theluyuan/RWKV-Runner
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

WebGPU (Python) strategy

Browse Source
This commit is contained in:
josc146 2023-12-14 20:39:42 +08:00
parent 2f45e9c33a
commit 18d4b2304e
2 changed files with 10 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
backend-python/rwkv_pip/webgpu/model.py vendored
View File

@ -12,8 +12,13 @@ except ModuleNotFoundError:
class RWKV: class RWKV:
def __init__(self, model_path: str, strategy=None): def __init__(self, model_path: str, strategy: str = None):
self.model = wrp.v5.Model(model_path, turbo=False) self.model = wrp.v5.Model(
model_path,
turbo=False,
quant=32 if "i8" in strategy else None,
quant_nf4=26 if "i4" in strategy else None,
)
self.w = {} # fake weight self.w = {} # fake weight
self.w["emb.weight"] = [0] * wrp.peek_info(model_path).num_vocab self.w["emb.weight"] = [0] * wrp.peek_info(model_path).num_vocab

6
frontend/src/pages/Configs.tsx
View File

@ -282,7 +282,7 @@ const Configs: FC = observer(() => {
selectedConfig.modelParameters.device !== 'Custom' && <Labeled label={t('Precision')} selectedConfig.modelParameters.device !== 'Custom' && <Labeled label={t('Precision')}
desc={t('int8 uses less VRAM, but has slightly lower quality. fp16 has higher quality.')} desc={t('int8 uses less VRAM, but has slightly lower quality. fp16 has higher quality.')}
content={ content={
<Dropdown disabled={selectedConfig.modelParameters.device === 'WebGPU (Python)'} <Dropdown
style={{ minWidth: 0 }} className="grow" style={{ minWidth: 0 }} className="grow"
value={selectedConfig.modelParameters.precision} value={selectedConfig.modelParameters.precision}
selectedOptions={[selectedConfig.modelParameters.precision]} selectedOptions={[selectedConfig.modelParameters.precision]}
@ -296,8 +296,8 @@ const Configs: FC = observer(() => {
{selectedConfig.modelParameters.device !== 'CPU' && selectedConfig.modelParameters.device !== 'MPS' && {selectedConfig.modelParameters.device !== 'CPU' && selectedConfig.modelParameters.device !== 'MPS' &&
<Option>fp16</Option>} <Option>fp16</Option>}
{selectedConfig.modelParameters.device !== 'CPU (rwkv.cpp)' && <Option>int8</Option>} {selectedConfig.modelParameters.device !== 'CPU (rwkv.cpp)' && <Option>int8</Option>}
{selectedConfig.modelParameters.device === 'WebGPU' && <Option>nf4</Option>} {selectedConfig.modelParameters.device.startsWith('WebGPU') && <Option>nf4</Option>}
{selectedConfig.modelParameters.device !== 'CPU (rwkv.cpp)' && selectedConfig.modelParameters.device !== 'WebGPU' && {selectedConfig.modelParameters.device !== 'CPU (rwkv.cpp)' && !selectedConfig.modelParameters.device.startsWith('WebGPU') &&
<Option>fp32</Option>} <Option>fp32</Option>}
{selectedConfig.modelParameters.device === 'CPU (rwkv.cpp)' && <Option>Q5_1</Option>} {selectedConfig.modelParameters.device === 'CPU (rwkv.cpp)' && <Option>Q5_1</Option>}
</Dropdown> </Dropdown>