### Launch the Soup Web UI Source: https://trysoup.dev/docs/web-ui Install the UI dependencies and start the dashboard server. ```bash pip install "soup-cli[ui]" soup ui # Opens http://127.0.0.1:7860 in your browser # Prints auth token to console ``` -------------------------------- ### Installation and Training Commands Source: https://trysoup.dev/docs/asr-fine-tuning Commands to install the necessary dependencies and execute the training process. ```bash pip install "soup-cli[train,audio]" soup recipes use whisper-tiny-asr # or whisper-base-asr / whisper-large-v3-asr soup train ``` -------------------------------- ### Install and Start Soup MCP Server Source: https://trysoup.dev/docs/mcp-server Install the required dependencies and start the server using stdio transport. The --allow-mutating flag is required to enable plan-only tools. ```bash pip install "soup-cli[mcp]" # official mcp SDK (mcp>=1.2.0,<2), lazy-imported soup mcp serve # stdio; read-only tools soup mcp serve --allow-mutating # also list the 2 plan-only mutating tools ``` -------------------------------- ### Install soup-cli Source: https://trysoup.dev/docs/fine-tune-gemma-3-qlora Install the necessary training stack and optimization dependencies. ```bash pip install "soup-cli[train,fast]" ``` -------------------------------- ### Deploy with Transformers Backend Source: https://trysoup.dev/docs/serving Install the Transformers serving package and start the server on a specified port. ```bash pip install "soup-cli[serve]" soup serve --model ./output --port 8000 ``` -------------------------------- ### Install Unsloth Backend Source: https://trysoup.dev/docs/backends Install the fast training dependencies for reduced VRAM usage. ```bash pip install "soup-cli[fast]" ``` -------------------------------- ### Install and use experiment trackers Source: https://trysoup.dev/docs/v0538-10-remote-hubs Install the trackers bundle and specify the tracker during training. ```bash pip install "soup-cli[trackers]" soup train --tracker mlflow # or swanlab | trackio ``` -------------------------------- ### Initialize Audio Fine-Tuning Source: https://trysoup.dev/docs/multimodal Install the audio dependencies and initialize the project template. ```bash pip install "soup-cli[audio]" soup init --template audio ``` -------------------------------- ### Initialize Vision Fine-Tuning Source: https://trysoup.dev/docs/multimodal Install the vision dependencies and initialize the project template. ```bash pip install "soup-cli[vision]" soup init --template vision ``` -------------------------------- ### Install Performance Packages Source: https://trysoup.dev/docs/backends Install optional packages to enable Liger, FlashAttention, and Ring FlashAttention. ```bash pip install "soup-cli[liger]" # Liger Kernel pip install flash-attn --no-build-isolation # FlashAttention pip install "soup-cli[ring-attn]" # Ring FlashAttention ``` -------------------------------- ### Run Quick Demo Source: https://trysoup.dev/docs/getting-started Execute a full demo workflow to test installation and training capabilities. ```bash soup quickstart # Creates sample data + config + trains TinyLlama soup quickstart --dry-run # Just create files without training ``` -------------------------------- ### Install MLX Backend Source: https://trysoup.dev/docs/backends Install the necessary dependencies for Apple Silicon support. ```bash pip install "soup-cli[mlx]" ``` -------------------------------- ### Install Soup CLI Source: https://trysoup.dev/docs/getting-started Install the core CLI or specific dependency stacks for training, serving, and other features. ```bash # Light core: CLI + config + data tools, no PyTorch pip install soup-cli # Add the training stack (torch, transformers, peft, trl, datasets, …) pip install "soup-cli[train]" # Everything (train + serve + ui + data + mcp) in one shot pip install "soup-cli[all]" # Or from GitHub (latest dev) pip install git+https://github.com/MakazhanAlpamys/Soup.git ``` -------------------------------- ### Enable Quantization-Aware Training (QAT) Source: https://trysoup.dev/docs/backends Install the QAT extension and configure the training parameters. ```bash pip install "soup-cli[qat]" ``` ```yaml training: quantization: 4bit quantization_aware: true # Enable QAT ``` -------------------------------- ### Install soup-cli with extras Source: https://trysoup.dev/docs/lean-install-live-wiring Commands to install the base CLI or include additional dependencies for training and other features. ```bash pip install soup-cli # light: CLI + config + data tools, no PyTorch pip install "soup-cli[train]" # add the training stack to fine-tune pip install "soup-cli[all]" # train + serve + ui + data + mcp ``` -------------------------------- ### Start inference server Source: https://trysoup.dev/docs/cli-reference Starts an OpenAI-compatible inference server with support for multiple backends and speculative decoding. ```bash soup serve [--model ] [--backend transformers|vllm|sglang] [--port ] [--host ] [--tensor-parallel ] [--gpu-memory ] [--speculative-decoding ] [--num-speculative-tokens ] [--adapters =] ``` -------------------------------- ### Serve model with vLLM Source: https://trysoup.dev/docs/fine-tune-qwen-3 Install the serving dependencies and launch the model with adapters using the vLLM backend. ```bash pip install "soup-cli[serve-fast]" soup serve --model Qwen/Qwen3-8B --adapters qwen3=./runs/qwen3/latest --backend vllm --port 8000 ``` -------------------------------- ### Train using a recipe Source: https://trysoup.dev/docs/tool-calling Commands to initialize and start training with a pre-defined tool-calling recipe. ```bash soup recipes use qwen3-8b-tools soup train ``` -------------------------------- ### Install Cut Cross-Entropy Support Source: https://trysoup.dev/docs/faq Installation command for the Cut Cross-Entropy feature to optimize large-vocab model training. ```bash pip install "soup-cli[cce]" ``` -------------------------------- ### Install and run ASR tasks Source: https://trysoup.dev/docs/cli-reference Install the necessary dependencies and execute ASR inference using the Whisper model. ```bash pip install "soup-cli[train,audio]" soup recipes use whisper-tiny-asr # task: asr · format: asr · fits a 4 GB GPU soup infer --task asr --model ./out --input eval.jsonl --output preds.jsonl [--audio-dir ./data/audio] [--asr-language en] [--asr-task transcribe] ``` -------------------------------- ### Deploy with SGLang Backend Source: https://trysoup.dev/docs/serving Install the SGLang serving package and configure multi-GPU support. ```bash pip install "soup-cli[sglang]" soup serve --model ./output --backend sglang # Multi-GPU soup serve --model ./output --backend sglang --tensor-parallel 2 ``` -------------------------------- ### soup serve Source: https://trysoup.dev/docs/cli-reference Starts an OpenAI-compatible inference server for model serving. ```APIDOC ## soup serve ### Description Start an OpenAI-compatible inference server. ### Usage `soup serve [--model ] [--backend transformers|vllm|sglang] [--port ] [--host ] [--tensor-parallel ] [--gpu-memory ] [--speculative-decoding ] [--num-speculative-tokens ] [--adapters =]` ``` -------------------------------- ### Configure Quantization for Training Source: https://trysoup.dev/docs/quant-menu Examples of setting quantization parameters for GPTQ checkpoints and FSDP with QLoRA. ```yaml # Train LoRA on top of a pre-quantized GPTQ checkpoint: base: TheBloke/Llama-2-7B-Chat-GPTQ training: quantization: gptq # or: awq, hqq:4bit, aqlm, eetq, mxfp4, fp8 # FSDP + QLoRA — set quant_storage: training: quantization: 4bit bnb_4bit_quant_storage: bfloat16 ``` -------------------------------- ### Advise output format Source: https://trysoup.dev/docs/advise Example of the structured output returned by the advise command. ```text SFT (recommended) why: - 4,213 rows (above _MIN_ROWS_FOR_TRAINING=50) - no preference pairs detected - no reasoning traces — GRPO ruled out - tone-shift goal — SFT is the right baseline next: soup autopilot --data data.jsonl --task sft --goal "polite customer support chat" ``` -------------------------------- ### Install Optional Extras Source: https://trysoup.dev/docs/getting-started Install specific functional modules for tasks like fast training, inference backends, data processing, and hardware-specific optimizations. ```bash pip install "soup-cli[train]" # Training stack: torch, transformers, peft, trl, … pip install "soup-cli[fast]" # Unsloth 2-5x training speedup pip install "soup-cli[mlx]" # Apple Silicon backend (M1–M4) pip install "soup-cli[serve]" # FastAPI inference server pip install "soup-cli[serve-fast]" # vLLM backend pip install "soup-cli[sglang]" # SGLang backend pip install "soup-cli[eval]" # lm-evaluation-harness pip install "soup-cli[data]" # MinHash deduplication pip install "soup-cli[data-pro]" # langdetect + Presidio PII pip install "soup-cli[vision]" # Pillow for vision fine-tuning pip install "soup-cli[audio]" # librosa + soundfile pip install "soup-cli[qat]" # Quantization-aware training pip install "soup-cli[liger]" # Liger Kernel fused ops pip install "soup-cli[onnx]" # ONNX export pip install "soup-cli[tensorrt]" # TensorRT-LLM export pip install "soup-cli[awq]" # AWQ / GPTQ quantized export pip install "soup-cli[ui]" # Web UI dashboard pip install "soup-cli[tui]" # Full-screen Textual dashboard pip install "soup-cli[trackers]" # MLflow / SwanLab / Trackio logging pip install "soup-cli[deepspeed]" # ZeRO distributed training pip install "soup-cli[ring-attn]" # Ring FlashAttention pip install "soup-cli[remote]" # Remote datasets (s3 / gs / az / oci) pip install "soup-cli[sign]" # ed25519 adapter / attestation signing pip install "soup-cli[compile]" # DSPy / GEPA / TextGrad prompt compiler pip install "soup-cli[carbon]" # codecarbon energy / CO2 tracking pip install "soup-cli[modal]" # Serverless cloud GPU (soup train --cloud modal) pip install "soup-cli[mcp]" # MCP server (soup mcp serve) for coding agents ``` -------------------------------- ### Axolotl and Soup configuration examples Source: https://trysoup.dev/docs/migrate-from-axolotl Comparison of a standard Axolotl configuration file and its corresponding output after running the migration tool. ```yaml base_model: mistralai/Mistral-7B-v0.3 load_in_4bit: true adapter: qlora datasets: - path: tatsu-lab/alpaca type: alpaca sequence_len: 2048 micro_batch_size: 2 gradient_accumulation_steps: 8 num_epochs: 3 learning_rate: 0.0002 lora_r: 16 lora_alpha: 32 flash_attention: true ``` ```yaml base: mistralai/Mistral-7B-v0.3 task: sft backend: unsloth # root-level, NOT under training data: train: tatsu-lab/alpaca format: alpaca max_length: 2048 # sequence length lives under data training: quantization: 4bit # the key is quantization, not quant epochs: 3 lr: 2.0e-4 # the key is lr, not learning_rate batch_size: 2 gradient_accumulation_steps: 8 lora: r: 16 # LoRA turns on when r > 0; there is no enabled flag alpha: 32 ``` -------------------------------- ### Initialize Iterative DPO Source: https://trysoup.dev/docs/loop-hardening Start the iterative DPO driver using a base model from the registry. ```bash soup iterative-dpo --base-model registry://policy-v3 \ ``` -------------------------------- ### Configure OpenTelemetry Tracing Source: https://trysoup.dev/docs/speculative-decoding Install the required SDK and enable OTLP tracing for per-request spans. ```bash pip install opentelemetry-sdk opentelemetry-exporter-otlp soup serve --model ./output \ --trace --trace-endpoint http://localhost:4317 ``` -------------------------------- ### Start model training Source: https://trysoup.dev/docs/cli-reference Initiates a training run with support for various distributed training backends and HuggingFace integration. ```bash soup train [--config ] [--resume auto|] [--wandb] [--tensorboard] \ [--deepspeed zero2|zero3|zero2_offload|zero3_offload|zero++|zero_pp] [--fsdp full_shard|shard_grad|full_offload] \ [--gpus auto|] [--gate ] \ [--push-as ] [--hf-resume] [--yes] ``` -------------------------------- ### Install shell completions Source: https://trysoup.dev/docs/preflight-tooling Generates shell completion scripts for bash, zsh, and fish. ```bash soup completions bash | sudo tee /etc/bash_completion.d/soup soup completions zsh > ~/.zsh/completions/_soup soup completions fish > ~/.config/fish/completions/soup.fish ``` -------------------------------- ### Full Training Configuration Example Source: https://trysoup.dev/docs/streaming-scaling A comprehensive YAML configuration for training a Llama-3.1-8B-Instruct model using layer streaming and 4-bit quantization. ```yaml base: meta-llama/Llama-3.1-8B-Instruct task: sft backend: transformers data: train: ./data.jsonl format: alpaca max_length: 512 training: epochs: 3 lr: 2e-5 batch_size: 4 # raise this first; the pre-flight sizes it gradient_accumulation_steps: 2 # then accumulate for the rest quantization: 4bit # NF4: ~4x smaller store gradient_checkpointing: true stream_layers: true stream_source: auto # RAM, falling back to NVMe stream_buffers: 2 lora: r: 64 alpha: 16 output: ./output ``` -------------------------------- ### ASR Configuration File Source: https://trysoup.dev/docs/asr-fine-tuning Example YAML configuration for fine-tuning a Whisper model. ```yaml base: openai/whisper-tiny task: asr data: train: ./data/train.jsonl format: asr # rows: {"audio": "clip.wav", "text": "hello world"} audio_dir: ./data/audio # audio paths resolve here (containment-checked) training: epochs: 3 lr: 1e-4 batch_size: 2 asr_language: en # optional; pins + persists the decoder prefix asr_task: transcribe # transcribe | translate asr_lora: true # LoRA on q/v projections; default is full fine-tune lora: { r: 16, alpha: 32, target_modules: [q_proj, v_proj] } output: ./out ``` -------------------------------- ### Deploy with vLLM Backend Source: https://trysoup.dev/docs/serving Install the vLLM serving package and configure high-throughput options like tensor parallelism and GPU memory limits. ```bash pip install "soup-cli[serve-fast]" soup serve --model ./output --backend vllm # Multi-GPU with tensor parallelism soup serve --model ./output --backend vllm --tensor-parallel 2 # Control GPU memory usage soup serve --model ./output --backend vllm --gpu-memory 0.8 ``` -------------------------------- ### Launch Mobile-Accessible Web UI Source: https://trysoup.dev/docs/v0539-live-ux Commands to start the Web UI with public access and optional authentication tokens. ```bash soup ui --public # bind 0.0.0.0 + print phone QR soup ui --public --auth-token MY_TOKEN # override the auto-generated bearer ``` -------------------------------- ### Enable Ring FlashAttention Source: https://trysoup.dev/docs/multi-gpu-deepspeed Installation and configuration steps to support extremely long sequence lengths. ```bash pip install "soup-cli[ring-attn]" ``` ```yaml data: max_length: 131072 training: use_ring_attention: true ``` -------------------------------- ### Autopilot Output Example Source: https://trysoup.dev/docs/autopilot Displays the generated configuration decisions and justifications provided by the autopilot process. ```text ╭─ Autopilot Decisions ─────────────────────────────────╮ │ ✓ Quantization: 4bit │ │ reason: 8B model needs ~5GB in 4bit, leaves 19GB │ │ │ │ ✓ PEFT: LoRA r=16, alpha=32 │ │ reason: 15k samples — r=16 balances capacity / │ │ overfitting risk │ │ │ │ ✓ Batch size: 4 × grad_accum 8 = effective 32 │ │ ✓ Learning rate: 2e-4 │ │ ✓ Epochs: 2 │ │ ✓ Max length: 2048 (p95=1820 + 10% margin) │ │ ✓ Flash Attention v2 (Ampere GPU) │ │ ✓ Liger Kernel (modern Llama arch) │ │ ✓ Forgetting detection (mini_mmlu) │ │ ✓ Checkpoint intelligence (judge metric) │ │ │ │ Estimated time: 1h 42min │ │ Estimated VRAM: 18.2GB / 24GB ✓ │ ╰────────────────────────────────────────────────────────╯ ``` -------------------------------- ### Manage plugins via CLI Source: https://trysoup.dev/docs/plugins Commands for listing, installing, enabling, and disabling plugins in the Soup environment. ```bash soup plugins list soup plugins install ./my_plugin soup plugins enable my-trainer soup plugins disable my-trainer ``` -------------------------------- ### Augmenting Datasets via CLI Source: https://trysoup.dev/docs/data-augment Basic usage examples for rephrasing, translating, and applying styles to JSONL datasets. ```bash # Rephrase each example 3x soup data augment \ --input data.jsonl \ --output augmented.jsonl \ --strategy rephrase \ --count 3 # Translate to other languages soup data augment --input data.jsonl --strategy translate --lang ru,zh,es # Rewrite in multiple styles soup data augment --input data.jsonl --strategy style --styles formal,casual,technical ``` -------------------------------- ### Launch the TUI dashboard Source: https://trysoup.dev/docs/observability Install the TUI component and launch the dashboard with custom refresh rates and run limits. ```bash pip install "soup-cli[tui]" soup tui --refresh 1.0 --limit 50 ``` -------------------------------- ### Define full project configuration structure Source: https://trysoup.dev/docs/configuration A comprehensive example of the YAML configuration file structure, including base model, data paths, and training hyperparameters. ```yaml base: meta-llama/Llama-3.1-8B-Instruct # HuggingFace model ID (required) task: sft # Training task # backend: unsloth # 2-5x faster (pip install "soup-cli[fast]") # modality: text # text, vision, or audio data: train: ./data/train.jsonl # Path to training data format: alpaca # Data format (auto-detected if omitted) val_split: 0.1 # Validation split ratio max_length: 2048 # Max sequence length (64-1048576) # image_dir: ./data/images # For vision modality # audio_dir: ./data/audio # For audio modality training: epochs: 3 lr: 2e-5 batch_size: auto # auto or integer quantization: 4bit # none, 4bit, 8bit # quantization_aware: false # Enable QAT # optimizer: adamw_8bit # gradient_checkpointing: true # stream_layers: false # Train or align a model bigger than VRAM (v0.72, BETA) # stream_source: auto # auto | ram | disk # stream_buffers: 2 # 2-8 lora: r: 64 alpha: 16 dropout: 0.05 # target_modules: auto # Auto-detected per model # use_dora: false # Weight-decomposed LoRA output: ./output ``` -------------------------------- ### VRAM Pre-flight Forecast Output Source: https://trysoup.dev/docs/streaming-scaling Example of the VRAM usage forecast printed by the system before starting a training run. ```text peak VRAM ~0.48 GB at batch 2 x seq 256 (logits 0.35 GB) free VRAM 3.46 GB forecast 5685-8361 tok/s — a compute-bound bound, not a promise (from a GEMM ceiling measured on this card now @ 862 MHz) ``` -------------------------------- ### Initialize Project Configuration Source: https://trysoup.dev/docs/getting-started Create a new project configuration using the interactive wizard or a predefined template. ```bash # Interactive wizard soup init # Or use a template soup init --template chat ``` -------------------------------- ### Configure MCP Client Source: https://trysoup.dev/docs/mcp-server Add the Soup server to your MCP client configuration file. Use the second example to enable mutating tools. ```json { "mcpServers": { "soup": { "command": "soup", "args": ["mcp", "serve"] } } } ``` ```json { "mcpServers": { "soup": { "command": "soup", "args": ["mcp", "serve", "--allow-mutating"] } } } ``` -------------------------------- ### Use pre-defined training recipes Source: https://trysoup.dev/docs/multi-gpu Load a pre-configured recipe and launch the training process. ```bash soup recipes use llama3-70b-fsdp2 soup train --gpus 8 ``` -------------------------------- ### Training Execution Commands Source: https://trysoup.dev/docs/training Command-line interface usage for starting, resuming, and configuring training runs with various logging and hardware acceleration options. ```bash # Start training soup train --config soup.yaml # Resume from checkpoint soup train --config soup.yaml --resume auto soup train --config soup.yaml --resume ./output/checkpoint-500 # With W&B logging soup train --config soup.yaml --wandb # With TensorBoard soup train --config soup.yaml --tensorboard # With DeepSpeed (multi-GPU) soup train --config soup.yaml --deepspeed zero2 # With FSDP2 soup train --config soup.yaml --fsdp full_shard # Skip confirmation soup train --config soup.yaml --yes # Continual learning: rehearse an old dataset so the new task # does not erase it (v0.71.36, sft/pretrain only) soup train --config new_task.yaml --replay old_task.jsonl --replay-ratio 0.1 ``` -------------------------------- ### Full Layer Streaming Configuration Source: https://trysoup.dev/docs/layer-streaming A complete YAML configuration example for enabling layer streaming with 4-bit quantization and gradient checkpointing. ```yaml base: meta-llama/Llama-3.1-8B-Instruct task: sft backend: transformers data: train: ./data.jsonl format: alpaca max_length: 512 val_split: 0.1 training: epochs: 3 lr: 2e-5 batch_size: 4 # v0.72.3: bigger batches amortise the weight read gradient_accumulation_steps: 2 # v0.72.3: values above 1 are allowed quantization: 4bit # NF4; 'none' streams bf16 instead gradient_checkpointing: true # handled per-layer by the streamer stream_layers: true stream_source: auto # RAM when it fits, NVMe disk when it does not stream_buffers: 2 lora: r: 64 alpha: 16 output: ./output ``` -------------------------------- ### Install additional CLI extras Source: https://trysoup.dev/docs/v0538-10-remote-hubs Install optional dependencies for Bayesian optimization or data processing tasks. ```bash pip install "soup-cli[mix]" pip install "soup-cli[data-pro]" ``` -------------------------------- ### Use a Recipe Source: https://trysoup.dev/docs/recipes Initializes a recipe into a local configuration file, optionally specifying a custom output path. ```bash # Copy recipe to soup.yaml soup recipes use llama3-sft # Custom output path soup recipes use qwen3-dpo --output my-config.yaml ``` -------------------------------- ### Initialize compliance templates Source: https://trysoup.dev/docs/compliance-pack Use these commands to bootstrap a project with specific regulatory compliance configurations. ```bash soup init --template hipaa # Protected Health Information soup init --template soc2 # SOC 2 Trust Services Criteria soup init --template eu-ai-act # EU AI Act Annex XI/XII soup init --template sr-11-7 # SR 11-7 Model Risk Management ``` -------------------------------- ### Install Soup CLI with DeepSpeed support Source: https://trysoup.dev/docs/multi-gpu-deepspeed Install the necessary dependencies to enable DeepSpeed functionality within the Soup CLI environment. ```bash pip install "soup-cli[train,deepspeed]" ``` -------------------------------- ### GET /health Source: https://trysoup.dev/docs/serving Performs a health check on the inference server. ```APIDOC ## GET /health ### Description Checks the health status of the inference server. ### Method GET ### Endpoint /health ``` -------------------------------- ### Configure Web UI startup options Source: https://trysoup.dev/docs/web-ui Run the UI on a specific port without automatically opening the browser. ```bash # Custom port, don't auto-open browser soup ui --port 8080 --no-browser ``` -------------------------------- ### GET /v1/models Source: https://trysoup.dev/docs/serving Lists the models currently available on the server. ```APIDOC ## GET /v1/models ### Description Retrieves a list of models available for inference. ### Method GET ### Endpoint /v1/models ``` -------------------------------- ### soup chat Source: https://trysoup.dev/docs/cli-reference Starts an interactive chat session with a trained model. ```APIDOC ## soup chat ### Description Interactive chat with a trained model. ### Usage `soup chat [--model ] [--base ] [--temperature ] [--max-tokens ]` ``` -------------------------------- ### Initialize Project Templates Source: https://trysoup.dev/docs/configuration Commands to initialize a new project using built-in templates for various fine-tuning and training tasks. ```bash soup init --template chat # Conversational fine-tune soup init --template code # Code generation soup init --template medical # Domain expert soup init --template reasoning # GRPO reasoning (DeepSeek-R1 style) soup init --template vision # Vision/multimodal fine-tune soup init --template audio # Audio/speech fine-tune soup init --template kto # KTO unpaired preference soup init --template orpo # ORPO (no reference model) soup init --template simpo # SimPO length-normalized preference soup init --template ipo # IPO regularized preference soup init --template bco # BCO binary classifier preference (v0.40) soup init --template rlhf # Full RLHF pipeline (SFT -> RM -> PPO) soup init --template pretrain # Continued pre-training on raw text soup init --template moe # MoE fine-tuning (ScatterMoE LoRA) soup init --template longcontext # 128k+ context fine-tuning soup init --template embedding # Sentence embedding fine-tuning soup init --template tool-calling # Function / tool-calling fine-tune (v0.25) ``` ```bash soup init --template hipaa # Protected Health Information soup init --template soc2 # SOC 2 Trust Services Criteria soup init --template eu-ai-act # EU AI Act Annex XI/XII soup init --template sr-11-7 # SR 11-7 Model Risk Management ``` -------------------------------- ### soup init Source: https://trysoup.dev/docs/cli-reference Initializes a configuration file using an interactive wizard or a specified template. ```APIDOC ## soup init ### Description Create a config file with interactive wizard or template. ### Usage `soup init [--template