Add files using upload-large-folder tool

build/lib/opencompass/configs/models/judge_llm/auto_j/hf_autoj_bilingual_6b.py

@@ -0,0 +1,24 @@
+from opencompass.models import HuggingFaceCausalLM
+
+'''
+This is a bilingual 6B version of Auto-J.
+It is trained on both the original training data
+and its Chinese translation, which can be find in
+https://huggingface.co/GAIR/autoj-bilingual-6b
+'''
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='autoj-bilingual-6b',
+        path='GAIR/autoj-bilingual-6b',
+        tokenizer_path='GAIR/autoj-bilingual-6b',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/auto_j/hf_autoj_eng_13b.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='autoj-13b',
+        path='GAIR/autoj-13b',
+        tokenizer_path='GAIR/autoj-13b',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/auto_j/hf_autoj_eng_13b_4bit.py

@@ -0,0 +1,23 @@
+from opencompass.models import HuggingFaceCausalLM
+
+'''
+#This is a 4bits quantized version of Auto-J by using AutoGPTQ,
+which is available on huggingface-hub:
+https://huggingface.co/GAIR/autoj-13b-GPTQ-4bits
+'''
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='autoj-13b-GPTQ-4bits',
+        path='GAIR/autoj-13b-GPTQ-4bits',
+        tokenizer_path='GAIR/autoj-13b-GPTQ-4bits',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/auto_j/hf_autoj_scen_classifier.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='autoj-scenario-classifier',
+        path='GAIR/autoj-scenario-classifier',
+        tokenizer_path='GAIR/autoj-scenario-classifier',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/judgelm/hf_judgelm_13b_v1.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='judgelm-13b-v1-hf',
+        path='BAAI/JudgeLM-13B-v1.0',
+        tokenizer_path='BAAI/JudgeLM-13B-v1.0',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/judgelm/hf_judgelm_33b_v1.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='judgelm-33b-v1-hf',
+        path='BAAI/JudgeLM-33B-v1.0',
+        tokenizer_path='BAAI/JudgeLM-33B-v1.0',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/judgelm/hf_judgelm_7b_v1.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='judgelm-7b-v1-hf',
+        path='BAAI/JudgeLM-7B-v1.0',
+        tokenizer_path='BAAI/JudgeLM-7B-v1.0',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/pandalm/hf_alpaca_pandalm_7b_v1.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='alpaca-pandalm-7b-v1-hf',
+        path='WeOpenML/PandaLM-Alpaca-7B-v1',
+        tokenizer_path='WeOpenML/PandaLM-Alpaca-7B-v1',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/judge_llm/pandalm/hf_pandalm_7b_v1.py

@@ -0,0 +1,18 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [dict(
+        type=HuggingFaceCausalLM,
+        abbr='pandalm-7b-v1-hf',
+        path='WeOpenML/PandaLM-7B-v1',
+        tokenizer_path='WeOpenML/PandaLM-7B-v1',
+        tokenizer_kwargs=dict(padding_side='left',
+                              truncation_side='left',
+                              trust_remote_code=True,
+                              use_fast=False,),
+        max_out_len=1024,
+        max_seq_len=4096,
+        batch_size=8,
+        model_kwargs=dict(device_map='auto', trust_remote_code=True),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )]

build/lib/opencompass/configs/models/qwen2_5/hf_qwen_2_5_32b.py

@@ -0,0 +1,12 @@
+from opencompass.models import HuggingFaceBaseModel
+
+models = [
+    dict(
+        type=HuggingFaceBaseModel,
+        abbr='qwen2.5-32b-hf',
+        path='Qwen/Qwen2.5-32B',
+        max_out_len=1024,
+        batch_size=8,
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/hf_qwen_2_5_7b.py

@@ -0,0 +1,12 @@
+from opencompass.models import HuggingFaceBaseModel
+
+models = [
+    dict(
+        type=HuggingFaceBaseModel,
+        abbr='qwen2.5-7b-hf',
+        path='Qwen/Qwen2.5-7B',
+        max_out_len=1024,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_0_5b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-0.5b-instruct-turbomind',
+        path='Qwen/Qwen2.5-0.5B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_14b.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModel
+
+models = [
+    dict(
+        type=TurboMindModel,
+        abbr='qwen2.5-14b-turbomind',
+        path='Qwen/Qwen2.5-14B',
+        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,
+        batch_size=16,
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_14b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-14b-instruct-turbomind',
+        path='Qwen/Qwen2.5-14B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_1_5b.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModel
+
+models = [
+    dict(
+        type=TurboMindModel,
+        abbr='qwen2.5-1.5b-turbomind',
+        path='Qwen/Qwen2.5-1.5B',
+        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,
+        batch_size=16,
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_1_5b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-1.5b-instruct-turbomind',
+        path='Qwen/Qwen2.5-1.5B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_32b.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModel
+
+models = [
+    dict(
+        type=TurboMindModel,
+        abbr='qwen2.5-32b-turbomind',
+        path='Qwen/Qwen2.5-32B',
+        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,
+        batch_size=16,
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_32b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-32b-instruct-turbomind',
+        path='Qwen/Qwen2.5-32B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_3b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-3b-instruct-turbomind',
+        path='Qwen/Qwen2.5-3B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_72b.py

@@ -0,0 +1,17 @@
+from opencompass.models import TurboMindModel
+
+models = [
+    dict(
+        type=TurboMindModel,
+        abbr='qwen2.5-72b-turbomind',
+        path='Qwen/Qwen2.5-72B',
+        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,
+        batch_size=16,
+        run_cfg=dict(num_gpus=4),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_72b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-72b-instruct-turbomind',
+        path='Qwen/Qwen2.5-72B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_7b.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModel
+
+models = [
+    dict(
+        type=TurboMindModel,
+        abbr='qwen2.5-7b-turbomind',
+        path='Qwen/Qwen2.5-7B',
+        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,
+        batch_size=16,
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/lmdeploy_qwen2_5_7b_instruct.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen2.5-7b-instruct-turbomind',
+        path='Qwen/Qwen2.5-7B-Instruct',
+        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),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_0_5b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-0.5b-instruct-vllm',
+        path='Qwen/Qwen2.5-0.5B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=1, gpu_memory_utilization=0.5),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_14b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-14b-instruct-vllm',
+        path='Qwen/Qwen2.5-14B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=2),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_14b_instruct_128k.py

@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-14b-instruct-vllm',
+        path='Qwen/Qwen2.5-14B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=4,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_1_5b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-1.5b-instruct-vllm',
+        path='Qwen/Qwen2.5-1.5B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=1, gpu_memory_utilization=0.5),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_32b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-32b-instruct-vllm',
+        path='Qwen/Qwen2.5-32B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=2),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_32b_instruct_128k.py

@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-32b-instruct-vllm',
+        path='Qwen/Qwen2.5-32B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=8,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=8),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_3b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-3b-instruct-vllm',
+        path='Qwen/Qwen2.5-3B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=1, gpu_memory_utilization=0.5),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_72b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2_5-72b-instruct-vllm',
+        path='Qwen/Qwen2.5-72B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=4),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_72b_instruct_128k.py

@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2_5-72b-instruct-vllm',
+        path='Qwen/Qwen2.5-72B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=8,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=8),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_7b_instruct.py

@@ -0,0 +1,14 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-7b-instruct-vllm',
+        path='Qwen/Qwen2.5-7B-Instruct',
+        model_kwargs=dict(tensor_parallel_size=1),
+        max_out_len=4096,
+        batch_size=16,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/qwen2_5/vllm_qwen2_5_7b_instruct_128k.py

@@ -0,0 +1,21 @@
+from opencompass.models import VLLMwithChatTemplate
+
+models = [
+    dict(
+        type=VLLMwithChatTemplate,
+        abbr='qwen2.5-7b-instruct-vllm',
+        path='Qwen/Qwen2.5-7B-Instruct',
+        model_kwargs=dict(
+            tensor_parallel_size=4,
+            rope_scaling={
+                'factor': 4.0,
+                'original_max_position_embeddings': 32768,
+                'rope_type': 'yarn'
+            },
+        ),
+        max_out_len=4096,
+        batch_size=1,
+        generation_kwargs=dict(temperature=0),
+        run_cfg=dict(num_gpus=4),
+    )
+]

build/lib/opencompass/configs/models/qwen3/lmdeploy_qwen3_0_6b.py

@@ -0,0 +1,19 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+from opencompass.utils.text_postprocessors import extract_non_reasoning_content
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='qwen_3_0.6b_thinking-turbomind',
+        path='Qwen/Qwen3-0.6B',
+        engine_config=dict(session_len=32768, max_batch_size=16, tp=1),
+        gen_config=dict(
+            top_k=20, temperature=0.6, top_p=0.95, do_sample=True, enable_thinking=True
+        ),
+        max_seq_len=32768,
+        max_out_len=32000,
+        batch_size=16,
+        run_cfg=dict(num_gpus=1),
+        pred_postprocessor=dict(type=extract_non_reasoning_content)
+    ),
+]

build/lib/opencompass/configs/models/qwq/lmdeploy_qwq_32b.py

@@ -0,0 +1,17 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+from opencompass.utils.text_postprocessors import extract_non_reasoning_content
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='QwQ-32B',
+        path='Qwen/QwQ-32B',
+        engine_config=dict(session_len=32768, max_batch_size=16, tp=2),
+        gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=8192),
+        max_seq_len=32768,
+        max_out_len=8192,
+        batch_size=16,
+        run_cfg=dict(num_gpus=2),
+        pred_postprocessor=dict(type=extract_non_reasoning_content)
+    )
+]

build/lib/opencompass/configs/models/qwq/lmdeploy_qwq_32b_preview.py

@@ -0,0 +1,15 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='QwQ-32B-Preview',
+        path='Qwen/QwQ-32B-Preview',
+        engine_config=dict(session_len=32768, max_batch_size=16, tp=2),
+        gen_config=dict(top_k=1, temperature=1e-6, top_p=0.9, max_new_tokens=8192),
+        max_seq_len=32768,
+        max_out_len=8192,
+        batch_size=16,
+        run_cfg=dict(num_gpus=2),
+    )
+]

build/lib/opencompass/configs/models/rwkv/rwkv5_3b.py

@@ -0,0 +1,25 @@
+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='rwkv-5-3b',
+        path='RWKV/rwkv-5-world-3b',
+        tokenizer_path='RWKV/rwkv-5-world-3b',
+        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,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_padding=True,
+        batch_size=16,
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]

build/lib/opencompass/configs/models/skywork/hf_skywork_13b.py

@@ -0,0 +1,12 @@
+from opencompass.models import HuggingFaceBaseModel
+
+models = [
+    dict(
+        type=HuggingFaceBaseModel,
+        abbr='skywork-13b-hf',
+        path='Skywork/Skywork-13B-base',
+        max_out_len=1024,
+        batch_size=8,
+        run_cfg=dict(num_gpus=1),
+    )
+]

build/lib/opencompass/configs/models/skywork/lmdeploy_skywork_o1_open_llama3_1_8b_instruct.py

@@ -0,0 +1,16 @@
+from opencompass.models import TurboMindModelwithChatTemplate
+
+models = [
+    dict(
+        type=TurboMindModelwithChatTemplate,
+        abbr='Skywork-o1-Open-Llama-3_1-8B-turbomind',
+        path='Skywork/Skywork-o1-Open-Llama-3.1-8B',
+        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=4096),
+        max_seq_len=16384,
+        max_out_len=8192,
+        batch_size=16,
+        run_cfg=dict(num_gpus=1),
+        stop_words=['<|end_of_text|>', '<|eot_id|>'],
+    )
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_13b_base_v1.py

@@ -0,0 +1,21 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-13b-base-v1-hf',
+        path='TigerResearch/tigerbot-13b-base-v1',
+        tokenizer_path='TigerResearch/tigerbot-13b-base-v1',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=2, num_procs=1),
+    ),
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_13b_base_v2.py

@@ -0,0 +1,21 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-13b-base-v2-hf',
+        path='TigerResearch/tigerbot-13b-base',
+        tokenizer_path='TigerResearch/tigerbot-13b-base',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=2, num_procs=1),
+    ),
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_13b_chat_v1.py

@@ -0,0 +1,29 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role='HUMAN', begin='\n\n### Instruction:\n'),
+        dict(role='BOT', begin='\n\n### Response:\n', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-13b-chat-v1-hf',
+        path='TigerResearch/tigerbot-13b-chat-v1',
+        tokenizer_path='TigerResearch/tigerbot-13b-chat-v1',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        meta_template=_meta_template,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=2, num_procs=1),
+    )
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_13b_chat_v2.py

@@ -0,0 +1,29 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role='HUMAN', begin='\n\n### Instruction:\n'),
+        dict(role='BOT', begin='\n\n### Response:\n', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-13b-chat-v2-hf',
+        path='TigerResearch/tigerbot-13b-chat',
+        tokenizer_path='TigerResearch/tigerbot-13b-chat',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        meta_template=_meta_template,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=2, num_procs=1),
+    )
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_70b_base.py

@@ -0,0 +1,24 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-70b-base-v1-hf',
+        path='TigerResearch/tigerbot-70b-base',
+        tokenizer_path='TigerResearch/tigerbot-70b-base',
+        model_kwargs=dict(
+            trust_remote_code=True,
+            device_map='auto',
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    ),
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_70b_chat_v2.py

@@ -0,0 +1,29 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role='HUMAN', begin='\n\n### Instruction:\n'),
+        dict(role='BOT', begin='\n\n### Response:\n', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-70b-chat-v2-hf',
+        path='TigerResearch/tigerbot-70b-chat-v2',
+        tokenizer_path='TigerResearch/tigerbot-70b-chat-v2',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        meta_template=_meta_template,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_70b_chat_v3.py

@@ -0,0 +1,32 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role='HUMAN', begin='\n\n### Instruction:\n'),
+        dict(role='BOT', begin='\n\n### Response:\n', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-70b-chat-v3-hf',
+        path='TigerResearch/tigerbot-70b-chat-v3',
+        tokenizer_path='TigerResearch/tigerbot-70b-chat-v3',
+        model_kwargs=dict(
+            trust_remote_code=True,
+            device_map='auto',
+        ),
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        meta_template=_meta_template,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=4, num_procs=1),
+    )
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_7b_base.py

@@ -0,0 +1,21 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-base-7b-hf',
+        path='TigerResearch/tigerbot-7b-base',
+        tokenizer_path='TigerResearch/tigerbot-7b-base',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    ),
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_7b_base_v3.py

@@ -0,0 +1,21 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-7b-base-v3-hf',
+        path='TigerResearch/tigerbot-7b-base',
+        tokenizer_path='TigerResearch/tigerbot-7b-base',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    ),
+]

build/lib/opencompass/configs/models/tigerbot/hf_tigerbot_7b_chat_v3.py

@@ -0,0 +1,29 @@
+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    round=[
+        dict(role='HUMAN', begin='\n\n### Instruction:\n'),
+        dict(role='BOT', begin='\n\n### Response:\n', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='tigerbot-7b-chat-v3-hf',
+        path='TigerResearch/tigerbot-7b-chat',
+        tokenizer_path='TigerResearch/tigerbot-7b-chat',
+        tokenizer_kwargs=dict(
+            padding_side='left',
+            truncation_side='left',
+            trust_remote_code=True,
+        ),
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        meta_template=_meta_template,
+        model_kwargs=dict(trust_remote_code=True, device_map='auto'),
+        run_cfg=dict(num_gpus=1, num_procs=1),
+    )
+]