.20%</li><li>14323: ~0.20%</li><li>14324: ~0.20%</li><li>14326: ~0.20%</li><li>14327: ~0.20%</li><li>14328: ~0.20%</li><li>14329: ~0.20%</li><li>14330: ~0.20%</li><li>14331: ~0.20%</li><li>14333: ~0.20%</li><li>14334: ~0.20%</li><li>14335: ~0.20%</li><li>14340: ~0.20%</li><li>14345: ~0.20%</li><li>14346: ~0.20%</li><li>14350: ~0.20%</li><li>14352: ~0.20%</li><li>14355: ~0.20%</li><li>14357: ~0.20%</li><li>14358: ~0.20%</li><li>14359: ~0.20%</li><li>14362: ~0.20%</li><li>14365: ~0.20%</li><li>14366: ~0.20%</li><li>14369: ~0.20%</li><li>14372: ~0.20%</li><li>14374: ~0.20%</li><li>14376: ~0.20%</li><li>14377: ~0.20%</li><li>14378: ~0.20%</li><li>14380: ~0.20%</li><li>14381: ~0.20%</li><li>14382: ~0.20%</li><li>14384: ~0.20%</li><li>14387: ~0.20%</li><li>14388: ~0.20%</li><li>14390: ~0.20%</li><li>14391: ~0.20%</li><li>14394: ~0.20%</li><li>14395: ~0.20%</li><li>14399: ~0.20%</li><li>14405: ~0.20%</li><li>14406: ~0.20%</li><li>14407: ~0.20%</li><li>14408: ~0.20%</li><li>14409: ~0.20%</li><li>14413: ~0.20%</li><li>14414: ~0.20%</li><li>14415: ~0.20%</li><li>14416: ~0.20%</li><li>14417: ~0.20%</li><li>14418: ~0.20%</li><li>14420: ~0.20%</li><li>14421: ~0.20%</li><li>14423: ~0.20%</li><li>14424: ~0.20%</li><li>14427: ~0.20%</li><li>14428: ~0.20%</li><li>14433: ~0.20%</li><li>14435: ~0.20%</li><li>14437: ~0.20%</li><li>14439: ~0.20%</li><li>14440: ~0.20%</li><li>14444: ~0.20%</li><li>14445: ~0.20%</li><li>14450: ~0.20%</li><li>14452: ~0.20%</li><li>14453: ~0.20%</li><li>14454: ~0.20%</li><li>14455: ~0.20%</li><li>14457: ~0.20%</li><li>14458: ~0.20%</li><li>14459: ~0.20%</li><li>14461: ~0.20%</li><li>14463: ~0.20%</li><li>14464: ~0.20%</li><li>14466: ~0.20%</li><li>14467: ~0.20%</li><li>14468: ~0.20%</li><li>14469: ~0.20%</li><li>14472: ~0.20%</li><li>14473: ~0.20%</li><li>14475: ~0.20%</li><li>14480: ~0.20%</li><li>14484: ~0.20%</li><li>14488: ~0.20%</li><li>14490: ~0.20%</li><li>14491: ~0.20%</li><li>14492: ~0.20%</li><li>14493: ~0.20%</li><li>14494: ~0.20%</li><li>14499: ~0.20%</li><li>14501: ~0.20%</li><li>14508: ~0.20%</li><li>14509: ~0.20%</li><li>14510: ~0.20%</li><li>14512: ~0.20%</li><li>14514: ~0.20%</li><li>14516: ~0.20%</li><li>14518: ~0.20%</li><li>14519: ~0.20%</li><li>14520: ~0.20%</li><li>14523: ~0.20%</li><li>14524: ~0.20%</li><li>14526: ~0.20%</li><li>14527: ~0.20%</li><li>14529: ~0.20%</li><li>14530: ~0.20%</li><li>14532: ~0.20%</li><li>14536: ~0.20%</li><li>14537: ~0.20%</li><li>14538: ~0.20%</li><li>14539: ~0.20%</li><li>14540: ~0.20%</li><li>14543: ~0.20%</li><li>14544: ~0.20%</li><li>14545: ~0.20%</li><li>14547: ~0.20%</li><li>14553: ~0.20%</li><li>14562: ~0.20%</li><li>14567: ~0.20%</li><li>14569: ~0.20%</li><li>14570: ~0.20%</li><li>14573: ~0.20%</li><li>14576: ~0.20%</li><li>14577: ~0.20%</li><li>14578: ~0.20%</li><li>14579: ~0.20%</li><li>14581: ~0.20%</li><li>14584: ~0.20%</li><li>14588: ~0.20%</li></ul>	<ul><li>min: 5 tokens</li><li>mean: 8.97 tokens</li><li>max: 19 tokens</li></ul>	<ul><li>min: 17 tokens</li><li>mean: 31.88 tokens</li><li>max: 32 tokens</li></ul>	<ul><li>min: 20 tokens</li><li>mean: 31.97 tokens</li><li>max: 32 tokens</li></ul>	<ul><li>min: 20 tokens</li><li>mean: 31.95 tokens</li><li>max: 32 tokens</li></ul>	<ul><li>min: 24 tokens</li><li>mean: 31.98 tokens</li><li>max: 32 tokens</li></ul>	<ul><li>min: 24 tokens</li><li>mean: 31.96 tokens</li><li>max: 32 tokens</li></ul>

• Samples:

  query_id	query	positive	negative_1	negative_2	negative_3	negative_4
  <code>13501</code>	<code>age of cats first heat cycle</code>	<code>Cats can have their first heat cycle between 4-6 months of age and will go in to heat approximately 3 times a year. Each litter typically has between 4 and 6 kittens. Dogs go into heat at about 4-6 months of age and typically have two heat cycles per year. Each litter has, on average, between 4 and 10 puppies. A dog’s heat cycle lasts a total of 30 days. The vagina will swell during the first 10 days. There will be bloody discharge during the next 10 days.</code>	<code>1 Cats can become pregnant during their first heat cycle, and they do not discriminate when it comes to finding an available male — they will mate with their parents or siblings. Cats can go back into heat soon after giving birth, according to Dr. Debra Primovic, DVM.</code>	<code>How Early Can My Cat or Dog Get Pregnant? The practice of early age spay/neuter, which was endorsed by the American Veterinary Medical Association (AVMA) in 2006, generally refers to dogs or cats which are at least two pounds and/ or two months of age at the time they are altered.</code>	<code>1 It does add to pet overpopulation. 2 Cats can become pregnant during their first heat cycle, and they do not discriminate when it comes to finding an available male — they will mate with their parents or siblings. Cats can go back into heat soon after giving birth, according to Dr. Debra Primovic, DVM.</code>	<code>Have your cat spayed to prevent reproduction. You can spay as early as 8 weeks of age, but check with your veterinarian. If a cat is already in heat, some vets may wait until the cycle is over before performing the spay. Cats do not need to have a litter of kittens before they are spayed.</code>
  <code>13501</code>	<code>age of cats first heat cycle</code>	<code>Cats can have their first heat cycle between 4-6 months of age and will go in to heat approximately 3 times a year. Each litter typically has between 4 and 6 kittens. Dogs go into heat at about 4-6 months of age and typically have two heat cycles per year. Each litter has, on average, between 4 and 10 puppies. A dog’s heat cycle lasts a total of 30 days. The vagina will swell during the first 10 days. There will be bloody discharge during the next 10 days.</code>	<code>Also any accidental escape can result in a pregnancy, unlike a dog that must already be “in heat,” to be at risk of pregnancy when she gets loose. The age at which dogs begin to come into estrus varies with size and breed, however many dogs can become pregnant at five months.</code>	<code>How young can my dog or cat get pregnant? This question is especially important for those who share their lives with cats. Cats can actually become pregnant as young as four months of age, having a litter when they are six months old.</code>	<code>No documented problems make early age sterilization, or sterilization before the first heat cycle, ill advised, especially when contrasted with the significant dangers of mammary tumors, pyometra or the tragedy of homelessness.</code>	<code>Have your cat spayed to prevent reproduction. You can spay as early as 8 weeks of age, but check with your veterinarian. If a cat is already in heat, some vets may wait until the cycle is over before performing the spay. Remember: Cats do not need to have a litter of kittens before they are spayed.</code>
  <code>13502</code>	<code>age of cecily tynan</code>	<code>Cecily Tynan was born on the 19th of March 1969, which was a Wednesday. Cecily Tynan will be turning 49 in only 332 days from today. Cecily Tynan is 48 years old. To be more precise (and nerdy), the current age as of right now is 17521 days or (even more geeky) 420504 hours. That's a lot of hours!</code>	<code>Inducted into the Black Filmmakers Hall of Fame in 1977. Pictured on a $3.25 postage stamp issued by the island of Nevis on 1 January 2014. Was misreported as being a decade younger than she actually was until The New York Times found out her real age in 2013.</code>	<code>Mini Bio (1) Cicely Tyson was born in Harlem, New York City, where she was raised by her devoutly religious parents, from the Caribbean island of Nevis. Her mother, Theodosia, was a domestic, and her father, William Tyson, was a carpenter and painter.</code>	<code>0 Affair 2 Married 0 Children. An American television reporter who is currently working for WPVI-TV. She is also widely known as the host of Saturday evening public affairs program named Primetime Weekend. She achieved degree in journalism and politics in 1991 from Washington and Lee University.</code>	<code>Facts of Cecily Tynan. Cecily Tynan has been one of the top level TV reported for long period of time and she has managed to do so with the help of her immense talent and passion for her job. She was born in the year 1969 on 19th of March and this makes her 45 years of age right now.</code>

• Loss: <code>pylate.losses.contrastive.Contrastive</code>

Evaluation Dataset

msmarco-train

• Dataset: msmarco-train at 6853021
• Size: 10,000 evaluation samples
• Columns: <code>query_id</code>, <code>query</code>, <code>positive</code>, <code>negative_1</code>, <code>negative_2</code>, <code>negative_3</code>, and <code>negative_4</code>
• Approximate statistics based on the first 1000 samples:

  	query_id	query	positive	negative_1	negative_2	negative_3	negative_4
  type	int	string	string	string	string	string	string

The model card is truncated. Read the rest on Hugging Face.