feat: integrate tests and scripts

[ganler]

No description provided.

[Copilot]

Pull Request Overview

This pull request integrates various utility functions, test scripts, and helper scripts for model testing and dataset visualization. The changes add comprehensive tooling for working with language models through different APIs and frameworks.

Key changes include:

• Utility functions for batched LLM inference using LiteLLM with retry logic and cost tracking
• Test scripts for validating model behavior across different frameworks (VLLM, HuggingFace, OpenAI API)
• Helper scripts for model deployment and dataset visualization

Reviewed Changes

Copilot reviewed 10 out of 11 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
utils/litellm.py	Core utility for batched LLM inference with retry logic and cost tracking
utils/init.py	Utility module initialization with batch splitting function
test/test_vllm_model.py	VLLM model testing script with various prompt scenarios
test/test_openai_api.py	OpenAI API latency testing with concurrent requests
test/test_model_8k_prompt.py	Long context testing script for 8K token prompts
test/test_hf_model.py	HuggingFace model testing with batch generation
test/test_bedrock_sonnet.py	Simple Bedrock Claude model test
test/check_chat_template.py	Chat template validation utility
script/push_model_hub.py	Model deployment script to HuggingFace Hub
script/cat_convo.py	Dataset conversation visualization tool

Comments suppressed due to low confidence (1)

utils/init.py:8

• Variable name 'l' is ambiguous and could be confused with the number '1'. Consider using a more descriptive name like 'length' or 'total_items'.

    l = len(iterable)