- API docs
- CLI
- Integration guides
- Blog
- How machines learn to understand words: a guide to embeddings in NLP
- Prompt-based learning with Transformers
- Efficient Transformers II: knowledge distillation & fine-tuning
- Efficient Transformers I: attention mechanisms
- Deep hierarchical unsupervised intent modelling: getting value without training data
- Fixing annotating bias with Communications Mining
- Active learning: better ML models in less time
- It's all in the numbers - assessing model performance with metrics
- Why model validation is important
- Comparing Communications Mining and Google AutoML for conversational data intelligence
Command reference
re config
command. These include subcommands to create, read, update, delete contexts, as well as setting a default named context and
utilities.
re config [OPTIONS] <SUBCOMMAND>
re config [OPTIONS] <SUBCOMMAND>
COMMAND | DESCRIPTION |
---|---|
re config add | Create or modify a named context in the configuration file |
re config current | Display the current context in use |
re config delete | Delete the specified context from the reinfer configuration file |
re config get-token | Print the token for a given context or the current one if unspecified |
re config ls | List all the available contexts in the configuration file |
re config use | Set the default context to use when none is specified for a command |
To see the available subcommands or help with a given subcommand
re config help
re config help
Create or modify a named context in the configuration file
re config add [FLAGS] [OPTIONS]
re config add [FLAGS] [OPTIONS]
NAME | DESCRIPTION |
---|---|
-n, --name <name> | The name of the context that will be created or updated |
-e, --endpoint <endpoint> | The Communications Mining cluster endpoint that will be used for this context |
-t, --token <token> | The Communications Mining API token that will be used for this context |
--proxy <proxy> | URL for an HTTP proxy that will be used for all requests if specified |
-k, --accept-invalid-certificates | Whether to accept invalid TLS certificates. You should never have to use this, you running a Communications Mining cluster locally for development |
# Add a new context called `acme`. The cli will interactively ask to provide an API token.
re config add --name acme --endpoint https://acme.reinfer.io
# Using `--name` with a context that already exists will modify it instead.# For example, the next command changes the endpoint of the `acme` context.
re config add --name acme --endpoint https://acme-dev.reinfer.io
# Create a context specifying all information as arguments.
re config add -n staging -e https://localhost:9443 -k -t MYTOKEN
# If options are not provided, the cli will interactively prompt user to enter them
re config add
# Standard output:
* Context name: my-context
I A new context `my-context` will be created
* Enter API token [none]: SECRETTOKEN
W Be careful, API tokens are stored in cleartext in /home/marius/.config/reinfer/contexts.json
* Endpoint [https://reinfer.io/]: https://acme.reinfer.io
I New context `my-context` was created
# Add a new context called `acme`. The cli will interactively ask to provide an API token.
re config add --name acme --endpoint https://acme.reinfer.io
# Using `--name` with a context that already exists will modify it instead.# For example, the next command changes the endpoint of the `acme` context.
re config add --name acme --endpoint https://acme-dev.reinfer.io
# Create a context specifying all information as arguments.
re config add -n staging -e https://localhost:9443 -k -t MYTOKEN
# If options are not provided, the cli will interactively prompt user to enter them
re config add
# Standard output:
* Context name: my-context
I A new context `my-context` will be created
* Enter API token [none]: SECRETTOKEN
W Be careful, API tokens are stored in cleartext in /home/marius/.config/reinfer/contexts.json
* Endpoint [https://reinfer.io/]: https://acme.reinfer.io
I New context `my-context` was created
Print the current context in use to standard output
re config current
re config current
# Print the currently used context to stdout
re config current
# Save the current context to an environment variable
export REINFER_CONTEXT=`re config current`
# Print the currently used context to stdout
re config current
# Save the current context to an environment variable
export REINFER_CONTEXT=`re config current`
Delete the specified context from the reinfer configuration file
re config delete [names]...
re config delete [names]...
# Delete context acme
re config delete acme
# Delete multiple contexts
re config delete acme other-context
# Delete context acme
re config delete acme
# Delete multiple contexts
re config delete acme other-context
Print the API token for a given context or the current one if unspecified to standard output
re config get-token [name]
re config get-token [name]
# Print the API token for the current context to stdout
re config get-token
# Save the API token for the current context in an environment variable
export REINFER_TOKEN=`re config get-token`
# Same, but get the API token for a specific context (acme)
export REINFER_TOKEN=`re config get-token`
# Print the API token for the current context to stdout
re config get-token
# Save the API token for the current context in an environment variable
export REINFER_TOKEN=`re config get-token`
# Same, but get the API token for a specific context (acme)
export REINFER_TOKEN=`re config get-token`
List all the available contexts in the configuration file
re config ls [OPTIONS]
re config ls [OPTIONS]
OPTION | DESCRIPTION |
---|---|
--tokens | Show API tokens (by default tokens are hidden) |
Subcommands for creating new resources
Usage
re create [OPTIONS] <SUBCOMMAND>
re create [OPTIONS] <SUBCOMMAND>
SUBCOMMAND | DESCRIPTION |
---|---|
re create annotations | Create or update annotations |
re create bucket | Create a new bucket |
re create comments | Create or update comments |
re create dataset | Create a new dataset |
re create emails | Create or update emails |
re create project | Create a new project |
re create source | Create a new source |
re create stream-exception | Create a new stream exception |
re create user | Create a new user (note: no welcome email will be sent) |
OPTION | DESCRIPTION |
---|---|
-h , --help | Prints help information |
-V , --version | Prints version information |
To see the available subcommands or help with a given subcommand
re create help
re create help
re get [OPTIONS] <SUBCOMMAND>
re get [OPTIONS] <SUBCOMMAND>
SUBCOMMAND | DESCRIPTION |
---|---|
re get buckets | List the available buckets |
re get comment | Get a single comment from a source |
re get comments | Download all comments from a source |
re get current-user | Get the user associated with the API token in use |
re get datasets | List the available datasets |
re get projects | List the available projects |
re get sources | List the available sources |
re get stream-comments | Fetch comments from a stream |
re get streams | List the available streams for a dataset |
re get users | List the available users |
To see the available subcommands or help with a given subcommand
re get help
re get help
OPTION | DESCRIPTION |
---|---|
-h , --help | Prints help information |
-V , --version | Prints version information |
List the available buckets.
Usage
re get buckets [bucket-name]
re get buckets [bucket-name]
Get a single comment from a source.
Usage
re get comment [OPTIONS] --source <source> <comment-id>
re get comment [OPTIONS] --source <source> <comment-id>
OPTION | DESCRIPTION |
---|---|
--source <source> | (Required) Source name or id |
-f, --file <path> | Path where to write comments as JSON. If not specified, stdout will be used |
<comment-id> | Comment id |
Download many comments from a source
re get comments [FLAGS] [OPTIONS] <source>
re get comments [FLAGS] [OPTIONS] <source>
OPTION | DESCRIPTION |
---|---|
<source> | (Required) Source name or id |
-d, --dataset <dataset> | Dataset name or id |
-f, --file <path> | Path where to write comments as JSON. If not specified, stdout will be used. |
--no-progress | Don't display a progress bar (only applicable when --file is used) |
--from-timestamp <from-timestamp> | Starting timestamp for comments to retrieve (inclusive) |
--predictions <include-predictions> | Save predicted labels and general fields for each comment |
--reviewed-only <reviewed-only> | Only download reviewed comments |
--to-timestamp <to-timestamp> | Ending timestamp for comments to retrieve (inclusive) |
Get the user associated with the API token in use
Usage
re get current-user
re get current-user
List the available datasets
Usage
re get datasets [dataset]
re get datasets [dataset]
OPTION | DESCRIPTION |
---|---|
<dataset> | If specified, only list this dataset (name or id) |
List the available projects
Usage
re get projects [dataset]
re get projects [dataset]
OPTION | DESCRIPTION |
---|---|
<dataset> | If specified, only list this dataset (name or id) |
List the available sources
Usage
re get sources [source]
re get sources [source]
OPTION | DESCRIPTION |
---|---|
<source> | If specified, only list this dataset (name or id) |
Fetch comments from a stream
Usage
re get stream-comments [FLAGS] [OPTIONS] --stream <stream>
re get stream-comments [FLAGS] [OPTIONS] --stream <stream>
OPTION | DESCRIPTION |
---|---|
--stream <stream> | (Required) The full stream name <owner>/<dataset>/<stream> |
--individual-advance | If set, the command will acknowledge each comment in turn, rather than full batches |
--listen <listen> | If set, the command will run forever polling every N seconds and advancing the stream |
--size <size> | The max number of comments to return per batch [default: 16] |
List the available streams for a dataset
Usage
re get streams --dataset <dataset>
re get streams --dataset <dataset>
OPTION | DESCRIPTION |
---|---|
-d, --dataset <dataset> | (Required) The dataset name or id |
Subcommands for deleting resources
re delete <SUBCOMMAND>
re delete <SUBCOMMAND>
SUBCOMMAND | DESCRIPTION |
---|---|
re delete bucket | Delete a bucket |
re delete bulk | Delete all comments in a given time range |
re delete comments | Delete comments by id in a source |
re delete dataset | Delete a dataset |
re delete help | Prints this message or the help of the given subcommand(s) |
re delete project | Delete a project |
re delete source | Delete a source |
re delete user | Delete a user |
FLAGS | DESCRIPTION |
---|---|
-h , --help | Prints help information |
-V , --version | Prints version information |
To see the available subcommands or help with a given subcommand
re delete help