[Feature] Add Qwen1.5 MoE 7b and Mixtral 8x22b model configs (#1123)

* added qwen moe and mixtral 8x22 model configs

* updated README files news section

README.md

@@ -70,6 +70,8 @@ Just like a compass guides us on our journey, OpenCompass will guide you through
 
 ## 🚀 What's New <a><img width="35" height="20" src="https://user-images.githubusercontent.com/12782558/212848161-5e783dd6-11e8-4fe0-bbba-39ffb77730be.png"></a>
 
+- **\[2024.05.08\]** We supported the evaluation of 4 MoE models: [Mixtral-8x22B-v0.1](configs/models/mixtral/hf_mixtral_8x22b_v0_1.py), [Mixtral-8x22B-Instruct-v0.1](configs/models/mixtral/hf_mixtral_8x22b_instruct_v0_1.py), [Qwen1.5-MoE-A2.7B](configs/models/qwen/hf_qwen1_5_moe_a2_7b.py), [Qwen1.5-MoE-A2.7B-Chat](configs/models/qwen/hf_qwen1_5_moe_a2_7b_chat.py). Try them out now!
+- **\[2024.04.30\]** We supported evaluating a model's compression efficiency by calculating its Bits per Character (BPC) metric on an [external corpora](configs/datasets/llm_compression/README.md) ([official paper](https://github.com/hkust-nlp/llm-compression-intelligence)). Check out the [llm-compression](configs/eval_llm_compression.py) evaluation config now! 🔥🔥🔥
 - **\[2024.04.29\]** We report the performance of several famous LLMs on the common benchmarks, welcome to [documentation](https://opencompass.readthedocs.io/en/latest/user_guides/corebench.html) for more information! 🔥🔥🔥.
 - **\[2024.04.26\]** We deprecated the multi-madality evaluating function from OpenCompass, related implement has moved to [VLMEvalKit](https://github.com/open-compass/VLMEvalKit), welcome to use! 🔥🔥🔥.
 - **\[2024.04.26\]** We supported the evaluation of [ArenaHard](configs/eval_subjective_arena_hard.py)  welcome to try!🔥🔥🔥.

README_zh-CN.md

@@ -69,6 +69,8 @@
 
 ## 🚀 最新进展 <a><img width="35" height="20" src="https://user-images.githubusercontent.com/12782558/212848161-5e783dd6-11e8-4fe0-bbba-39ffb77730be.png"></a>
 
+- **\[2024.05.08\]** 我们支持了以下四个MoE模型的评测配置文件: [Mixtral-8x22B-v0.1](configs/models/mixtral/hf_mixtral_8x22b_v0_1.py), [Mixtral-8x22B-Instruct-v0.1](configs/models/mixtral/hf_mixtral_8x22b_instruct_v0_1.py), [Qwen1.5-MoE-A2.7B](configs/models/qwen/hf_qwen1_5_moe_a2_7b.py), [Qwen1.5-MoE-A2.7B-Chat](configs/models/qwen/hf_qwen1_5_moe_a2_7b_chat.py) 。欢迎试用!
+- **\[2024.04.30\]** 我们支持了计算模型在给定[数据集](configs/datasets/llm_compression/README.md)上的压缩率（Bits per Character）的评测方法（[官方文献](https://github.com/hkust-nlp/llm-compression-intelligence)）。欢迎试用[llm-compression](configs/eval_llm_compression.py)评测集! 🔥🔥🔥
 - **\[2024.04.26\]** 我们报告了典型LLM在常用基准测试上的表现，欢迎访问[文档](https://opencompass.readthedocs.io/zh-cn/latest/user_guides/corebench.html)以获取更多信息！🔥🔥🔥.
 - **\[2024.04.26\]** 我们废弃了 OpenCompass 进行多模态大模型评测的功能，相关功能转移至 [VLMEvalKit](https://github.com/open-compass/VLMEvalKit)，推荐使用！🔥🔥🔥.
 - **\[2024.04.26\]** 我们支持了 [ArenaHard评测](configs/eval_subjective_arena_hard.py) 欢迎试用！🔥🔥🔥.

configs/models/mixtral/hf_mixtral_8x22b_instruct_v0_1.py

+from opencompass.models import HuggingFaceCausalLM
+
+
+_meta_template = dict(
+    begin="<s>",
+    round=[
+        dict(role="HUMAN", begin='[INST] ', end=' [/INST]'),
+        dict(role="BOT", begin="", end='</s> ', generate=True),
+    ],
+)
+
+models = [
+    dict(
+        abbr='mixtral-8x22b-instruct-v0.1',
+        type=HuggingFaceCausalLM,
+        path='mistralai/Mixtral-8x22B-Instruct-v0.1',
+        tokenizer_path='mistralai/Mixtral-8x22B-Instruct-v0.1',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True,
+        ),
+        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=2, num_procs=1),
+        batch_padding=True,
+    )
+]

configs/models/mixtral/hf_mixtral_8x22b_v0_1.py

+from opencompass.models import HuggingFaceCausalLM
+
+
+models = [
+    dict(
+        abbr='mixtral-8x22b-v0.1',
+        type=HuggingFaceCausalLM,
+        path='mistralai/Mixtral-8x22B-v0.1',
+        tokenizer_path='mistralai/Mixtral-8x22B-v0.1',
+        model_kwargs=dict(
+            device_map='auto',
+            trust_remote_code=True,
+        ),
+        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=2, num_procs=1),
+    )
+]

configs/models/qwen/hf_qwen1_5_moe_a2_7b.py

+from opencompass.models import HuggingFaceCausalLM
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen1.5-moe-a2-7b-hf',
+        path="Qwen/Qwen1.5-MoE-A2.7B",
+        tokenizer_path='Qwen/Qwen1.5-MoE-A2.7B',
+        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_qwen1_5_moe_a2_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),
+    ],
+)
+
+models = [
+    dict(
+        type=HuggingFaceCausalLM,
+        abbr='qwen1.5-moe-a2-7b-chat-hf',
+        path="Qwen/Qwen1.5-MoE-A2.7B-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,
+        max_out_len=100,
+        max_seq_len=2048,
+        batch_size=8,
+        run_cfg=dict(num_gpus=4, num_procs=1),
+        end_str='<|im_end|>',
+        batch_padding=True,
+    )
+]