API Types
scope3ai.api.types
Classes:
-
CountryCode– -
DataType– -
DebugInfo– -
Details– -
Error– -
Family–Core AI model families from various organizations:
-
GPU– -
GPUResponse– -
GridMix– -
Image– -
ImpactBigQueryError– -
ImpactBigQueryRequest– -
ImpactBigQueryResponse– -
ImpactMetrics– -
ImpactRequest– -
ImpactResponse– -
ImpactRow– -
Model– -
ModelResponse– -
ModeledRow– -
Node– -
NodeResponse– -
PredictionStep– -
RegionCode– -
Scope3AIContext–Context helper for scope3ai.
-
StatusResponse– -
Task–Common types of AI/ML models and their primary functions:
CountryCode
Bases: RootModel[str]
Attributes:
-
root(Annotated[str, Field(description='Two-letter country code as defined by ISO 3166-1 alpha-2', examples=['US'], max_length=2, min_length=2, pattern='^[A-Z]{2}$')]) –
DataType
DebugInfo
Bases: BaseModel
Attributes:
-
grid_mix(Optional[GridMix]) – -
hardware_node(Optional[Node]) – -
model(Optional[Model]) – -
model_config– -
steps(Annotated[Optional[List[PredictionStep]], Field(max_length=100)]) –
steps
class-attribute
instance-attribute
steps: Annotated[Optional[List[PredictionStep]], Field(max_length=100)] = None
Details
Error
Bases: BaseModel
Attributes:
-
code(Optional[str]) – -
details(Optional[Details]) – -
message(str) – -
model_config–
Family
Bases: str, Enum
Core AI model families from various organizations: - Commercial models from major AI companies - Open source model families - Research/academic model families - Models may appear in multiple categories if they have both commercial and open source variants
Attributes:
-
bert– -
bloom– -
claude– -
command– -
dall_e– -
dolly– -
embed– -
falcon– -
galactica– -
gemini– -
gpt– -
gpt_j– -
gpt_neo– -
llama– -
mistral– -
mixtral– -
mpt– -
opt– -
palm– -
phi– -
pythia– -
roberta– -
stable_diffusion– -
stable_lm– -
t5– -
whisper–
GPU
Bases: BaseModel
Attributes:
-
created_at(datetime) – -
created_by(Annotated[str, Field(description='ID of the user who created the node (admin or owner only)')]) – -
customer_id(Annotated[int, Field(description='ID of the customer who owns this node (visible to admins only)')]) – -
embodied_emissions_kgco2e(Annotated[float, Field(examples=[282.1])]) – -
embodied_water_mlh2o(Annotated[float, Field(examples=[181.1])]) – -
id(Annotated[str, Field(examples=['a100_40gb'])]) – -
max_power_w(Annotated[float, Field(examples=[700])]) – -
model_config– -
name(Annotated[str, Field(examples=['NVIDIA A100 40GB'])]) – -
ols_coefficient_gpu_count(Annotated[float, Field(examples=[11.4])]) – -
ols_intercept(Annotated[float, Field(examples=[11.4])]) – -
performance_ratio_to_h200(Annotated[float, Field(examples=[1.5])]) – -
updated_at(datetime) –
created_by
instance-attribute
created_by: Annotated[str, Field(description='ID of the user who created the node (admin or owner only)')]
customer_id
instance-attribute
customer_id: Annotated[int, Field(description='ID of the customer who owns this node (visible to admins only)')]
embodied_emissions_kgco2e
instance-attribute
embodied_water_mlh2o
instance-attribute
ols_coefficient_gpu_count
instance-attribute
performance_ratio_to_h200
instance-attribute
GPUResponse
GridMix
Bases: BaseModel
Attributes:
-
country(CountryCode) – -
gco2e_per_kwh(Annotated[float, Field(examples=[475], ge=0.0, le=2000.0)]) – -
model_config– -
region(Optional[RegionCode]) –
gco2e_per_kwh
instance-attribute
Image
Bases: RootModel[str]
Attributes:
-
root(Annotated[str, Field(examples=['1024x1024'], pattern='^(\\d{1,4})x(\\d{1,4})$')]) –
ImpactBigQueryError
Bases: BaseModel
Attributes:
-
errorMessage(Annotated[str, Field(description='Error message for BigQuery', examples=["Invalid request format: missing required field 'calls'"])]) –
ImpactBigQueryRequest
Bases: BaseModel
Attributes:
-
caller(Annotated[str, Field(description='Full resource name of the BigQuery job', examples=['//bigquery.googleapis.com/projects/myproject/jobs/myproject:US.bquxjob_5b4c112c_17961fafeaf'])]) – -
calls(Annotated[List[Call], Field(max_length=1000)]) – -
requestId(Annotated[str, Field(description='Unique identifier for the request', examples=['124ab1c'])]) – -
sessionUser(Annotated[str, Field(description='Email of the user executing the BigQuery query', examples=['user@company.com'])]) – -
userDefinedContext(Annotated[Optional[Dict[str, Any]], Field(description='User-defined context from BigQuery')]) –
caller
instance-attribute
caller: Annotated[str, Field(description='Full resource name of the BigQuery job', examples=['//bigquery.googleapis.com/projects/myproject/jobs/myproject:US.bquxjob_5b4c112c_17961fafeaf'])]
requestId
instance-attribute
requestId: Annotated[str, Field(description='Unique identifier for the request', examples=['124ab1c'])]
ImpactBigQueryResponse
Bases: BaseModel
Attributes:
-
errorMessage(Optional[str]) – -
replies(Annotated[List[str], Field(description='Array of impact metric results', max_length=1000, min_length=0)]) –
ImpactMetrics
Bases: BaseModel
Attributes:
-
embodied_emissions_gco2e(Annotated[float, Field(examples=[0.81])]) – -
embodied_water_ml(Annotated[float, Field(examples=[1.32])]) – -
model_config– -
usage_emissions_gco2e(Annotated[float, Field(examples=[0.81])]) – -
usage_energy_wh(Annotated[float, Field(examples=[0.13])]) – -
usage_water_ml(Annotated[float, Field(examples=[1.32])]) –
embodied_emissions_gco2e
instance-attribute
usage_emissions_gco2e
instance-attribute
ImpactRequest
Bases: BaseModel
Attributes:
-
model_config– -
rows(Annotated[List[ImpactRow], Field(max_length=1000)]) –
ImpactResponse
Bases: BaseModel
Attributes:
-
has_errors(Annotated[bool, Field(examples=[False])]) – -
model_config– -
rows(Annotated[List[ModeledRow], Field(max_length=1000)]) – -
total_energy_wh(Annotated[Optional[float], Field(examples=[0.13])]) – -
total_gco2e(Annotated[Optional[float], Field(examples=[0.81])]) – -
total_mlh2o(Annotated[Optional[float], Field(examples=[1.32])]) –
total_energy_wh
class-attribute
instance-attribute
total_gco2e
class-attribute
instance-attribute
ImpactRow
Bases: BaseModel
Attributes:
-
application_id(Annotated[Optional[str], Field(description='The application to attribute this call to')]) – -
client_id(Annotated[Optional[str], Field(description='The client to attribute this call to')]) – -
cloud_id(Annotated[Optional[str], Field(description='The ID of the cloud', examples=['aws'])]) – -
cloud_instance_id(Annotated[Optional[str], Field(description='The instance type in the cloud', examples=['p4d.24xlarge'])]) – -
cloud_region(Annotated[Optional[str], Field(description='The region of cloud hosting', examples=['us-central1'])]) – -
country(Optional[CountryCode]) – -
currency(Annotated[Optional[str], Field(description='The currency for the cost data', examples=['USD'])]) – -
environment(Annotated[Optional[str], Field(description='Environment (prod/production indicates production)', examples=['staging'])]) – -
input_audio_seconds(Annotated[Optional[float], Field(description='the duration of audio input in seconds', examples=[60], ge=0.0, le=100000.0)]) – -
input_images(Annotated[Optional[List[Image]], Field(max_length=100)]) – -
input_steps(Annotated[Optional[int], Field(description='the number of steps to use in the model', examples=[50], ge=1, le=10000)]) – -
input_tokens(Annotated[Optional[int], Field(description='the number of input (or prompt) tokens', examples=[128], ge=0, le=100000000)]) – -
integration_source(Annotated[Optional[str], Field(description='The integration used to source the data', examples=['litellm'])]) – -
managed_service_id(Annotated[Optional[str], Field(description='The ID of a managed service provider', examples=['aws-bedrock'])]) – -
model_config– -
model_family(Annotated[Optional[str], Field(description='The family of the model', examples=['llama'])]) – -
model_hugging_face_path(Annotated[Optional[str], Field(description='The Hugging Face path of the model', examples=['meta/llama31_8b'])]) – -
model_id(Annotated[Optional[str], Field(description='The ID of the model requested', examples=['llama_31_8b'])]) – -
model_name(Annotated[Optional[str], Field(description='The name of the model', examples=['LLaMa v3.1 8B'])]) – -
model_used_id(Annotated[Optional[str], Field(description='The ID of the model that did the inference', examples=['llama_31_8b_0125'])]) – -
node_id(Annotated[Optional[str], Field(description='The ID of a custom or global node', examples=['aws:p4d.24xlarge'])]) – -
output_audio_seconds(Annotated[Optional[float], Field(description='the duration of audio output in seconds', examples=[60], ge=0.0, le=100000.0)]) – -
output_audio_tokens(Annotated[Optional[int], Field(description='the number of audio tokens in the output', examples=[2300], ge=0, le=100000000)]) – -
output_images(Annotated[Optional[List[Image]], Field(description='a list of output image sizes', max_length=100)]) – -
output_tokens(Annotated[Optional[int], Field(description='the number of output (or completion) tokens', examples=[128], ge=0, le=100000000)]) – -
output_video_frames(Annotated[Optional[int], Field(description='the number of video frames (frame rate x duration)', examples=[60], ge=0, le=100000000)]) – -
output_video_resolution(Annotated[Optional[int], Field(description='the resolution of the video in number of lines (for instance, 1080 for 1080p)', examples=[1080])]) – -
processing_duration_ms(Annotated[Optional[float], Field(description='The time taken in processing the request (as measured at execution)', examples=[238])]) – -
project_id(Annotated[Optional[str], Field(description='The project to attribute this call to')]) – -
region(Optional[RegionCode]) – -
request_cost(Annotated[Optional[float], Field(description='The cost to execute this request', examples=[0.18])]) – -
request_duration_ms(Annotated[Optional[float], Field(description='The time the request took (as measured by client or proxy)', examples=[283])]) – -
request_id(Annotated[Optional[str], Field(description='The unique identifier of this request')]) – -
session_id(Annotated[Optional[str], Field(description='The ID of the session (multiple requests)')]) – -
task(Annotated[Optional[Task], Field(examples=['text-generation'])]) – -
trace_id(Annotated[Optional[str], Field(description='The trace ID of the request (multiple requests in one task)')]) – -
utc_datetime(Annotated[Optional[datetime], Field(description='The start time of the request in UTC', examples=['2022-01-01T00:00:00Z'])]) –
application_id
class-attribute
instance-attribute
application_id: Annotated[Optional[str], Field(description='The application to attribute this call to')] = None
client_id
class-attribute
instance-attribute
client_id: Annotated[Optional[str], Field(description='The client to attribute this call to')] = None
cloud_id
class-attribute
instance-attribute
cloud_id: Annotated[Optional[str], Field(description='The ID of the cloud', examples=['aws'])] = None
cloud_instance_id
class-attribute
instance-attribute
cloud_instance_id: Annotated[Optional[str], Field(description='The instance type in the cloud', examples=['p4d.24xlarge'])] = None
cloud_region
class-attribute
instance-attribute
cloud_region: Annotated[Optional[str], Field(description='The region of cloud hosting', examples=['us-central1'])] = None
currency
class-attribute
instance-attribute
currency: Annotated[Optional[str], Field(description='The currency for the cost data', examples=['USD'])] = None
environment
class-attribute
instance-attribute
environment: Annotated[Optional[str], Field(description='Environment (prod/production indicates production)', examples=['staging'])] = None
input_audio_seconds
class-attribute
instance-attribute
input_audio_seconds: Annotated[Optional[float], Field(description='the duration of audio input in seconds', examples=[60], ge=0.0, le=100000.0)] = None
input_images
class-attribute
instance-attribute
input_images: Annotated[Optional[List[Image]], Field(max_length=100)] = None
input_steps
class-attribute
instance-attribute
input_steps: Annotated[Optional[int], Field(description='the number of steps to use in the model', examples=[50], ge=1, le=10000)] = None
input_tokens
class-attribute
instance-attribute
input_tokens: Annotated[Optional[int], Field(description='the number of input (or prompt) tokens', examples=[128], ge=0, le=100000000)] = None
integration_source
class-attribute
instance-attribute
integration_source: Annotated[Optional[str], Field(description='The integration used to source the data', examples=['litellm'])] = None
managed_service_id
class-attribute
instance-attribute
managed_service_id: Annotated[Optional[str], Field(description='The ID of a managed service provider', examples=['aws-bedrock'])] = None
model_family
class-attribute
instance-attribute
model_family: Annotated[Optional[str], Field(description='The family of the model', examples=['llama'])] = None
model_hugging_face_path
class-attribute
instance-attribute
model_hugging_face_path: Annotated[Optional[str], Field(description='The Hugging Face path of the model', examples=['meta/llama31_8b'])] = None
model_id
class-attribute
instance-attribute
model_id: Annotated[Optional[str], Field(description='The ID of the model requested', examples=['llama_31_8b'])] = None
model_name
class-attribute
instance-attribute
model_name: Annotated[Optional[str], Field(description='The name of the model', examples=['LLaMa v3.1 8B'])] = None
model_used_id
class-attribute
instance-attribute
model_used_id: Annotated[Optional[str], Field(description='The ID of the model that did the inference', examples=['llama_31_8b_0125'])] = None
node_id
class-attribute
instance-attribute
node_id: Annotated[Optional[str], Field(description='The ID of a custom or global node', examples=['aws:p4d.24xlarge'])] = None
output_audio_seconds
class-attribute
instance-attribute
output_audio_seconds: Annotated[Optional[float], Field(description='the duration of audio output in seconds', examples=[60], ge=0.0, le=100000.0)] = None
output_audio_tokens
class-attribute
instance-attribute
output_audio_tokens: Annotated[Optional[int], Field(description='the number of audio tokens in the output', examples=[2300], ge=0, le=100000000)] = None
output_images
class-attribute
instance-attribute
output_images: Annotated[Optional[List[Image]], Field(description='a list of output image sizes', max_length=100)] = None
output_tokens
class-attribute
instance-attribute
output_tokens: Annotated[Optional[int], Field(description='the number of output (or completion) tokens', examples=[128], ge=0, le=100000000)] = None
output_video_frames
class-attribute
instance-attribute
output_video_frames: Annotated[Optional[int], Field(description='the number of video frames (frame rate x duration)', examples=[60], ge=0, le=100000000)] = None
output_video_resolution
class-attribute
instance-attribute
output_video_resolution: Annotated[Optional[int], Field(description='the resolution of the video in number of lines (for instance, 1080 for 1080p)', examples=[1080])] = None
processing_duration_ms
class-attribute
instance-attribute
processing_duration_ms: Annotated[Optional[float], Field(description='The time taken in processing the request (as measured at execution)', examples=[238])] = None
project_id
class-attribute
instance-attribute
project_id: Annotated[Optional[str], Field(description='The project to attribute this call to')] = None
request_cost
class-attribute
instance-attribute
request_cost: Annotated[Optional[float], Field(description='The cost to execute this request', examples=[0.18])] = None
request_duration_ms
class-attribute
instance-attribute
request_duration_ms: Annotated[Optional[float], Field(description='The time the request took (as measured by client or proxy)', examples=[283])] = None
request_id
class-attribute
instance-attribute
request_id: Annotated[Optional[str], Field(description='The unique identifier of this request')] = None
session_id
class-attribute
instance-attribute
session_id: Annotated[Optional[str], Field(description='The ID of the session (multiple requests)')] = None
task
class-attribute
instance-attribute
task: Annotated[Optional[Task], Field(examples=['text-generation'])] = None
Model
Bases: BaseModel
Attributes:
-
aliases(Annotated[List[str], Field(description='List of aliases for this model; must be globally-unique with id', examples=[['claude-latest', 'claude-3-sonnet-current']], max_length=100)]) – -
benchmark_model_id(Annotated[Optional[str], Field(examples=['GPTJ-6B'])]) – -
created_at(datetime) – -
created_by(Annotated[str, Field(description='ID of the user who created the node (admin or owner only)')]) – -
customer_id(Annotated[int, Field(description='ID of the customer who owns this node (visible to admins only)')]) – -
datatype(Annotated[Optional[DataType], Field(examples=['fp8'])]) – -
estimated_requests_per_day(Annotated[Optional[float], Field(examples=[1013.1])]) – -
estimated_use_life_days(Annotated[Optional[float], Field(examples=[1013.1])]) – -
family(Annotated[Optional[str], Field(examples=['gpt'])]) – -
fine_tuned_from_model_id(Annotated[Optional[str], Field(examples=['llama_31_8b'])]) – -
hugging_face_path(Annotated[Optional[str], Field(examples=['EleutherAI/gpt-neo-2.7B'])]) – -
id(Annotated[str, Field(examples=['gpt-4-turbo'])]) – -
model_config– -
name(Annotated[Optional[str], Field(examples=['GPT-4 Turbo'])]) – -
number_of_experts(Annotated[Optional[int], Field(examples=[7])]) – -
params_per_expert_billions(Annotated[Optional[float], Field(examples=[8])]) – -
task(Annotated[Optional[Task], Field(examples=['text-generation'])]) – -
tensor_parallelism(Annotated[Optional[int], Field(examples=[1])]) – -
total_params_billions(Annotated[Optional[float], Field(examples=[175])]) – -
training_embodied_emissions_kgco2e(Annotated[Optional[float], Field(examples=[11013.1])]) – -
training_embodied_water_l(Annotated[Optional[float], Field(examples=[11013.1])]) – -
training_usage_emissions_kgco2e(Annotated[Optional[float], Field(examples=[1013.1])]) – -
training_usage_energy_kwh(Annotated[Optional[float], Field(examples=[1013.1])]) – -
training_usage_water_l(Annotated[Optional[float], Field(examples=[1013.1])]) – -
updated_at(datetime) –
aliases
instance-attribute
aliases: Annotated[List[str], Field(description='List of aliases for this model; must be globally-unique with id', examples=[['claude-latest', 'claude-3-sonnet-current']], max_length=100)]
benchmark_model_id
class-attribute
instance-attribute
created_by
instance-attribute
created_by: Annotated[str, Field(description='ID of the user who created the node (admin or owner only)')]
customer_id
instance-attribute
customer_id: Annotated[int, Field(description='ID of the customer who owns this node (visible to admins only)')]
datatype
class-attribute
instance-attribute
datatype: Annotated[Optional[DataType], Field(examples=['fp8'])] = None
estimated_requests_per_day
class-attribute
instance-attribute
estimated_use_life_days
class-attribute
instance-attribute
family
class-attribute
instance-attribute
fine_tuned_from_model_id
class-attribute
instance-attribute
hugging_face_path
class-attribute
instance-attribute
name
class-attribute
instance-attribute
number_of_experts
class-attribute
instance-attribute
params_per_expert_billions
class-attribute
instance-attribute
task
class-attribute
instance-attribute
task: Annotated[Optional[Task], Field(examples=['text-generation'])] = None
tensor_parallelism
class-attribute
instance-attribute
total_params_billions
class-attribute
instance-attribute
training_embodied_emissions_kgco2e
class-attribute
instance-attribute
training_embodied_water_l
class-attribute
instance-attribute
training_usage_emissions_kgco2e
class-attribute
instance-attribute
training_usage_energy_kwh
class-attribute
instance-attribute
training_usage_water_l
class-attribute
instance-attribute
ModelResponse
Bases: BaseModel
Attributes:
-
model_config– -
models(Annotated[List[Model], Field(max_length=100)]) –
ModeledRow
Bases: BaseModel
Attributes:
-
debug(Optional[DebugInfo]) – -
error(Optional[Error]) – -
fine_tuning_impact(Optional[ImpactMetrics]) – -
inference_impact(Optional[ImpactMetrics]) – -
model_config– -
total_impact(ImpactMetrics) – -
training_impact(Optional[ImpactMetrics]) –
fine_tuning_impact
class-attribute
instance-attribute
fine_tuning_impact: Optional[ImpactMetrics] = None
inference_impact
class-attribute
instance-attribute
inference_impact: Optional[ImpactMetrics] = None
Node
Bases: NodeCreateRequest
Attributes:
-
average_utilization_rate(Annotated[Optional[float], Field(examples=[0.8], ge=0.0, le=1.0)]) – -
cloud_id(Annotated[Optional[str], Field(examples=['aws'])]) – -
cloud_instance_id(Annotated[Optional[str], Field(examples=['a2-highgpu-1g'])]) – -
cpu_count(Annotated[int, Field(examples=[2], ge=1, le=10000)]) – -
created_at(datetime) – -
created_by(Annotated[Optional[str], Field(description='ID of the user who created the node (admin or owner only)')]) – -
customer_id(Annotated[Optional[int], Field(description='ID of the customer who owns this node (visible to admins only)')]) – -
embodied_emissions_kgco2e_ex_gpu(Annotated[Optional[float], Field(examples=[2500], ge=0.0, le=100000.0)]) – -
embodied_water_l_ex_gpu(Annotated[Optional[float], Field(examples=[2500], ge=0.0, le=100000.0)]) – -
gpu_count(Annotated[int, Field(examples=[8], ge=0, le=10000)]) – -
gpu_id(Annotated[str, Field(examples=['a100_40gb'])]) – -
id(Annotated[str, Field(examples=['my-custom-node-1'], max_length=64, min_length=3, pattern='^[a-z0-9-]+$')]) – -
idle_power_w_ex_gpu(Annotated[Optional[float], Field(examples=[100], ge=0.0, le=10000.0)]) – -
managed_service_id(Annotated[Optional[str], Field(examples=['aws-bedrock'])]) – -
updated_at(datetime) – -
use_life_years(Annotated[Optional[float], Field(examples=[5], ge=1.0, le=30.0)]) –
average_utilization_rate
class-attribute
instance-attribute
cloud_id
class-attribute
instance-attribute
cloud_instance_id
class-attribute
instance-attribute
created_by
class-attribute
instance-attribute
created_by: Annotated[Optional[str], Field(description='ID of the user who created the node (admin or owner only)')] = None
customer_id
class-attribute
instance-attribute
customer_id: Annotated[Optional[int], Field(description='ID of the customer who owns this node (visible to admins only)')] = None
embodied_emissions_kgco2e_ex_gpu
class-attribute
instance-attribute
embodied_emissions_kgco2e_ex_gpu: Annotated[Optional[float], Field(examples=[2500], ge=0.0, le=100000.0)] = None
embodied_water_l_ex_gpu
class-attribute
instance-attribute
embodied_water_l_ex_gpu: Annotated[Optional[float], Field(examples=[2500], ge=0.0, le=100000.0)] = None
id
instance-attribute
id: Annotated[str, Field(examples=['my-custom-node-1'], max_length=64, min_length=3, pattern='^[a-z0-9-]+$')]
idle_power_w_ex_gpu
class-attribute
instance-attribute
managed_service_id
class-attribute
instance-attribute
NodeResponse
PredictionStep
Bases: BaseModel
Attributes:
-
description(str) – -
duration_ms(float) – -
inferences(int) – -
model_config–
RegionCode
Bases: RootModel[str]
Attributes:
-
root(Annotated[str, Field(description='Two-letter region code as defined by ISO 3166-1 alpha-2', examples=['NY'], max_length=2, min_length=2, pattern='^[A-Z]{2}$')]) –
Scope3AIContext
Bases: BaseModel
Context helper for scope3ai.
Context can be retrieve from object with scope3.context, i.e. response.scope3ai. This object contains the impact request and response data.
Methods:
Attributes:
-
impact(Optional[ModeledRow]) – -
request(Optional[ImpactRow]) –
impact
class-attribute
instance-attribute
impact: Optional[ModeledRow] = Field(None, description='The impact response. Use `wait_impact` to wait for the response, or configure `scope3.sync_mode` to True')
request
class-attribute
instance-attribute
request: Optional[ImpactRow] = Field(None, description='The impact request information. Contains `trace_id` and `record_id`')
StatusResponse
Task
Bases: str, Enum
Common types of AI/ML models and their primary functions: - Text-based models for natural language processing - Vision models for image analysis and generation - Audio models for speech and sound processing - Multimodal models that combine different types of inputs/outputs - Specialized models for specific use cases
Attributes:
-
anomaly_detection– -
audio_classification– -
chat– -
clustering– -
face_detection– -
facial_recognition– -
image_classification– -
image_generation– -
image_segmentation– -
image_to_text– -
multimodal_embedding– -
multimodal_generation– -
music_generation– -
named_entity_recognition– -
object_detection– -
question_answering– -
recommendation_system– -
reinforcement_learning– -
sentiment_analysis– -
speaker_identification– -
speech_to_text– -
style_transfer– -
summarization– -
text_classification– -
text_embedding– -
text_generation– -
text_to_image– -
text_to_speech– -
time_series_forecasting– -
translation– -
visual_question_answering–