Compare commits
10 Commits
lorne-test
...
main
Author | SHA1 | Date | |
---|---|---|---|
0ca66e886a | |||
9f8a583857 | |||
6b4e39e5be | |||
145f80849f | |||
45a5f81c8c | |||
50e8912f73 | |||
af2d3fae90 | |||
71a1559dca | |||
3a6e2ceac2 | |||
27f6ba306f |
|
@ -24,7 +24,7 @@ dataset:
|
||||||
snr_step: 3
|
snr_step: 3
|
||||||
|
|
||||||
# Number of iterations (signal recordings) per modulation and SNR combination
|
# Number of iterations (signal recordings) per modulation and SNR combination
|
||||||
num_iterations: 100
|
num_iterations: 10
|
||||||
|
|
||||||
# Modulation scheme settings; keys must match the `modulation_types` list above
|
# Modulation scheme settings; keys must match the `modulation_types` list above
|
||||||
# Each entry includes:
|
# Each entry includes:
|
||||||
|
@ -57,7 +57,7 @@ training:
|
||||||
batch_size: 256
|
batch_size: 256
|
||||||
|
|
||||||
# Number of complete passes through the training dataset during training
|
# Number of complete passes through the training dataset during training
|
||||||
epochs: 30
|
epochs: 5
|
||||||
|
|
||||||
# Learning rate: step size for weight updates after each batch
|
# Learning rate: step size for weight updates after each batch
|
||||||
# Recommended range for fine-tuning: 1e-6 to 1e-4
|
# Recommended range for fine-tuning: 1e-6 to 1e-4
|
||||||
|
|
0
helpers/__init__.py
Normal file
0
helpers/__init__.py
Normal file
|
@ -21,7 +21,7 @@ def convert_to_onnx(ckpt_path: str, fp16: bool = False) -> None:
|
||||||
|
|
||||||
in_channels = 2
|
in_channels = 2
|
||||||
batch_size = 1
|
batch_size = 1
|
||||||
slice_length = int(1024 / dataset_cfg.num_slices)
|
slice_length = int(dataset_cfg.recording_length / dataset_cfg.num_slices)
|
||||||
num_classes = len(dataset_cfg.modulation_types)
|
num_classes = len(dataset_cfg.modulation_types)
|
||||||
|
|
||||||
model = RFClassifier(
|
model = RFClassifier(
|
||||||
|
@ -42,7 +42,7 @@ def convert_to_onnx(ckpt_path: str, fp16: bool = False) -> None:
|
||||||
model.eval()
|
model.eval()
|
||||||
|
|
||||||
# Generate random sample data
|
# Generate random sample data
|
||||||
base, ext = os.path.splitext(os.path.basename(ckpt_path))
|
base, _ = os.path.splitext(os.path.basename(ckpt_path))
|
||||||
if fp16:
|
if fp16:
|
||||||
output_path = os.path.join("onnx_files", f"{base}.onnx")
|
output_path = os.path.join("onnx_files", f"{base}.onnx")
|
||||||
sample_input = torch.from_numpy(
|
sample_input = torch.from_numpy(
|
||||||
|
|
|
@ -90,7 +90,7 @@ def split_recording(
|
||||||
snippet_list = []
|
snippet_list = []
|
||||||
|
|
||||||
for data, md in recording_list:
|
for data, md in recording_list:
|
||||||
C, N = data.shape
|
_, N = data.shape
|
||||||
L = N // num_snippets
|
L = N // num_snippets
|
||||||
for i in range(num_snippets):
|
for i in range(num_snippets):
|
||||||
start = i * L
|
start = i * L
|
||||||
|
|
Loading…
Reference in New Issue
Block a user