Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Update model classes to use correct LlamaMLAConfig

Fix: Update config class to inherit from LlamaConfig

Fix: Update model classes to use correct LlamaMLAConfig

Fix: Align model_type to 'llamamla'

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Fix: Add complete model architecture to mla.py

Update modeling_llamamla.py

Update config.json

Update config.json

Update modeling_llamamla.py

Upload mla.py

Upload modeling_llamamla.py

Adding tokenizer files

Initial upload of Qwen2.5-7B-Instruct-deepseek with custom MLA architecture

initial commit