-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathfastchat.sh
45 lines (39 loc) · 1.99 KB
/
fastchat.sh
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
log_dir="your log_dir"
python3 -m fastchat.serve.controller --host localhost --port 23241 >$log_dir/controller.log 2>$log_dir/controller.err &
sleep 10
CUDA_VISIBLE_DEVICES=0 python3 -m fastchat.serve.multi_model_worker \
--model-path ./LLMs/Xwin-LM-13B-V0.1 \
--model-names Xwin-LM-13B-V0.1 \
--model-path ./LLMs/Mistral-7B-Instruct-v0.2 \
--model-names Mistral \
--worker-address http://localhost:23251 \
--controller-address http://localhost:23241 \
--host localhost --port 23251 >$log_dir/1.log 2>$log_dir/1.err &
CUDA_VISIBLE_DEVICES=1 python3 -m fastchat.serve.multi_model_worker \
--model-path ./LLMs/WizardLM-13B-V1.2 \
--model-names WizardLM-13B-V1.2 \
--model-path ./LLMs/zephyr-7b-beta \
--model-names zephyr-7b-beta \
--worker-address http://localhost:23252 \
--controller-address http://localhost:23241 \
--host localhost --port 23252 >$log_dir/2.log 2>$log_dir/2.err &
CUDA_VISIBLE_DEVICES=2 python3 -m fastchat.serve.model_worker \
--model-path ./LLMs/vicuna-33b-v1.3 \
--model-name vicuna-33b-v1.3 \
--worker-address http://localhost:23353 \
--controller-address http://localhost:23241 \
--host localhost --port 23353 >$log_dir/3.log 2>$log_dir/3.err &
CUDA_VISIBLE_DEVICES=3 python3 -m fastchat.serve.model_worker \
--model-path ./LLMs/bge-large-en-v1.5 \
--model-name bge-large-en-v1.5 \
--worker-address http://localhost:23253 \
--controller-address http://localhost:23241 \
--host localhost --port 23253 >$log_dir/4.log 2>$log_dir/4.err &
CUDA_VISIBLE_DEVICES=3 python3 -m fastchat.serve.model_worker \
--model-path ./LLMs/Yi-34B-Chat \
--model-name Yi-34B-Chat \
--worker-address http://localhost:23354 \
--controller-address http://localhost:23241 \
--host localhost --port 23354 >$log_dir/5.log 2>$log_dir/5.err &
# After the above command runs, run the following command
# python3 -m fastchat.serve.openai_api_server --host localhost --port 23261 --controller-address http://localhost:23241