Merge remote-tracking branch 'refs/remotes/origin/dev' into dev

modules/llamacpp_model.py

@@ -49,6 +49,7 @@ class LlamaCppModel:
             'n_batch': shared.args.n_batch,
             'use_mmap': not shared.args.no_mmap,
             'use_mlock': shared.args.mlock,
+            'low_vram': shared.args.low_vram,
             'n_gpu_layers': shared.args.n_gpu_layers
         }
 

modules/loaders.py

@@ -34,6 +34,7 @@ loaders_and_params = {
         'n_batch',
         'threads',
         'no_mmap',
+        'low_vram',
         'mlock',
         'llama_cpp_seed',
     ],

modules/shared.py

@@ -120,6 +120,7 @@ parser.add_argument('--use_double_quant', action='store_true', help='use_double_
 parser.add_argument('--threads', type=int, default=0, help='Number of threads to use.')
 parser.add_argument('--n_batch', type=int, default=512, help='Maximum number of prompt tokens to batch together when calling llama_eval.')
 parser.add_argument('--no-mmap', action='store_true', help='Prevent mmap from being used.')
+parser.add_argument('--low-vram', action='store_true', help='Low VRAM Mode')
 parser.add_argument('--mlock', action='store_true', help='Force the system to keep the model in RAM.')
 parser.add_argument('--cache-capacity', type=str, help='Maximum cache capacity. Examples: 2000MiB, 2GiB. When provided without units, bytes will be assumed.')
 parser.add_argument('--n-gpu-layers', type=int, default=0, help='Number of layers to offload to the GPU.')

modules/ui.py

@@ -57,6 +57,7 @@ def list_model_elements():
         'threads',
         'n_batch',
         'no_mmap',
+        'low_vram',
         'mlock',
         'n_gpu_layers',
         'n_ctx',