jiangzhs/OpenCompass
1
0
Fork 0
mirror of https://github.com/open-compass/opencompass.git synced 2025-05-30 16:03:24 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Feature] Update the max_out_len for many models (#1559)

Browse Source
This commit is contained in:
Songyang Zhang 2024-09-24 21:52:28 +08:00 committed by GitHub
parent a2e9bc0c41
commit e7681943f3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
74 changed files with 420 additions and 186 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
configs/models/chatglm/lmdeploy_glm4_9b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='glm-4-9b-chat-turbomind',
path='THUDM/glm-4-9b-chat',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
configs/models/deepseek/lmdeploy_deepseek_67b_chat.py
View File

@ -7,8 +7,8 @@ models = [
path='deepseek-ai/deepseek-llm-67b-chat',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9),
max_seq_len=7168,
max_out_len=1024,
max_seq_len=8192,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

4
configs/models/deepseek/lmdeploy_deepseek_7b_chat.py
View File

@ -7,8 +7,8 @@ models = [
path='deepseek-ai/deepseek-llm-7b-chat',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9),
max_seq_len=7168,
max_out_len=1024,
max_seq_len=8192,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_5_1_8b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-1_8b-chat-turbomind',
path='internlm/internlm2_5-1_8b-chat',
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=2048,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_5_20b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-20b-chat-turbomind',
path='internlm/internlm2_5-20b-chat',
engine_config=dict(session_len=8192, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=2048,
engine_config=dict(session_len=16384, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_5_7b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-7b-chat-turbomind',
path='internlm/internlm2_5-7b-chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_chat_1_8b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-1.8b-turbomind',
path='internlm/internlm2-chat-1_8b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_chat_20b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-20b-turbomind',
path='internlm/internlm2-chat-20b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

8
configs/models/hf_internlm/lmdeploy_internlm2_chat_7b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-7b-turbomind',
path='internlm/internlm2-chat-7b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
configs/models/hf_internlm/lmdeploy_internlm_chat_20b.py
View File

@ -6,9 +6,9 @@ models = [
abbr='internlm-chat-20b-turbomind',
path='internlm/internlm-chat-20b',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

4
configs/models/hf_internlm/lmdeploy_internlm_chat_7b.py
View File

@ -6,9 +6,9 @@ models = [
abbr='internlm-chat-7b-turbomind',
path='internlm/internlm-chat-7b',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

2
configs/models/hf_llama/hf_llama3_1_70b_instruct.py
View File

@ -5,7 +5,7 @@ models = [
type=HuggingFacewithChatTemplate,
abbr='llama-3_1-70b-instruct-hf',
path='meta-llama/Meta-Llama-3.1-70B-Instruct',
max_out_len=1024,
max_out_len=4096,
batch_size=8,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

12
configs/models/hf_llama/hf_llama3_1_8b.py Normal file
View File

@ -0,0 +1,12 @@
from opencompass.models import HuggingFaceBaseModel
models = [
dict(
type=HuggingFaceBaseModel,
abbr='llama-3_1-8b-hf',
path='meta-llama/Meta-Llama-3.1-8B-Instruct',
max_out_len=1024,
batch_size=8,
run_cfg=dict(num_gpus=1),
)
]

4
configs/models/hf_llama/lmdeploy_llama2_13b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-13b-chat-turbomind',
path='meta-llama/Llama-2-13b-chat-hf',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
configs/models/hf_llama/lmdeploy_llama2_70b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-70b-chat-turbomind',
path='meta-llama/Llama-2-70b-chat-hf',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

4
configs/models/hf_llama/lmdeploy_llama2_7b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-7b-chat-turbomind',
path='meta-llama/Llama-2-7b-chat-hf',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

6
configs/models/hf_llama/lmdeploy_llama3_1_70b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3_1-70b-instruct-turbomind',
path='meta-llama/Meta-Llama-3.1-70B-Instruct',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
configs/models/hf_llama/lmdeploy_llama3_1_8b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3_1-8b-instruct-turbomind',
path='meta-llama/Meta-Llama-3.1-8B-Instruct',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
configs/models/hf_llama/lmdeploy_llama3_70b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3-70b-instruct-turbomind',
path='meta-llama/Meta-Llama-3-70B-Instruct',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
configs/models/hf_llama/lmdeploy_llama3_8b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3-8b-instruct-turbomind',
path='meta-llama/Meta-Llama-3-8B-Instruct',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

15
configs/models/mistral/lmdeploy_mistral_7b_instruct_v0_3.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='mistral-7b-instruct-v0.3-turbomind',
path='mistralai/Mistral-7B-Instruct-v0.3',
engine_config=dict(session_len=32768, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=32768,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
configs/models/mistral/lmdeploy_mixtral_large_instruct_2407.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='mixtral-large-instruct-2407-turbomind',
path='mistralai/Mistral-Large-Instruct-2407',
engine_config=dict(session_len=32768, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=32768,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
)
]

8
configs/models/qwen/lmdeploy_qwen1_5_110b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-110b-chat-turbomind',
path='Qwen/Qwen1.5-110B-Chat',
engine_config=dict(session_len=7168, max_batch_size=8, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16834, max_batch_size=8, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16834,
max_out_len=4096,
batch_size=8,
run_cfg=dict(num_gpus=4),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_14b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-14b-chat-turbomind',
path='Qwen/Qwen1.5-14B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_1_8b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-1.8b-chat-turbomind',
path='Qwen/Qwen1.5-1.8B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_32b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-32b-chat-turbomind',
path='Qwen/Qwen1.5-32B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_4b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-4b-chat-turbomind',
path='Qwen/Qwen1.5-4B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_72b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-72b-chat-turbomind',
path='Qwen/Qwen1.5-72B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen1_5_7b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-7b-chat-turbomind',
path='Qwen/Qwen1.5-7B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
configs/models/qwen/lmdeploy_qwen2_1_5b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-1.5b-instruct-turbomind',
path='Qwen/Qwen2-1.5B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
configs/models/qwen/lmdeploy_qwen2_72b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-72b-instruct-turbomind',
path='Qwen/Qwen2-72B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

8
configs/models/qwen/lmdeploy_qwen2_7b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-7b-instruct-turbomind',
path='Qwen/Qwen2-7B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

15
configs/models/yi/lmdeploy_yi_1_5_34b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-34b-chat-turbomind',
path='01-ai/Yi-1.5-34B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)
]

15
configs/models/yi/lmdeploy_yi_1_5_6b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-6b-chat-turbomind',
path='01-ai/Yi-1.5-6B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
configs/models/yi/lmdeploy_yi_1_5_9b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-9b-chat-turbomind',
path='01-ai/Yi-1.5-9B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
configs/models/yi/lmdeploy_yi_34b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-34b-chat-turbomind',
path='01-ai/Yi-34B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)
]

15
configs/models/yi/lmdeploy_yi_6b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-6b-chat-turbomind',
path='01-ai/Yi-6B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

4
opencompass/configs/models/chatglm/lmdeploy_glm4_9b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='glm-4-9b-chat-turbomind',
path='THUDM/glm-4-9b-chat',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
opencompass/configs/models/deepseek/lmdeploy_deepseek_67b_chat.py
View File

@ -7,8 +7,8 @@ models = [
path='deepseek-ai/deepseek-llm-67b-chat',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9),
max_seq_len=7168,
max_out_len=1024,
max_seq_len=8192,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

4
opencompass/configs/models/deepseek/lmdeploy_deepseek_7b_chat.py
View File

@ -7,8 +7,8 @@ models = [
path='deepseek-ai/deepseek-llm-7b-chat',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9),
max_seq_len=7168,
max_out_len=1024,
max_seq_len=8192,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_5_1_8b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-1_8b-chat-turbomind',
path='internlm/internlm2_5-1_8b-chat',
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=2048,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_5_20b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-20b-chat-turbomind',
path='internlm/internlm2_5-20b-chat',
engine_config=dict(session_len=8192, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=8192,
max_out_len=2048,
engine_config=dict(session_len=16384, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_5_7b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2_5-7b-chat-turbomind',
path='internlm/internlm2_5-7b-chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_chat_1_8b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-1.8b-turbomind',
path='internlm/internlm2-chat-1_8b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_chat_20b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-20b-turbomind',
path='internlm/internlm2-chat-20b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

8
opencompass/configs/models/hf_internlm/lmdeploy_internlm2_chat_7b.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='internlm2-chat-7b-turbomind',
path='internlm/internlm2-chat-7b',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=8192, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=8192,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
opencompass/configs/models/hf_internlm/lmdeploy_internlm_chat_20b.py
View File

@ -6,9 +6,9 @@ models = [
abbr='internlm-chat-20b-turbomind',
path='internlm/internlm-chat-20b',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)

4
opencompass/configs/models/hf_internlm/lmdeploy_internlm_chat_7b.py
View File

@ -6,9 +6,9 @@ models = [
abbr='internlm-chat-7b-turbomind',
path='internlm/internlm-chat-7b',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

2
opencompass/configs/models/hf_llama/hf_llama3_1_70b_instruct.py
View File

@ -5,7 +5,7 @@ models = [
type=HuggingFacewithChatTemplate,
abbr='llama-3_1-70b-instruct-hf',
path='meta-llama/Meta-Llama-3.1-70B-Instruct',
max_out_len=1024,
max_out_len=4096,
batch_size=8,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

12
opencompass/configs/models/hf_llama/hf_llama3_1_8b.py Normal file
View File

@ -0,0 +1,12 @@
from opencompass.models import HuggingFaceBaseModel
models = [
dict(
type=HuggingFaceBaseModel,
abbr='llama-3_1-8b-hf',
path='meta-llama/Meta-Llama-3.1-8B-Instruct',
max_out_len=1024,
batch_size=8,
run_cfg=dict(num_gpus=1),
)
]

4
opencompass/configs/models/hf_llama/lmdeploy_llama2_13b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-13b-chat-turbomind',
path='meta-llama/Llama-2-13b-chat-hf',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

4
opencompass/configs/models/hf_llama/lmdeploy_llama2_70b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-70b-chat-turbomind',
path='meta-llama/Llama-2-70b-chat-hf',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

4
opencompass/configs/models/hf_llama/lmdeploy_llama2_7b_chat.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-2-7b-chat-turbomind',
path='meta-llama/Llama-2-7b-chat-hf',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=1024,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

6
opencompass/configs/models/hf_llama/lmdeploy_llama3_1_70b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3_1-70b-instruct-turbomind',
path='meta-llama/Meta-Llama-3.1-70B-Instruct',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
opencompass/configs/models/hf_llama/lmdeploy_llama3_1_8b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3_1-8b-instruct-turbomind',
path='meta-llama/Meta-Llama-3.1-8B-Instruct',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
opencompass/configs/models/hf_llama/lmdeploy_llama3_70b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3-70b-instruct-turbomind',
path='meta-llama/Meta-Llama-3-70B-Instruct',
engine_config=dict(max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

6
opencompass/configs/models/hf_llama/lmdeploy_llama3_8b_instruct.py
View File

@ -6,9 +6,9 @@ models = [
abbr='llama-3-8b-instruct-turbomind',
path='meta-llama/Meta-Llama-3-8B-Instruct',
engine_config=dict(max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|end_of_text|>', '<|eot_id|>'],

15
opencompass/configs/models/mistral/lmdeploy_mistral_7b_instruct_v0_3.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='mistral-7b-instruct-v0.3-turbomind',
path='mistralai/Mistral-7B-Instruct-v0.3',
engine_config=dict(session_len=32768, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=32768,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
opencompass/configs/models/mistral/lmdeploy_mixtral_large_instruct_2407.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='mixtral-large-instruct-2407-turbomind',
path='mistralai/Mistral-Large-Instruct-2407',
engine_config=dict(session_len=32768, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=32768,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
)
]

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_110b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-110b-chat-turbomind',
path='Qwen/Qwen1.5-110B-Chat',
engine_config=dict(session_len=7168, max_batch_size=8, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16834, max_batch_size=8, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16834,
max_out_len=4096,
batch_size=8,
run_cfg=dict(num_gpus=4),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_14b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-14b-chat-turbomind',
path='Qwen/Qwen1.5-14B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_1_8b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-1.8b-chat-turbomind',
path='Qwen/Qwen1.5-1.8B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_32b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-32b-chat-turbomind',
path='Qwen/Qwen1.5-32B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=2),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_4b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-4b-chat-turbomind',
path='Qwen/Qwen1.5-4B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_72b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-72b-chat-turbomind',
path='Qwen/Qwen1.5-72B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen1_5_7b_chat.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen1.5-7b-chat-turbomind',
path='Qwen/Qwen1.5-7B-Chat',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
stop_words=['<|im_end|>', '<|im_start|>'],

8
opencompass/configs/models/qwen/lmdeploy_qwen2_1_5b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-1.5b-instruct-turbomind',
path='Qwen/Qwen2-1.5B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

8
opencompass/configs/models/qwen/lmdeploy_qwen2_72b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-72b-instruct-turbomind',
path='Qwen/Qwen2-72B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=4),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=4),
)

8
opencompass/configs/models/qwen/lmdeploy_qwen2_7b_instruct.py
View File

@ -5,10 +5,10 @@ models = [
type=TurboMindModelwithChatTemplate,
abbr='qwen2-7b-instruct-turbomind',
path='Qwen/Qwen2-7B-Instruct',
engine_config=dict(session_len=7168, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=1024),
max_seq_len=7168,
max_out_len=1024,
engine_config=dict(session_len=16384, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=4096),
max_seq_len=16384,
max_out_len=4096,
batch_size=16,
run_cfg=dict(num_gpus=1),
)

15
opencompass/configs/models/yi/lmdeploy_yi_1_5_34b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-34b-chat-turbomind',
path='01-ai/Yi-1.5-34B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)
]

15
opencompass/configs/models/yi/lmdeploy_yi_1_5_6b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-6b-chat-turbomind',
path='01-ai/Yi-1.5-6B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
opencompass/configs/models/yi/lmdeploy_yi_1_5_9b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-1.5-9b-chat-turbomind',
path='01-ai/Yi-1.5-9B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]

15
opencompass/configs/models/yi/lmdeploy_yi_34b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-34b-chat-turbomind',
path='01-ai/Yi-34B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=2),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=2),
)
]

15
opencompass/configs/models/yi/lmdeploy_yi_6b_chat.py Normal file
View File

@ -0,0 +1,15 @@
from opencompass.models import TurboMindModelwithChatTemplate
models = [
dict(
type=TurboMindModelwithChatTemplate,
abbr='yi-6b-chat-turbomind',
path='01-ai/Yi-6B-Chat',
engine_config=dict(session_len=4096, max_batch_size=16, tp=1),
gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=2048),
max_seq_len=4096,
max_out_len=2048,
batch_size=16,
run_cfg=dict(num_gpus=1),
)
]