test: add adapter unit tests for Llama

[mukund1985]

Adds a focused test suite for LlamaArchitectureAdapter per issue #1302.

What this adds

test_llama_adapter.py (22 tests)

• Config flags: normalization_type="RMS", positional_embedding_type="rotary", final_rms=True, gated_mlp=True, attn_only=False, uses_rms_norm=True
• GQA support via n_key_value_heads: propagation to cfg, presence in default_config when set, and absence when not set
• Component mapping: top-level keys, bridge types, and HF module paths (model.embed_tokens, model.rotary_emb, model.layers, model.norm, lm_head)
• Block submodules: ln1/ln2/attn/mlp bridge types and HF paths (input_layernorm, post_attention_layernorm, self_attn, mlp)
• Attention Q/K/V/O and MLP gate/in/out projections: keys, HF paths, and LinearBridge types
• Weight conversion: exact weight_processing_conversions key set

All 22 tests pass locally.

Checklist

• I have commented my code, particularly in hard-to-understand areas
• My changes generate no new warnings
• New and existing unit tests pass locally with my changes
• I have not rewritten tests relating to key interfaces which would affect backward compatibility