mmarch/text-generation-webui
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

added no_mmap & mlock parameters to llama.cpp and removed llamacpp_model_alternative (#1649)

Browse source 
---------

Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>
This commit is contained in:
Ahmed Said 2023-05-03 00:25:28 +03:00 • committed by GitHub
parent 2f1a2846d1
commit fbcd32988e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
5 changed files with 50 additions and 126 deletions
Download patch file Download diff file Expand all files Collapse all files

6
modules/shared.py
View file

@ -120,8 +120,10 @@ parser.add_argument('--sdp-attention', action='store_true', help="Use torch 2.0'
parser.add_argument('--trust-remote-code', action='store_true', help="Set trust_remote_code=True while loading a model. Necessary for ChatGLM.")
# llama.cpp
parser.add_argument('--threads', type=int, default=0, help='Number of threads to use in llama.cpp.')
parser.add_argument('--n_batch', type=int, default=8, help='Processing batch size for llama.cpp.')
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('--mlock', action='store_true', help='Force the system to keep the model in RAM.')
# GPTQ
parser.add_argument('--wbits', type=int, default=0, help='Load a pre-quantized model with specified precision in bits. 2, 3, 4 and 8 are supported.')