Linux support for BF16 (#36)

- linux support gguf-imat-lossless-for-BF16-linux.py (9e58050ed7fc995caff97ad81f63f8c7dc99416d)


Co-authored-by: Virt <[email protected]>

gguf-imat-lossless-for-BF16-linux.py

@@ -0,0 +1,207 @@
+import os
+from posix import system
+import requests
+import zipfile
+import subprocess
+import shutil
+import platform
+from huggingface_hub import snapshot_download
+
+def clone_or_update_llama_cpp():
+    print("Preparing...")
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+    os.chdir(base_dir)
+    if not os.path.exists("llama.cpp"):
+        subprocess.run(["git", "clone", "--depth", "1", "https://github.com/ggerganov/llama.cpp"])
+    else:
+        os.chdir("llama.cpp")
+        subprocess.run(["git", "pull"])
+    os.chdir(base_dir)
+    print("The 'llama.cpp' repository is ready.")
+
+def build_or_rebuild_llama_cpp():
+    answer = input("Build or Rebuild? yes/no\n")
+    answer = answer.strip().lower()
+    if answer == "yes":
+        os.chdir("llama.cpp")
+        subprocess.run(["make", "clean"])
+        subprocess.run(["make", "-j", "8", "GGML_CUDA=1"])
+    elif answer == "no":
+        print("No action")
+
+def download_llama_release():
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+    dl_dir = os.path.join(base_dir, "bin", "dl")
+    if not os.path.exists(dl_dir):
+        os.makedirs(dl_dir)
+
+    os.chdir(dl_dir)
+    latest_release_url = "https://github.com/ggerganov/llama.cpp/releases/latest"
+    response = requests.get(latest_release_url)
+    if response.status_code == 200:
+        latest_release_tag = response.url.split("/")[-1]
+        download_url = f"https://github.com/ggerganov/llama.cpp/releases/download/{latest_release_tag}/llama-{latest_release_tag}-bin-win-cuda-cu12.2.0-x64.zip"
+        response = requests.get(download_url)
+        if response.status_code == 200:
+            with open(f"llama-{latest_release_tag}-bin-win-cuda-cu12.2.0-x64.zip", "wb") as f:
+                f.write(response.content)
+            with zipfile.ZipFile(f"llama-{latest_release_tag}-bin-win-cuda-cu12.2.0-x64.zip", "r") as zip_ref:
+                zip_ref.extractall(os.path.join(base_dir, "bin"))
+            print("Downloading latest 'llama.cpp' prebuilt Windows binaries...")
+            print("Download and extraction completed successfully.")
+            return latest_release_tag
+        else:
+            print("Failed to download the release file.")
+    else:
+        print("Failed to fetch the latest release information.")
+
+def download_cudart_if_necessary(latest_release_tag):
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+    cudart_dl_dir = os.path.join(base_dir, "bin", "dl")
+    if not os.path.exists(cudart_dl_dir):
+        os.makedirs(cudart_dl_dir)
+
+    cudart_zip_file = os.path.join(cudart_dl_dir, "cudart-llama-bin-win-cu12.2.0-x64.zip")
+    cudart_extracted_files = ["cublas64_12.dll", "cublasLt64_12.dll", "cudart64_12.dll"]
+
+    if all(os.path.exists(os.path.join(base_dir, "bin", file)) for file in cudart_extracted_files):
+        print("Cuda resources already exist. Skipping download.")
+    else:
+        cudart_download_url = f"https://github.com/ggerganov/llama.cpp/releases/download/{latest_release_tag}/cudart-llama-bin-win-cu12.2.0-x64.zip"
+        response = requests.get(cudart_download_url)
+        if response.status_code == 200:
+            with open(cudart_zip_file, "wb") as f:
+                f.write(response.content)
+            with zipfile.ZipFile(cudart_zip_file, "r") as zip_ref:
+                zip_ref.extractall(os.path.join(base_dir, "bin"))
+            print("Preparing 'cuda' resources...")
+            print("Download and extraction of cudart completed successfully.")
+        else:
+            print("Failed to download the cudart release file.")
+
+def download_model_repo():
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+    models_dir = os.path.join(base_dir, "models")
+
+    if not os.path.exists(models_dir):
+        os.makedirs(models_dir)
+
+    model_id = input("Enter the model ID to download (e.g., huggingface/transformers): ")
+    model_name = model_id.split("/")[-1]
+    model_dir = os.path.join(models_dir, model_name)
+
+    gguf_dir = os.path.join(base_dir, "models", f"{model_name}-GGUF")
+    gguf_model_path = os.path.join(gguf_dir, f"{model_name}-F16.gguf")
+    imatrix_file_name = input("Enter the name of the imatrix.txt file (default: imatrix.txt): ").strip() or "imatrix.txt"
+    delete_model_dir = input("Remove HF model folder after converting original model to GGUF? (yes/no) (default: no): ").strip().lower()
+
+    if os.path.exists(gguf_model_path):
+        create_imatrix(base_dir, gguf_dir, gguf_model_path, model_name, imatrix_file_name)
+    else:
+        if os.path.exists(model_dir):
+            print("Model repository already exists. Using existing repository.")
+
+            convert_model_to_gguf_bf16(base_dir, model_dir, model_name, delete_model_dir, imatrix_file_name)
+
+        else:
+            revision = input("Enter the revision (branch, tag, or commit) to download (default: main): ") or "main"
+
+            print("Downloading model repository...")
+            snapshot_download(repo_id=model_id, local_dir=model_dir, revision=revision)
+            print("Model repository downloaded successfully.")
+
+            convert_model_to_gguf_bf16(base_dir, model_dir, model_name, delete_model_dir, imatrix_file_name)
+
+def convert_model_to_gguf_bf16(base_dir, model_dir, model_name, delete_model_dir, imatrix_file_name):
+    convert_script = os.path.join(base_dir, "llama.cpp", "convert_hf_to_gguf.py")
+    gguf_dir = os.path.join(base_dir, "models", f"{model_name}-GGUF")
+    gguf_model_path = os.path.join(gguf_dir, f"{model_name}-BF16.gguf")
+
+    if not os.path.exists(gguf_dir):
+        os.makedirs(gguf_dir)
+
+    if not os.path.exists(gguf_model_path):
+        subprocess.run(["python", convert_script, model_dir, "--outfile", gguf_model_path, "--outtype", "bf16"])
+
+    convert_model_to_gguf_f16(base_dir, model_dir, model_name, delete_model_dir, imatrix_file_name)
+
+def convert_model_to_gguf_f16(base_dir, model_dir, model_name, delete_model_dir, imatrix_file_name):
+    convert_script = os.path.join(base_dir, "llama.cpp", "convert_hf_to_gguf.py")
+    gguf_dir = os.path.join(base_dir, "models", f"{model_name}-GGUF")
+    gguf_model_path = os.path.join(gguf_dir, f"{model_name}-F16.gguf")
+
+    if not os.path.exists(gguf_dir):
+        os.makedirs(gguf_dir)
+
+    if not os.path.exists(gguf_model_path):
+        subprocess.run(["python", convert_script, model_dir, "--outfile", gguf_model_path, "--outtype", "f16"])
+
+        if delete_model_dir == 'yes' or delete_model_dir == 'y':
+            shutil.rmtree(model_dir)
+            print(f"Original model directory '{model_dir}' deleted.")
+        else:
+            print(f"Original model directory '{model_dir}' was not deleted. You can remove it manually.")
+
+    create_imatrix(base_dir, gguf_dir, gguf_model_path, model_name, imatrix_file_name)
+
+def create_imatrix(base_dir, gguf_dir, gguf_model_path, model_name, imatrix_file_name):
+    if platform.system() == "Linux":
+        imatrix_exe = os.path.join(base_dir, "llama.cpp", "llama-imatrix")
+    elif platform.system() == "Windows":
+        imatrix_exe = os.path.join(base_dir, "bin", "llama-imatrix.exe")
+    imatrix_output_src = os.path.join(gguf_dir, "imatrix.dat")
+    imatrix_output_dst = os.path.join(gguf_dir, "imatrix.dat")
+    if not os.path.exists(imatrix_output_dst):
+        try:
+            subprocess.run([imatrix_exe, "-m", gguf_model_path, "-f", os.path.join(base_dir, "imatrix", imatrix_file_name), "-ngl", "7"], cwd=gguf_dir)
+            shutil.move(imatrix_output_src, imatrix_output_dst)
+            print("imatrix.dat moved successfully.")
+        except Exception as e:
+            print("Error occurred while moving imatrix.dat:", e)
+    else:
+        print("imatrix.dat already exists in the GGUF folder.")
+
+    quantize_models(base_dir, model_name)
+
+def quantize_models(base_dir, model_name):
+    gguf_dir = os.path.join(base_dir, "models", f"{model_name}-GGUF")
+    bf16_gguf_path = os.path.join(gguf_dir, f"{model_name}-BF16.gguf")
+
+    quantization_options = [
+        "IQ3_M", "IQ3_XXS",
+        "Q4_K_M", "Q4_K_S", "IQ4_XS",
+        "Q5_K_M", "Q5_K_S",
+        "Q6_K",
+        "Q8_0"
+    ]
+
+    for quant_option in quantization_options:
+        quantized_gguf_name = f"{model_name}-{quant_option}-imat.gguf"
+        quantized_gguf_path = os.path.join(gguf_dir, quantized_gguf_name)
+        if platform.system() == "Linux":
+            quantize_command = os.path.join(base_dir, "llama.cpp", "llama-quantize")
+        elif platform.system() == "Windows":
+            quantize_command = os.path.join(base_dir, "bin", "llama-quantize.exe")
+        imatrix_path = os.path.join(gguf_dir, "imatrix.dat")
+
+        subprocess.run([quantize_command, "--imatrix", imatrix_path,
+                        bf16_gguf_path, quantized_gguf_path, quant_option], cwd=gguf_dir)
+        print(f"Model quantized with {quant_option} option.")
+
+
+if platform.system() == "Linux":
+    def main():
+        clone_or_update_llama_cpp()
+        build_or_rebuild_llama_cpp()
+        download_model_repo()
+        print("Finished preparing resources.")
+elif platform.system() == "Windows":
+    def main():
+        clone_or_update_llama_cpp()
+        latest_release_tag = download_llama_release()
+        download_cudart_if_necessary(latest_release_tag)
+        download_model_repo()
+        print("Finished preparing resources.")
+
+if __name__ == "__main__":
+    main()