huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-11 04:14:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: adjust typos and docker build

Browse Source
This commit is contained in:
drbh 2024-02-07 03:24:28 +00:00
parent 5b30a425f6
commit deed8e8154
2 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
Dockerfile
View File

@ -154,17 +154,17 @@ COPY server/Makefile-vllm Makefile
# Build specific version of vllm
RUN make build-vllm-cuda
# Build megablocks
FROM kernel-builder as megablocks-builder
RUN pip install git+https://github.com/OlivierDehaene/megablocks@181709df192de9a941fdf3a641cdc65a0462996e
# Build mamba kernels
FROM kernel-builder as mamba-builder
WORKDIR /usr/src
COPY server/Makefile-selective-scan Makefile
RUN make build-all
# Build megablocks
FROM kernel-builder as megablocks-builder
RUN pip install git+https://github.com/OlivierDehaene/megablocks@181709df192de9a941fdf3a641cdc65a0462996e
# Text Generation Inference base image
FROM nvidia/cuda:12.1.0-base-ubuntu20.04 as base

2
server/text_generation_server/models/custom_modeling/mamba_modeling.py
View File

@ -137,7 +137,7 @@ class MambaBlock(nn.Module):
z = _z[i:i+1, -1, :]
x_db = self.x_proj(x)
dt, B, C = torch.split(x_db, [self.dt_rank, self.d_state, self.d_state], dim=-1)
df = self.dt_proj_no_bias(x)
dt = F.linear(dt, self.dt_proj.weight)
y = selective_state_update(
ssm_state[i:i+1,:,:], x, dt, self.negA, B, C, self.D, z=z, dt_bias=self.dt_proj.bias, dt_softplus=True
)