[Sync] Merge branch 'dev' into zfz/update-keyset-demo (#876)

configs/models/qwen/hf_qwen2_beta_0_5b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-0.5b-hf',
+        path="Qwen/Qwen2-beta-0_5B",
+        tokenizer_path='Qwen/Qwen2-beta-0_5B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_0_5b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-0.5b-chat-hf',
+        path="Qwen/Qwen2-beta-0_5B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen2_beta_14b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-14b-hf',
+        path="Qwen/Qwen2-beta-14B",
+        tokenizer_path='Qwen/Qwen2-beta-14B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_14b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-14b-chat-hf',
+        path="Qwen/Qwen2-beta-14B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen2_beta_1_8b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-1.8b-hf',
+        path="Qwen/Qwen2-beta-1_8B",
+        tokenizer_path='Qwen/Qwen2-beta-1_8B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_1_8b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-1.8b-chat-hf',
+        path="Qwen/Qwen2-beta-1_8B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen2_beta_4b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-4b-hf',
+        path="Qwen/Qwen2-beta-4B",
+        tokenizer_path='Qwen/Qwen2-beta-4B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_4b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-4b-chat-hf',
+        path="Qwen/Qwen2-beta-4B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen2_beta_72b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-72b-hf',
+        path="Qwen/Qwen2-beta-72B",
+        tokenizer_path='Qwen/Qwen2-beta-72B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_72b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-72b-chat-hf',
+        path="Qwen/Qwen2-beta-72B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=4, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen2_beta_7b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-7b-hf',
+        path="Qwen/Qwen2-beta-7B",
+        tokenizer_path='Qwen/Qwen2-beta-7B',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen2_beta_7b_chat.py

+from opencompass.models import HuggingFaceCausalLM
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen2-beta-7b-chat-hf',
+        path="Qwen/Qwen2-beta-7B-Chat",
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+            use_fast=False,
+        ),
+        meta_template=_meta_template,
+        pad_token_id=151645,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
+    )
+]

configs/models/qwen/hf_qwen_14b.py

@@ -17,6 +17,7 @@ models = [
             use_fast=False,
         ),
         pad_token_id=151643,
+        min_out_len=1,
         max_out_len=100,
         max_seq_len=2048,
         batch_size=8,

configs/models/qwen/hf_qwen_1_8b.py

@@ -17,6 +17,7 @@ models = [
             use_fast=False,
         ),
         pad_token_id=151643,
+        min_out_len=1,
         max_out_len=100,
         max_seq_len=2048,
         batch_size=8,

configs/models/qwen/hf_qwen_1_8b_chat.py

@@ -5,6 +5,7 @@ _meta_template = dict(
         dict(role="HUMAN", begin='\n<|im_start|>user\n', end='<|im_end|>'),
         dict(role="BOT", begin="\n<|im_start|>assistant\n", end='<|im_end|>', generate=True),
     ],
+    eos_token_id=151645,
 )
 
 models = [
@@ -21,12 +22,14 @@ models = [
             padding_side='left',
             truncation_side='left',
             trust_remote_code=True,
-            use_fast=False,),
+            use_fast=False,
+        ),
         pad_token_id=151643,
         max_out_len=100,
         max_seq_len=2048,
         batch_size=8,
         meta_template=_meta_template,
         run_cfg=dict(num_gpus=1, num_procs=1),
+        end_str='<|im_end|>',
     )
 ]

configs/models/qwen/hf_qwen_72b.py

@@ -17,6 +17,7 @@ models = [
             use_fast=False,
         ),
         pad_token_id=151643,
+        min_out_len=1,
         max_out_len=100,
         max_seq_len=2048,
         batch_size=8,

configs/models/qwen/hf_qwen_7b.py

@@ -17,6 +17,7 @@ models = [
             use_fast=False,
         ),
         pad_token_id=151643,
+        min_out_len=1,
         max_out_len=100,
         max_seq_len=2048,
         batch_size=8,

configs/models/qwen/vllm_qwen2_beta_72b.py

+from opencompass.models import VLLM
+
+
+models = [
+    dict(
+        type=VLLM,
+        abbr='qwen2-beta-72b-vllm',
+        path="Qwen/Qwen2-beta-72B",
+        model_kwargs=dict(tensor_parallel_size=4),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=32,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )
+]

configs/models/qwen/vllm_qwen2_beta_72b_chat.py

+from opencompass.models import VLLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role="HUMAN", begin='<|im_start|>user\n', end='<|im_end|>\n'),
+        dict(role="BOT", begin="<|im_start|>assistant\n", end='<|im_end|>\n', generate=True),
+    ],
+    eos_token_id=151645,
+)
+
+models = [
+    dict(
+        type=VLLM,
+        abbr='qwen2-beta-72b-chat-vllm',
+        path="Qwen/Qwen2-beta-72B-Chat",
+        model_kwargs=dict(tensor_parallel_size=4),
+        meta_template=_meta_template,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=32,
+        generation_kwargs=dict(temperature=0),
+        end_str='<|im_end|>',
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )
+]

configs/models/yi/hf_yi_6b_200k.py

 from opencompass.models import HuggingFace
 
-
 models = [
     dict(
         type=HuggingFace,