Updated Colab to use internal downloader

fixed model command (+1 squashed commits)

Squashed commits:

[a4d8fd9f1] tryout new colab (+1 squashed commits)

Squashed commits:

[c97333d44] tryout new colab

colab.ipynb

@@ -53,6 +53,7 @@
     "ContextSize = 4096 #@param [4096,8192] {allow-input: true}\n",
     "FlashAttention = True #@param {type:\"boolean\"}\n",
     "Multiplayer = False #@param {type:\"boolean\"}\n",
+    "DeleteExistingModels = False #@param {type:\"boolean\"}\n",
     "FACommand = \"\"\n",
     "MPCommand = \"\"\n",
     "#@markdown <hr>\n",
@@ -84,7 +85,7 @@
     "#@markdown Only select the following box if regular cloudflare tunnel fails to work. It will generate an inferior localtunnel tunnel, which you can use after entering a password.\n",
     "MakeLocalTunnelFallback = False #@param {type:\"boolean\"}\n",
     "\n",
-    "import os\n",
+    "import os, glob\n",
     "if not os.path.isfile(\"/opt/bin/nvidia-smi\"):\n",
     "  raise RuntimeError(\"⚠️Colab did not give you a GPU due to usage limits, this can take a few hours before they let you back in. Check out https://lite.koboldai.net for a free alternative (that does not provide an API link but can load KoboldAI saves and chat cards) or subscribe to Colab Pro for immediate access.⚠️\")\n",
     "\n",
@@ -113,23 +114,23 @@
     "\n",
     "%cd /content\n",
     "if Mmproj and LoadVisionMMProjector:\n",
-    "  VCommand = \"--mmproj vmodel.gguf\"\n",
+    "  VCommand = f\"--mmproj {Mmproj}\"\n",
     "else:\n",
     "  SCommand = \"\"\n",
     "if ImgModel and LoadImgModel:\n",
-    "  SCommand = \"--sdmodel imodel.gguf --sdthreads 4 --sdquant --sdclamped\"\n",
+    "  SCommand = f\"--sdmodel {ImgModel} --sdthreads 4 --sdquant --sdclamped\"\n",
     "else:\n",
     "  SCommand = \"\"\n",
     "if SpeechModel and LoadSpeechModel:\n",
-    "  WCommand = \"--whispermodel wmodel.bin\"\n",
+    "  WCommand = f\"--whispermodel {SpeechModel}\"\n",
     "else:\n",
     "  WCommand = \"\"\n",
     "if TTSModel and WavTokModel and LoadTTSModel:\n",
-    "  TTSCommand = \"--ttsmodel ttsmodel.bin --ttswavtokenizer ttswavtok.bin --ttsgpu\"\n",
+    "  TTSCommand = f\"--ttsmodel {TTSModel} --ttswavtokenizer {WavTokModel} --ttsgpu\"\n",
     "else:\n",
     "  TTSCommand = \"\"\n",
     "if EmbeddingsModel and LoadEmbeddingsModel:\n",
-    "  ECommand = \"--embeddingsmodel emodel.bin\"\n",
+    "  ECommand = f\"--embeddingsmodel {EmbeddingsModel}\"\n",
     "else:\n",
     "  ECommand = \"\"\n",
     "if FlashAttention:\n",
@@ -141,27 +142,26 @@
     "else:\n",
     "  MPCommand = \"\"\n",
     "\n",
+    "ModelCommand = f\"--model {Model}\"\n",
+    "\n",
+    "if DeleteExistingModels:\n",
+    "  print(\"Deleting all cached models to redownload...\")\n",
+    "  patterns = ['*.gguf', '*.bin']\n",
+    "  for pattern in patterns:\n",
+    "      for file_path in glob.glob(pattern):\n",
+    "          try:\n",
+    "              os.remove(file_path)\n",
+    "              print(f\"Deleted: {file_path}\")\n",
+    "          except Exception as e:\n",
+    "              print(f\"Failed to delete {file_path}: {e}\")\n",
+    "\n",
+    "\n",
     "!echo Downloading KoboldCpp, please wait...\n",
     "!wget -O dlfile.tmp https://kcpplinux.concedo.workers.dev && mv dlfile.tmp koboldcpp_linux\n",
     "!test -f koboldcpp_linux && echo Download Successful || echo Download Failed\n",
     "!chmod +x ./koboldcpp_linux\n",
     "!apt update\n",
     "!apt install aria2 -y\n",
-    "# simple fix for a common URL mistake\n",
-    "if \"https://huggingface.co/\" in Model and \"/blob/main/\" in Model:\n",
-    "    Model = Model.replace(\"/blob/main/\", \"/resolve/main/\")\n",
-    "!aria2c -x 10 -o model.gguf --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $Model\n",
-    "if VCommand:\n",
-    "  !aria2c -x 10 -o vmodel.gguf --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $Mmproj\n",
-    "if SCommand:\n",
-    "  !aria2c -x 10 -o imodel.gguf --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $ImgModel\n",
-    "if WCommand:\n",
-    "  !aria2c -x 10 -o wmodel.bin --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $SpeechModel\n",
-    "if TTSCommand:\n",
-    "  !aria2c -x 10 -o ttsmodel.bin --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $TTSModel\n",
-    "  !aria2c -x 10 -o ttswavtok.bin --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $WavTokModel\n",
-    "if ECommand:\n",
-    "  !aria2c -x 10 -o emodel.bin --summary-interval=5 --download-result=default --allow-overwrite=true --file-allocation=none $EmbeddingsModel\n",
     "\n",
     "if MakeLocalTunnelFallback:\n",
     "  import urllib\n",
@@ -175,7 +175,7 @@
     "  print(f\"Please open the above link, and input the password '{ltpw}'\\nYour KoboldCpp will start shortly...\")\n",
     "  print(\"=================\")\n",
     "  !sleep 10\n",
-    "!./koboldcpp_linux model.gguf --usecublas 0 mmq --chatcompletionsadapter AutoGuess --multiuser --gpulayers $Layers --contextsize $ContextSize --websearch --quiet --remotetunnel $FACommand $MPCommand $VCommand $SCommand $WCommand $TTSCommand $ECommand $SavGdriveCommand\n"
+    "!./koboldcpp_linux $ModelCommand --usecublas 0 mmq --chatcompletionsadapter AutoGuess --multiuser --gpulayers $Layers --contextsize $ContextSize --websearch --quiet --remotetunnel $FACommand $MPCommand $VCommand $SCommand $WCommand $TTSCommand $ECommand $SavGdriveCommand\n"
    ]
   }
  ],