JJiang02 commited on
Commit
5418fda
·
verified ·
1 Parent(s): 1534fe9

Link reference data generation pipeline

Browse files
Files changed (1) hide show
  1. README.md +7 -8
README.md CHANGED
@@ -19,10 +19,7 @@ ResNet-50 weights associated with
19
  [*A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency*](https://arxiv.org/abs/2502.11965).
20
 
21
  - Code: [GREAT-ISAC/CSI-CLIP](https://github.com/GREAT-ISAC/CSI-CLIP)
22
-
23
- Released by **[GREAT Wireless AI](https://github.com/GREAT-ISAC)**. The
24
- `GREAT-ISAC` GitHub handle is retained for continuity with existing papers and
25
- community links.
26
  - Paper: [arXiv:2502.11965](https://arxiv.org/abs/2502.11965)
27
  - IEEE: [ICMLCN 2025](https://ieeexplore.ieee.org/abstract/document/11140262/)
28
 
@@ -36,7 +33,7 @@ an optimizer, scheduler, training data, or a downstream task head.
36
  | `model.safetensors` | Model-only pre-trained weights | Recommended for safe standalone loading and feature extraction |
37
  | `model.pth` | Model-only PyTorch checkpoint with a `model` key | Compatibility with the existing fine-tuning scripts |
38
 
39
- These are foundation-model pre-training weights, not task-specific checkpoints
40
  for positioning, beam management, or LOS/NLOS classification. A downstream
41
  head must be trained before task-level inference.
42
 
@@ -100,8 +97,10 @@ python load_pretrained.py \
100
 
101
  The model was pre-trained on simulated DeepMIMO CSI covering multiple wireless
102
  scenarios. Generated training arrays are not included in this model repository.
103
- A **reproducible data-generation pipeline** will be released separately with
104
- the simulator configuration, preprocessing, and split definitions.
 
 
105
 
106
  ## Intended use
107
 
@@ -123,7 +122,7 @@ the simulator configuration, preprocessing, and split definitions.
123
 
124
  ## License
125
 
126
- The original CSI-CLIP code, these model weights, and the accompanying original
127
  data-generation scripts are released under the
128
  [Creative Commons Attribution-NonCommercial 4.0 International](https://creativecommons.org/licenses/by-nc/4.0/)
129
  license (**CC BY-NC 4.0**). Attribution is required and commercial use is not
 
19
  [*A MIMO Wireless Channel Foundation Model via CIR-CSI Consistency*](https://arxiv.org/abs/2502.11965).
20
 
21
  - Code: [GREAT-ISAC/CSI-CLIP](https://github.com/GREAT-ISAC/CSI-CLIP)
22
+ - Reference data generation: [GREAT-ISAC/Channel-Simulation-Data](https://github.com/GREAT-ISAC/Channel-Simulation-Data)
 
 
 
23
  - Paper: [arXiv:2502.11965](https://arxiv.org/abs/2502.11965)
24
  - IEEE: [ICMLCN 2025](https://ieeexplore.ieee.org/abstract/document/11140262/)
25
 
 
33
  | `model.safetensors` | Model-only pre-trained weights | Recommended for safe standalone loading and feature extraction |
34
  | `model.pth` | Model-only PyTorch checkpoint with a `model` key | Compatibility with the existing fine-tuning scripts |
35
 
36
+ These are channel-foundation-model pre-training weights, not task-specific checkpoints
37
  for positioning, beam management, or LOS/NLOS classification. A downstream
38
  head must be trained before task-level inference.
39
 
 
97
 
98
  The model was pre-trained on simulated DeepMIMO CSI covering multiple wireless
99
  scenarios. Generated training arrays are not included in this model repository.
100
+ A **reproducible, model-compatible reference data-generation pipeline** is
101
+ available in [Channel Simulation Data](https://github.com/GREAT-ISAC/Channel-Simulation-Data).
102
+ Its committed O1_60 configuration is one runnable reference example; it does
103
+ not reconstruct the complete multi-scenario checkpoint training data.
104
 
105
  ## Intended use
106
 
 
122
 
123
  ## License
124
 
125
+ The original CSI-CLIP code, these model weights, and the repository-owned
126
  data-generation scripts are released under the
127
  [Creative Commons Attribution-NonCommercial 4.0 International](https://creativecommons.org/licenses/by-nc/4.0/)
128
  license (**CC BY-NC 4.0**). Attribution is required and commercial use is not