Add support for CohereForAI/c4ai-command-r-v01 tool template

[xenova]

This PR adds the missing language features to support the new function calling template by Cohere, for their new https://huggingface.co/CohereForAI/c4ai-command-r-v01 model. This includes:

• Tuple types
• Iterator unpacking in for loops
• object.items() built-in
• | items and length filters for objects

cc @drbh for TGI e2e test case
cc @Rocketknight1 for transformers
cc @Wauplin for chat widget

[pcuenca]

Very cool!

[julien-c]

very cool, let's make this big when it's out

[xenova]

Now it's compatible with Transformers.js too! 😎 e.g.,

import { AutoTokenizer } from "@xenova/transformers";

const tokenizer = await AutoTokenizer.from_pretrained("Xenova/c4ai-command-r-v01-tokenizer")

// define conversation input:
const conversation = [
  { role: "user", content: "Whats the biggest penguin in the world?" }
]
// define documents to ground on:
const documents = [
  { title: "Tall penguins", text: "Emperor penguins are the tallest growing up to 122 cm in height." },
  { title: "Penguin habitats", text: "Emperor penguins only live in Antarctica." }
]

// render the RAG prompt as a string:
const grounded_generation_prompt = tokenizer.apply_chat_template(
  conversation,
  {
    chat_template: "rag",
    tokenize: false,
    add_generation_prompt: true,

    documents,
    citation_mode: "accurate", // or "fast"
  }
)
console.log(grounded_generation_prompt);

I'm ready to launch whenever. Might be good to align with @Rocketknight1 on this? If it will take too long, I can do a soft launch today too.