import os import shutil import subprocess import signal os.environ["GRADIO_ANALYTICS_ENABLED"] = "False" import gradio as gr from huggingface_hub import create_repo, HfApi from huggingface_hub import snapshot_download from huggingface_hub import whoami from huggingface_hub import ModelCard from gradio_huggingfacehub_search import HuggingfaceHubSearch from apscheduler.schedulers.background import BackgroundScheduler from textwrap import dedent HF_TOKEN = os.environ.get("HF_TOKEN") def generate_importance_matrix(model_path, train_data_path): imatrix_command = f"./llama-imatrix -m ../{model_path} -f {train_data_path} -ngl 99 --output-frequency 10" os.chdir("llama.cpp") print(f"Current working directory: {os.getcwd()}") print(f"Files in the current directory: {os.listdir('.')}") if not os.path.isfile(f"../{model_path}"): raise Exception(f"Model file not found: {model_path}") print("Running imatrix command...") process = subprocess.Popen(imatrix_command, shell=True) try: process.wait(timeout=60) # added wait except subprocess.TimeoutExpired: print("Imatrix computation timed out. Sending SIGINT to allow graceful termination...") process.send_signal(signal.SIGINT) try: process.wait(timeout=5) # grace period except subprocess.TimeoutExpired: print("Imatrix proc still didn't term. Forecfully terming process...") process.kill() os.chdir("..") print("Importance matrix generation completed.") def split_upload_model(model_path, repo_id, oauth_token: gr.OAuthToken | None, split_max_tensors=256, split_max_size=None): if oauth_token.token is None: raise ValueError("You have to be logged in.") split_cmd = f"llama.cpp/llama-gguf-split --split --split-max-tensors {split_max_tensors}" if split_max_size: split_cmd += f" --split-max-size {split_max_size}" split_cmd += f" {model_path} {model_path.split('.')[0]}" print(f"Split command: {split_cmd}") result = subprocess.run(split_cmd, shell=True, capture_output=True, text=True) print(f"Split command stdout: {result.stdout}") print(f"Split command stderr: {result.stderr}") if result.returncode != 0: raise Exception(f"Error splitting the model: {result.stderr}") print("Model split successfully!") sharded_model_files = [f for f in os.listdir('.') if f.startswith(model_path.split('.')[0])] if sharded_model_files: print(f"Sharded model files: {sharded_model_files}") api = HfApi(token=oauth_token.token) for file in sharded_model_files: file_path = os.path.join('.', file) print(f"Uploading file: {file_path}") try: api.upload_file( path_or_fileobj=file_path, path_in_repo=file, repo_id=repo_id, ) except Exception as e: raise Exception(f"Error uploading file {file_path}: {e}") else: raise Exception("No sharded files found.") print("Sharded model has been uploaded successfully!") def process_model(model_id, q_method, use_imatrix, imatrix_q_method, private_repo, train_data_file, split_model, split_max_tensors, split_max_size, oauth_token: gr.OAuthToken | None): if oauth_token.token is None: raise ValueError("You must be logged in to use GGUF-my-repo") model_name = model_id.split('/')[-1] fp16 = f"{model_name}.fp16.gguf" try: api = HfApi(token=oauth_token.token) dl_pattern = ["*.md", "*.json", "*.model"] pattern = ( "*.safetensors" if any( file.path.endswith(".safetensors") for file in api.list_repo_tree( repo_id=model_id, recursive=True, ) ) else "*.bin" ) dl_pattern += pattern api.snapshot_download(repo_id=model_id, local_dir=model_name, local_dir_use_symlinks=False, allow_patterns=dl_pattern) print("Model downloaded successfully!") print(f"Current working directory: {os.getcwd()}") print(f"Model directory contents: {os.listdir(model_name)}") conversion_script = "convert_hf_to_gguf.py" fp16_conversion = f"python llama.cpp/{conversion_script} {model_name} --outtype f16 --outfile {fp16}" result = subprocess.run(fp16_conversion, shell=True, capture_output=True) print(result) if result.returncode != 0: raise Exception(f"Error converting to fp16: {result.stderr}") print("Model converted to fp16 successfully!") print(f"Converted model path: {fp16}") imatrix_path = "llama.cpp/imatrix.dat" if use_imatrix: if train_data_file: train_data_path = train_data_file.name else: train_data_path = "groups_merged.txt" #fallback calibration dataset print(f"Training data file path: {train_data_path}") if not os.path.isfile(train_data_path): raise Exception(f"Training data file not found: {train_data_path}") generate_importance_matrix(fp16, train_data_path) else: print("Not using imatrix quantization.") username = whoami(oauth_token.token)["name"] quantized_gguf_name = f"{model_name.lower()}-{imatrix_q_method.lower()}-imat.gguf" if use_imatrix else f"{model_name.lower()}-{q_method.lower()}.gguf" quantized_gguf_path = quantized_gguf_name if use_imatrix: quantise_ggml = f"./llama.cpp/llama-quantize --imatrix {imatrix_path} {fp16} {quantized_gguf_path} {imatrix_q_method}" else: quantise_ggml = f"./llama.cpp/llama-quantize {fp16} {quantized_gguf_path} {q_method}" result = subprocess.run(quantise_ggml, shell=True, capture_output=True) if result.returncode != 0: raise Exception(f"Error quantizing: {result.stderr}") print(f"Quantized successfully with {imatrix_q_method if use_imatrix else q_method} option!") print(f"Quantized model path: {quantized_gguf_path}") # Create empty repo new_repo_url = api.create_repo(repo_id=f"{username}/{model_name}-{imatrix_q_method if use_imatrix else q_method}-GGUF", exist_ok=True, private=private_repo) new_repo_id = new_repo_url.repo_id print("Repo created successfully!", new_repo_url) try: card = ModelCard.load(model_id, token=oauth_token.token) except: card = ModelCard("") if card.data.tags is None: card.data.tags = [] card.data.tags.append("llama-cpp") card.data.tags.append("gguf-my-repo") card.data.base_model = model_id card.text = dedent( f""" # {new_repo_id} Asalamu Alaikum! This model was converted to GGUF format from [`{model_id}`](https://huggingface.co/{model_id}) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space. Refer to the [original model card](https://huggingface.co/{model_id}) for more details on the model. ## Description (per [TheBloke](https://huggingface.co/TheBloke)) This repo contains GGUF format model files. These files were quantised using ggml-org/gguf-my-repo [https://huggingface.co/spaces/ggml-org/gguf-my-repo] ### About GGUF (per [TheBloke](https://huggingface.co/TheBloke)) GGUF is a new format introduced by the llama.cpp team on August 21st 2023. It is a replacement for GGML, which is no longer supported by llama.cpp. Here is an incomplete list of clients and libraries that are known to support GGUF: * [llama.cpp](https://github.com/ggerganov/llama.cpp). The source project for GGUF. Offers a CLI and a server option. * [text-generation-webui](https://github.com/oobabooga/text-generation-webui), the most widely used web UI, with many features and powerful extensions. Supports GPU acceleration. * [KoboldCpp](https://github.com/LostRuins/koboldcpp), a fully featured web UI, with GPU accel across all platforms and GPU architectures. Especially good for story telling. * [GPT4All](https://gpt4all.io/index.html), a free and open source local running GUI, supporting Windows, Linux and macOS with full GPU accel. * [LM Studio](https://lmstudio.ai/), an easy-to-use and powerful local GUI for Windows and macOS (Silicon), with GPU acceleration. Linux available, in beta as of 27/11/2023. * [LoLLMS Web UI](https://github.com/ParisNeo/lollms-webui), a great web UI with many interesting and unique features, including a full model library for easy model selection. * [Faraday.dev](https://faraday.dev/), an attractive and easy to use character-based chat GUI for Windows and macOS (both Silicon and Intel), with GPU acceleration. * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python), a Python library with GPU accel, LangChain support, and OpenAI-compatible API server. * [candle](https://github.com/huggingface/candle), a Rust ML framework with a focus on performance, including GPU support, and ease of use. * [ctransformers](https://github.com/marella/ctransformers), a Python library with GPU accel, LangChain support, and OpenAI-compatible AI server. Note, as of time of writing (November 27th 2023), ctransformers has not been updated in a long time and does not support many recent models. ## Compatibility These quantised GGUFv2 files are compatible with llama.cpp from August 27th 2023 onwards, as of commit [d0cee0d](https://github.com/ggerganov/llama.cpp/commit/d0cee0d36d5be95a0d9088b674dbb27354107221) They are also compatible with many third party UIs and libraries - please see the list at the top of this README. ## Explanation of quantisation methods
Click to see details The new methods available are: * GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw) * GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw. * GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw. * GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw * GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw Refer to the Provided Files table below to see what files use which methods, and how.
## Provided Files (Not Including iMatrix Quantization) | Quant method | Bits | Example Size | Max RAM required | Use case | | ---- | ---- | ---- | ---- | ----- | | Q2_K | 2 | 2.72 GB| 5.22 GB | significant quality loss - not recommended for most purposes | | Q3_K_S | 3 | 3.16 GB| 5.66 GB | very small, high quality loss | | Q3_K_M | 3 | 3.52 GB| 6.02 GB | very small, high quality loss | | Q3_K_L | 3 | 3.82 GB| 6.32 GB | small, substantial quality loss | | Q4_0 | 4 | 4.11 GB| 6.61 GB | legacy; small, very high quality loss - prefer using Q3_K_M | | Q4_K_S | 4 | 4.14 GB| 6.64 GB | small, greater quality loss | | Q4_K_M | 4 | 4.37 GB| 6.87 GB | medium, balanced quality - recommended | | Q5_0 | 5 | 5.00 GB| 7.50 GB | legacy; medium, balanced quality - prefer using Q4_K_M | | Q5_K_S | 5 | 5.00 GB| 7.50 GB | large, low quality loss - recommended | | Q5_K_M | 5 | 5.13 GB| 7.63 GB | large, very low quality loss - recommended | | Q6_K | 6 | 5.94 GB| 8.44 GB | very large, extremely low quality loss | | Q8_0 | 8 | 7.70 GB| 10.20 GB | very large, extremely low quality loss - not recommended | **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead. --- ## Use with llama.cpp Install llama.cpp through brew (works on Mac and Linux) ```bash brew install llama.cpp ``` Invoke the llama.cpp server or the CLI. ### CLI: ```bash llama-cli --hf-repo {new_repo_id} --hf-file {quantized_gguf_name} -p "The meaning to life and the universe is" ``` ### Server: ```bash llama-server --hf-repo {new_repo_id} --hf-file {quantized_gguf_name} -c 2048 ``` Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well. Step 1: Clone llama.cpp from GitHub. ``` git clone https://github.com/ggerganov/llama.cpp ``` Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux). ``` cd llama.cpp && LLAMA_CURL=1 make ``` Step 3: Run inference through the main binary. ``` ./llama-cli --hf-repo {new_repo_id} --hf-file {quantized_gguf_name} -p "The meaning to life and the universe is" ``` or ``` ./llama-server --hf-repo {new_repo_id} --hf-file {quantized_gguf_name} -c 2048 ``` """ ) card.save(f"README.md") if split_model: split_upload_model(quantized_gguf_path, new_repo_id, oauth_token, split_max_tensors, split_max_size) else: try: print(f"Uploading quantized model: {quantized_gguf_path}") api.upload_file( path_or_fileobj=quantized_gguf_path, path_in_repo=quantized_gguf_name, repo_id=new_repo_id, ) except Exception as e: raise Exception(f"Error uploading quantized model: {e}") imatrix_path = "llama.cpp/imatrix.dat" if os.path.isfile(imatrix_path): try: print(f"Uploading imatrix.dat: {imatrix_path}") api.upload_file( path_or_fileobj=imatrix_path, path_in_repo="imatrix.dat", repo_id=new_repo_id, ) except Exception as e: raise Exception(f"Error uploading imatrix.dat: {e}") api.upload_file( path_or_fileobj=f"README.md", path_in_repo=f"README.md", repo_id=new_repo_id, ) print(f"Uploaded successfully with {imatrix_q_method if use_imatrix else q_method} option!") return ( f'Find your repo here', "llama.png", ) except Exception as e: return (f"Error: {e}", "error.png") finally: shutil.rmtree(model_name, ignore_errors=True) print("Folder cleaned up successfully!") css="""/* Custom CSS to allow scrolling */ .gradio-container {overflow-y: auto;} """ # Create Gradio interface with gr.Blocks(css=css) as demo: gr.Markdown("You must be logged in to use GGUF-my-repo.") gr.LoginButton(min_width=250) model_id = HuggingfaceHubSearch( label="Hub Model ID", placeholder="Search for model id on Huggingface", search_type="model", ) q_method = gr.Dropdown( ["Q2_K", "Q3_K_S", "Q3_K_M", "Q3_K_L", "Q4_0", "Q4_K_S", "Q4_K_M", "Q5_0", "Q5_K_S", "Q5_K_M", "Q6_K", "Q8_0"], label="Quantization Method", info="GGML quantization type", value="Q8_0", filterable=False, visible=True ) imatrix_q_method = gr.Dropdown( ["IQ3_M", "IQ3_XXS", "Q4_K_M", "Q4_K_S", "IQ4_NL", "IQ4_XS", "Q5_K_M", "Q5_K_S"], label="Imatrix Quantization Method", info="GGML imatrix quants type", value="IQ4_NL", filterable=False, visible=False ) use_imatrix = gr.Checkbox( value=False, label="Use Imatrix Quantization", info="Use importance matrix for quantization." ) private_repo = gr.Checkbox( value=True, label="Private Repo", info="Create a private repo under your username." ) train_data_file = gr.File( label="Training Data File", file_types=["txt"], visible=False ) split_model = gr.Checkbox( value=False, label="Split Model", info="Shard the model using gguf-split." ) split_max_tensors = gr.Number( value=256, label="Max Tensors per File", info="Maximum number of tensors per file when splitting model.", visible=False ) split_max_size = gr.Textbox( label="Max File Size", info="Maximum file size when splitting model (--split-max-size). May leave empty to use the default.", visible=False ) def update_visibility(use_imatrix): return gr.update(visible=not use_imatrix), gr.update(visible=use_imatrix), gr.update(visible=use_imatrix) use_imatrix.change( fn=update_visibility, inputs=use_imatrix, outputs=[q_method, imatrix_q_method, train_data_file] ) iface = gr.Interface( fn=process_model, inputs=[ model_id, q_method, use_imatrix, imatrix_q_method, private_repo, train_data_file, split_model, split_max_tensors, split_max_size, ], outputs=[ gr.Markdown(label="output"), gr.Image(show_label=False), ], title="Asalamu Alaikum! Create your own GGUF Quantizations, B̶L̶A̶Z̶I̶N̶G̶L̶Y̶ ̶F̶A̶S̶T̶ ⚡! (Hey it's free!)", description="The space takes a HuggingFace repo as an input, quantizes it and creates a private repo containing the selected quant under your HF user namespace.", api_name=False ) def update_split_visibility(split_model): return gr.update(visible=split_model), gr.update(visible=split_model) split_model.change( fn=update_split_visibility, inputs=split_model, outputs=[split_max_tensors, split_max_size] ) def restart_space(): HfApi().restart_space(repo_id="ggml-org/gguf-my-repo", token=HF_TOKEN, factory_reboot=True) scheduler = BackgroundScheduler() scheduler.add_job(restart_space, "interval", seconds=21600) scheduler.start() # Launch the interface demo.queue(default_concurrency_limit=1, max_size=5).launch(debug=True, show_api=False)