Time Series Forecasting
Transformers
Safetensors
fela-pdm
feature-extraction
fela
fourier-neural-operator
fno
cpu
on-device
predictive-maintenance
time-series
anomaly-detection
custom_code
Instructions to use lowdown-labs/fela-pdm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use lowdown-labs/fela-pdm with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("lowdown-labs/fela-pdm", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
File size: 3,075 Bytes
031d0d7 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 | {
"model_type": "fela-pdm",
"library_name": "pytorch",
"architectures": [
"FelaPdmModel"
],
"auto_map": {
"AutoConfig": "configuration_pdm.FelaPdmConfig",
"AutoModel": "modeling_pdm.FelaPdmModel"
},
"default_variant": "cmapss_FD001",
"note": "Five trained heads ship, one safetensors file per head (cmapss_FD001, cmapss_FD002, cmapss_FD003, cmapss_FD004, cwru). Pick variants[<name>] to match the head; load_model(dir, variant=<name>) resolves <name>.safetensors.",
"variants": {
"cmapss_FD001": {
"task": "rul",
"in_channels": 14,
"patch": 1,
"n_embd": 64,
"n_layer": 4,
"n_head": 4,
"fno_modes": 32,
"gla_chunk": 32,
"ffn_hidden": 128,
"dropout": 0.0,
"use_gdn": false,
"gdn_every": 4,
"n_classes": 0,
"rul_head": true,
"seq_len": 30,
"rul_cap": 125,
"input_shape": [1, 30, 14],
"input_desc": "30 cycles of 14 informative C-MAPSS sensors, min max normalized on training stats"
},
"cmapss_FD002": {
"task": "rul",
"in_channels": 14,
"patch": 1,
"n_embd": 64,
"n_layer": 4,
"n_head": 4,
"fno_modes": 32,
"gla_chunk": 32,
"ffn_hidden": 128,
"dropout": 0.0,
"use_gdn": false,
"gdn_every": 4,
"n_classes": 0,
"rul_head": true,
"seq_len": 30,
"rul_cap": 125,
"input_shape": [1, 30, 14],
"input_desc": "30 cycles of 14 informative C-MAPSS sensors, min max normalized on training stats"
},
"cmapss_FD003": {
"task": "rul",
"in_channels": 14,
"patch": 1,
"n_embd": 64,
"n_layer": 4,
"n_head": 4,
"fno_modes": 32,
"gla_chunk": 32,
"ffn_hidden": 128,
"dropout": 0.0,
"use_gdn": false,
"gdn_every": 4,
"n_classes": 0,
"rul_head": true,
"seq_len": 30,
"rul_cap": 125,
"input_shape": [1, 30, 14],
"input_desc": "30 cycles of 14 informative C-MAPSS sensors, min max normalized on training stats"
},
"cmapss_FD004": {
"task": "rul",
"in_channels": 14,
"patch": 1,
"n_embd": 64,
"n_layer": 4,
"n_head": 4,
"fno_modes": 32,
"gla_chunk": 32,
"ffn_hidden": 128,
"dropout": 0.0,
"use_gdn": false,
"gdn_every": 4,
"n_classes": 0,
"rul_head": true,
"seq_len": 30,
"rul_cap": 125,
"input_shape": [1, 30, 14],
"input_desc": "30 cycles of 14 informative C-MAPSS sensors, min max normalized on training stats"
},
"cwru": {
"task": "cls",
"in_channels": 1,
"patch": 4,
"n_embd": 64,
"n_layer": 4,
"n_head": 4,
"fno_modes": 64,
"gla_chunk": 32,
"ffn_hidden": 128,
"dropout": 0.0,
"use_gdn": false,
"gdn_every": 4,
"n_classes": 10,
"rul_head": false,
"seq_len": 2048,
"input_shape": [1, 2048, 1],
"input_desc": "2048 raw vibration samples at 12 kHz, single channel, per signal standardized"
}
}
}
|