diff --git a/latent.ipynb b/latent.ipynb
index f691d4e..03c6eba 100644
--- a/latent.ipynb
+++ b/latent.ipynb
@@ -3,7 +3,6 @@
{
"cell_type": "markdown",
"metadata": {
- "colab_type": "text",
"id": "view-in-github"
},
"source": [
@@ -58,6 +57,32 @@
"#@markdown Changelog 1.6 - ViT-L conditioning for latenet diffusion, adds noising and scaling during advanced scheduling phases, fixes linear ETA, adss LAION models"
]
},
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "fKgwAFDyLeNW"
+ },
+ "source": [
+ "## Check GPU "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "cellView": "form",
+ "id": "IK9tMRnfLXLF"
+ },
+ "outputs": [],
+ "source": [
+ "#@title Check GPU Status\n",
+ "import subprocess\n",
+ "\n",
+ "nvidiasmi_output = subprocess.run(['nvidia-smi'], stdout=subprocess.PIPE).stdout.decode('utf-8')\n",
+ "print('\\n'.join(nvidiasmi_output.splitlines()[1:3]))\n",
+ "print('\\n'.join(nvidiasmi_output.splitlines()[7:12]))"
+ ]
+ },
{
"cell_type": "markdown",
"metadata": {
@@ -196,7 +221,6 @@
"cell_type": "code",
"execution_count": null,
"metadata": {
- "cellView": "form",
"id": "cNHvQBhzyXCI"
},
"outputs": [],
@@ -267,31 +291,10 @@
"!cp $model_path/GFPGANv1.3.pth GFPGAN/experiments/pretrained_models/GFPGANv1.3.pth\n"
]
},
- {
- "cell_type": "markdown",
- "metadata": {
- "id": "ThxmCePqt1mt"
- },
- "source": [
- "Let's also check what type of GPU we've got."
- ]
- },
{
"cell_type": "code",
"execution_count": null,
"metadata": {
- "id": "jbL2zJ7Pt7Jl"
- },
- "outputs": [],
- "source": [
- "!nvidia-smi"
- ]
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "cellView": "form",
"id": "BPnyd-XUKbfE"
},
"outputs": [],
@@ -380,12 +383,12 @@
" sd = torch.load(ckpt, map_location=\"cuda\")[\"state_dict\"]\n",
" m, u = model.load_state_dict(sd, strict = False)\n",
" \n",
- " if(latent_diffusion_model == \"finetuned\"): \n",
+ " elif(latent_diffusion_model == \"finetuned\"): \n",
" sd = torch.load(f\"{model_path}/txt2img-f8-large-jack000-finetuned-fp16.ckpt\",map_location=\"cuda\")\n",
" m, u = model.load_state_dict(sd, strict = False)\n",
" #model.model = model.model.half().eval().to(device)\n",
" \n",
- " if(latent_diffusion_model == \"ongo (fine tuned in art)\"):\n",
+ " elif(latent_diffusion_model == \"ongo (fine tuned in art)\"):\n",
" del sd \n",
" sd_finetuned = torch.load(f\"{model_path}/ongo.pt\")\n",
" sd_finetuned[\"input_blocks.0.0.weight\"] = sd_finetuned[\"input_blocks.0.0.weight\"][:,0:4,:,:]\n",
@@ -394,7 +397,7 @@
" torch.cuda.empty_cache()\n",
" gc.collect()\n",
"\n",
- " if(latent_diffusion_model == \"erlich (fine tuned in logos)\"):\n",
+ " elif(latent_diffusion_model == \"erlich (fine tuned in logos)\"):\n",
" del sd \n",
" sd_finetuned = torch.load(f\"{model_path}/erlich.pt\")\n",
" sd_finetuned[\"input_blocks.0.0.weight\"] = sd_finetuned[\"input_blocks.0.0.weight\"][:,0:4,:,:]\n",
@@ -404,11 +407,9 @@
" gc.collect()\n",
"\n",
" if len(m) > 0 and verbose:\n",
- " print(\"missing keys:\")\n",
- " print(m)\n",
+ " print(f\"missing keys:\\n{m}\")\n",
" if len(u) > 0 and verbose:\n",
- " print(\"unexpected keys:\")\n",
- " print(u)\n",
+ " print(f\"unexpected keys:\\n{u}\")\n",
"\n",
" model.requires_grad_(False).half().eval().to('cuda')\n",
" return model\n",
@@ -424,7 +425,6 @@
"cell_type": "code",
"execution_count": null,
"metadata": {
- "cellView": "form",
"id": "HY_7vvnPThzS"
},
"outputs": [],
@@ -665,8 +665,7 @@
" if grad_center: grad = centralized_grad(grad, use_gc=True, gc_conv_only=False)\n",
" mag = grad.square().mean().sqrt()\n",
" if mag==0 or torch.isnan(mag):\n",
- " print(\"ERROR\")\n",
- " print(t)\n",
+ " print(f\"ERROR\\n{t}\")\n",
" return(grad)\n",
" if t>=0:\n",
" if active_function == \"softsign\":\n",
@@ -1060,7 +1059,7 @@
" score_corrector = score_corrector,\n",
" corrector_kwargs = score_corrector_setting,\n",
" x0_adjust_fn = dynamic_thresholding,\n",
- " clip_embed = target_embeds[\"ViT-L-14--openai\"] if \"ViT-L-14--openai\" in clip_list else None\n",
+ " clip_embed = target_embeds[\"ViT-L-14--openai\"].mean(0, keepdim = True) if \"ViT-L-14--openai\" in clip_list else None\n",
" )\n",
" #x_T = samples_ddim.clamp(-6,6)\n",
" x_T = samples_ddim\n",
@@ -1336,10 +1335,17 @@
"use_cond_fn = True\n",
"\n",
"#Custom cut schedules and super-resolution. Check out the guide on how to use it a https://multimodal.art/majestydiffusion\n",
+ "#Maximum step is 25, larger - create artifacts\n",
"custom_schedule_setting = [\n",
- " [50,1000,8],\n",
+ " [200,1000,8],\n",
" \"gfpgan:1.5\",\"scale:.9\",\"noise:.55\",\n",
- " [50,200,5],\n",
+ " [1,200,5],\n",
+ "]\n",
+ "\n",
+ "#For init image\n",
+ "custom_schedule_setting_init = [\n",
+ " [500,1000,10],\n",
+ " [1,400,8],\n",
"]\n",
" \n",
"#Cut settings\n",
@@ -1448,7 +1454,7 @@
"#Negative prompts for Latent Diffusion\n",
"latent_negatives = [\"\"]\n",
"\n",
- "image_prompts = []"
+ "image_prompts = [\"\"]"
]
},
{
@@ -1485,16 +1491,19 @@
"augment_cuts=True #@param{type:'boolean'}\n",
"\n",
"#@markdown\n",
- "\n",
+ "def NoneToNone(par): #fix for None(str)\n",
+ " if(par == '' or par == 'none' or par == 'None'):\n",
+ " par = None\n",
+ " return par\n",
"#@markdown ### Init image settings\n",
"#@markdown `init_image` requires the path of an image to use as init to the model\n",
"init_image = None #@param{type: 'string'}\n",
- "if(init_image == '' or init_image == 'None'):\n",
- " init_image = None\n",
+ "init_image = NoneToNone(init_image)\n",
"#@markdown `starting_timestep`: How much noise do you want to add to your init image for it to then be difused by the model\n",
"starting_timestep = 0.9 #@param{type: 'number'}\n",
"#@markdown `init_mask` is a mask same width and height as the original image with the color black indicating where to inpaint\n",
"init_mask = None #@param{type: 'string'}\n",
+ "init_mask = NoneToNone(init_mask)\n",
"#@markdown `init_scale` controls how much the init image should influence the final result. Experiment with values around `1000`\n",
"init_scale = 1000 #@param{type: 'integer'}\n",
"init_brightness = 0.0 #@param{type: 'number'}\n",
@@ -1504,7 +1513,7 @@
"#@markdown\n",
"\n",
"#@markdown ### Custom saved settings\n",
- "#@markdown If you choose custom saved settings, the settings set by the preset overrule some of your choices. You can still modify the settings not in the preset. Check what each preset modifies here\n",
+ "#@markdown If you choose custom saved settings, the settings set by the preset overrule some of your choices. You can still modify the settings not in the preset. Check what each preset modifies here\n",
"custom_settings = 'path/to/settings.cfg' #@param{type:'string'}\n",
"settings_library = 'None (use settings defined above)' #@param [\"None (use settings defined above)\", \"default\", \"defaults_v1_3\", \"dango233_princesses\", \"the_other_zippy_defaults\", \"makeitrad_defaults\"]\n",
"if(settings_library != 'None (use settings defined above)'):\n",
@@ -1542,13 +1551,22 @@
" for advanced_setting in advanced_settings:\n",
" global_var_scope[advanced_setting[0]] = eval(advanced_setting[1])\n",
"\n",
- "if(((init_image is not None) and (init_image != 'None') and (init_image != '')) and starting_timestep != 1 and custom_schedule_setting[0][1] == 1000):\n",
- " custom_schedule_setting[0] = [custom_schedule_setting[0][0], int(custom_schedule_setting[0][1]*starting_timestep), custom_schedule_setting[0][2]]\n",
- "\n",
+ "if((init_image is not None) and starting_timestep != 1 and custom_schedule_setting[0][1] == 1000):\n",
+ " custom_schedules = custom_schedule_setting_init[:] #Use cfg for init_image\n",
+ " if int(custom_schedules[0][1]*starting_timestep) <= custom_schedules[0][0]:\n",
+ " print(f\"Due to the fact that starting_timestep is too small, we skip the first step.\")\n",
+ " if len(custom_schedules) <= 1:\n",
+ " print(\"There are no more steps left, increase starting_timestep or add additional steps!\")\n",
+ " exit(0)\n",
+ " del custom_schedules[0]\n",
+ " else:\n",
+ " custom_schedules[0] = [custom_schedules[0][0], int(custom_schedules[0][1]*starting_timestep), custom_schedules[0][2]]\n",
+ " custom_schedules = set_custom_schedules(custom_schedules)\n",
+ "else:\n",
+ " custom_schedules = set_custom_schedules(custom_schedule_setting)\n",
"prompts = clip_prompts\n",
"opt.prompt = latent_prompts\n",
"opt.uc = latent_negatives\n",
- "custom_schedules = set_custom_schedules(custom_schedule_setting)\n",
"aes_scale = aesthetic_loss_scale\n",
"try: \n",
" clip_guidance_schedule\n",
@@ -1564,6 +1582,10 @@
" opt.W = (width//64)*64;\n",
" opt.H = (height//64)*64;\n",
" if opt.W != width or opt.H != height:\n",
+ " if opt.W == 0:\n",
+ " opt.W= 64\n",
+ " if opt.H == 0:\n",
+ " opt.H = 64\n",
" print(f'Changing output size to {opt.W}x{opt.H}. Dimensions must by multiples of 64.')\n",
"\n",
" opt.mag_mul = opt_mag_mul \n",
@@ -1671,14 +1693,11 @@
"accelerator": "GPU",
"colab": {
"collapsed_sections": [
- "xEVSOJ4f0B21",
- "VpR9JhyCu5iq",
- "N_Di3xFSXGWe",
- "xEVSOJ4f0B21",
- "WOAs3ZvLlktt"
+ "WOAs3ZvLlktt",
+ "xEVSOJ4f0B21"
],
"machine_shape": "hm",
- "name": "Latent Majesty Diffusion v1.6",
+ "name": "Latent Majesty Diffusion v1.6.ipynb",
"private_outputs": true,
"provenance": []
},