Add video call screenshot to README

Fix voice clone, path defaults, README rewrite; add pydub dep

Rewrite README in English with codev-style layout; add Xet install guide

Update README: add omni-o HF conversion, call server, and frontend

feat: add interactive mode to eval_vlm.py with auto <image> insertion, update README

feat: add VLM SFT mini configs, resume training, dataset streaming, and eval fixes

refactor: rename VLM configs, add VLM SFT configs, and improve model loading

refactor: reorganize dataset into lm/vlm/vam subdirectories

docs: add convert command for native_hf and note about tokenizer compat

docs: update README and docs for model_type rename, RoPE guard fix, config paths

rename minimind→omni in model_type, eval/convert scripts, and docs

fix eval_llm: add --tokenizer_path/--native flags, auto-detect num_hidden_layers from checkpoint; update README with correct inference commands

reorganize configs into subdirs lm/vlm/vam, remove runs/, add interview docs for pretrain & full_sft

refactor(runs): rename lm/vlm/vam.sh to train_*

refactor(configs): flatten configs/model/ into configs/

Set base model hidden_size to 512 across all configs for 8G VRAM

Update docs for tokenizer training script and checkpoint save path

Replace root trainer/ entry points with runs/ shell scripts

Flatten src layout: drop omni namespace, modules live directly under src/

Move LoRA into models/lm to match modality-scoped layout

Move MiniMindModel backbone into models/lm, keep core pure-components only

Restructure models into lm/vlm/vam sub-packages by modality

Fuse minimind/minimind-v/minimind-o into omni with layered architecture

Initial commit: omni package with src layout and pyproject config