-
Notifications
You must be signed in to change notification settings - Fork 0
/
Dockerfile
63 lines (49 loc) · 3.12 KB
/
Dockerfile
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
FROM nvcr.io/nvidia/pytorch:24.12-py3
RUN chown root:root /usr/lib
RUN apt update -y && apt install -y build-essential curl openssh-server openssh-client pdsh tmux
RUN pip install --upgrade pip wheel
RUN pip install \
accelerate \
deepspeed \
openai \
mistral_common \
msgspec \
peft \
pyarrow \
sentencepiece \
tiktoken \
transformers \
trl
RUN pip install stanford-stk --no-deps
RUN pip install \
aioprometheus \
fastapi \
fschat[model_worker,webui] \
gguf \
lm-format-enforcer \
llmcompressor \
outlines \
partial_json_parser \
prometheus-fastapi-instrumentator \
ray==2.34.0 \
typer \
uvicorn[standard]
RUN pip uninstall -y pynvml
RUN pip install nvidia-ml-py
RUN mkdir /packages/
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/flash_attn-2.7.2.post1-cp312-cp312-linux_aarch64.whl /packages/flash_attn-2.7.2.post1-cp312-cp312-linux_aarch64.whl
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/flash_attn-2.7.2.post1-cp312-cp312-linux_x86_64.whl /packages/flash_attn-2.7.2.post1-cp312-cp312-linux_x86_64.whl
RUN pip install --no-deps --no-index --upgrade --find-links /packages flash-attn
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/xformers-0.0.30%2B46a02df6.d20250103-cp312-cp312-linux_aarch64.whl /packages/xformers-0.0.30+46a02df6.d20250103-cp312-cp312-linux_aarch64.whl
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/xformers-0.0.30%2B46a02df6.d20250103-cp312-cp312-linux_x86_64.whl /packages/xformers-0.0.30+46a02df6.d20250103-cp312-cp312-linux_x86_64.whl
RUN pip install --no-deps --no-index --find-links /packages xformers
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/megablocks-0.7.0-cp312-cp312-linux_aarch64.whl /packages/megablocks-0.7.0-cp312-cp312-linux_aarch64.whl
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/megablocks-0.7.0-cp312-cp312-linux_x86_64.whl /packages/megablocks-0.7.0-cp312-cp312-linux_x86_64.whl
RUN pip install --no-deps --no-index --find-links /packages megablocks
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/bitsandbytes-0.45.1.dev0-cp312-cp312-linux_aarch64.whl /packages/bitsandbytes-0.45.1.dev0-cp312-cp312-linux_aarch64.whl
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/bitsandbytes-0.45.1.dev0-cp312-cp312-linux_x86_64.whl /packages/bitsandbytes-0.45.1.dev0-cp312-cp312-linux_x86_64.whl
RUN pip install --no-deps --no-index --find-links /packages bitsandbytes
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/vllm-0.6.6.post1%2Bcu126-cp312-cp312-linux_aarch64.whl /packages/vllm-0.6.6.post1+cu126-cp312-cp312-linux_aarch64.whl
ADD https://static.abacus.ai/pypi/abacusai/gh200-llm/pytorch-2412-cuda126/vllm-0.6.6.post1%2Bcu126-cp312-cp312-linux_x86_64.whl /packages/vllm-0.6.6.post1+cu126-cp312-cp312-linux_x86_64.whl
RUN pip install --no-deps --no-index --find-links /packages vllm
RUN rm -r /packages