 draff
|
28fd4fc970
Change wording to be consistent with other args
|
vor 2 Jahren |
|
draff
|
804486214b
Re-implement --load-in-4bit and update --llama-bits arg description
|
vor 2 Jahren |
|
draff
|
e6c631aea4
Replace --load-in-4bit with --llama-bits
|
vor 2 Jahren |
 oobabooga
|
026d60bd34
Remove default preset that didn't do anything
|
vor 2 Jahren |
|
oobabooga
|
828a524f9a
Add LLaMA 4-bit support
|
vor 2 Jahren |
|
oobabooga
|
153dfeb4dd
Add --rwkv-cuda-on parameter, bump rwkv version
|
vor 2 Jahren |
|
oobabooga
|
bf56b6c1fb
Load settings.json without the need for --settings settings.json
|
vor 2 Jahren |
|
oobabooga
|
c33715ad5b
Move towards HF LLaMA implementation
|
vor 2 Jahren |
|
oobabooga
|
ed8b35efd2
Add --pin-weight parameter for FlexGen
|
vor 2 Jahren |
|
oobabooga
|
ea5c5eb3da
Add LLaMA support
|
vor 2 Jahren |
|
oobabooga
|
169209805d
Model-aware prompts and presets
|
vor 2 Jahren |
|
oobabooga
|
a2a3e8f797
Add --rwkv-strategy parameter
|
vor 2 Jahren |
|
oobabooga
|
f3da6dcc8f
Merge pull request #149 from oobabooga/RWKV
|
vor 2 Jahren |
|
oobabooga
|
a1429d1607
Add default extensions to the settings
|
vor 2 Jahren |
|
oobabooga
|
ebc64a408c
RWKV support prototype
|
vor 2 Jahren |
|
oobabooga
|
7c2babfe39
Rename greed to "generation attempts"
|
vor 2 Jahren |
|
oobabooga
|
7a527a5581
Move "send picture" into an extension
|
vor 2 Jahren |
|
oobabooga
|
65326b545a
Move all gradio elements to shared (so that extensions can use them)
|
vor 2 Jahren |
|
oobabooga
|
f6f792363b
Separate command-line params by spaces instead of commas
|
vor 2 Jahren |
 luis
|
5abdc99a7c
gpu-memory arg change
|
vor 2 Jahren |
|
oobabooga
|
b78561fba6
Minor bug fix
|
vor 2 Jahren |
|
oobabooga
|
2e86a1ec04
Move chat history into shared module
|
vor 2 Jahren |
|
oobabooga
|
e46c43afa6
Move some stuff from server.py to modules
|
vor 2 Jahren |
|
oobabooga
|
1dacd34165
Further refactor
|
vor 2 Jahren |
|
oobabooga
|
ce7feb3641
Further refactor
|
vor 2 Jahren |
|
oobabooga
|
98af4bfb0d
Refactor the code to make it more modular
|
vor 2 Jahren |
