Mark text_encoder-llm-tail complete

Add text_encoder LLM tail (nf4, layers 51-63 + bf16 lm_head) for the prompt assistant

Add text_encoder LLM tail (int8, layers 51-63 + bf16 lm_head) for the prompt assistant

quantized text_encoder (int8) from memory

quantized transformer_ref (nvfp4) from memory

quantized transformer (nvfp4) from memory

Delete quantized

quantized transformer_ref (nvfp4) from memory

quantized transformer (nvfp4) from memory