Short Answer

LoRA vs full FT on 7B + 10k examples: order of magnitude cost comparison

Extending a 4k-context model to 32k via FT: how RoPE scaling fits in

Closed-API fine-tuning in 2025-2026: options and trade-offs vs open-weight FT

Why is chat-template integrity the most common 'silent quality killer' in FT?

How do you detect catastrophic forgetting after a domain fine-tune?

AdaLoRA: how does adaptive rank allocation across layers work?

What does W8A8 vs W4A16 mean and which regime is appropriate for memory-bound decode?

How does PagedAttention enable prefix sharing across requests and what data structure makes it cheap?

Define TTFT, TPOT and throughput. Which should a chatbot vs a batch summarizer optimize?

Why does tree-based speculative decoding outperform sequential draft chains at the same target-pass cost?

Why does an agent loop with 5 tool calls have ~6× the latency of a single completion and how do you mitigate?

Why does streaming (SSE) improve perceived latency without changing total decode time?

How does the serving stack actually decide when to stop generation and what bugs come from mis-set stop sequences?

How does vanilla speculative decoding work and why is verification a single forward pass?

Given draft acceptance rate α and target/draft cost ratio c, when does speculative decoding actually win?

What is RadixAttention in SGLang and how does it differ from vLLM's prefix sharing?

When would you choose vLLM, TensorRT-LLM, SGLang, or TGI for a production serving deployment?

Apply the roofline model to LLM inference and identify where decode and prefill sit on it.

Given a TPM rate limit, how do you compute supportable concurrent users for a feature?

When would you choose QAT over PTQ for inference quantization and what's the cost?

How do Anthropic and OpenAI prompt-cache pricing models work and what counts as a cache hit?

Why does prefill saturate compute while decode is bottlenecked on memory bandwidth?

Why does prefill's O(seq^2) attention cost saturate compute rather than bandwidth?

Decompose the cost of a single API call into its components and explain which dominates.

Compare tensor, pipeline and expert parallelism for inference serving.