update all README.md

README.md

@@ -62,7 +62,7 @@ cd BioM3_PenCL
 ```bash
 python run_PenCL_inference.py \
     --json_path "stage1_config.json" \
-    --model_path "BioM3_PenCL_epoch20.bin"
+    --model_path "./weights/PenCL/BioM3_PenCL_epoch20.bin"
 ```
 
 ### Example Input Data

weights/Facilitator/README.md

@@ -0,0 +1,35 @@
+
+---
+
+### **`weights/Facilitator/README.md`**
+
+```markdown
+# Facilitator Pre-trained Weights
+
+This folder will contain the pre-trained weights for the **Facilitator** model. The Facilitator model is part of the BioM3 pipeline and serves as a key component for further alignment or generation tasks.
+
+---
+
+## **Downloading Pre-trained Weights**
+
+The Google Drive link for downloading the Facilitator pre-trained weights will be added here soon.
+
+---
+
+## **File Details**
+
+- **File Name**: Facilitator pre-trained weights (TBD).
+- **Description**: Pre-trained weights for the Facilitator model.
+
+---
+
+## **Usage**
+
+Once available, the pre-trained weights can be loaded as follows:
+
+```python
+import torch
+model = YourFacilitatorModel()  # Replace with your model class
+model.load_state_dict(torch.load("weights/Facilitator/Facilitator_weights.bin", map_location="cpu"))
+model.eval()
+

weights/PenCL/README.md

@@ -0,0 +1,69 @@
+
+---
+
+### **`weights/PenCL/README.md`**
+
+```markdown
+# PenCL Pre-trained Weights
+
+This folder contains the pre-trained weights for the **PenCL** model (Stage 1 of BioM3). The PenCL model aligns protein sequences and text descriptions to compute joint latent embeddings.
+
+---
+
+## **Downloading Pre-trained Weights**
+
+To download the **PenCL epoch 20 pre-trained weights** as a `.bin` file from Google Drive, use the following command:
+
+```bash
+pip install gdown
+gdown --id 1Lup7Xqwa1NjJpoM2uvvBAdghoM-fecEj -O BioM3_PenCL_epoch20.bin
+
+---
+
+## **Usage**
+
+Once available, the pre-trained weights can be loaded as follows:
+
+```python
+import json
+import torch
+from argparse import Namespace
+import Stage1_source.model as mod
+
+# Step 1: Load JSON Configuration
+def load_json_config(json_path):
+    """
+    Load a JSON configuration file and return it as a dictionary.
+    """
+    with open(json_path, "r") as f:
+        config = json.load(f)
+    return config
+
+# Step 2: Convert JSON Dictionary to Namespace
+def convert_to_namespace(config_dict):
+    """
+    Recursively convert a dictionary to an argparse Namespace.
+    """
+    for key, value in config_dict.items():
+        if isinstance(value, dict):
+            config_dict[key] = convert_to_namespace(value)
+    return Namespace(**config_dict)
+
+if __name__ == '__main__':
+    # Path to configuration and weights
+    config_path = "stage1_config.json"
+    model_weights_path = "weights/PenCL/BioM3_PenCL_epoch20.bin"
+
+    # Load Configuration
+    print("Loading configuration...")
+    config_dict = load_json_config(config_path)
+    config_args = convert_to_namespace(config_dict)
+
+    # Load Model
+    print("Loading pre-trained model weights...")
+    model = mod.pfam_PEN_CL(args=config_args)  # Initialize the model with arguments
+    model.load_state_dict(torch.load(model_weights_path, map_location="cpu"))
+    model.eval()
+    print("Model loaded successfully with weights!")
+
+

weights/ProteoScribe/README.md

@@ -0,0 +1,35 @@
+
+---
+
+### **`weights/ProteoScribe/README.md`**
+
+```markdown
+# ProteoScribe Pre-trained Weights
+
+This folder will contain the pre-trained weights for the **ProteoScribe** model. ProteoScribe enables advanced functional annotation or protein generation tasks.
+
+---
+
+## **Downloading Pre-trained Weights**
+
+The Google Drive link for downloading the ProteoScribe pre-trained weights will be added here soon.
+
+---
+
+## **File Details**
+
+- **File Name**: ProteoScribe pre-trained weights (TBD).
+- **Description**: Pre-trained weights for the ProteoScribe model.
+
+---
+
+## **Usage**
+
+Once available, you can load the weights into your model using PyTorch:
+
+```python
+import torch
+model = YourProteoScribeModel()  # Replace with your model class
+model.load_state_dict(torch.load("weights/ProteoScribe/ProteoScribe_weights.bin", map_location="cpu"))
+model.eval()
+

weights/README.md

@@ -0,0 +1,27 @@
+# Weights Directory
+
+This folder contains the pre-trained weights for the **BioM3** project models. The weights are stored as `.bin` files for different components of the BioM3 pipeline:
+
+1. **PenCL**: Pre-trained weights for the PenCL model (Stage 1).
+2. **Facilitator**: Pre-trained weights for the Facilitator model (Stage 2).
+3. **ProteoScribe**: Pre-trained weights for the ProteoScribe model (Stage 3).
+
+---
+
+## **Purpose**
+
+The weights provided here enable users to quickly load and run inference with the pre-trained models for text-protein sequence alignment, functional annotation, and other tasks.
+
+Each subfolder includes:
+- Instructions for downloading the desired `.bin` files.
+- Information on integrating the weights into your workflows.
+
+---
+
+### **Prerequisites**
+
+To download pre-trained weights, you must install `gdown`:
+
+```bash
+pip install gdown
+