From 6b0756506b75d03afc8aa2f4535800c91dff9528 Mon Sep 17 00:00:00 2001 From: Concedo <39025047+LostRuins@users.noreply.github.com> Date: Thu, 4 Jul 2024 15:34:08 +0800 Subject: [PATCH] improvements to model downloader and chat completions adapter loader --- colab.ipynb | 24 +++++++----------------- koboldcpp.py | 30 ++++++++++++++++++++++-------- 2 files changed, 29 insertions(+), 25 deletions(-) diff --git a/colab.ipynb b/colab.ipynb index c19c47434..b87209aa2 100644 --- a/colab.ipynb +++ b/colab.ipynb @@ -48,12 +48,12 @@ "source": [ "#@title v-- Enter your model below and then click this to start Koboldcpp\r\n", "\r\n", - "Model = \"https://huggingface.co/KoboldAI/LLaMA2-13B-Tiefighter-GGUF/resolve/main/LLaMA2-13B-Tiefighter.Q4_K_S.gguf\" #@param [\"https://huggingface.co/KoboldAI/LLaMA2-13B-Tiefighter-GGUF/resolve/main/LLaMA2-13B-Tiefighter.Q4_K_S.gguf\",\"https://huggingface.co/Sao10K/Fimbulvetr-11B-v2-GGUF/resolve/main/Fimbulvetr-11B-v2-Test-14.q4_K_M.gguf\",\"https://huggingface.co/TheBloke/MythoMax-L2-13B-GGUF/resolve/main/mythomax-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/ReMM-SLERP-L2-13B-GGUF/resolve/main/remm-slerp-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/Xwin-LM-13B-v0.2-GGUF/resolve/main/xwin-lm-13b-v0.2.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/Stheno-L2-13B-GGUF/resolve/main/stheno-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/MythoMax-L2-Kimiko-v2-13B-GGUF/resolve/main/mythomax-l2-kimiko-v2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/airoboros-mistral2.2-7B-GGUF/resolve/main/airoboros-mistral2.2-7b.Q4_K_S.gguf\",\"https://huggingface.co/concedo/KobbleTinyV2-1.1B-GGUF/resolve/main/KobbleTiny-Q4_K.gguf\",\"https://huggingface.co/grimjim/kukulemon-7B-GGUF/resolve/main/kukulemon-7B.Q8_0.gguf\"]{allow-input: true}\r\n", + "Model = \"https://huggingface.co/KoboldAI/LLaMA2-13B-Tiefighter-GGUF/resolve/main/LLaMA2-13B-Tiefighter.Q4_K_S.gguf\" #@param [\"https://huggingface.co/KoboldAI/LLaMA2-13B-Tiefighter-GGUF/resolve/main/LLaMA2-13B-Tiefighter.Q4_K_S.gguf\",\"https://huggingface.co/KoboldAI/LLaMA2-13B-Estopia-GGUF/resolve/main/LLaMA2-13B-Estopia.Q4_K_S.gguf\",\"https://huggingface.co/Sao10K/Fimbulvetr-11B-v2-GGUF/resolve/main/Fimbulvetr-11B-v2-Test-14.q4_K_M.gguf\",\"https://huggingface.co/TheBloke/MythoMax-L2-13B-GGUF/resolve/main/mythomax-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/ReMM-SLERP-L2-13B-GGUF/resolve/main/remm-slerp-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/Xwin-LM-13B-v0.2-GGUF/resolve/main/xwin-lm-13b-v0.2.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/Stheno-L2-13B-GGUF/resolve/main/stheno-l2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/MythoMax-L2-Kimiko-v2-13B-GGUF/resolve/main/mythomax-l2-kimiko-v2-13b.Q4_K_M.gguf\",\"https://huggingface.co/TheBloke/airoboros-mistral2.2-7B-GGUF/resolve/main/airoboros-mistral2.2-7b.Q4_K_S.gguf\",\"https://huggingface.co/concedo/KobbleTinyV2-1.1B-GGUF/resolve/main/KobbleTiny-Q4_K.gguf\",\"https://huggingface.co/grimjim/kukulemon-7B-GGUF/resolve/main/kukulemon-7B.Q8_0.gguf\",\"https://huggingface.co/mradermacher/LemonKunoichiWizardV3-GGUF/resolve/main/LemonKunoichiWizardV3.Q4_K_M.gguf\",\"https://huggingface.co/Lewdiculous/Kunoichi-DPO-v2-7B-GGUF-Imatrix/resolve/main/Kunoichi-DPO-v2-7B-Q4_K_M-imatrix.gguf\",\"https://huggingface.co/mradermacher/L3-8B-Stheno-v3.2-i1-GGUF/resolve/main/L3-8B-Stheno-v3.2.i1-Q4_K_M.gguf\",\"https://huggingface.co/Lewdiculous/Llama-3-Lumimaid-8B-v0.1-OAS-GGUF-IQ-Imatrix/resolve/main/v2-Llama-3-Lumimaid-8B-v0.1-OAS-Q4_K_M-imat.gguf\",\"https://huggingface.co/bartowski/NeuralDaredevil-8B-abliterated-GGUF/resolve/main/NeuralDaredevil-8B-abliterated-Q4_K_M.gguf\",\"https://huggingface.co/bartowski/L3-8B-Lunaris-v1-GGUF/resolve/main/L3-8B-Lunaris-v1-Q4_K_M.gguf\",\"https://huggingface.co/mradermacher/L3-Umbral-Mind-RP-v2.0-8B-GGUF/resolve/main/L3-Umbral-Mind-RP-v2.0-8B.Q4_K_M.gguf\"]{allow-input: true}\r\n", "Layers = 99 #@param [99]{allow-input: true}\r\n", "ContextSize = 4096 #@param [4096] {allow-input: true}\r\n", "#@markdown
\r\n", "LoadLLaVAmmproj = False #@param {type:\"boolean\"}\r\n", - "LLaVAmmproj = \"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-13b-mmproj-v1.5.Q4_1.gguf\" #@param [\"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-13b-mmproj-v1.5.Q4_1.gguf\",\"https://huggingface.co/koboldcpp/mmproj/resolve/main/mistral-7b-mmproj-v1.5-Q4_1.gguf\",\"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-7b-mmproj-v1.5-Q4_0.gguf\"]{allow-input: true}\r\n", + "LLaVAmmproj = \"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-13b-mmproj-v1.5.Q4_1.gguf\" #@param [\"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-13b-mmproj-v1.5.Q4_1.gguf\",\"https://huggingface.co/koboldcpp/mmproj/resolve/main/mistral-7b-mmproj-v1.5-Q4_1.gguf\",\"https://huggingface.co/koboldcpp/mmproj/resolve/main/llama-7b-mmproj-v1.5-Q4_0.gguf\",\"https://huggingface.co/koboldcpp/mmproj/resolve/main/LLaMA3-8B_mmproj-Q4_1.gguf\"]{allow-input: true}\r\n", "VCommand = \"\"\r\n", "#@markdown
\r\n", "LoadImgModel = False #@param {type:\"boolean\"}\r\n", @@ -87,6 +87,9 @@ "!chmod +x ./koboldcpp_linux\r\n", "!apt update\r\n", "!apt install aria2 -y\r\n", + "# simple fix for a common URL mistake\r\n", + "if \"https://huggingface.co/\" in Model and \"/blob/main/\" in Model: \r\n", + " Model = Model.replace(\"/blob/main/\", \"/resolve/main/\")\r\n", "!aria2c -x 10 -o model.gguf --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $Model\r\n", "if VCommand:\r\n", " !aria2c -x 10 -o vmodel.gguf --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $LLaVAmmproj\r\n", @@ -99,21 +102,8 @@ } ], "metadata": { - "accelerator": "GPU", - "colab": { - "cell_execution_strategy": "setup", - "gpuType": "T4", - "include_colab_link": true, - "private_outputs": true, - "provenance": [] - }, - "kernelspec": { - "display_name": "Python 3", - "name": "python3" - }, - "language_info": { - "name": "python" - } + "language_info": {}, + "orig_nbformat": 3 }, "nbformat": 4, "nbformat_minor": 0 diff --git a/koboldcpp.py b/koboldcpp.py index 90e44f8c4..f86881a3f 100644 --- a/koboldcpp.py +++ b/koboldcpp.py @@ -3302,16 +3302,27 @@ def main(launch_args,start_server=True): # try to read chat completions adapter if args.chatcompletionsadapter: + ccadapter_path = None + adapt_dir = os.path.join(os.path.abspath(os.path.dirname(__file__)), 'kcpp_adapters') + adapt_dir = adapt_dir if os.path.isdir(adapt_dir) else None if isinstance(args.chatcompletionsadapter, str) and os.path.exists(args.chatcompletionsadapter): - print(f"Loading Chat Completions Adapter...") - with open(args.chatcompletionsadapter, 'r') as f: + ccadapter_path = os.path.abspath(args.chatcompletionsadapter) + elif isinstance(args.chatcompletionsadapter, str) and adapt_dir: + filename = args.chatcompletionsadapter + if not filename.endswith(".json"): + filename += ".json" + premade_adapt_path = os.path.join(adapt_dir,filename) + if os.path.exists(premade_adapt_path): + ccadapter_path = os.path.abspath(premade_adapt_path) + if ccadapter_path: + print(f"Loading Chat Completions Adapter: {ccadapter_path}") + with open(ccadapter_path, 'r') as f: global chatcompl_adapter chatcompl_adapter = json.load(f) print(f"Chat Completions Adapter Loaded") else: print(f"Warning: Chat Completions Adapter {args.chatcompletionsadapter} invalid or not found.") - if args.model_param and args.model_param!="": if args.model_param.endswith("?download=true"): args.model_param = args.model_param.replace("?download=true","") @@ -3320,13 +3331,16 @@ def main(launch_args,start_server=True): mdlfilename = os.path.basename(args.model_param) #check if file already exists if mdlfilename: - if not os.path.exists(mdlfilename): - print(f"Downloading model from external URL at {args.model_param}") - subprocess.run(f"curl -fL {args.model_param} -o {mdlfilename}", shell=True, capture_output=True, text=True, check=True, encoding='utf-8') - print(f"Download {mdlfilename} completed...", flush=True) + if os.path.exists(mdlfilename) and os.path.getsize(mdlfilename) > 10000000: #10MB trigger + print(f"Model file {mdlfilename} already exists, not redownloading.") args.model_param = mdlfilename else: - print(f"Model file {mdlfilename} already exists, not redownloading.") + dl_url = args.model_param + if "https://huggingface.co/" in dl_url and "/blob/main/" in dl_url: + dl_url = dl_url.replace("/blob/main/", "/resolve/main/") + print(f"Downloading model from external URL at {dl_url}") + subprocess.run(f"curl -fL {dl_url} -o {mdlfilename}", shell=True, capture_output=True, text=True, check=True, encoding='utf-8') + print(f"Download {mdlfilename} completed...", flush=True) args.model_param = mdlfilename # sanitize and replace the default vanity name. remember me....