Upload 10 files

MODEL_LICENSE.md

@@ -0,0 +1,53 @@
+Tongyi Qianwen LICENSE AGREEMENT
+
+Tongyi Qianwen Release Date: August 3, 2023
+
+By clicking to agree or by using or distributing any portion or element of the Tongyi Qianwen Materials, you will be deemed to have recognized and accepted the content of this Agreement, which is effective immediately.
+
+1. Definitions
+    a. This Tongyi Qianwen LICENSE AGREEMENT (this "Agreement") shall mean the terms and conditions for use, reproduction, distribution and modification of the Materials as defined by this Agreement.
+    b. "We"(or "Us") shall mean Alibaba Cloud.
+    c. "You" (or "Your") shall mean a natural person or legal entity exercising the rights granted by this Agreement and/or using the Materials for any purpose and in any field of use.
+    d. "Third Parties" shall mean individuals or legal entities that are not under common control with Us or You.
+    e. "Tongyi Qianwen" shall mean the large language models (including Qwen model and Qwen-Chat model), and software and algorithms, consisting of trained model weights, parameters (including optimizer states), machine-learning model code, inference-enabling code, training-enabling code, fine-tuning enabling code and other elements of the foregoing distributed by Us.
+    f. "Materials" shall mean, collectively, Alibaba Cloud's proprietary Tongyi Qianwen and Documentation (and any portion thereof) made available under this Agreement.
+    g. "Source" form shall mean the preferred form for making modifications, including but not limited to model source code, documentation source, and configuration files.
+    h. "Object" form shall mean any form resulting from mechanical transformation or translation of a Source form, including but not limited to compiled object code, generated documentation,
+ and conversions to other media types.
+
+2. Grant of Rights
+You are granted a non-exclusive, worldwide, non-transferable and royalty-free limited license under Alibaba Cloud's intellectual property or other rights owned by Us embodied in the Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the Materials.
+
+3. Redistribution
+You may reproduce and distribute copies of the Materials or derivative works thereof in any medium, with or without modifications, and in Source or Object form, provided that You meet the following conditions:
+    a. You shall give any other recipients of the Materials or derivative works a copy of this Agreement;
+    b. You shall cause any modified files to carry prominent notices stating that You changed the files;
+    c. You shall retain in all copies of the Materials that You distribute the following attribution notices within a "Notice" text file distributed as a part of such copies: "Tongyi Qianwen is licensed under the Tongyi Qianwen LICENSE AGREEMENT, Copyright (c) Alibaba Cloud. All Rights Reserved."; and
+    d. You may add Your own copyright statement to Your modifications and may provide additional or different license terms and conditions for use, reproduction, or distribution of Your modifications, or for any such derivative works as a whole, provided Your use, reproduction, and distribution of the work otherwise complies with the terms and conditions of this Agreement.
+
+4. Restrictions
+If you are commercially using the Materials, and your product or service has more than 100 million monthly active users, You shall request a license from Us. You cannot exercise your rights under this Agreement without our express authorization.
+
+5. Rules of use
+    a. The Materials may be subject to export controls or restrictions in China, the United States or other countries or regions. You shall comply with applicable laws and regulations in your use of the Materials.
+    b. You can not use the Materials or any output therefrom to improve any other large language model (excluding Tongyi Qianwen or derivative works thereof).
+
+6. Intellectual Property
+    a. We retain ownership of all intellectual property rights in and to the Materials and derivatives made by or for Us. Conditioned upon compliance with the terms and conditions of this Agreement, with respect to any derivative works and modifications of the Materials that are made by you, you are and will be the owner of such derivative works and modifications.
+    b. No trademark license is granted to use the trade names, trademarks, service marks, or product names of Us, except as required to fulfill notice requirements under this Agreement or as required for reasonable and customary use in describing and redistributing the Materials.
+    c. If you commence a lawsuit or other proceedings (including a cross-claim or counterclaim in a lawsuit) against Us or any entity alleging that the Materials or any output therefrom, or any part of the foregoing, infringe any intellectual property or other right owned or licensable by you, then all licences granted to you under this Agreement shall terminate as of the date such lawsuit or other proceeding is commenced or brought.
+
+7. Disclaimer of Warranty and Limitation of Liability
+
+    a. We are not obligated to support, update, provide training for, or develop any further version of the Tongyi Qianwen Materials or to grant any license thereto.
+    b. THE MATERIALS ARE PROVIDED "AS IS" WITHOUT ANY EXPRESS OR IMPLIED WARRANTY OF ANY KIND INCLUDING WARRANTIES OF MERCHANTABILITY, NONINFRINGEMENT, OR FITNESS FOR A PARTICULAR PURPOSE. WE MAKE NO WARRANTY AND ASSUME NO RESPONSIBILITY FOR THE SAFETY OR STABILITY OF THE MATERIALS AND ANY OUTPUT THEREFROM.
+    c. IN NO EVENT SHALL WE BE LIABLE TO YOU FOR ANY DAMAGES, INCLUDING, BUT NOT LIMITED TO ANY DIRECT, OR INDIRECT, SPECIAL OR CONSEQUENTIAL DAMAGES ARISING FROM YOUR USE OR INABILITY TO USE THE MATERIALS OR ANY OUTPUT OF IT, NO MATTER HOW IT’S CAUSED.
+    d. You will defend, indemnify and hold harmless Us from and against any claim by any third party arising out of or related to your use or distribution of the Materials.
+
+8. Survival and Termination.
+    a. The term of this Agreement shall commence upon your acceptance of this Agreement or access to the Materials and will continue in full force and effect until terminated in accordance with the terms and conditions herein.
+    b. We may terminate this Agreement if you breach any of the terms or conditions of this Agreement. Upon termination of this Agreement, you must delete and cease use of the Materials. Sections 7 and 9 shall survive the termination of this Agreement.
+
+9. Governing Law and Jurisdiction.
+    a. This Agreement and any dispute arising out of or relating to it will be governed by the laws of China, without regard to conflict of law principles, and the UN Convention on Contracts for the International Sale of Goods does not apply to this Agreement.
+    b. The People's Courts in Hangzhou City shall have exclusive jurisdiction over any dispute arising out of this Agreement.

README.md

@@ -1,5 +1,81 @@
----
-license: other
-license_name: tongyi-qwen
-license_link: LICENSE
----
+## CodeFuse-CGE-Large
+CodeFuse-CGE-Large is the Large version of the CodeFuse-CGE family which is fine-tuned based on CodeQwen1.5-7B. CodeFuse-CGE-Large is distinguish on text2code task for it's powerful ability of capturing the semantic relationship between code and text.  
+  
+This model has the following notable features:  
+● Instruction-tuning is enabled for both query and code snippet sides.  
+● The model obtains sentence-level and code-level representations through a layer of cross-attention computation module.  
+● The model has a smaller dimensional size without significant degradation in performance.
+
+Model Configuration  
+Model Size: 7B  
+Embedding Dimension: 1024  
+Hidden Layers: 32  
+Max Input Tokens: 1024  
+
+
+Requirements  
+```
+flash_attn==2.4.2
+torch==2.1.0
+accelerate==0.28.0
+transformers==4.39.2 
+vllm=0.5.3
+```
+
+
+## How to Use
+### transformers
+```
+from transformers import AutoTokenizer, AutoModel
+
+model_name_or_path = "CodeFuse-CGE-Large"
+model = AutoModel.from_pretrained(model_name_or_path, trust_remote_code=True)
+tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True, truncation_side='right', padding_side='right')
+model = model.to(torch.bfloat16)
+
+if torch.cuda.is_available():
+    device = 'cuda'
+else:
+    device = 'cpu'
+model.to(device)
+
+prefix_dict =  {'python':{'query':'Retrieve the Python code that solves the following query:', 'passage':'Python code:'},
+                'java':{'query':'Retrieve the Java code that solves the following query:', 'passage':'Java code:'},
+                'go':{'query':'Retrieve the Go code that solves the following query:', 'passage':'Go code:'},
+                'c++':{'query':'Retrieve the C++ code that solves the following query:', 'passage':'C++ code:'},
+                'javascript':{'query':'Retrieve the Javascript code that solves the following query:', 'passage':'Javascript code:'},
+                'php':{'query':'Retrieve the PHP code that solves the following query:', 'passage':'PHP code:'},
+                'ruby':{'query':'Retrieve the Ruby code that solves the following query:', 'passage':'Ruby code:'},
+                'default':{'query':'Retrieve the code that solves the following query:', 'passage':'Code:'}
+                }
+
+text = ["Writes a Boolean to the stream.",
+        "def writeBoolean(self, n): t = TYPE_BOOL_TRUE if n is False: t = TYPE_BOOL_FALSE self.stream.write(t)"]
+text[0] += prefix_dict['python']['query']
+text[0] += prefix_dict['python']['passage']
+embed = model.encode(tokenizer, text)
+score = embed[0] @ embed[1].T
+print("score", score)
+
+```
+
+## Benchmark the Performance
+We use MRR metric to evaluate the ability on text2code retrieval tasks: AdvTest, CosQA, CSN
+
+
+![result](./result.png)
+
+## Acknowledgement
+Thanks to the authors of open-sourced datasets, including CSN, Adv, CoSQA.
+
+## License
+Since CodeFuse-CGE-Large is fine-tuned based on CodeQwen1.5-7B model, our usage license follows the same terms as that of CodeQwen1.5-7B model.
+
+## 加入我们
+我们是平台技术事业群AI Native团队，负责蚂蚁蚂蚁集团平台工程的智能化，团队成立3年多以来，支持了蚂蚁集团云计算基础设施智能化运维的升级改造。团队的Mission是，通过世界级的技术创新和影响，构建有广泛用户的算法服务和平台，支撑内外部产品和业务落地。团队秉承创新基因，在支撑业务落地的同时，推动技术影响。3年以来在ICLR、NeurIPS、KDD、ACL等顶会发表论文20余篇，创新业务结果获得两次蚂蚁技术最高奖T-Star，1次蚂蚁集团最高奖SuperMA。开源项目CodeFuse获得4K点赞(2024年2月)，Huggingface和modelscope上模型累积下载量超过150万次。
+
+我们正在寻找行业中的佼佼者加入我们的团队！如果您希望在一个充满活力、创新和卓越文化的环境中发展您的职业生涯，欢迎您查看我们的社招&校招机会，加入我们，一起创造下一个行业里程碑。
+
+校招：https://hrrecommend.antgroup.com/guide.html?code=8uoP5mlus5DqQYbE_EnqcE2FD5JZH21MwvMUIb9mb6X3osXPuBraG54SyM8GLn_7
+
+社招：https://talent.antgroup.com/off-campus-position?positionId=1933830

config.json

@@ -0,0 +1,45 @@
+{
+  "architectures": [
+    "CodeFuse_CGE_Large"
+  ],
+  "auto_map": {
+    "AutoConfig": "configuration_codefuse_cge_large.CodeFuseCGELargeConfig",
+    "AutoModel": "modeling_codefuse_cge_large.CodeFuse_CGE_Large"
+  },
+  "attention_dropout": 0.0,
+  "bos_token_id": 2,
+  "embedding_dim": 4096,
+  "embedding_method": "pma",
+  "encoder_mode": "post_normal",
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "inf_seq_length": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 13440,
+  "keep_max_layer": 32,
+  "max_position_embeddings": 65536,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 32,
+  "num_encoder_layers": 0,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 4,
+  "padding_side": "right",
+  "pma_ln": true,
+  "pma_norm": false,
+  "pma_norm_mode": "post_normal",
+  "pma_num_heads": 32,
+  "compress_dim": 1024,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 1000000,
+  "rotary_emb_base": 1000000,
+  "seq_length": 65536,
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.39.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 92416
+}

configuration.json

@@ -0,0 +1 @@
+{"task":"sentence-embedding"}

configuration_codefuse_cge_large.py

@@ -0,0 +1,76 @@
+from transformers import PretrainedConfig
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+
+class CodeFuseCGELargeConfig(PretrainedConfig):
+    model_type = "qwen2"
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+        self,
+        vocab_size=151936,
+        hidden_size=4096,
+        intermediate_size=22016,
+        num_hidden_layers=32,
+        num_attention_heads=32,
+        num_key_value_heads=32,
+        hidden_act="silu",
+        max_position_embeddings=32768,
+        initializer_range=0.02,
+        rms_norm_eps=1e-6,
+        use_cache=True,
+        tie_word_embeddings=False,
+        rope_theta=10000.0,
+        use_sliding_window=False,
+        sliding_window=4096,
+        max_window_layers=28,
+        attention_dropout=0.0,
+        embedding_method="pma",
+        inf_seq_length=1024,
+        padding_side="right",
+        compress_dim=1024,
+        keep_max_layer=32,
+        pma_num_heads=32,
+        pma_ln=True,
+        pma_norm=False,
+        pma_norm_mode="post_normal",
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.use_sliding_window = use_sliding_window
+        self.sliding_window = sliding_window if use_sliding_window else None
+        self.max_window_layers = max_window_layers
+
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+
+        self.num_key_value_heads = num_key_value_heads
+        self.hidden_act = hidden_act
+        self.initializer_range = initializer_range
+        self.rms_norm_eps = rms_norm_eps
+        self.use_cache = use_cache
+        self.rope_theta = rope_theta
+        self.attention_dropout = attention_dropout
+
+        self.embedding_method = embedding_method
+        self.inf_seq_length = inf_seq_length
+        self.padding_side = padding_side
+        self.compress_dim = compress_dim
+        self.keep_max_layer = keep_max_layer
+        self.pma_num_heads = pma_num_heads
+        self.pma_ln = pma_ln
+        self.pma_norm = pma_norm
+        self.pma_norm_mode = pma_norm_mode
+
+        super().__init__(
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs,
+        )

model.safetensors.index.json

@@ -0,0 +1,407 @@
+{
+  "metadata": {
+    "total_size": 29043118080
+  },
+  "weight_map": {
+    "mha_pma.S": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_k.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_k.weight": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_o.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_o.weight": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_q.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_q.weight": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_v.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.fc_v.weight": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.ln0.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.ln0.weight": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.ln1.bias": "model-00006-of-00006.safetensors",
+    "mha_pma.mab.ln1.weight": "model-00006-of-00006.safetensors",
+    "plm_model.lm_head.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.embed_tokens.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.input_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.k_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.q_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.v_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.input_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.k_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.q_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.v_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.10.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.10.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.10.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.10.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.10.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.10.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.11.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.11.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.12.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.13.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.14.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.input_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.15.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.16.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.16.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.16.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.16.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.k_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.q_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.v_proj.bias": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.16.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
+    "plm_model.model.layers.17.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.17.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.18.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.19.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.2.input_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.k_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.q_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.v_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.20.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.20.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.input_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.21.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.22.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.22.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.22.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.22.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.k_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.q_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.v_proj.bias": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.22.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
+    "plm_model.model.layers.23.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.23.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.24.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.25.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.26.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.input_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.27.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.input_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.28.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.28.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.28.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.28.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.k_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.q_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.v_proj.bias": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.28.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
+    "plm_model.model.layers.29.input_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.k_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.q_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.v_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.29.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.3.input_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.k_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.q_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.v_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.30.input_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.k_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.q_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.v_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.30.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.input_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.k_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.q_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.v_proj.bias": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.31.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
+    "plm_model.model.layers.4.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.4.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.4.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.4.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.4.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.k_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.q_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.v_proj.bias": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
+    "plm_model.model.layers.5.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.5.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.6.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.7.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.8.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.input_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.k_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.q_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.v_proj.bias": "model-00002-of-00006.safetensors",
+    "plm_model.model.layers.9.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
+    "plm_model.model.norm.weight": "model-00006-of-00006.safetensors"
+  }
+}

modeling_codefuse_cge_large.py

@@ -0,0 +1,245 @@
+from transformers import Qwen2Config
+import inspect
+import math
+import os
+import warnings
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.nn.functional as F
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
+from transformers import PretrainedConfig
+
+from transformers.activations import ACT2FN
+from transformers.cache_utils import Cache, DynamicCache
+from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask, _prepare_4d_causal_attention_mask_for_sdpa, _prepare_4d_attention_mask, _prepare_4d_attention_mask_for_sdpa
+from transformers.modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast, SequenceClassifierOutputWithPast
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import (
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    is_flash_attn_2_available,
+    is_flash_attn_greater_or_equal_2_10,
+    logging,
+    replace_return_docstrings,
+)
+import numpy as np
+from transformers import Qwen2Config
+from transformers import Qwen2ForCausalLM
+import inspect
+import math
+import os
+import warnings
+from typing import List, Optional, Tuple, Union
+from tqdm import tqdm, trange
+import torch
+import torch.nn.functional as F
+import torch.utils.checkpoint
+from torch import nn
+from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
+
+from transformers.activations import ACT2FN
+from transformers.cache_utils import Cache, DynamicCache
+from transformers.modeling_attn_mask_utils import _prepare_4d_causal_attention_mask, _prepare_4d_causal_attention_mask_for_sdpa, _prepare_4d_attention_mask, _prepare_4d_attention_mask_for_sdpa
+from transformers.modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast, SequenceClassifierOutputWithPast
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import (
+    add_start_docstrings,
+    add_start_docstrings_to_model_forward,
+    is_flash_attn_2_available,
+    is_flash_attn_greater_or_equal_2_10,
+    logging,
+    replace_return_docstrings,
+)
+import numpy as np
+import torch
+import os
+import argparse
+import json
+from tqdm import tqdm
+from typing import cast, List, Union, Tuple
+from transformers import AutoTokenizer, AutoModel  # pylint: disable=C0413
+from peft import LoraConfig, get_peft_model, TaskType
+import time
+import torch.nn.functional as F
+import sys
+import time
+import torch 
+import torch.nn as nn
+import torch.nn.functional as F
+import numpy as np
+from tqdm import tqdm, trange
+from collections import defaultdict
+from transformers import AutoTokenizer, AutoModel, AutoModelForCausalLM, AutoConfig
+import torch.distributed as dist
+from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+import sys
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import math
+import re
+
+
+class MAB_POST(nn.Module):
+    def __init__(self, dim_Q, dim_K, dim_V, num_heads, ln=False):
+        super(MAB_POST, self).__init__()
+        self.dim_V = dim_V
+        self.num_heads = num_heads
+        self.fc_q = nn.Linear(dim_Q, dim_V)
+        self.fc_k = nn.Linear(dim_K, dim_V)
+        self.fc_v = nn.Linear(dim_K, dim_V)
+        if ln:
+            self.ln0 = nn.LayerNorm(dim_V)
+            self.ln1 = nn.LayerNorm(dim_V)
+        self.fc_o = nn.Linear(dim_V, dim_V)
+        nn.init.xavier_uniform_(self.fc_q.weight)
+        nn.init.xavier_uniform_(self.fc_k.weight)
+        nn.init.xavier_uniform_(self.fc_v.weight)
+        nn.init.xavier_uniform_(self.fc_o.weight)
+
+    def forward(self, Q, K, pad_mask=None):
+
+        Q_ = self.fc_q(Q)
+        K_, V_ = self.fc_k(K), self.fc_v(K)
+
+        dim_split = self.dim_V // self.num_heads
+        Q_ = torch.cat(Q_.split(dim_split, 2), 0)
+        K_ = torch.cat(K_.split(dim_split, 2), 0)
+        V_ = torch.cat(V_.split(dim_split, 2), 0)
+
+        pad_mask = pad_mask.unsqueeze(1).repeat(self.num_heads, Q.size(1), 1)
+        score = Q_.bmm(K_.transpose(1,2))/math.sqrt(self.dim_V)
+        score = score.masked_fill(pad_mask == 0, -1e12)
+        A = torch.softmax(score, 2)
+        A = A * pad_mask
+        O = torch.cat(A.bmm(V_).split(Q.size(0), 0), 2) 
+        O = Q + O
+        O = O if getattr(self, 'ln0', None) is None else self.ln0(O)
+        O = O + F.relu(self.fc_o(O))
+        O = O if getattr(self, 'ln1', None) is None else self.ln1(O)
+        return O
+
+
+class PMA(nn.Module):
+    def __init__(self, dim, compress_dim, num_heads, num_seeds, ln=False, pma_mode=None):
+        super(PMA, self).__init__()
+        self.S = nn.Parameter(torch.Tensor(1, num_seeds, compress_dim))
+        nn.init.xavier_uniform_(self.S)
+        if pma_mode == 'post_normal':
+            self.mab = MAB_POST(compress_dim, dim, compress_dim, num_heads, ln=ln)
+        elif pma_mode == 'pre_normal':
+            self.mab = MAB_PRE_NORMAL(compress_dim, dim, compress_dim, num_heads, ln=ln)
+        elif pma_mode == 'pre_gptj':
+            self.mab = MAB_PRE_GPTJ(compress_dim, dim, compress_dim, num_heads, ln=ln)
+        else:
+            raise ValueError(f"Error, the pma_mode {pma_mode} is not implemented !")
+
+    def forward(self, X, pad_mask):
+        if self.S.dtype != torch.bfloat16:
+            X = X.float()
+        return self.mab(self.S.repeat(X.size(0), 1, 1), X, pad_mask)
+
+
+class CodeFuse_CGE_Large(PreTrainedModel):
+
+    def __init__(self, config):
+        super().__init__(config)
+        self.plm_model = Qwen2ForCausalLM(config)
+        self.embedding_method = config.embedding_method
+        self.inf_seq_length = config.inf_seq_length
+        self.padding_side = config.padding_side
+
+        self.keep_max_layer = config.keep_max_layer
+        self.emb_dim = self.plm_model.model.embed_tokens.weight.size(1)
+        self.num_heads = config.pma_num_heads
+        self.ln = config.pma_ln
+        self.norm = config.pma_norm
+        self.compress_dim = config.compress_dim
+        self.pma_mode = config.pma_norm_mode
+        self.mha_pma = PMA(self.emb_dim, self.compress_dim, self.num_heads, 1, ln=self.ln, pma_mode=self.pma_mode)
+
+    def last_embedding(self, A, index):
+        bs, seq, emb = A.size()
+        res = A[torch.arange(bs), index, :]
+        return res
+
+    def mean_embedding(self, A, mask):
+        bs, seq, emb = A.size()
+        res = (A * (mask.unsqueeze(-1))).sum(1) / (mask.sum(1).unsqueeze(-1))
+        return res
+
+    def weighted_embedding(self, A, mask):
+        weights = (torch.arange(start=1, end=A.size(1) + 1).unsqueeze(0).unsqueeze(-1).expand(A.size()).float()).to(A.device)
+        input_mask_expanded = (mask.squeeze(1).unsqueeze(-1).expand(A.size()).float()).to(A.device)
+        sum_embedding = torch.sum(A * input_mask_expanded * weights, dim=1)
+        sum_mask = torch.sum(input_mask_expanded * weights, dim=1)
+        weighted_embedding = sum_embedding / sum_mask
+        return weighted_embedding
+
+    def pma_embedding(self, A, mask):
+        res = self.mha_pma(A, mask).squeeze(1)
+        return res
+
+    def get_sentence_embedding(self, embedding_method, **inputs):
+        outputs = self.plm_model(inputs['input_ids'], inputs['attention_mask'], output_hidden_states=True)
+        if embedding_method == 'last':
+            embedding = outputs.hidden_states[self.keep_max_layer]
+            index = inputs['attention_mask'].sum(-1).long() - 1
+            res_embedding = self.last_embedding(embedding, index)
+        elif embedding_method == 'mean':
+            embedding = outputs.hidden_states[self.keep_max_layer]
+            res_embedding = self.mean_embedding(embedding, inputs['attention_mask'])
+        elif embedding_method == 'weighted':
+            embedding = outputs.hidden_states[self.keep_max_layer]
+            res_embedding = self.weighted_embedding(embedding, inputs['attention_mask'])
+        elif embedding_method == 'pma':
+            embedding = outputs.hidden_states[self.keep_max_layer]
+            attention_mask = inputs['attention_mask']
+            res_embedding = self.pma_embedding(embedding, attention_mask)
+        else:
+            logger.debug('Error, no {} way to obtain embbedings'.format(embedding_method))
+
+        if not self.norm:
+            res_embedding = torch.nn.functional.normalize(res_embedding, p=2.0, dim=-1, eps=1e-12, out=None)
+        return res_embedding
+
+
+    def encode(self, tokenizer, sentences, batch_size=32, convert_to_numpy=True,
+            convert_to_tensor=False, show_progress_bar=True, max_seq_length=None, **kwargs):
+
+        if max_seq_length is None:
+            max_seq_length = self.inf_seq_length
+
+        input_is_string = False        
+        if isinstance(sentences, str) or not hasattr(sentences, "__len__"):
+            sentences = [sentences]
+            input_is_string = True
+
+        all_embeddings = []
+        length_sorted_idx = np.argsort([-len(s) for s in sentences])
+        sentences_sorted = [sentences[idx] for idx in length_sorted_idx]
+        with torch.no_grad():
+            for start_index in trange(0, len(sentences), batch_size, desc="Batches", disable=not show_progress_bar):
+                sentences_batch = sentences_sorted[start_index: start_index + batch_size]
+                with torch.no_grad():
+                    inputs = tokenizer(sentences_batch, padding=True, truncation=True, max_length=max_seq_length, add_special_tokens=False, return_tensors='pt').to(self.plm_model.device)
+                    embeddings = self.get_sentence_embedding(self.embedding_method, **inputs)
+                embeddings = embeddings.detach()
+                if convert_to_numpy:
+                    if embeddings.dtype == torch.bfloat16:
+                        embeddings = embeddings.cpu().to(torch.float32)
+                    else:
+                        embeddings = embeddings.cpu()
+                all_embeddings.extend(embeddings)
+        all_embeddings = [all_embeddings[idx] for idx in np.argsort(length_sorted_idx)]
+        if convert_to_tensor:
+            all_embeddings = torch.stack(all_embeddings)
+        elif convert_to_numpy:
+            all_embeddings = np.asarray([emb.numpy() for emb in all_embeddings])
+
+        if input_is_string:
+            all_embeddings = all_embeddings[0]
+        return all_embeddings

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:656b66a920a54bc45e8e06dc587691ab3c0b2930b9ae56d5fa31e72db2f3bff3
+size 1423961

tokenizer_config.json

@@ -0,0 +1,19 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "add_prefix_space": false,
+  "additional_special_tokens": ["<|im_start|>", "<|im_end|>", "<fim_prefix>", "<fim_middle>", "<fim_suffix>", "<fim_pad>"],
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "legacy": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<fim_pad>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false,
+  "add_prefix_space": true,
+  "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}"
+}