aggiornato readme e documentazione con ai

AGENTS.md

@@ -29,8 +29,8 @@ gcc -o classificatore_mnist classificatore.c -lm && ./classificatore_mnist
 # Memory leak detection
 valgrind --leak-check=full ./classificatore_mnist
 
-# Run pre-compiled binary (50 epochs)
-./classificatore_mnist_50_epoche
+# Run pre-compiled binary 
+./classificatore
 ```
 
 ### Running a Single Test
@@ -88,10 +88,11 @@ gcc -o test_xor classificatore.c -lm && ./test_xor
 - Validate function inputs at entry points
 
 ## Key Constants (from percettroni.h)
-- `LRE = 0.1` (learning rate)
+- `LRE = 0.01` (learning rate)
 - `soglia_sigmoide = 0.5` (sigmoid threshold)
-- `file_pesi = "rete_pesi.bin"` (model weights file)
-- `SOFTMAX = 1` (use softmax for multi-class prediction)
+- `file_pesi = "rete_mnist.bin"` (model weights file)
+- `TOLLERANZA = 99.5` (accuracy tolerance for early stopping)
+- `FUNZIONE_ATTIVAZIONE = 1` (0=sigmoid, 1=ReLU, 2=step function)
 
 ## Dataset Configuration
 In `percettroni.h`, include the desired dataset manager:
@@ -106,6 +107,13 @@ No formal test framework. Use these approaches:
 3. Monitor epoch error rates in training output
 4. Check memory leaks with valgrind
 
+## Linting and Type Checking
+No formal linting or type checking tools are configured for this C project. Code quality is maintained through:
+- Manual code review
+- Compilation warnings (use `-Wall -Wextra` flags if needed)
+- Valgrind for memory issues
+- Consistent adherence to the style guidelines below
+
 ## Project Structure
 - `percettroni.h` - Core neural network (header-only library)
 - `classificatore.c` - Main classifier program