- Machine State Classification
- Activity Detection
UserComponentParameters
Hide properties
Hide properties
MachineStateJobUserConfig
required
Per-worker machine state classification job configuration.
Hide properties
Hide properties
PositiveInt
default:"8"
Inference batch size. Must be an integer
1 or higher.ClassifierConfig
Optional K-Nearest Neighbor (KNN) classifier over embeddings configuration.
Show properties
Show properties
string
default:"euclidean"
The type of metric to use. One of
cityblock, cosine, euclidean, haversine, l1, l2, manhattan, nan_euclidean.PositiveInt
default:"5"
KNN count. Must be at least
3.boolean
default:"false"
Whether or not to normalize embeddings before classification.
string
default:"uniform"
The uniform distance weighting method to use with KNN:
distance or uniform.PositiveInt
default:"150"
The number of iterations to process between forced flushes. Must be at least
1.string
The type of encoder model to use for the Machine State Classification task:
omega_1_3_surface, omega_1_3_power_drive, omega_1_4_base, omega_1_5_base.ReaderConfig
required
Windowed reader configuration options.
Show properties
Show properties
string[]
required
The names of the columns to include in the analysis.
PositiveInt
required
The step between consecutive windows. Must be at least
1.string
required
The name of the column in the input data that contains the timestamps.
PositiveInt
required
The number of samples in each window. Must bet at least
1.integer
default:"1"
Number of worker pods to run in parallel for a single job. Must be set to
1.Example configuration
worker:
config:
batch_size: 32
classifier_config:
metric: euclidean
n_neighbors: 5
normalize_embeddings: false
weights: uniform
flush_every_n_iteration: 150
model_type: omega_1_4_base
reader_config:
data_columns:
- c1
- c2
- c3
step_size: 1024
timestamp_column: timestamp
window_size: 64
parallelism: 1
UserComponentParameters
Hide properties
Hide properties
UserInferenceConfig
required
Per-worker activity detection inference configuration.
Hide properties
Hide properties
GenerationConfig
Generation hyperparameters.
Show properties
Show properties
boolean
default:"true"
Whether or not to enable stochastic sampling. When
false, decoding is greedy.integer
default:"256"
The upper bound on tokens generated, per sample. Must be at least
1.float
default:"1.0"
Penalty applied to previously-generated tokens. Must be greater than
0 and no greater than 2.0. Values above 1.0 discourage repetition.float
default:"0.7"
Softmax temperature for sampling. Must be greater than
0 and no greater than 2.0. The higher this value, the more diverse.float
default:"0.8"
Nucleus-sampling threshold. Keeps the smallest token set whose cumulative probability is greater than or equal to this value. Must be greater than
0 and no greater than 1.0.integer
default:"20"
Top-k cutoff. Set to
0 to disable cutoff.integer
default:"32"
The maximum number of video frames sampled per video input. Extra frames are dropped uniformly. Must be between 1 and 64.
string
default:"newton/c:2.5.1-8b-base"
Specifies the model bundle to use. Standard permitted values are:
newton/c:2.3.0-7b-basenewton/c:2.4.0-7b-basenewton/c:2.5.0-8b-basenewton/c:2.5.1-8b-basenewton/c:2.6.0-8b-bf16-basenewton/c:2.6.0-8b-fp8-basenewton/c:2.6.0-30b-a3b-fp8-basenewton/visual-audio-c:2.5.0-v2-8bnewton/visual-audio-c:2.6.0-v2-8b-fp8
integer
default:"1"
Number of worker pods to run in parallel for a single job. Must be set to
1.Example configuration
worker:
config:
generation:
do_sample: true
max_new_tokens: 256
repetition_penalty: 1
temperature: 0.7
top_k: 20
top_p: 0.8
max_video_frames: 32
model_variant: newton/c:2.5.1-8b-base
parallelism: 1