text generation webui
v1.16
Une interface utilisateur Web Gradio pour les grands modèles de langage.
Son objectif est de devenir le AUTOMATIC1111/stable-diffusion-webui de génération de texte.
 |  |
|---|---|
 |  |
instruct , chat-instruct et chat , avec des modèles d'invite automatiques dans chat-instruct .installer_files autonome qui n'interfère pas avec l'environnement système.start_linux.sh , start_windows.bat , start_macos.sh ou start_wsl.bat .http://localhost:7860 . Pour redémarrer l'interface utilisateur Web plus tard, exécutez simplement le même script start_ . Si vous devez réinstaller, supprimez le dossier installer_files créé lors de l'installation et exécutez à nouveau le script.
Vous pouvez utiliser des indicateurs de ligne de commande, comme ./start_linux.sh --help , ou les ajouter à CMD_FLAGS.txt (comme --api pour activer l'utilisation de l'API). Pour mettre à jour le projet, exécutez update_wizard_linux.sh , update_wizard_windows.bat , update_wizard_macos.sh ou update_wizard_wsl.bat .
Le script utilise Miniconda pour configurer un environnement Conda dans le dossier installer_files .
Si jamais vous avez besoin d'installer quelque chose manuellement dans l'environnement installer_files , vous pouvez lancer un shell interactif à l'aide du script cmd : cmd_linux.sh , cmd_windows.bat , cmd_macos.sh ou cmd_wsl.bat .
start_ , update_wizard_ ou cmd_ ) en tant qu'administrateur/root.extensions_reqs pour votre système d'exploitation. A la fin, ce script installera les principales exigences du projet pour s'assurer qu'elles prévalent en cas de conflits de versions.GPU_CHOICE , USE_CUDA118 , LAUNCH_AFTER_INSTALL et INSTALL_EXTENSIONS . Par exemple : GPU_CHOICE=A USE_CUDA118=FALSE LAUNCH_AFTER_INSTALL=FALSE INSTALL_EXTENSIONS=TRUE ./start_linux.sh .Recommandé si vous avez une certaine expérience avec la ligne de commande.
https://docs.conda.io/en/latest/miniconda.html
Sous Linux ou WSL, il peut être installé automatiquement avec ces deux commandes (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
| Système | GPU | Commande |
|---|---|---|
| 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 | Processeur uniquement | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cpu |
| Linux | DMLA | 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 | N'importe lequel | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 |
| Fenêtres | Nvidia | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 --index-url https://download.pytorch.org/whl/cu121 |
| Fenêtres | Processeur uniquement | pip3 install torch==2.4.1 torchvision==0.19.1 torchaudio==2.4.1 |
Les commandes à jour peuvent être trouvées ici : https://pytorch.org/get-started/locally/.
Pour NVIDIA, vous devez également installer les bibliothèques d'exécution CUDA :
conda install -y -c "nvidia/label/cuda-12.1.1" cuda-runtime
Si vous avez besoin nvcc pour compiler manuellement une bibliothèque, remplacez la commande ci-dessus par
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>
Fichier de pré-requis à utiliser :
| GPU | Processeur | fichier d'exigences à utiliser |
|---|---|---|
| Nvidia | a AVX2 | requirements.txt |
| Nvidia | pas d'AVX2 | requirements_noavx2.txt |
| DMLA | a AVX2 | requirements_amd.txt |
| DMLA | pas d'AVX2 | requirements_amd_noavx2.txt |
| Processeur uniquement | a AVX2 | requirements_cpu_only.txt |
| Processeur uniquement | pas d'AVX2 | requirements_cpu_only_noavx2.txt |
| Pomme | Intel | requirements_apple_intel.txt |
| Pomme | Pomme Silicium | requirements_apple_silicon.txt |
conda activate textgen
cd text-generation-webui
python server.py
Ensuite, naviguez jusqu'à
http://localhost:7860/?__theme=dark
Utilisez requirements_cpu_only.txt ou requirements_cpu_only_noavx2.txt dans la commande ci-dessus.
Installez manuellement llama-cpp-python à l'aide de la commande appropriée à votre matériel : Installation depuis PyPI.
LLAMA_HIPBLAS=on .Installez manuellement AutoGPTQ : Installation.
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
--load-in-8bit , vous devrez peut-être rétrograder comme ceci :pip install bitsandbytes==0.38.1pip install https://github.com/jllllll/bitsandbytes-windows-webui/raw/main/bitsandbytes-0.38.1-py3-none-any.whl Les requirements*.txt ci-dessus contiennent diverses roues précompilées via GitHub Actions. Si vous souhaitez compiler les choses manuellement, ou si vous en avez besoin parce qu'aucune roue adaptée n'est disponible pour votre matériel, vous pouvez utiliser requirements_nowheels.txt puis installer manuellement les chargeurs souhaités.
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
De temps en temps, les requirements*.txt changent. Pour mettre à jour, utilisez ces commandes :
conda activate textgen
cd text-generation-webui
pip install -r <requirements file that you have used> --upgrade
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
Les modèles doivent être placés dans le dossier text-generation-webui/models . Ils sont généralement téléchargés depuis Hugging Face.
models . Exemple: text-generation-webui
└── models
└── llama-2-13b-chat.Q4_K_M.gguf
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
Dans les deux cas, vous pouvez utiliser l'onglet "Modèle" de l'interface utilisateur pour télécharger automatiquement le modèle depuis Hugging Face. Il est également possible de le télécharger via la ligne de commande avec
python download-model.py organization/model
Exécutez python download-model.py --help pour voir toutes les options.
https://colab.research.google.com/github/oobabooga/text-Generation-webui/blob/main/Colab-TextGen-GPU.ipynb
En août 2023, Andreessen Horowitz (a16z) a accordé une généreuse subvention pour encourager et soutenir mon travail indépendant sur ce projet. Je suis extrêmement reconnaissant pour leur confiance et leur reconnaissance.
