feat(gemma): initial implementation of Gemma model and configs

- Add core Gemma model (architecture, attention, GeGLU, RoPE, RMSNorm, etc)
- Add configs for training and generation: gemma_train.json, gemma_generate.json
- Add Gemma notebook for exploratory analysis and demonstration
- Add __init__.py for Gemma submodule
- Update run_llm_experiment.py to support Gemma experiment configs

test(gemma): add comprehensive unit tests for Gemma

- Test forward pass (with/without cache)
- Test autoregressive generation (greedy, top-k, top-p)
- Test shape correctness and max sequence length errors
- Test multi-layer stack and token embeddings

docs: add documentation notebook for Gemma usage and analysis

Closes: #issue (if applicable)

experiments/llm_only/configs/gemma_train.json

@@ -0,0 +1,28 @@
+{
+    "bpe_tokenizer": "checkpoints/bpe_tokenizer.json",
+    "bpe_vocab_size": 1000,
+    "bpe_special_tokens": ["<pad>", "<unk>", "<bos>", "<eos>"],
+    "test_prompts": ["Open source AI", "What is Llama?"],
+    "model_config": {
+      "vocab_size": null,
+      "embed_dim": 256,
+      "num_q_heads": 4,
+      "num_kv_heads": 2,
+      "head_size": 64,
+      "num_layers": 4,
+      "max_position_embeddings": 512,
+      "num_experts": 8,
+      "top_k_experts": 2,
+      "window_size": 16,
+      "dropout": 0.1
+    },
+    "model_weights": "checkpoints/gemma-bpe/model.pt",
+    "model_config_path": "checkpoints/gemma-bpe/config.json",
+    "training": {
+      "learning_rate": 0.0003,
+      "batch_size": 2,
+      "num_epochs": 3,
+      "warmup_steps": 50
+    },
+    "log_path": "checkpoints/gemma_only_training_logs.json"
+  }