FastChat
FastChat copied to clipboard
lm-sys
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
In qa_browser.py gradio has been having problems with versions and missing files. 
Hello LMSys Team, First, thank you for providing the Chatbot Arena as an invaluable public research tool. While conducting long-form narrative stress tests on your platform, I was able to...
Is the development happening somewhere else? This hasn't been touched for 5 months....
For the current gemma template, the template is such that the marks the end of each turn and the last token in the sequence. However, for [Gemma 2 (Table 4)](https://arxiv.org/pdf/2408.00118),...
Iād like to request evaluation of a new Hugging Face model: š https://huggingface.co/rfcoder0/qwen3-4b-custom-Sile ā¢ Model size: 4B parameters ā¢ Training hardware: RTX 3060 + RTX 3070 (consumer GPUs) ā¢ Performance:...
API Access (OpenAI-compatible) Base URL: https://api.pqn.ai/v1 Auth: Authorization: Bearer Models: pqn-ai/pqn-ai-v1 Rate limit: 30 req/min (pilot) Context length: 32,768 tokens (Max output: 2,048 tokens) Throughput (A100 80G): ~110 tok/s @...
# Request: Add Pey Yar (PQN.AI) to Chatbot Arena ## Model Information **Model Name**: Pey Yar (PQN.AI) - v1 **Technical ID**: pqn-ai/pqn-ai-v1 **Display Name**: Ł¾Ū ŪŲ§Ų± (Pey Yar) - PQN.AI...
code here report bug: ModuleNotFoundError: No module named 'fastchat.eval' ``` result = subprocess.run([ "python", "-m", "fastchat.eval.run_judger", "--model-list", config["name"], "--question-file", self.test_file, "--answer-dir", str(self.output_dir) ], capture_output=True, text=True) if result.returncode != 0: print(f"ā...
Hello, team! In my first and third rounds of Battle Mode, I wrongly voted for "both images are bad" because I initially thought the generations didn't have the girl's head....
This pull request makes a minor change by adding a new file named `Slut` to the repository. No additional content or modifications are included in this change. ## Why are...
