Support block-modular architecture

[oleksost]

✨ Description

This draft PR addresses #242 by introducing a flexible, modular configuration system for hybrid model architectures.

TODOs:

• [ ] add more testing to make sure legacy behaviour is well supported
• [ ] implement weight sharing and support for layer-specific learning rates.
• [ ] make sure model serialization/conversion qworks as expected
• [ ] review and unify naming conventions (block, layer) across codebase.
• [ ] clean up & test

model:
  base_model:
    cross_entropy_impl: fused
    blocks:
      bob:
        type: transformer
        hidden_size: 512
        
      mamba:
        type: discrete_mamba2
        state_size: 16
        expansion_factor: 2
        hidden_size: 512
        
    hybrid_block_layout: ["bob", "mamba"]
    num_layers: 4

🔍 Type of change

Class hierarchy in the config system:

• started moving functionality specific to BaseBlock into BaseBlockConfig in layers/common
• transformer and SSM layer configs inherit from BaseBlockConfig, both holding functionality specific to their dedicated blocks (TransformerLayer, LlambaBlock)

Block-specific hyperparameters & tensor space defintion:

• HybridBlockConfigs implemented under models/hybrid/config allowing block-specific hyperparameters definition
• the names of the elements in the tensor space now include block suffixes; no suffixes are used in the case of non-hybrid GPT models
• still supports legacy behaviour with blocks defined using lists like [t,m2d,m] & non-hybrid GPT models

Select all that apply:

• [ ] 🐛 Bug fix (non-breaking change that addresses a specific issue)
• [x] 🚀 New feature (non-breaking change that adds functionality)
• [ ] ⚠️ Breaking change (a change that could affect existing functionality)
• [ ] 📈 Performance improvement/optimization (improves speed, memory usage, or efficiency)
• [x] 🛠️ Code refactor (non-functional changes that improve code readability, structure, etc.)
• [ ] 📦 Dependency bump (updates dependencies, including Dockerfile or package changes)
• [ ] 📝 Documentation change (updates documentation, including new content or typo fixes)
• [ ] 🔧 Infrastructure/Build change (affects build process, CI/CD, or dependencies)

📝 Changes

✅ Checklist

Make sure the following tasks are completed before submitting the PR:

General

• [x] 📜 I have read and followed the contributing guidelines.
• [x] 🏷️ I am using a clear and descriptive PR title that summarizes the key change or feature introduced.
• [ ] 🎉 The functionality is complete, and I have tested the changes.
• [ ] 📝 I have updated the documentation if needed.
• [ ] ⚠️ The change does not introduce any new issues (e.g., runtime warnings, type checker errors, linting problems, unhandled edge cases).
• [x] 🧩 I have commented my code, especially in hard-to-understand areas.

Dependencies and Configuration

• [ ] 🐋 I have updated the Docker configuration or dependencies, if applicable.
• [ ] 🔄 I have ensured compatibility with the existing setup after dependency changes.

Testing

• [ ] 🧪 I have added or updated tests to cover my changes.
• [ ] ✔️ New and existing tests pass locally with my changes.
• [ ] 🚦 I have tested these changes on GPUs and verified training stability.
• [ ] 🏋️ I have tested the changes on realistic training workloads, if applicable.

Performance Impact

• [ ] 📊 I have run benchmarks where applicable to evaluate the performance impact.
• [ ] ✅ The benchmarks show no performance regression.
• [ ] 🚀 The benchmarks indicate a potential performance improvement.
• [ ] ⚠️ The benchmarks indicate a potential performance degradation.
• [ ] 📈 I have provided benchmark results and detailed any performance impact below, if applicable.

---