jing xu

作者你好，能否提供arm架构上的binary呀

In the Llama-2-7b-chat-hf-q4f32_1-1k model, the number of tokens in the prefill is 36 when inputting 'hello'.

2

Why is the number of tokens in the prompt's output different from the actual number of tokens produced by the tokenizer? I used the LLaMA 2 tokenizer, and the prompt...

I tried to compare a specific model (such as llama 3B) between Web-LLM and local (MLC-LLM) environments, and found that under the same parameters, i.e., without making any changes, the...

wasm optimization？

1

When I use the web-llm instance (path: /web-llm/examples/simple-chat), and observe the source file (@mlc-ai/web-llm/lib/index.js), I notice that there is a lot of interaction with wasm files, which makes reading the...

When using the perf tool to analyze JavaScript application performance, the source functions in the generated flame graph may have the following prefix identifiers: JS:*, JS:+, and JS:^. What do...