Add community evaluation results for GPQA, MMLU-PRO, SWE-BENCH_VERIFIED, HLE, TERMINAL-BENCH-2.0

Update README

Update README

Update benchmark figure

Update technical report

Update technical report

Update technical report link

Update inference/generate.py

Upload config.json

Release DeepSeek-V4

use sys.path for encoding

Upload folder using huggingface_hub

add inference code

Add files using upload-large-folder tool

initial commit