Hugging Face's logo

emilwin
/
reranker-ms-marco-sympathy-docs

Text Ranking
sentence-transformers
Safetensors
English
bert
cross-encoder
reranker
Generated from Trainer
dataset_size:18884
loss:BinaryCrossEntropyLoss
loss:CachedMultipleNegativesRankingLoss
Eval Results
Model card Files
xet
Community

Reranker model trained on Sympathy Documentation

This is a Cross Encoder model finetuned from cross-encoder/ms-marco-MiniLM-L6-v2 on the query-doc, anc-pos-neg and anc-pos-neg-2 datasets using the sentence-transformers library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.

Model Details

Model Description

  • Model Type: Cross Encoder
  • Base model: cross-encoder/ms-marco-MiniLM-L6-v2
  • Maximum Sequence Length: 512 tokens
  • Number of Output Labels: 1 label
  • Training Datasets:
    • query-doc
    • anc-pos-neg
    • anc-pos-neg-2
  • Language: en
  • License: apache-2.0

Model Sources

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import CrossEncoder

# Download from the 🤗 Hub
model = CrossEncoder("emilwin/reranker-ms-marco-sympathy-docs")
# Get scores for pairs of texts
pairs = [
    ['What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?', '# API\n\nADAF API\n========\n\nAPI for working with the ADAF type.\n\nImport this module like this:\n\n```\nfrom sympathy.api import adaf\n\n```\n\nThe ADAF structure\n------------------\n\nAn ADAF consists of three parts: meta data, results, and timeseries.\n\nMeta data contains information about the data in the ADAF. Stuff like when,\nwhere and how it was measured or what parameter values were used to generated\nit. A general guideline is that the meta data should be enough to (at least in\ntheory) reproduce the data in the ADAF.\n\nResults and timeseries contain the actual data. Results are always scalar\nwhereas the timeseries can have any number of values.\n\nTimeseries can come in several systems and each system can contain several\nrasters. Each raster in turn has one basis and any number of timeseries. So\nfor example an experiment where some signals are sampled at 100Hz and others\nare sampled only once per second would have (at least) two rasters. A basis\ndoesn’t have to be uniform but can have samples only every now and then.\n\n'],
    ['What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?', '# Node\n\nTable to ADAF\n=============\n\nConvert a Table into an ADAF, placing its content in the specified container.\n\nDocumentation\n-------------\n\nThe target container in the ADAF is specified in the configuration GUI. If the\ntimeseries container is chosen it is necessary to specify the column in the\nTable which will be the time basis signal in the ADAF. You can also specify\nthe name of the system and raster containers.\n\nSee also Working with ADAF for tips about how to use these conversion\nnodes.\n\n'],
    ['What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?', '# Node\n\nSelect columns in ADAF with structure Table\n===========================================\n\nSelect the columns to keep in ADAF using selection table created by ADAF structure to table\n\nDocumentation\n-------------\n\nUse this node if you’re only interested in some of the data in an ADAF\ne.g. for performance reasons.\n\nThe Table/Tables argument should have four columns, which must be named\nType, System, Raster, and Parameter. These columns hold the names of the\ncorresponding fields in the ADAF/ADAFs.\n\nDefinition\n----------\n\n### Input ports\n\n> **selection**\n> :   Type: table\n> ADAF structure selection\n> \n> **data**\n> :   Type: adaf\n> ADAF data matched with selection\n\n### Output ports\n\n> **data**\n> :   Type: adaf\n> ADAF data after selection\n\n### Configuration\n\n> **Remove selected columns** (complement)\n> :   When enabled, the selected columns will be removed. When disabled, the non\\-selected columns will be removed.\n\n### Related nodes\n\n* Select columns in ADAFs with structure Table\n* Select columns in ADAFs with structure Tables\n* ADAF structure to Table\n* Select categories in ADAFs\n\n### Examples\n\nThis node is used in the following example flows:\n\n* SelectColumns.syx\n\n### Implementation\n\n*class* node\\_select\\_adaf\\_columns.SelectColumnsADAFWithTable\\[source]\n\n'],
    ['What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?', '# Node\n\nSelect categories in ADAFs\n==========================\n\nSelect what catgories to exist in the output ADAFs.\n\nDocumentation\n-------------\n\nA selector of categories in ADAFs can be used to drop parts of ADAFs.\nThe main reason to do this is when the ADAFs contain data that is no longer\nneeded further along a workflow. Dropping the unnecessary data can then be\nused as a way to try to optimize the workflow.\n\nDefinition\n----------\n\n### Input ports\n\n> **port1**\n> :   Type: \\[adaf]\n> Input ADAFs\n\n### Output ports\n\n> **port3**\n> :   Type: \\[adaf]\n> ADAFs with selected categories\n\n### Configuration\n\n> **Select meta group** (select\\_meta)\n> :   Select the meta group for inclusion in the output.\n> \n> **Select specific rasters:** (select\\_rasters)\n> :   Select specific rasters for inclusion in the output.\n> \n> **Select result group** (select\\_res)\n> :   Select the result group for inclusion in the output.\n\n### Related nodes\n\n* ADAF to Table\n\n### Examples\n\nThis node is used in the following example flows:\n\n* SelectCategoryInADAFs.syx\n\n### Implementation\n\n*class* node\\_category\\_selector.CategorySelectorMultiple\\[source]\n\n'],
    ['In Sympathy: What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?', '# API\n\nADAF API\n========\n\nAPI for working with the ADAF type.\n\nImport this module like this:\n\n```\nfrom sympathy.api import adaf\n\n```\n\nThe ADAF structure\n------------------\n\nAn ADAF consists of three parts: meta data, results, and timeseries.\n\nMeta data contains information about the data in the ADAF. Stuff like when,\nwhere and how it was measured or what parameter values were used to generated\nit. A general guideline is that the meta data should be enough to (at least in\ntheory) reproduce the data in the ADAF.\n\nResults and timeseries contain the actual data. Results are always scalar\nwhereas the timeseries can have any number of values.\n\nTimeseries can come in several systems and each system can contain several\nrasters. Each raster in turn has one basis and any number of timeseries. So\nfor example an experiment where some signals are sampled at 100Hz and others\nare sampled only once per second would have (at least) two rasters. A basis\ndoesn’t have to be uniform but can have samples only every now and then.\n\n'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility?',
    [
        '# API\n\nADAF API\n========\n\nAPI for working with the ADAF type.\n\nImport this module like this:\n\n```\nfrom sympathy.api import adaf\n\n```\n\nThe ADAF structure\n------------------\n\nAn ADAF consists of three parts: meta data, results, and timeseries.\n\nMeta data contains information about the data in the ADAF. Stuff like when,\nwhere and how it was measured or what parameter values were used to generated\nit. A general guideline is that the meta data should be enough to (at least in\ntheory) reproduce the data in the ADAF.\n\nResults and timeseries contain the actual data. Results are always scalar\nwhereas the timeseries can have any number of values.\n\nTimeseries can come in several systems and each system can contain several\nrasters. Each raster in turn has one basis and any number of timeseries. So\nfor example an experiment where some signals are sampled at 100Hz and others\nare sampled only once per second would have (at least) two rasters. A basis\ndoesn’t have to be uniform but can have samples only every now and then.\n\n',
        '# Node\n\nTable to ADAF\n=============\n\nConvert a Table into an ADAF, placing its content in the specified container.\n\nDocumentation\n-------------\n\nThe target container in the ADAF is specified in the configuration GUI. If the\ntimeseries container is chosen it is necessary to specify the column in the\nTable which will be the time basis signal in the ADAF. You can also specify\nthe name of the system and raster containers.\n\nSee also Working with ADAF for tips about how to use these conversion\nnodes.\n\n',
        '# Node\n\nSelect columns in ADAF with structure Table\n===========================================\n\nSelect the columns to keep in ADAF using selection table created by ADAF structure to table\n\nDocumentation\n-------------\n\nUse this node if you’re only interested in some of the data in an ADAF\ne.g. for performance reasons.\n\nThe Table/Tables argument should have four columns, which must be named\nType, System, Raster, and Parameter. These columns hold the names of the\ncorresponding fields in the ADAF/ADAFs.\n\nDefinition\n----------\n\n### Input ports\n\n> **selection**\n> :   Type: table\n> ADAF structure selection\n> \n> **data**\n> :   Type: adaf\n> ADAF data matched with selection\n\n### Output ports\n\n> **data**\n> :   Type: adaf\n> ADAF data after selection\n\n### Configuration\n\n> **Remove selected columns** (complement)\n> :   When enabled, the selected columns will be removed. When disabled, the non\\-selected columns will be removed.\n\n### Related nodes\n\n* Select columns in ADAFs with structure Table\n* Select columns in ADAFs with structure Tables\n* ADAF structure to Table\n* Select categories in ADAFs\n\n### Examples\n\nThis node is used in the following example flows:\n\n* SelectColumns.syx\n\n### Implementation\n\n*class* node\\_select\\_adaf\\_columns.SelectColumnsADAFWithTable\\[source]\n\n',
        '# Node\n\nSelect categories in ADAFs\n==========================\n\nSelect what catgories to exist in the output ADAFs.\n\nDocumentation\n-------------\n\nA selector of categories in ADAFs can be used to drop parts of ADAFs.\nThe main reason to do this is when the ADAFs contain data that is no longer\nneeded further along a workflow. Dropping the unnecessary data can then be\nused as a way to try to optimize the workflow.\n\nDefinition\n----------\n\n### Input ports\n\n> **port1**\n> :   Type: \\[adaf]\n> Input ADAFs\n\n### Output ports\n\n> **port3**\n> :   Type: \\[adaf]\n> ADAFs with selected categories\n\n### Configuration\n\n> **Select meta group** (select\\_meta)\n> :   Select the meta group for inclusion in the output.\n> \n> **Select specific rasters:** (select\\_rasters)\n> :   Select specific rasters for inclusion in the output.\n> \n> **Select result group** (select\\_res)\n> :   Select the result group for inclusion in the output.\n\n### Related nodes\n\n* ADAF to Table\n\n### Examples\n\nThis node is used in the following example flows:\n\n* SelectCategoryInADAFs.syx\n\n### Implementation\n\n*class* node\\_category\\_selector.CategorySelectorMultiple\\[source]\n\n',
        '# API\n\nADAF API\n========\n\nAPI for working with the ADAF type.\n\nImport this module like this:\n\n```\nfrom sympathy.api import adaf\n\n```\n\nThe ADAF structure\n------------------\n\nAn ADAF consists of three parts: meta data, results, and timeseries.\n\nMeta data contains information about the data in the ADAF. Stuff like when,\nwhere and how it was measured or what parameter values were used to generated\nit. A general guideline is that the meta data should be enough to (at least in\ntheory) reproduce the data in the ADAF.\n\nResults and timeseries contain the actual data. Results are always scalar\nwhereas the timeseries can have any number of values.\n\nTimeseries can come in several systems and each system can contain several\nrasters. Each raster in turn has one basis and any number of timeseries. So\nfor example an experiment where some signals are sampled at 100Hz and others\nare sampled only once per second would have (at least) two rasters. A basis\ndoesn’t have to be uniform but can have samples only every now and then.\n\n',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Evaluation

Metrics

Cross Encoder Reranking

Metric Value
map 0.3233 (+0.1907)
mrr@10 0.3233 (+0.2013)
ndcg@10 0.3488 (+0.1993)

Cross Encoder Reranking

  • Datasets: NanoMSMARCO_R100, NanoNFCorpus_R100 and NanoNQ_R100
  • Evaluated with CrossEncoderRerankingEvaluator with these parameters:
    {
    "at_k": 10,
    "always_rerank_positives": true
}
Metric NanoMSMARCO_R100 NanoNFCorpus_R100 NanoNQ_R100
map 0.5604 (+0.0708) 0.3633 (+0.1023) 0.6359 (+0.2163)
mrr@10 0.5468 (+0.0693) 0.5569 (+0.0570) 0.6529 (+0.2262)
ndcg@10 0.6088 (+0.0683) 0.3953 (+0.0703) 0.6934 (+0.1928)

Cross Encoder Nano BEIR

  • Dataset: NanoBEIR_R100_mean
  • Evaluated with CrossEncoderNanoBEIREvaluator with these parameters:
    {
    "dataset_names": [
        "msmarco",
        "nfcorpus",
        "nq"
    ],
    "rerank_k": 100,
    "at_k": 10,
    "always_rerank_positives": true
}
Metric Value
map 0.5199 (+0.1298)
mrr@10 0.5855 (+0.1175)
ndcg@10 0.5658 (+0.1105)

Training Details

Training Datasets

query-doc

  • Dataset: query-doc
  • Size: 15,230 training samples
  • Columns: query, document, and label
  • Approximate statistics based on the first 1000 samples:
    query document label
    type string string int
    details
    • min: 31 characters
    • mean: 150.61 characters
    • max: 242 characters
    • min: 163 characters
    • mean: 1891.69 characters
    • max: 12851 characters
    • 0: ~74.10%
    • 1: ~25.90%
  • Samples:
    query document label
    What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility? # API

    ADAF API
    ========

    API for working with the ADAF type.

    Import this module like this:

    <br>from sympathy.api import adaf<br><br>

    The ADAF structure
    ------------------

    An ADAF consists of three parts: meta data, results, and timeseries.

    Meta data contains information about the data in the ADAF. Stuff like when,
    where and how it was measured or what parameter values were used to generated
    it. A general guideline is that the meta data should be enough to (at least in
    theory) reproduce the data in the ADAF.

    Results and timeseries contain the actual data. Results are always scalar
    whereas the timeseries can have any number of values.

    Timeseries can come in several systems and each system can contain several
    rasters. Each raster in turn has one basis and any number of timeseries. So
    for example an experiment where some signals are sampled at 100Hz and others
    are sampled only once per second would have (at least) two rasters. A basis
    doesn’t have to be uniform but can have samples on...    		 1
    What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility? # Node

    Table to ADAF
    =============

    Convert a Table into an ADAF, placing its content in the specified container.

    Documentation
    -------------

    The target container in the ADAF is specified in the configuration GUI. If the
    timeseries container is chosen it is necessary to specify the column in the
    Table which will be the time basis signal in the ADAF. You can also specify
    the name of the system and raster containers.

    See also Working with ADAF for tips about how to use these conversion
    nodes.

    		 0
    What are the three primary parts of an ADAF, and how does the meta data differ functionally from the results and timeseries in terms of reproducibility? # Node

    Select columns in ADAF with structure Table
    ===========================================

    Select the columns to keep in ADAF using selection table created by ADAF structure to table

    Documentation
    -------------

    Use this node if you’re only interested in some of the data in an ADAF
    e.g. for performance reasons.

    The Table/Tables argument should have four columns, which must be named
    Type, System, Raster, and Parameter. These columns hold the names of the
    corresponding fields in the ADAF/ADAFs.

    Definition
    ----------

    ### Input ports

    > selection
    > : Type: table
    > ADAF structure selection
    >
    > data
    > : Type: adaf
    > ADAF data matched with selection

    ### Output ports

    > data
    > : Type: adaf
    > ADAF data after selection

    ### Configuration

    > Remove selected columns (complement)
    > : When enabled, the selected columns will be removed. When disabled, the non-selected columns will be removed.

    ### Related nodes

    * Select columns in ADAFs with structure Table
    * Sel...    		 0
  • Loss: BinaryCrossEntropyLoss with these parameters:
    {
    "activation_fn": "torch.nn.modules.linear.Identity",
    "pos_weight": 3
}

anc-pos-neg

  • Dataset: anc-pos-neg
  • Size: 2,435 training samples
  • Columns: anchor, positive, and negative
  • Approximate statistics based on the first 1000 samples:
    anchor positive negative
    type string string string
    details
    • min: 14 characters
    • mean: 158.34 characters
    • max: 317 characters
    • min: 163 characters
    • mean: 2474.63 characters
    • max: 20435 characters
    • min: 214 characters
    • mean: 1745.03 characters
    • max: 20435 characters
  • Samples:
    anchor positive negative
    - Retrieve the time (t) and signal (y) values of "Voltage"? # API ADAF API

    Accessing the data
    ------------------

    The adaf.ADAF object has two members called meta and rescontaining the meta data and results respectively. Both are sympathy.api.adaf.Groupobjects.

    Example of how to use meta (res is completely analogous):
    : <br>>>> from sympathy.api import adaf<br>>>> import numpy as np<br>>>> f = adaf.ADAF()<br>>>> f.meta.create_column(<br>... 'Duration', np.array([3]), {'unit': 'h'})<br>>>> f.meta.create_column(<br>... 'Relative humidity', np.array([63]), {'unit': '%'})<br>>>> print(f.meta['Duration'].value())<br>[3]<br>>>> print(f.meta['Duration'].attr['unit'])<br><br>

    Timeseries can be accessed in two different ways. Either via the membersys or via the member ts. Using sys is generally recommended sincets handles multiple timeseries with the same name across different rasters
    poorly.

    Example of how to use sys:
    : ```
    >>> f.sys.create('Measurement system')
    >>> f.sys['Measurement system'].create('Raster1')
    >>> f.sys['Measurement system']['Raster...    		 # API ADAF API

    Class sympathy.api.adaf.Timeseries
    ----------------------------------

    class sympathy.api.adaf.Timeseries(node, data, name: str)
    : Class representing a timeseries. The values in the timeseries can be
    accessed as a numpy array via the member y. The timeseries is also
    connected to a time basis whose values can be accessed as a numpy array
    via the property t.

    The timeseries can also have any number of attributes. The methodssympathy.api.adaf.Timeseries.unit and sympathy.api.adaf.Timeseries.description retrieve those two attributes. To get
    all attributes use the method sympathy.api.adaf.Timeseries.get_attributes.

    basis() → sympathy.typeutils.adaf.Column
    : Return the timeseries data basis as a sympathy.api.adaf.Column.

    description() → str
    : Return the description attribute or an empty string if it is not set.

    property dtype*: dtype*
    : dtype of timeseries.

    get_attributes() → Dict[str, int \
    How can you add a custom attribute (e.g., {'description': 'Indicates system health'}) to a signal named "Status" in a raster, and how would you later retrieve this attribute? # API ADAF API

    Accessing the data
    ------------------

    The adaf.ADAF object has two members called meta and rescontaining the meta data and results respectively. Both are sympathy.api.adaf.Groupobjects.

    Example of how to use meta (res is completely analogous):
    : <br>>>> from sympathy.api import adaf<br>>>> import numpy as np<br>>>> f = adaf.ADAF()<br>>>> f.meta.create_column(<br>... 'Duration', np.array([3]), {'unit': 'h'})<br>>>> f.meta.create_column(<br>... 'Relative humidity', np.array([63]), {'unit': '%'})<br>>>> print(f.meta['Duration'].value())<br>[3]<br>>>> print(f.meta['Duration'].attr['unit'])<br><br>

    Timeseries can be accessed in two different ways. Either via the membersys or via the member ts. Using sys is generally recommended sincets handles multiple timeseries with the same name across different rasters
    poorly.

    Example of how to use sys:
    : ```
    >>> f.sys.create('Measurement system')
    >>> f.sys['Measurement system'].create('Raster1')
    >>> f.sys['Measurement system']['Raster...    		 # API ADAF API

    Class sympathy.api.adaf.RasterN
    -------------------------------

    class sympathy.api.adaf.RasterN(record, system: str, name: str)
    : Represents a raster with a single time basis and any number of timeseries
    columns.

    property attr*: Attributes*
    : Raster level attributes.

    basis_column() → sympathy.typeutils.adaf.Column
    : Return the time basis for this raster. The returned object is of typesympathy.api.adaf.Column.

    create_basis(data: ndarray, *attributes: Dict[str, int \
    How can you add a custom attribute (e.g., {'description': 'Indicates system health'}) to a signal named "Status" in a raster, and how would you later retrieve this attribute? # API ADAF API

    Accessing the data
    ------------------

    The adaf.ADAF object has two members called meta and rescontaining the meta data and results respectively. Both are sympathy.api.adaf.Groupobjects.

    Example of how to use meta (res is completely analogous):
    : <br>>>> from sympathy.api import adaf<br>>>> import numpy as np<br>>>> f = adaf.ADAF()<br>>>> f.meta.create_column(<br>... 'Duration', np.array([3]), {'unit': 'h'})<br>>>> f.meta.create_column(<br>... 'Relative humidity', np.array([63]), {'unit': '%'})<br>>>> print(f.meta['Duration'].value())<br>[3]<br>>>> print(f.meta['Duration'].attr['unit'])<br><br>

    Timeseries can be accessed in two different ways. Either via the membersys or via the member ts. Using sys is generally recommended sincets handles multiple timeseries with the same name across different rasters
    poorly.

    Example of how to use sys:
    : ```
    >>> f.sys.create('Measurement system')
    >>> f.sys['Measurement system'].create('Raster1')
    >>> f.sys['Measurement system']['Raster...    		 # API ADAF API

    Class sympathy.api.adaf.Timeseries
    ----------------------------------

    class sympathy.api.adaf.Timeseries(node, data, name: str)
    : Class representing a timeseries. The values in the timeseries can be
    accessed as a numpy array via the member y. The timeseries is also
    connected to a time basis whose values can be accessed as a numpy array
    via the property t.

    The timeseries can also have any number of attributes. The methodssympathy.api.adaf.Timeseries.unit and sympathy.api.adaf.Timeseries.description retrieve those two attributes. To get
    all attributes use the method sympathy.api.adaf.Timeseries.get_attributes.

    basis() → sympathy.typeutils.adaf.Column
    : Return the timeseries data basis as a sympathy.api.adaf.Column.

    description() → str
    : Return the description attribute or an empty string if it is not set.

    property dtype*: dtype*
    : dtype of timeseries.

    get_attributes() → Dict[str, int \
  • Loss: CachedMultipleNegativesRankingLoss with these parameters:
    {
    "scale": 10.0,
    "num_negatives": 4,
    "activation_fn": "torch.nn.modules.activation.Sigmoid",
    "mini_batch_size": 32
}

anc-pos-neg-2

  • Dataset: anc-pos-neg-2
  • Size: 1,219 training samples
  • Columns: anchor, positive, and negative
  • Approximate statistics based on the first 1000 samples:
    anchor positive negative
    type string string string
    details
    • min: 0 characters
    • mean: 89.95 characters
    • max: 143 characters
    • min: 496 characters
    • mean: 1890.31 characters
    • max: 7315 characters
    • min: 160 characters
    • mean: 2026.14 characters
    • max: 12851 characters
  • Samples:
    anchor positive negative
    Is it possible to run a Node.js script or environment from within a Python program? # Nodes in python

    Working with nodes
    ------------------

    Nodes store the changes made during configure and when the parameters are
    changed. They produce a list of data elements when executed and expect a list of
    data elements as input, this makes it possible to easily connect the data
    between nodes. Note that the ordering of inputs and outputs is important and
    should match the declaration order in the node definition.

    The code example below demonstrates how to use the result produced by one node as
    input for another.

    <br>random_table = library.node('Random Table')<br>rt_output = random_table.execute()<br><br>item_to_list = library.node('Item to List')<br>itl_output = item_to_list.execute(rt_output)<br><br>assert itl_output[0][0].equal_to(rt_output[0])<br><br>
    The code example below demonstrates how to use the result produced by multiple
    nodes as input for another.

    ```
    random_table0 = library.node('Random Table')
    rt_output0 = random_table.execute()

    random_table1 = library.node('Random Table')
    rt_outpu...

    Nodes
    =====

    A node is defined as a Python class which inherits fromsympathy.api.node.Node. All node definitions should be in files with
    filenames matching node_*.py and be placed in the nodes folder of a node
    library. See Libraries for information about where to put nodes in your
    library. Nodes can be placed in subfolders and multiple nodes can be defined in
    the same file.

    Node definition
    ---------------

    The following class variables make up the definition of a node.

    Note

    The fields name and nodeid are needed to generate the node. If any
    of these two are missing any attempt at creating this node stops immediately
    without any error message. This can be a good way of e.g. creating a
    superclass for multiple node classes.

    name
    : Required.

    The name of the node, is what the user will rely on to identify the node. It
    will show in the library view and in the node’s tooltip. It will also be used
    as the default label of any instance of the node in a flow.

    Try to keep the ...
    Is it possible to run a Node.js script or environment from within a Python program? # Nodes in python

    Working with nodes
    ------------------

    Nodes store the changes made during configure and when the parameters are
    changed. They produce a list of data elements when executed and expect a list of
    data elements as input, this makes it possible to easily connect the data
    between nodes. Note that the ordering of inputs and outputs is important and
    should match the declaration order in the node definition.

    The code example below demonstrates how to use the result produced by one node as
    input for another.

    <br>random_table = library.node('Random Table')<br>rt_output = random_table.execute()<br><br>item_to_list = library.node('Item to List')<br>itl_output = item_to_list.execute(rt_output)<br><br>assert itl_output[0][0].equal_to(rt_output[0])<br><br>
    The code example below demonstrates how to use the result produced by multiple
    nodes as input for another.

    ```
    random_table0 = library.node('Random Table')
    rt_output0 = random_table.execute()

    random_table1 = library.node('Random Table')
    rt_outpu...    		 # Nodes in python

    Reference
    ---------

    exception sympathy.app.interactive.InteractiveNotNodeError[source]

    class sympathy.app.interactive.SyiLibrary(context, library, name_library, paths)[source]
    : A library of nodes that can be configured and executed in Python code.

    Should not be instantiated directly. Instead call sympathy.app.interactive.load_library.

    node(nid, fuzzy_names=True) → sympathy.app.interactive.SyiNode[source]
    : Attempt to find nid in the library.

    Argument nid can be either a node id or a node name. If no matching
    node can be found a KeyError is raised.

    If fuzzy_names is True (the default) and nid doesn’t match any
    node exactly, it is used as a pattern that the node name must match.
    The characters of the pattern must appear in the node name in the same
    order as in the pattern, but must not be of the same case, and may have
    other characters in between them. If multiple nodes match the pattern a
    KeyError is raised.

    nodeids() ...
    Is it possible to run a Node.js script or environment from within a Python program? # Nodes in python

    Working with nodes
    ------------------

    Nodes store the changes made during configure and when the parameters are
    changed. They produce a list of data elements when executed and expect a list of
    data elements as input, this makes it possible to easily connect the data
    between nodes. Note that the ordering of inputs and outputs is important and
    should match the declaration order in the node definition.

    The code example below demonstrates how to use the result produced by one node as
    input for another.

    <br>random_table = library.node('Random Table')<br>rt_output = random_table.execute()<br><br>item_to_list = library.node('Item to List')<br>itl_output = item_to_list.execute(rt_output)<br><br>assert itl_output[0][0].equal_to(rt_output[0])<br><br>
    The code example below demonstrates how to use the result produced by multiple
    nodes as input for another.

    ```
    random_table0 = library.node('Random Table')
    rt_output0 = random_table.execute()

    random_table1 = library.node('Random Table')
    rt_outpu...    		 # API

    Datasource API
    ==============

    API for working with the Datasource type.

    Import this module like this:

    <br>from sympathy.api import datasource<br><br>
    Class datasource.Datasource
    -----------------------------

    class sympathy.api.datasource.Datasource(*filename: str \
  • Loss: CachedMultipleNegativesRankingLoss with these parameters:
    {
    "scale": 10.0,
    "num_negatives": 4,
    "activation_fn": "torch.nn.modules.activation.Sigmoid",
    "mini_batch_size": 32
}

Training Hyperparameters

Non-Default Hyperparameters

  • eval_strategy: epoch
  • per_device_train_batch_size: 64
  • per_device_eval_batch_size: 64
  • learning_rate: 0.0001
  • num_train_epochs: 10
  • warmup_ratio: 0.1
  • dataloader_num_workers: 4
  • load_best_model_at_end: True
  • multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand
  • overwrite_output_dir: False
  • do_predict: False
  • eval_strategy: epoch
  • prediction_loss_only: True
  • per_device_train_batch_size: 64
  • per_device_eval_batch_size: 64
  • per_gpu_train_batch_size: None
  • per_gpu_eval_batch_size: None
  • gradient_accumulation_steps: 1
  • eval_accumulation_steps: None
  • torch_empty_cache_steps: None
  • learning_rate: 0.0001
  • weight_decay: 0.0
  • adam_beta1: 0.9
  • adam_beta2: 0.999
  • adam_epsilon: 1e-08
  • max_grad_norm: 1.0
  • num_train_epochs: 10
  • max_steps: -1
  • lr_scheduler_type: linear
  • lr_scheduler_kwargs: {}
  • warmup_ratio: 0.1
  • warmup_steps: 0
  • log_level: passive
  • log_level_replica: warning
  • log_on_each_node: True
  • logging_nan_inf_filter: True
  • save_safetensors: True
  • save_on_each_node: False
  • save_only_model: False
  • restore_callback_states_from_checkpoint: False
  • no_cuda: False
  • use_cpu: False
  • use_mps_device: False
  • seed: 42
  • data_seed: None
  • jit_mode_eval: False
  • use_ipex: False
  • bf16: False
  • fp16: False
  • fp16_opt_level: O1
  • half_precision_backend: auto
  • bf16_full_eval: False
  • fp16_full_eval: False
  • tf32: None
  • local_rank: 0
  • ddp_backend: None
  • tpu_num_cores: None
  • tpu_metrics_debug: False
  • debug: []
  • dataloader_drop_last: False
  • dataloader_num_workers: 4
  • dataloader_prefetch_factor: None
  • past_index: -1
  • disable_tqdm: False
  • remove_unused_columns: True
  • label_names: None
  • load_best_model_at_end: True
  • ignore_data_skip: False
  • fsdp: []
  • fsdp_min_num_params: 0
  • fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
  • fsdp_transformer_layer_cls_to_wrap: None
  • accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
  • deepspeed: None
  • label_smoothing_factor: 0.0
  • optim: adamw_torch
  • optim_args: None
  • adafactor: False
  • group_by_length: False
  • length_column_name: length
  • ddp_find_unused_parameters: None
  • ddp_bucket_cap_mb: None
  • ddp_broadcast_buffers: False
  • dataloader_pin_memory: True
  • dataloader_persistent_workers: False
  • skip_memory_metrics: True
  • use_legacy_prediction_loop: False
  • push_to_hub: False
  • resume_from_checkpoint: None
  • hub_model_id: None
  • hub_strategy: every_save
  • hub_private_repo: None
  • hub_always_push: False
  • hub_revision: None
  • gradient_checkpointing: False
  • gradient_checkpointing_kwargs: None
  • include_inputs_for_metrics: False
  • include_for_metrics: []
  • eval_do_concat_batches: True
  • fp16_backend: auto
  • push_to_hub_model_id: None
  • push_to_hub_organization: None
  • mp_parameters:
  • auto_find_batch_size: False
  • full_determinism: False
  • torchdynamo: None
  • ray_scope: last
  • ddp_timeout: 1800
  • torch_compile: False
  • torch_compile_backend: None
  • torch_compile_mode: None
  • include_tokens_per_second: False
  • include_num_input_tokens_seen: False
  • neftune_noise_alpha: None
  • optim_target_modules: None
  • batch_eval_metrics: False
  • eval_on_start: False
  • use_liger_kernel: False
  • liger_kernel_config: None
  • eval_use_gather_object: False
  • average_tokens_across_devices: False
  • prompts: None
  • batch_sampler: batch_sampler
  • multi_dataset_batch_sampler: round_robin
  • router_mapping: {}
  • learning_rate_mapping: {}

Training Logs

Click to expand
Epoch Step Training Loss sydoc-tester_ndcg@10 NanoMSMARCO_R100_ndcg@10 NanoNFCorpus_R100_ndcg@10 NanoNQ_R100_ndcg@10 NanoBEIR_R100_mean_ndcg@10
-1 -1 - 0.2780 (+0.1285) 0.6686 (+0.1282) 0.3930 (+0.0680) 0.7599 (+0.2592) 0.6072 (+0.1518)
0.0167 1 1.1888 - - - - -
0.0333 2 1.8501 - - - - -
0.05 3 3.0206 - - - - -
0.0667 4 1.5729 - - - - -
0.0833 5 1.8201 - - - - -
0.1 6 2.7519 - - - - -
0.1167 7 1.7264 - - - - -
0.1333 8 1.9018 - - - - -
0.15 9 2.5682 - - - - -
0.1667 10 2.6998 - - - - -
0.1833 11 2.0299 - - - - -
0.2 12 2.7956 - - - - -
0.2167 13 0.6817 - - - - -
0.2333 14 1.838 - - - - -
0.25 15 2.2811 - - - - -
0.2667 16 1.3663 - - - - -
0.2833 17 2.0837 - - - - -
0.3 18 2.4574 - - - - -
0.3167 19 0.23 - - - - -
0.3333 20 1.8395 - - - - -
0.35 21 2.4167 - - - - -
0.3667 22 0.6286 - - - - -
0.3833 23 1.8573 - - - - -
0.4 24 2.3595 - - - - -
0.4167 25 0.5143 - - - - -
0.4333 26 1.4291 - - - - -
0.45 27 2.0018 - - - - -
0.4667 28 0.1993 - - - - -
0.4833 29 1.7079 - - - - -
0.5 30 1.9053 - - - - -
0.5167 31 0.6029 - - - - -
0.5333 32 1.4611 - - - - -
0.55 33 2.0044 - - - - -
0.5667 34 0.4241 - - - - -
0.5833 35 2.071 - - - - -
0.6 36 2.0503 - - - - -
0.6167 37 1.0458 - - - - -
0.6333 38 1.5994 - - - - -
0.65 39 1.868 - - - - -
0.6667 40 0.5284 - - - - -
0.6833 41 1.3488 - - - - -
0.7 42 1.9041 - - - - -
0.7167 43 0.5827 - - - - -
0.7333 44 1.3666 - - - - -
0.75 45 2.1058 - - - - -
0.7667 46 0.6255 - - - - -
0.7833 47 1.0372 - - - - -
0.8 48 2.2852 - - - - -
0.8167 49 0.5618 - - - - -
0.8333 50 1.1474 - - - - -
0.85 51 2.1265 - - - - -
0.8667 52 0.4827 - - - - -
0.8833 53 1.2651 - - - - -
0.9 54 1.8336 - - - - -
0.9167 55 0.7961 - - - - -
0.9333 56 1.0884 - - - - -
0.95 57 1.6975 - - - - -
0.9667 58 0.5475 - - - - -
0.9833 59 0.8953 - - - - -
1.0 60 1.8382 0.2914 (+0.1420) 0.6658 (+0.1254) 0.4003 (+0.0752) 0.7547 (+0.2540) 0.6069 (+0.1516)
1.0167 61 0.5987 - - - - -
1.0333 62 1.0246 - - - - -
1.05 63 1.6712 - - - - -
1.0667 64 0.4722 - - - - -
1.0833 65 1.1193 - - - - -
1.1 66 1.5013 - - - - -
1.1167 67 0.5394 - - - - -
1.1333 68 1.1887 - - - - -
1.15 69 1.7034 - - - - -
1.1667 70 0.4565 - - - - -
1.1833 71 1.2703 - - - - -
1.2 72 1.753 - - - - -
1.2167 73 0.3727 - - - - -
1.2333 74 0.8781 - - - - -
1.25 75 1.6562 - - - - -
1.2667 76 0.7796 - - - - -
1.2833 77 1.0529 - - - - -
1.3 78 1.5911 - - - - -
1.3167 79 0.3978 - - - - -
1.3333 80 0.8815 - - - - -
1.35 81 1.6555 - - - - -
1.3667 82 0.4231 - - - - -
1.3833 83 0.8421 - - - - -
1.4 84 1.78 - - - - -
1.4167 85 0.4604 - - - - -
1.4333 86 1.4535 - - - - -
1.45 87 1.5948 - - - - -
1.4667 88 1.0813 - - - - -
1.4833 89 0.9153 - - - - -
1.5 90 1.3446 - - - - -
1.5167 91 0.8085 - - - - -
1.5333 92 0.8611 - - - - -
1.55 93 2.0656 - - - - -
1.5667 94 0.8703 - - - - -
1.5833 95 1.0746 - - - - -
1.6 96 1.8937 - - - - -
1.6167 97 0.3555 - - - - -
1.6333 98 0.9181 - - - - -
1.65 99 1.666 - - - - -
1.6667 100 0.5811 - - - - -
1.6833 101 0.8751 - - - - -
1.7 102 1.4337 - - - - -
1.7167 103 0.5711 - - - - -
1.7333 104 0.8895 - - - - -
1.75 105 1.5261 - - - - -
1.7667 106 0.4124 - - - - -
1.7833 107 1.0844 - - - - -
1.8 108 1.3582 - - - - -
1.8167 109 0.6696 - - - - -
1.8333 110 1.014 - - - - -
1.85 111 1.8169 - - - - -
1.8667 112 0.4394 - - - - -
1.8833 113 0.8345 - - - - -
1.9 114 1.3999 - - - - -
1.9167 115 0.1797 - - - - -
1.9333 116 0.8217 - - - - -
1.95 117 1.2372 - - - - -
1.9667 118 0.3477 - - - - -
1.9833 119 0.9426 - - - - -
2.0 120 0.7439 0.3266 (+0.1771) 0.6720 (+0.1315) 0.4090 (+0.0840) 0.7295 (+0.2289) 0.6035 (+0.1482)
2.0167 121 0.5735 - - - - -
2.0333 122 1.0874 - - - - -
2.05 123 1.5375 - - - - -
2.0667 124 0.4699 - - - - -
2.0833 125 0.6828 - - - - -
2.1 126 1.1029 - - - - -
2.1167 127 0.2952 - - - - -
2.1333 128 0.7866 - - - - -
2.15 129 1.1173 - - - - -
2.1667 130 0.4053 - - - - -
2.1833 131 0.8136 - - - - -
2.2 132 1.1145 - - - - -
2.2167 133 0.2084 - - - - -
2.2333 134 0.6429 - - - - -
2.25 135 1.0727 - - - - -
2.2667 136 0.2806 - - - - -
2.2833 137 0.7038 - - - - -
2.3 138 1.3219 - - - - -
2.3167 139 0.3426 - - - - -
2.3333 140 0.939 - - - - -
2.35 141 1.3082 - - - - -
2.3667 142 0.4325 - - - - -
2.3833 143 0.8041 - - - - -
2.4 144 1.2372 - - - - -
2.4167 145 0.3477 - - - - -
2.4333 146 0.6534 - - - - -
2.45 147 0.9268 - - - - -
2.4667 148 0.1559 - - - - -
2.4833 149 0.8769 - - - - -
2.5 150 0.8099 - - - - -
2.5167 151 0.1916 - - - - -
2.5333 152 0.9749 - - - - -
2.55 153 0.8685 - - - - -
2.5667 154 0.4233 - - - - -
2.5833 155 0.7877 - - - - -
2.6 156 1.0647 - - - - -
2.6167 157 0.3441 - - - - -
2.6333 158 0.8019 - - - - -
2.65 159 0.8691 - - - - -
2.6667 160 0.2585 - - - - -
2.6833 161 0.7472 - - - - -
2.7 162 0.8618 - - - - -
2.7167 163 0.2301 - - - - -
2.7333 164 0.6078 - - - - -
2.75 165 0.8942 - - - - -
2.7667 166 0.3613 - - - - -
2.7833 167 0.6139 - - - - -
2.8 168 0.8171 - - - - -
2.8167 169 0.2423 - - - - -
2.8333 170 0.7126 - - - - -
2.85 171 0.8464 - - - - -
2.8667 172 0.2323 - - - - -
2.8833 173 0.5863 - - - - -
2.9 174 0.9001 - - - - -
2.9167 175 0.3677 - - - - -
2.9333 176 0.6953 - - - - -
2.95 177 0.816 - - - - -
2.9667 178 0.1606 - - - - -
2.9833 179 0.4495 - - - - -
3.0 180 0.5979 0.3271 (+0.1777) 0.6738 (+0.1333) 0.4114 (+0.0864) 0.7131 (+0.2125) 0.5994 (+0.1441)
3.0167 181 0.2455 - - - - -
3.0333 182 0.8384 - - - - -
3.05 183 0.7267 - - - - -
3.0667 184 0.8089 - - - - -
3.0833 185 0.5904 - - - - -
3.1 186 0.6173 - - - - -
3.1167 187 0.3746 - - - - -
3.1333 188 0.4729 - - - - -
3.15 189 0.7779 - - - - -
3.1667 190 0.323 - - - - -
3.1833 191 0.5322 - - - - -
3.2 192 0.6053 - - - - -
3.2167 193 0.4589 - - - - -
3.2333 194 0.5053 - - - - -
3.25 195 0.7136 - - - - -
3.2667 196 0.296 - - - - -
3.2833 197 0.631 - - - - -
3.3 198 0.8061 - - - - -
3.3167 199 0.2414 - - - - -
3.3333 200 0.6171 - - - - -
3.35 201 0.5376 - - - - -
3.3667 202 0.5552 - - - - -
3.3833 203 0.6648 - - - - -
3.4 204 0.7012 - - - - -
3.4167 205 0.4025 - - - - -
3.4333 206 0.5783 - - - - -
3.45 207 0.4234 - - - - -
3.4667 208 0.5073 - - - - -
3.4833 209 0.6345 - - - - -
3.5 210 0.6181 - - - - -
3.5167 211 0.2886 - - - - -
3.5333 212 0.4679 - - - - -
3.55 213 0.3889 - - - - -
3.5667 214 0.2376 - - - - -
3.5833 215 0.7177 - - - - -
3.6 216 0.4891 - - - - -
3.6167 217 0.3411 - - - - -
3.6333 218 0.8069 - - - - -
3.65 219 0.8119 - - - - -
3.6667 220 0.4792 - - - - -
3.6833 221 0.8323 - - - - -
3.7 222 0.7516 - - - - -
3.7167 223 0.2906 - - - - -
3.7333 224 0.5762 - - - - -
3.75 225 0.6405 - - - - -
3.7667 226 0.1347 - - - - -
3.7833 227 0.4869 - - - - -
3.8 228 0.5139 - - - - -
3.8167 229 0.2649 - - - - -
3.8333 230 0.7511 - - - - -
3.85 231 0.552 - - - - -
3.8667 232 0.2641 - - - - -
3.8833 233 0.3692 - - - - -
3.9 234 0.6599 - - - - -
3.9167 235 0.9202 - - - - -
3.9333 236 0.6013 - - - - -
3.95 237 0.6525 - - - - -
3.9667 238 0.3979 - - - - -
3.9833 239 0.5321 - - - - -
4.0 240 0.0005 0.3370 (+0.1876) 0.6507 (+0.1103) 0.4011 (+0.0760) 0.6923 (+0.1917) 0.5814 (+0.1260)
4.0167 241 0.1341 - - - - -
4.0333 242 0.5269 - - - - -
4.05 243 0.6917 - - - - -
4.0667 244 0.437 - - - - -
4.0833 245 0.5446 - - - - -
4.1 246 0.5892 - - - - -
4.1167 247 0.2742 - - - - -
4.1333 248 0.5049 - - - - -
4.15 249 0.7015 - - - - -
4.1667 250 0.2648 - - - - -
4.1833 251 0.5977 - - - - -
4.2 252 0.8432 - - - - -
4.2167 253 0.281 - - - - -
4.2333 254 0.5203 - - - - -
4.25 255 0.6649 - - - - -
4.2667 256 0.1843 - - - - -
4.2833 257 0.4616 - - - - -
4.3 258 0.3689 - - - - -
4.3167 259 0.2484 - - - - -
4.3333 260 0.4718 - - - - -
4.35 261 0.5886 - - - - -
4.3667 262 0.1984 - - - - -
4.3833 263 0.6351 - - - - -
4.4 264 0.4616 - - - - -
4.4167 265 0.3106 - - - - -
4.4333 266 0.5568 - - - - -
4.45 267 0.3814 - - - - -
4.4667 268 0.2351 - - - - -
4.4833 269 0.548 - - - - -
4.5 270 0.5559 - - - - -
4.5167 271 0.2272 - - - - -
4.5333 272 0.5367 - - - - -
4.55 273 0.4771 - - - - -
4.5667 274 0.5025 - - - - -
4.5833 275 0.4496 - - - - -
4.6 276 0.3119 - - - - -
4.6167 277 0.1054 - - - - -
4.6333 278 0.5954 - - - - -
4.65 279 0.5023 - - - - -
4.6667 280 0.1567 - - - - -
4.6833 281 0.5903 - - - - -
4.7 282 0.5529 - - - - -
4.7167 283 0.5897 - - - - -
4.7333 284 0.4256 - - - - -
4.75 285 0.3928 - - - - -
4.7667 286 0.2755 - - - - -
4.7833 287 0.5036 - - - - -
4.8 288 0.464 - - - - -
4.8167 289 0.1169 - - - - -
4.8333 290 0.6028 - - - - -
4.85 291 0.2327 - - - - -
4.8667 292 0.6823 - - - - -
4.8833 293 0.5122 - - - - -
4.9 294 0.4079 - - - - -
4.9167 295 0.4138 - - - - -
4.9333 296 0.6886 - - - - -
4.95 297 0.2706 - - - - -
4.9667 298 0.2255 - - - - -
4.9833 299 0.4051 - - - - -
5.0 300 0.4815 0.3403 (+0.1909) 0.6408 (+0.1003) 0.4042 (+0.0791) 0.7126 (+0.2119) 0.5858 (+0.1305)
5.0167 301 0.1022 - - - - -
5.0333 302 0.3965 - - - - -
5.05 303 0.3549 - - - - -
5.0667 304 0.4604 - - - - -
5.0833 305 0.4974 - - - - -
5.1 306 0.5253 - - - - -
5.1167 307 0.1403 - - - - -
5.1333 308 0.554 - - - - -
5.15 309 0.4808 - - - - -
5.1667 310 0.3776 - - - - -
5.1833 311 0.5058 - - - - -
5.2 312 0.5046 - - - - -
5.2167 313 0.0419 - - - - -
5.2333 314 0.5171 - - - - -
5.25 315 0.2989 - - - - -
5.2667 316 0.1901 - - - - -
5.2833 317 0.4728 - - - - -
5.3 318 0.5452 - - - - -
5.3167 319 0.3045 - - - - -
5.3333 320 0.4575 - - - - -
5.35 321 0.4383 - - - - -
5.3667 322 0.367 - - - - -
5.3833 323 0.6289 - - - - -
5.4 324 0.5697 - - - - -
5.4167 325 0.3275 - - - - -
5.4333 326 0.6355 - - - - -
5.45 327 0.2026 - - - - -
5.4667 328 0.3994 - - - - -
5.4833 329 0.6455 - - - - -
5.5 330 0.293 - - - - -
5.5167 331 0.6003 - - - - -
5.5333 332 0.46 - - - - -
5.55 333 0.291 - - - - -
5.5667 334 0.2577 - - - - -
5.5833 335 0.4286 - - - - -
5.6 336 0.5138 - - - - -
5.6167 337 0.4342 - - - - -
5.6333 338 0.7158 - - - - -
5.65 339 0.3723 - - - - -
5.6667 340 0.3464 - - - - -
5.6833 341 0.5797 - - - - -
5.7 342 0.3321 - - - - -
5.7167 343 0.4743 - - - - -
5.7333 344 0.4901 - - - - -
5.75 345 0.4753 - - - - -
5.7667 346 0.4173 - - - - -
5.7833 347 0.291 - - - - -
5.8 348 0.2717 - - - - -
5.8167 349 0.237 - - - - -
5.8333 350 0.5443 - - - - -
5.85 351 0.3157 - - - - -
5.8667 352 0.1993 - - - - -
5.8833 353 0.4968 - - - - -
5.9 354 0.4172 - - - - -
5.9167 355 0.1981 - - - - -
5.9333 356 0.4192 - - - - -
5.95 357 0.3236 - - - - -
5.9667 358 0.3602 - - - - -
5.9833 359 0.4311 - - - - -
6.0 360 0.4171 0.3336 (+0.1842) 0.6444 (+0.1040) 0.4074 (+0.0824) 0.7000 (+0.1994) 0.5840 (+0.1286)
6.0167 361 0.2868 - - - - -
6.0333 362 0.5633 - - - - -
6.05 363 0.4367 - - - - -
6.0667 364 0.4977 - - - - -
6.0833 365 0.6418 - - - - -
6.1 366 0.2547 - - - - -
6.1167 367 0.3511 - - - - -
6.1333 368 0.5132 - - - - -
6.15 369 0.3701 - - - - -
6.1667 370 0.2419 - - - - -
6.1833 371 0.3204 - - - - -
6.2 372 0.3631 - - - - -
6.2167 373 0.3157 - - - - -
6.2333 374 0.5016 - - - - -
6.25 375 0.297 - - - - -
6.2667 376 0.4432 - - - - -
6.2833 377 0.345 - - - - -
6.3 378 0.3711 - - - - -
6.3167 379 0.5635 - - - - -
6.3333 380 0.3848 - - - - -
6.35 381 0.1937 - - - - -
6.3667 382 0.1609 - - - - -
6.3833 383 0.4873 - - - - -
6.4 384 0.3656 - - - - -
6.4167 385 0.0947 - - - - -
6.4333 386 0.3603 - - - - -
6.45 387 0.4195 - - - - -
6.4667 388 0.2649 - - - - -
6.4833 389 0.3971 - - - - -
6.5 390 0.2258 - - - - -
6.5167 391 0.1702 - - - - -
6.5333 392 0.3994 - - - - -
6.55 393 0.3631 - - - - -
6.5667 394 0.1625 - - - - -
6.5833 395 0.375 - - - - -
6.6 396 0.3067 - - - - -
6.6167 397 0.116 - - - - -
6.6333 398 0.3915 - - - - -
6.65 399 0.2512 - - - - -
6.6667 400 0.5099 - - - - -
6.6833 401 0.3622 - - - - -
6.7 402 0.2473 - - - - -
6.7167 403 0.3713 - - - - -
6.7333 404 0.4604 - - - - -
6.75 405 0.4876 - - - - -
6.7667 406 0.0745 - - - - -
6.7833 407 0.4345 - - - - -
6.8 408 0.3579 - - - - -
6.8167 409 0.2141 - - - - -
6.8333 410 0.5035 - - - - -
6.85 411 0.2538 - - - - -
6.8667 412 0.329 - - - - -
6.8833 413 0.338 - - - - -
6.9 414 0.4243 - - - - -
6.9167 415 0.3974 - - - - -
6.9333 416 0.486 - - - - -
6.95 417 0.1896 - - - - -
6.9667 418 0.2265 - - - - -
6.9833 419 0.4796 - - - - -
7.0 420 0.7441 0.3388 (+0.1894) 0.6231 (+0.0827) 0.3935 (+0.0684) 0.6922 (+0.1916) 0.5696 (+0.1142)
7.0167 421 0.0353 - - - - -
7.0333 422 0.5483 - - - - -
7.05 423 0.4845 - - - - -
7.0667 424 0.4536 - - - - -
7.0833 425 0.3831 - - - - -
7.1 426 0.297 - - - - -
7.1167 427 0.1597 - - - - -
7.1333 428 0.5623 - - - - -
7.15 429 0.2996 - - - - -
7.1667 430 0.2648 - - - - -
7.1833 431 0.4407 - - - - -
7.2 432 0.2885 - - - - -
7.2167 433 0.2438 - - - - -
7.2333 434 0.4212 - - - - -
7.25 435 0.3673 - - - - -
7.2667 436 0.3299 - - - - -
7.2833 437 0.402 - - - - -
7.3 438 0.2375 - - - - -
7.3167 439 0.329 - - - - -
7.3333 440 0.5249 - - - - -
7.35 441 0.3656 - - - - -
7.3667 442 0.3228 - - - - -
7.3833 443 0.4069 - - - - -
7.4 444 0.37 - - - - -
7.4167 445 0.2823 - - - - -
7.4333 446 0.4723 - - - - -
7.45 447 0.2711 - - - - -
7.4667 448 0.0393 - - - - -
7.4833 449 0.5585 - - - - -
7.5 450 0.2636 - - - - -
7.5167 451 0.1146 - - - - -
7.5333 452 0.4453 - - - - -
7.55 453 0.3957 - - - - -
7.5667 454 0.5111 - - - - -
7.5833 455 0.3581 - - - - -
7.6 456 0.2948 - - - - -
7.6167 457 0.0755 - - - - -
7.6333 458 0.3249 - - - - -
7.65 459 0.4024 - - - - -
7.6667 460 0.1671 - - - - -
7.6833 461 0.4869 - - - - -
7.7 462 0.1798 - - - - -
7.7167 463 0.3332 - - - - -
7.7333 464 0.4123 - - - - -
7.75 465 0.2245 - - - - -
7.7667 466 0.3406 - - - - -
7.7833 467 0.3521 - - - - -
7.8 468 0.2257 - - - - -
7.8167 469 0.3469 - - - - -
7.8333 470 0.3765 - - - - -
7.85 471 0.2123 - - - - -
7.8667 472 0.4465 - - - - -
7.8833 473 0.3888 - - - - -
7.9 474 0.2459 - - - - -
7.9167 475 0.7323 - - - - -
7.9333 476 0.3495 - - - - -
7.95 477 0.2518 - - - - -
7.9667 478 0.1534 - - - - -
7.9833 479 0.2959 - - - - -
8.0 480 0.07 0.3409 (+0.1915) 0.6194 (+0.0790) 0.3933 (+0.0682) 0.6939 (+0.1933) 0.5689 (+0.1135)
8.0167 481 0.5044 - - - - -
8.0333 482 0.3476 - - - - -
8.05 483 0.254 - - - - -
8.0667 484 0.2724 - - - - -
8.0833 485 0.4188 - - - - -
8.1 486 0.1158 - - - - -
8.1167 487 0.1707 - - - - -
8.1333 488 0.3424 - - - - -
8.15 489 0.3508 - - - - -
8.1667 490 0.1103 - - - - -
8.1833 491 0.4909 - - - - -
8.2 492 0.1988 - - - - -
8.2167 493 0.1158 - - - - -
8.2333 494 0.4486 - - - - -
8.25 495 0.2352 - - - - -
8.2667 496 0.0265 - - - - -
8.2833 497 0.3565 - - - - -
8.3 498 0.4176 - - - - -
8.3167 499 0.1988 - - - - -
8.3333 500 0.5012 - - - - -
8.35 501 0.2685 - - - - -
8.3667 502 0.8838 - - - - -
8.3833 503 0.2845 - - - - -
8.4 504 0.172 - - - - -
8.4167 505 0.1257 - - - - -
8.4333 506 0.4394 - - - - -
8.45 507 0.3462 - - - - -
8.4667 508 0.1913 - - - - -
8.4833 509 0.3712 - - - - -
8.5 510 0.3224 - - - - -
8.5167 511 0.4246 - - - - -
8.5333 512 0.3068 - - - - -
8.55 513 0.3086 - - - - -
8.5667 514 0.5934 - - - - -
8.5833 515 0.3877 - - - - -
8.6 516 0.2269 - - - - -
8.6167 517 0.0762 - - - - -
8.6333 518 0.4297 - - - - -
8.65 519 0.3039 - - - - -
8.6667 520 0.112 - - - - -
8.6833 521 0.5505 - - - - -
8.7 522 0.2615 - - - - -
8.7167 523 0.3927 - - - - -
8.7333 524 0.5144 - - - - -
8.75 525 0.2332 - - - - -
8.7667 526 0.1296 - - - - -
8.7833 527 0.3209 - - - - -
8.8 528 0.2175 - - - - -
8.8167 529 0.1195 - - - - -
8.8333 530 0.5232 - - - - -
8.85 531 0.2233 - - - - -
8.8667 532 0.5163 - - - - -
8.8833 533 0.3405 - - - - -
8.9 534 0.2303 - - - - -
8.9167 535 0.3043 - - - - -
8.9333 536 0.5338 - - - - -
8.95 537 0.1804 - - - - -
8.9667 538 0.5183 - - - - -
8.9833 539 0.2846 - - - - -
9.0 540 0.0954 0.3488 (+0.1993) 0.6088 (+0.0683) 0.3953 (+0.0703) 0.6934 (+0.1928) 0.5658 (+0.1105)
9.0167 541 0.4875 - - - - -
9.0333 542 0.3688 - - - - -
9.05 543 0.3237 - - - - -
9.0667 544 0.0898 - - - - -
9.0833 545 0.2571 - - - - -
9.1 546 0.3119 - - - - -
9.1167 547 0.2481 - - - - -
9.1333 548 0.2996 - - - - -
9.15 549 0.4057 - - - - -
9.1667 550 0.4908 - - - - -
9.1833 551 0.585 - - - - -
9.2 552 0.2549 - - - - -
9.2167 553 0.0969 - - - - -
9.2333 554 0.4962 - - - - -
9.25 555 0.5536 - - - - -
9.2667 556 0.3017 - - - - -
9.2833 557 0.3386 - - - - -
9.3 558 0.1268 - - - - -
9.3167 559 0.2953 - - - - -
9.3333 560 0.4083 - - - - -
9.35 561 0.2145 - - - - -
9.3667 562 0.3205 - - - - -
9.3833 563 0.3553 - - - - -
9.4 564 0.2183 - - - - -
9.4167 565 0.2132 - - - - -
9.4333 566 0.4707 - - - - -
9.45 567 0.3248 - - - - -
9.4667 568 0.635 - - - - -
9.4833 569 0.3263 - - - - -
9.5 570 0.2805 - - - - -
9.5167 571 0.0421 - - - - -
9.5333 572 0.4996 - - - - -
9.55 573 0.2134 - - - - -
9.5667 574 0.0383 - - - - -
9.5833 575 0.5026 - - - - -
9.6 576 0.2033 - - - - -
9.6167 577 0.147 - - - - -
9.6333 578 0.381 - - - - -
9.65 579 0.2251 - - - - -
9.6667 580 0.2874 - - - - -
9.6833 581 0.3673 - - - - -
9.7 582 0.1544 - - - - -
9.7167 583 0.3899 - - - - -
9.7333 584 0.3182 - - - - -
9.75 585 0.3009 - - - - -
9.7667 586 0.0267 - - - - -
9.7833 587 0.3682 - - - - -
9.8 588 0.2009 - - - - -
9.8167 589 0.1356 - - - - -
9.8333 590 0.5001 - - - - -
9.85 591 0.1517 - - - - -
9.8667 592 0.2848 - - - - -
9.8833 593 0.3336 - - - - -
9.9 594 0.2787 - - - - -
9.9167 595 0.3367 - - - - -
9.9333 596 0.3952 - - - - -
9.95 597 0.2262 - - - - -
9.9667 598 0.355 - - - - -
9.9833 599 0.4903 - - - - -
10.0 600 0.0002 0.3435 (+0.1941) 0.6074 (+0.0669) 0.4011 (+0.0760) 0.6901 (+0.1894) 0.5662 (+0.1108)
-1 -1 - 0.3488 (+0.1993) 0.6088 (+0.0683) 0.3953 (+0.0703) 0.6934 (+0.1928) 0.5658 (+0.1105)
  • The bold row denotes the saved checkpoint.

Framework Versions

  • Python: 3.11.2
  • Sentence Transformers: 5.1.0
  • Transformers: 4.55.0
  • PyTorch: 2.7.0+cu126
  • Accelerate: 1.10.0
  • Datasets: 4.0.0
  • Tokenizers: 0.21.4

Citation

BibTeX

Sentence Transformers 

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}
Downloads last month
17
Safetensors
Model size
22.7M params
Tensor type
F32
·
Inference Providers NEW
Text Ranking
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for emilwin/reranker-ms-marco-sympathy-docs

Base model

microsoft/MiniLM-L12-H384-uncased
Quantized
cross-encoder/ms-marco-MiniLM-L12-v2
Quantized
cross-encoder/ms-marco-MiniLM-L6-v2
Finetuned
(25)
this model

Evaluation results