A Gradio web UI for Large Language Models.
Its goal is to become the AUTOMATIC1111/stable-diffusion-webui of text generation.
- Supports multiple text generation backends in one UI/API, including Transformers, llama.cpp, and ExLlamaV2. TensorRT-LLM, AutoGPTQ, AutoAWQ, HQQ, and AQLM are also supported but you need to install them manually.
- OpenAI-compatible API with Chat and Completions endpoints – see examples.
- Automatic prompt formatting using Jinja2 templates.
- Three chat modes:
instruct
,chat-instruct
, andchat
, with automatic prompt templates inchat-instruct
. - "Past chats" menu to quickly switch between conversations.
- Free-form text generation in the Default/Notebook tabs without being limited to chat turns. You can send formatted conversations from the Chat tab to these.
- Multiple sampling parameters and generation options for sophisticated text generation control.
- Switch between different models easily in the UI without restarting.
- Simple LoRA fine-tuning tool.
- Requirements installed in a self-contained
installer_files
directory that doesn't interfere with the system environment. - Extension support, with numerous built-in and user-contributed extensions available. See the wiki and extensions directory for details.
- Clone or download the repository.
- Run the script that matches your OS:
start_linux.sh
,start_windows.bat
,start_macos.sh
, orstart_wsl.bat
. - Select your GPU vendor when asked.
- Once the installation ends, browse to
http://localhost:7860
. - Have fun!
To restart the web UI later, just run the same start_
script. If you need to reinstall, delete the installer_files
folder created during setup and run the script again.
You can use command-line flags, like ./start_linux.sh --help
, or add them to CMD_FLAGS.txt
(such as --api
to enable API use). To update the project, run update_wizard_linux.sh
, update_wizard_windows.bat
, update_wizard_macos.sh
, or update_wizard_wsl.bat
.
Setup details and information about installing manually
The script uses Miniconda to set up a Conda environment in the installer_files
folder.
If you ever need to install something manually in the installer_files
environment, you can launch an interactive shell using the cmd script: cmd_linux.sh
, cmd_windows.bat
, cmd_macos.sh
, or cmd_wsl.bat
.
- There is no need to run any of those scripts (
start_
,update_wizard_
, orcmd_
) as admin/root. - To install the requirements for extensions, you can use the
extensions_reqs
script for your OS. At the end, this script will install the main requirements for the project to make sure that they take precedence in case of version conflicts. - For additional instructions about AMD and WSL setup, consult the documentation.
- For automated installation, you can use the
GPU_CHOICE
,USE_CUDA118
,LAUNCH_AFTER_INSTALL
, andINSTALL_EXTENSIONS
environment variables. For instance:GPU_CHOICE=A USE_CUDA118=FALSE LAUNCH_AFTER_INSTALL=FALSE INSTALL_EXTENSIONS=TRUE ./start_linux.sh
.
Recommended if you have some experience with the command-line.
https://docs.conda.io/en/latest/miniconda.html
On Linux or WSL, it can be automatically installed with these two commands (source):
curl -sL "https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh" > "Miniconda3.sh"
bash Miniconda3.sh
conda create -n textgen python=3.11
conda activate textgen
System | GPU | Command |
---|---|---|
Linux/WSL | NVIDIA | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu121 |
Linux/WSL | CPU only | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cpu |
Linux | AMD | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/rocm6.1 |
MacOS + MPS | Any | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 |
Windows | NVIDIA | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu121 |
Windows | CPU only | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 |
The up-to-date commands can be found here: https://pytorch.org/get-started/locally/.
For NVIDIA, you also need to install the CUDA runtime libraries:
conda install -y -c "nvidia/label/cuda-12.1.1" cuda-runtime
If you need nvcc
to compile some library manually, replace the command above with
conda install -y -c "nvidia/label/cuda-12.1.1" cuda
git clone https://github.com/oobabooga/text-generation-webui
cd text-generation-webui
pip install -r <requirements file according to table below>
Requirements file to use:
GPU | CPU | requirements file to use |
---|---|---|
NVIDIA | has AVX2 | requirements.txt |
NVIDIA | no AVX2 | requirements_noavx2.txt |
AMD | has AVX2 | requirements_amd.txt |
AMD | no AVX2 | requirements_amd_noavx2.txt |
CPU only | has AVX2 | requirements_cpu_only.txt |
CPU only | no AVX2 | requirements_cpu_only_noavx2.txt |
Apple | Intel | requirements_apple_intel.txt |
Apple | Apple Silicon | requirements_apple_silicon.txt |
conda activate textgen
cd text-generation-webui
python server.py
Then browse to
http://localhost:7860/?__theme=dark
-
Use
requirements_cpu_only.txt
orrequirements_cpu_only_noavx2.txt
in the command above. -
Manually install llama-cpp-python using the appropriate command for your hardware: Installation from PyPI.
- Use the
LLAMA_HIPBLAS=on
toggle. - Note the Windows remarks.
- Use the
-
Manually install AutoGPTQ: Installation.
- Perform the from-source installation - there are no prebuilt ROCm packages for Windows.
- For Kepler GPUs and older, you will need to install CUDA 11.8 instead of 12:
pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu118
conda install -y -c "nvidia/label/cuda-11.8.0" cuda-runtime
- bitsandbytes >= 0.39 may not work. In that case, to use
--load-in-8bit
, you may have to downgrade like this:- Linux:
pip install bitsandbytes==0.38.1
- Windows:
pip install https://github.com/jllllll/bitsandbytes-windows-webui/raw/main/bitsandbytes-0.38.1-py3-none-any.whl
- Linux:
The requirements*.txt
above contain various wheels precompiled through GitHub Actions. If you wish to compile things manually, or if you need to because no suitable wheels are available for your hardware, you can use requirements_nowheels.txt
and then install your desired loaders manually.
For NVIDIA GPU:
ln -s docker/{nvidia/Dockerfile,nvidia/docker-compose.yml,.dockerignore} .
For AMD GPU:
ln -s docker/{amd/Dockerfile,intel/docker-compose.yml,.dockerignore} .
For Intel GPU:
ln -s docker/{intel/Dockerfile,amd/docker-compose.yml,.dockerignore} .
For CPU only
ln -s docker/{cpu/Dockerfile,cpu/docker-compose.yml,.dockerignore} .
cp docker/.env.example .env
#Create logs/cache dir :
mkdir -p logs cache
# Edit .env and set:
# TORCH_CUDA_ARCH_LIST based on your GPU model
# APP_RUNTIME_GID your host user's group id (run `id -g` in a terminal)
# BUILD_EXTENIONS optionally add comma separated list of extensions to build
# Edit CMD_FLAGS.txt and add in it the options you want to execute (like --listen --cpu)
#
docker compose up --build
- You need to have Docker Compose v2.17 or higher installed. See this guide for instructions.
- For additional docker files, check out this repository.
From time to time, the requirements*.txt
change. To update, use these commands:
conda activate textgen
cd text-generation-webui
pip install -r <requirements file that you have used> --upgrade
List of command-line flags
usage: server.py [-h] [--multi-user] [--character CHARACTER] [--model MODEL] [--lora LORA [LORA ...]] [--model-dir MODEL_DIR] [--lora-dir LORA_DIR] [--model-menu] [--settings SETTINGS]
[--extensions EXTENSIONS [EXTENSIONS ...]] [--verbose] [--chat-buttons] [--idle-timeout IDLE_TIMEOUT] [--loader LOADER] [--cpu] [--auto-devices]
[--gpu-memory GPU_MEMORY [GPU_MEMORY ...]] [--cpu-memory CPU_MEMORY] [--disk] [--disk-cache-dir DISK_CACHE_DIR] [--load-in-8bit] [--bf16] [--no-cache] [--trust-remote-code]
[--force-safetensors] [--no_use_fast] [--use_flash_attention_2] [--use_eager_attention] [--load-in-4bit] [--use_double_quant] [--compute_dtype COMPUTE_DTYPE] [--quant_type QUANT_TYPE]
[--flash-attn] [--tensorcores] [--n_ctx N_CTX] [--threads THREADS] [--threads-batch THREADS_BATCH] [--no_mul_mat_q] [--n_batch N_BATCH] [--no-mmap] [--mlock]
[--n-gpu-layers N_GPU_LAYERS] [--tensor_split TENSOR_SPLIT] [--numa] [--logits_all] [--no_offload_kqv] [--cache-capacity CACHE_CAPACITY] [--row_split] [--streaming-llm]
[--attention-sink-size ATTENTION_SINK_SIZE] [--tokenizer-dir TOKENIZER_DIR] [--gpu-split GPU_SPLIT] [--autosplit] [--max_seq_len MAX_SEQ_LEN] [--cfg-cache] [--no_flash_attn]
[--no_xformers] [--no_sdpa] [--cache_8bit] [--cache_4bit] [--num_experts_per_token NUM_EXPERTS_PER_TOKEN] [--triton] [--no_inject_fused_mlp] [--no_use_cuda_fp16] [--desc_act]
[--disable_exllama] [--disable_exllamav2] [--wbits WBITS] [--groupsize GROUPSIZE] [--hqq-backend HQQ_BACKEND] [--cpp-runner] [--deepspeed] [--nvme-offload-dir NVME_OFFLOAD_DIR]
[--local_rank LOCAL_RANK] [--alpha_value ALPHA_VALUE] [--rope_freq_base ROPE_FREQ_BASE] [--compress_pos_emb COMPRESS_POS_EMB] [--listen] [--listen-port LISTEN_PORT]
[--listen-host LISTEN_HOST] [--share] [--auto-launch] [--gradio-auth GRADIO_AUTH] [--gradio-auth-path GRADIO_AUTH_PATH] [--ssl-keyfile SSL_KEYFILE] [--ssl-certfile SSL_CERTFILE]
[--subpath SUBPATH] [--api] [--public-api] [--public-api-id PUBLIC_API_ID] [--api-port API_PORT] [--api-key API_KEY] [--admin-key ADMIN_KEY] [--nowebui]
[--multimodal-pipeline MULTIMODAL_PIPELINE] [--model_type MODEL_TYPE] [--pre_layer PRE_LAYER [PRE_LAYER ...]] [--checkpoint CHECKPOINT] [--monkey-patch] [--no_inject_fused_attention]
Text generation web UI
options:
-h, --help show this help message and exit
Basic settings:
--multi-user Multi-user mode. Chat histories are not saved or automatically loaded. Warning: this is likely not safe for sharing publicly.
--character CHARACTER The name of the character to load in chat mode by default.
--model MODEL Name of the model to load by default.
--lora LORA [LORA ...] The list of LoRAs to load. If you want to load more than one LoRA, write the names separated by spaces.
--model-dir MODEL_DIR Path to directory with all the models.
--lora-dir LORA_DIR Path to directory with all the loras.
--model-menu Show a model menu in the terminal when the web UI is first launched.
--settings SETTINGS Load the default interface settings from this yaml file. See settings-template.yaml for an example. If you create a file called settings.yaml, this
file will be loaded by default without the need to use the --settings flag.
--extensions EXTENSIONS [EXTENSIONS ...] The list of extensions to load. If you want to load more than one extension, write the names separated by spaces.
--verbose Print the prompts to the terminal.
--chat-buttons Show buttons on the chat tab instead of a hover menu.
--idle-timeout IDLE_TIMEOUT Unload model after this many minutes of inactivity. It will be automatically reloaded when you try to use it again.
Model loader:
--loader LOADER Choose the model loader manually, otherwise, it will get autodetected. Valid options: Transformers, llama.cpp, llamacpp_HF, ExLlamav2_HF, ExLlamav2,
AutoGPTQ.
Transformers/Accelerate:
--cpu Use the CPU to generate text. Warning: Training on CPU is extremely slow.
--auto-devices Automatically split the model across the available GPU(s) and CPU.
--gpu-memory GPU_MEMORY [GPU_MEMORY ...] Maximum GPU memory in GiB to be allocated per GPU. Example: --gpu-memory 10 for a single GPU, --gpu-memory 10 5 for two GPUs. You can also set values
in MiB like --gpu-memory 3500MiB.
--cpu-memory CPU_MEMORY Maximum CPU memory in GiB to allocate for offloaded weights. Same as above.
--disk If the model is too large for your GPU(s) and CPU combined, send the remaining layers to the disk.
--disk-cache-dir DISK_CACHE_DIR Directory to save the disk cache to. Defaults to "cache".
--load-in-8bit Load the model with 8-bit precision (using bitsandbytes).
--bf16 Load the model with bfloat16 precision. Requires NVIDIA Ampere GPU.
--no-cache Set use_cache to False while generating text. This reduces VRAM usage slightly, but it comes at a performance cost.
--trust-remote-code Set trust_remote_code=True while loading the model. Necessary for some models.
--force-safetensors Set use_safetensors=True while loading the model. This prevents arbitrary code execution.
--no_use_fast Set use_fast=False while loading the tokenizer (it's True by default). Use this if you have any problems related to use_fast.
--use_flash_attention_2 Set use_flash_attention_2=True while loading the model.
--use_eager_attention Set attn_implementation= eager while loading the model.
bitsandbytes 4-bit:
--load-in-4bit Load the model with 4-bit precision (using bitsandbytes).
--use_double_quant use_double_quant for 4-bit.
--compute_dtype COMPUTE_DTYPE compute dtype for 4-bit. Valid options: bfloat16, float16, float32.
--quant_type QUANT_TYPE quant_type for 4-bit. Valid options: nf4, fp4.
llama.cpp:
--flash-attn Use flash-attention.
--tensorcores NVIDIA only: use llama-cpp-python compiled with tensor cores support. This may increase performance on newer cards.
--n_ctx N_CTX Size of the prompt context.
--threads THREADS Number of threads to use.
--threads-batch THREADS_BATCH Number of threads to use for batches/prompt processing.
--no_mul_mat_q Disable the mulmat kernels.
--n_batch N_BATCH Maximum number of prompt tokens to batch together when calling llama_eval.
--no-mmap Prevent mmap from being used.
--mlock Force the system to keep the model in RAM.
--n-gpu-layers N_GPU_LAYERS Number of layers to offload to the GPU.
--tensor_split TENSOR_SPLIT Split the model across multiple GPUs. Comma-separated list of proportions. Example: 60,40.
--numa Activate NUMA task allocation for llama.cpp.
--logits_all Needs to be set for perplexity evaluation to work. Otherwise, ignore it, as it makes prompt processing slower.
--no_offload_kqv Do not offload the K, Q, V to the GPU. This saves VRAM but reduces the performance.
--cache-capacity CACHE_CAPACITY Maximum cache capacity (llama-cpp-python). Examples: 2000MiB, 2GiB. When provided without units, bytes will be assumed.
--row_split Split the model by rows across GPUs. This may improve multi-gpu performance.
--streaming-llm Activate StreamingLLM to avoid re-evaluating the entire prompt when old messages are removed.
--attention-sink-size ATTENTION_SINK_SIZE StreamingLLM: number of sink tokens. Only used if the trimmed prompt does not share a prefix with the old prompt.
--tokenizer-dir TOKENIZER_DIR Load the tokenizer from this folder. Meant to be used with llamacpp_HF through the command-line.
ExLlamaV2:
--gpu-split GPU_SPLIT Comma-separated list of VRAM (in GB) to use per GPU device for model layers. Example: 20,7,7.
--autosplit Autosplit the model tensors across the available GPUs. This causes --gpu-split to be ignored.
--max_seq_len MAX_SEQ_LEN Maximum sequence length.
--cfg-cache ExLlamav2_HF: Create an additional cache for CFG negative prompts. Necessary to use CFG with that loader.
--no_flash_attn Force flash-attention to not be used.
--no_xformers Force xformers to not be used.
--no_sdpa Force Torch SDPA to not be used.
--cache_8bit Use 8-bit cache to save VRAM.
--cache_4bit Use Q4 cache to save VRAM.
--num_experts_per_token NUM_EXPERTS_PER_TOKEN Number of experts to use for generation. Applies to MoE models like Mixtral.
AutoGPTQ:
--triton Use triton.
--no_inject_fused_mlp Triton mode only: disable the use of fused MLP, which will use less VRAM at the cost of slower inference.
--no_use_cuda_fp16 This can make models faster on some systems.
--desc_act For models that do not have a quantize_config.json, this parameter is used to define whether to set desc_act or not in BaseQuantizeConfig.
--disable_exllama Disable ExLlama kernel, which can improve inference speed on some systems.
--disable_exllamav2 Disable ExLlamav2 kernel.
--wbits WBITS Load a pre-quantized model with specified precision in bits. 2, 3, 4 and 8 are supported.
--groupsize GROUPSIZE Group size.
HQQ:
--hqq-backend HQQ_BACKEND Backend for the HQQ loader. Valid options: PYTORCH, PYTORCH_COMPILE, ATEN.
TensorRT-LLM:
--cpp-runner Use the ModelRunnerCpp runner, which is faster than the default ModelRunner but doesn't support streaming yet.
DeepSpeed:
--deepspeed Enable the use of DeepSpeed ZeRO-3 for inference via the Transformers integration.
--nvme-offload-dir NVME_OFFLOAD_DIR DeepSpeed: Directory to use for ZeRO-3 NVME offloading.
--local_rank LOCAL_RANK DeepSpeed: Optional argument for distributed setups.
RoPE:
--alpha_value ALPHA_VALUE Positional embeddings alpha factor for NTK RoPE scaling. Use either this or compress_pos_emb, not both.
--rope_freq_base ROPE_FREQ_BASE If greater than 0, will be used instead of alpha_value. Those two are related by rope_freq_base = 10000 * alpha_value ^ (64 / 63).
--compress_pos_emb COMPRESS_POS_EMB Positional embeddings compression factor. Should be set to (context length) / (model's original context length). Equal to 1/rope_freq_scale.
Gradio:
--listen Make the web UI reachable from your local network.
--listen-port LISTEN_PORT The listening port that the server will use.
--listen-host LISTEN_HOST The hostname that the server will use.
--share Create a public URL. This is useful for running the web UI on Google Colab or similar.
--auto-launch Open the web UI in the default browser upon launch.
--gradio-auth GRADIO_AUTH Set Gradio authentication password in the format "username:password". Multiple credentials can also be supplied with "u1:p1,u2:p2,u3:p3".
--gradio-auth-path GRADIO_AUTH_PATH Set the Gradio authentication file path. The file should contain one or more user:password pairs in the same format as above.
--ssl-keyfile SSL_KEYFILE The path to the SSL certificate key file.
--ssl-certfile SSL_CERTFILE The path to the SSL certificate cert file.
--subpath SUBPATH Customize the subpath for gradio, use with reverse proxy
API:
--api Enable the API extension.
--public-api Create a public URL for the API using Cloudfare.
--public-api-id PUBLIC_API_ID Tunnel ID for named Cloudflare Tunnel. Use together with public-api option.
--api-port API_PORT The listening port for the API.
--api-key API_KEY API authentication key.
--admin-key ADMIN_KEY API authentication key for admin tasks like loading and unloading models. If not set, will be the same as --api-key.
--nowebui Do not launch the Gradio UI. Useful for launching the API in standalone mode.
Multimodal:
--multimodal-pipeline MULTIMODAL_PIPELINE The multimodal pipeline to use. Examples: llava-7b, llava-13b.
https://github.com/oobabooga/text-generation-webui/wiki
Models should be placed in the folder text-generation-webui/models
. They are usually downloaded from Hugging Face.
- GGUF models are a single file and should be placed directly into
models
. Example:
text-generation-webui
└── models
└── llama-2-13b-chat.Q4_K_M.gguf
- The remaining model types (like 16-bit transformers models and GPTQ models) are made of several files and must be placed in a subfolder. Example:
text-generation-webui
├── models
│ ├── lmsys_vicuna-33b-v1.3
│ │ ├── config.json
│ │ ├── generation_config.json
│ │ ├── pytorch_model-00001-of-00007.bin
│ │ ├── pytorch_model-00002-of-00007.bin
│ │ ├── pytorch_model-00003-of-00007.bin
│ │ ├── pytorch_model-00004-of-00007.bin
│ │ ├── pytorch_model-00005-of-00007.bin
│ │ ├── pytorch_model-00006-of-00007.bin
│ │ ├── pytorch_model-00007-of-00007.bin
│ │ ├── pytorch_model.bin.index.json
│ │ ├── special_tokens_map.json
│ │ ├── tokenizer_config.json
│ │ └── tokenizer.model
In both cases, you can use the "Model" tab of the UI to download the model from Hugging Face automatically. It is also possible to download it via the command-line with
python download-model.py organization/model
Run python download-model.py --help
to see all the options.
- Subreddit: https://www.reddit.com/r/Oobabooga/
- Discord: https://discord.gg/jwZCF2dPQN
In August 2023, Andreessen Horowitz (a16z) provided a generous grant to encourage and support my independent work on this project. I am extremely grateful for their trust and recognition.