From cb9fab3c2b49edfa0f14377914c4fdf06820cbcc Mon Sep 17 00:00:00 2001 From: xxl <505279206@qq.com> Date: Mon, 3 Mar 2025 14:38:15 +0800 Subject: [PATCH] first commit --- 1_Pooling/config.json | 10 + README.md | 124 +- config.json | 40 + config_sentence_transformers.json | 10 + configuration_gemma2.py | 156 +++ model-00001-of-00002.safetensors | 3 + model-00002-of-00002.safetensors | 3 + model.safetensors.index.json | 295 +++++ modeling_gemma2.py | 1398 ++++++++++++++++++++ modules.json | 14 + sentence_bert_config.json | 4 + special_tokens_map.json | 34 + tokenizer.json | 3 + tokenizer.model | 3 + tokenizer_config.json | 2013 +++++++++++++++++++++++++++++ 15 files changed, 4108 insertions(+), 2 deletions(-) create mode 100644 1_Pooling/config.json create mode 100644 config.json create mode 100644 config_sentence_transformers.json create mode 100644 configuration_gemma2.py create mode 100644 model-00001-of-00002.safetensors create mode 100644 model-00002-of-00002.safetensors create mode 100644 model.safetensors.index.json create mode 100644 modeling_gemma2.py create mode 100644 modules.json create mode 100644 sentence_bert_config.json create mode 100644 special_tokens_map.json create mode 100644 tokenizer.json create mode 100644 tokenizer.model create mode 100644 tokenizer_config.json diff --git a/1_Pooling/config.json b/1_Pooling/config.json new file mode 100644 index 0000000..aec2f7c --- /dev/null +++ b/1_Pooling/config.json @@ -0,0 +1,10 @@ +{ + "word_embedding_dimension": 2304, + "pooling_mode_cls_token": false, + "pooling_mode_mean_tokens": false, + "pooling_mode_max_tokens": false, + "pooling_mode_mean_sqrt_len_tokens": false, + "pooling_mode_weightedmean_tokens": false, + "pooling_mode_lasttoken": true, + "include_prompt": true +} \ No newline at end of file diff --git a/README.md b/README.md index 2514085..d58f7b8 100644 --- a/README.md +++ b/README.md @@ -1,3 +1,123 @@ -# SFR-Embedding-Code-2B_R +--- +license: cc-by-nc-4.0 +pipeline_tag: feature-extraction +tags: +- transformers +- sentence-transformers +- code +- retrieval +--- +

Salesforce/SFR-Embedding-Code-2B_R

+ +**SFR-Embedding by Salesforce Research.** + +The Salesforce/SFR-Embedding-Code is a generalist embedding model family for multilingual and multi-task code and Text retrieval. It demonstrates superior performance compared to various open-source code embedding models across multiple code retrieval tasks. + +Check out our [paper](https://arxiv.org/abs/2411.12644) for more details! + +We also offer 400M-size model [Salesforce/SFR-Embedding-Code-400_R](https://huggingface.co/Salesforce/SFR-Embedding-Code-400M_R) + +### Ethical Considerations +This release is for research purposes only in support of an academic paper. Our models, datasets, and code are not specifically designed or evaluated for all downstream purposes. We strongly recommend users evaluate and address potential concerns related to accuracy, safety, and fairness before deploying this model. We encourage users to consider the common limitations of AI, comply with applicable laws, and leverage best practices when selecting use cases, particularly for high-risk scenarios where errors or misuse could significantly impact people’s lives, rights, or safety. For further guidance on use cases, refer to our [AUP](https://www.salesforce.com/content/dam/web/en_us/www/documents/legal/Agreements/policies/ExternalFacing_Services_Policy.pdf) and [AI AUP](https://www.salesforce.com/content/dam/web/en_us/www/documents/legal/Agreements/policies/ai-acceptable-use-policy.pdf). + +### License Statement: +Users need to make their own assessment regarding any obligations or responsibilities under the corresponding licenses or terms and conditions pertaining to the original datasets and data. This release is for research purposes only in support of an academic paper. + +This released model is a fine-tuned version of Gemma and Gemma is provided under and subject to the Gemma Terms of Use found at ai.google.dev/gemma/terms. Additionally, the use of this model is restricted as set forth in the Gemma Prohibited Use Policy at ai.google.dev/gemma/prohibited_use_policy ("Prohibited Use Policy"), which is hereby incorporated by reference into this Agreement. + +### Performance on CoIR Benchmark +| Model | Model Size | CoIR AVG (NDCG@10) | +|-----------------------|------------|---------------------| +| **SFR-Embedding-Code** | 2B | 67.4 | +| CodeSage-Large-v2 | 1.3B | 64.2 | +| CodeSage-Large | 1.3B | 61.0 | +| **SFR-Embedding-Code** | 400M | 61.9 | +| CodeRankEmbed | 137M | 60.1 | +| CodeSage-Base | 356M | 57.5 | +| Voyage-Code-002 | - | 56.3 | +| CodeSage-Small | 130M | 54.4 | + + +SFR-Embedding Team († indicates co-leaders) +* Ye Liu +* Rui Meng +* Shafiq Rayhan Joty +* Silvio Savarese +* Caiming Xiong † +* Yingbo Zhou † +* Semih Yavuz † + +## How to run + +#### Transformers +```python +import torch.nn.functional as F +from transformers import AutoTokenizer, AutoModel + +# Each query needs to be accompanied by an corresponding instruction describing the task. +query_instruction_example = "Given Code or Text, retrieval relevant content" +queries = [ + "how to implement quick sort in Python?" +] + +# No instruction needed for retrieval passages +passages = [ + "def quick_sort(arr):\n if len(arr) <= 1:\n return arr\n pivot = arr[len(arr) // 2]\n left = [x for x in arr if x < pivot]\n middle = [x for x in arr if x == pivot]\n right = [x for x in arr if x > pivot]\n return quick_sort(left) + middle + quick_sort(right)", + "def bubble_sort(arr):\n n = len(arr)\n for i in range(n):\n for j in range(0, n-i-1):\n if arr[j] > arr[j+1]:\n arr[j], arr[j+1] = arr[j+1], arr[j]\n return arr" +] + +# load model with tokenizer +model = AutoModel.from_pretrained('Salesforce/SFR-Embedding-Code-2B_R', trust_remote_code=True) + +# get the embeddings +max_length = 32768 +query_embeddings = model.encode_queries(queries, instruction=query_instruction_example, max_length=max_length) +passage_embeddings = model.encode_corpus(passages, max_length=max_length) + +# normalize embeddings +query_embeddings = F.normalize(query_embeddings, p=2, dim=1) +passage_embeddings = F.normalize(passage_embeddings, p=2, dim=1) + +scores = (query_embeddings @ passage_embeddings.T) * 100 +print(scores.tolist()) +# [[69.26929473876953, 58.41606903076172]] +``` + +#### Sentence Transformers + +```python +from sentence_transformers import SentenceTransformer + +# Each query needs to be accompanied by an corresponding instruction describing the task. +query_instruction_example = "Instruct: Given Code or Text, retrieval relevant content\nQuery: " +queries = ["how to implement quick sort in Python?"] + +# No instruction needed for retrieval passages +passages = [ + "def quick_sort(arr):\n if len(arr) <= 1:\n return arr\n pivot = arr[len(arr) // 2]\n left = [x for x in arr if x < pivot]\n middle = [x for x in arr if x == pivot]\n right = [x for x in arr if x > pivot]\n return quick_sort(left) + middle + quick_sort(right)", + "def bubble_sort(arr):\n n = len(arr)\n for i in range(n):\n for j in range(0, n-i-1):\n if arr[j] > arr[j+1]:\n arr[j], arr[j+1] = arr[j+1], arr[j]\n return arr" +] + +# Load the Sentence Transformer model, including pooling +model = SentenceTransformer('Salesforce/SFR-Embedding-Code-2B_R', trust_remote_code=True) + +# Compute the embeddings for both queries and passages. Use 'prompt' for queries only +query_embeddings = model.encode(queries, prompt=query_instruction_example) +passage_embeddings = model.encode(passages) + +# Compute the similarities between the queries and passages +similarities = model.similarity(query_embeddings, passage_embeddings) +print(similarities) +# tensor([[0.6927, 0.5842]]) +``` + +### Citation +```bibtex +@article{liu2024codexembed, + title={CodeXEmbed: A Generalist Embedding Model Family for Multiligual and Multi-task Code Retrieval}, + author={Liu, Ye and Meng, Rui and Jot, Shafiq and Savarese, Silvio and Xiong, Caiming and Zhou, Yingbo and Yavuz, Semih}, + journal={arXiv preprint arXiv:2411.12644}, + year={2024} +} +``` -SFR-Embedding-Code-2B_R \ No newline at end of file diff --git a/config.json b/config.json new file mode 100644 index 0000000..afe4d63 --- /dev/null +++ b/config.json @@ -0,0 +1,40 @@ +{ + "_name_or_path": "Salesforce/SFR-Embedding-Code-2B_R", + "architectures": [ + "CodeXEmbedModel2B" + ], + "auto_map": { + "AutoConfig": "configuration_gemma2.CodeXEmbedConfig", + "AutoModel": "modeling_gemma2.CodeXEmbedModel2B" + }, + "attention_bias": false, + "attention_dropout": 0.0, + "attn_logit_softcapping": 50.0, + "bos_token_id": 2, + "cache_implementation": "hybrid", + "eos_token_id": [ + 1, + 107 + ], + "final_logit_softcapping": 30.0, + "head_dim": 256, + "hidden_act": "gelu_pytorch_tanh", + "hidden_activation": "gelu_pytorch_tanh", + "hidden_size": 2304, + "initializer_range": 0.02, + "intermediate_size": 9216, + "max_position_embeddings": 8192, + "model_type": "codexembed2b", + "num_attention_heads": 8, + "num_hidden_layers": 26, + "num_key_value_heads": 4, + "pad_token_id": 0, + "query_pre_attn_scalar": 256, + "rms_norm_eps": 1e-06, + "rope_theta": 10000.0, + "sliding_window": 4096, + "torch_dtype": "bfloat16", + "transformers_version": "4.45.1", + "use_cache": true, + "vocab_size": 256000 +} diff --git a/config_sentence_transformers.json b/config_sentence_transformers.json new file mode 100644 index 0000000..7f29d3b --- /dev/null +++ b/config_sentence_transformers.json @@ -0,0 +1,10 @@ +{ + "__version__": { + "sentence_transformers": "3.0.1", + "transformers": "4.41.2", + "pytorch": "2.3.0+cu121" + }, + "prompts": {}, + "default_prompt_name": null, + "similarity_fn_name": "cosine" +} \ No newline at end of file diff --git a/configuration_gemma2.py b/configuration_gemma2.py new file mode 100644 index 0000000..42e3960 --- /dev/null +++ b/configuration_gemma2.py @@ -0,0 +1,156 @@ +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# This file was automatically generated from . +# Do NOT edit this file manually as any edits will be overwritten by the generation of +# the file from the diff. If any change should be done, please apply the change to the +# diff.py file directly. +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# coding=utf-8 +# Copyright 2024 Google Inc. HuggingFace Inc. team. All rights reserved. +# +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +from transformers import PretrainedConfig + + +class CodeXEmbedConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`Gemma2Model`]. It is used to instantiate an Gemma2 + model according to the specified arguments, defining the model architecture. Instantiating a configuration with the + defaults will yield a similar configuration to that of the Gemma2-7B. + e.g. [google/gemma2-7b](https://huggingface.co/google/gemma2-7b) + Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the + documentation from [`PretrainedConfig`] for more information. + Args: + vocab_size (`int`, *optional*, defaults to 256000): + Vocabulary size of the Gemma2 model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`Gemma2Model`] + hidden_size (`int`, *optional*, defaults to 3072): + Dimension of the hidden representations. + intermediate_size (`int`, *optional*, defaults to 24576): + Dimension of the MLP representations. + num_hidden_layers (`int`, *optional*, defaults to 28): + Number of hidden layers in the Transformer decoder. + num_attention_heads (`int`, *optional*, defaults to 16): + Number of attention heads for each attention layer in the Transformer decoder. + num_key_value_heads (`int`, *optional*, defaults to 16): + This is the number of key_value heads that should be used to implement Grouped Query Attention. If + `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if + `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When + converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed + by meanpooling all the original heads within that group. For more details checkout [this + paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to + `num_attention_heads`. + head_dim (`int`, *optional*, defaults to 256): + The attention head dimension. + hidden_activation (`str` or `function`, *optional*, defaults to `"gelu_pytorch_tanh"`): + The non-linear activation function (function or string) in the decoder. + max_position_embeddings (`int`, *optional*, defaults to 8192): + The maximum sequence length that this model might ever be used with. + initializer_range (`float`, *optional*, defaults to 0.02): + The standard deviation of the truncated_normal_initializer for initializing all weight matrices. + rms_norm_eps (`float`, *optional*, defaults to 1e-06): + The epsilon used by the rms normalization layers. + use_cache (`bool`, *optional*, defaults to `True`): + Whether or not the model should return the last key/values attentions (not used by all models). Only + relevant if `config.is_decoder=True`. + pad_token_id (`int`, *optional*, defaults to 0): + Padding token id. + eos_token_id (`int`, *optional*, defaults to 1): + End of stream token id. + bos_token_id (`int`, *optional*, defaults to 2): + Beginning of stream token id. + tie_word_embeddings (`bool`, *optional*, defaults to `True`): + Whether to tie weight embeddings + rope_theta (`float`, *optional*, defaults to 10000.0): + The base period of the RoPE embeddings. + attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`): + Whether to use a bias in the query, key, value and output projection layers during self-attention. + attention_dropout (`float`, *optional*, defaults to 0.0): + The dropout ratio for the attention probabilities. + final_logit_softcapping (`float`, *optional*, defaults to 30.0): scaling factor when applying tanh softcapping on the logits. + attn_logit_softcapping (`float`, *optional*, defaults to 50.0): scaling factor when applying tanh softcapping on the attention scores. + query_pre_attn_scalar (`float`, *optional*, defaults to 224): scaling factor used on the attention scores + sliding_window (`int`, *optional*, defaults to 4096): in Gemma2, every other layer uses sliding window attention. This is the + size of the sliding window. + ```python + >>> from transformers import Gemma2Model, CodeXEmbedConfig + >>> # Initializing a Gemma2 gemma2-9b style configuration + >>> configuration = CodeXEmbedConfig() + >>> # Initializing a model from the gemma2-9b style configuration + >>> model = Gemma2Model(configuration) + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "codexembed2b" + keys_to_ignore_at_inference = ["past_key_values"] + + def __init__( + self, + vocab_size=256000, + hidden_size=3072, + intermediate_size=24576, + num_hidden_layers=28, + num_attention_heads=16, + num_key_value_heads=16, + head_dim=256, + hidden_activation="gelu_pytorch_tanh", + max_position_embeddings=8192, + initializer_range=0.02, + rms_norm_eps=1e-6, + use_cache=True, + pad_token_id=0, + eos_token_id=1, + bos_token_id=2, + tie_word_embeddings=True, + rope_theta=10000.0, + attention_bias=False, + attention_dropout=0.0, + final_logit_softcapping=30.0, + attn_logit_softcapping=50.0, + query_pre_attn_scalar=224, + sliding_window=4096, + **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.head_dim = head_dim + self.num_key_value_heads = num_key_value_heads + self.hidden_activation = hidden_activation + self.initializer_range = initializer_range + self.rms_norm_eps = rms_norm_eps + self.use_cache = use_cache + self.rope_theta = rope_theta + self.attention_bias = attention_bias + self.attention_dropout = attention_dropout + self.attn_logit_softcapping = attn_logit_softcapping + + super().__init__( + pad_token_id=pad_token_id, + bos_token_id=bos_token_id, + eos_token_id=eos_token_id, + tie_word_embeddings=tie_word_embeddings, + **kwargs, + ) + self.final_logit_softcapping = final_logit_softcapping + self.query_pre_attn_scalar = query_pre_attn_scalar + self.sliding_window = sliding_window + self.cache_implementation = "hybrid" + +MODEL_TYPE = "codexembed2b" +from transformers import AutoConfig +AutoConfig.register(MODEL_TYPE, CodeXEmbedConfig) \ No newline at end of file diff --git a/model-00001-of-00002.safetensors b/model-00001-of-00002.safetensors new file mode 100644 index 0000000..e5be53b --- /dev/null +++ b/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8768183f31afaadc29851e4d42bfdff0db564ef913341a1510ced73b33d1c611 +size 4988024144 diff --git a/model-00002-of-00002.safetensors b/model-00002-of-00002.safetensors new file mode 100644 index 0000000..00dfc2c --- /dev/null +++ b/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc2655272864197226ddcb9a8951470459d790cc43087e9180b8a5f975653d36 +size 240691624 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000..36d5501 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,295 @@ +{ + "metadata": { + "total_size": 5228683776 + }, + "weight_map": { + "embed_tokens.weight": "model-00001-of-00002.safetensors", + "layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.24.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.24.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.mlp.gate_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.mlp.up_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.25.post_feedforward_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.25.pre_feedforward_layernorm.weight": "model-00002-of-00002.safetensors", + "layers.25.self_attn.k_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.self_attn.q_proj.weight": "model-00002-of-00002.safetensors", + "layers.25.self_attn.v_proj.weight": "model-00002-of-00002.safetensors", + "layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.post_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.pre_feedforward_layernorm.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "norm.weight": "model-00002-of-00002.safetensors" + } +} diff --git a/modeling_gemma2.py b/modeling_gemma2.py new file mode 100644 index 0000000..61aa9c7 --- /dev/null +++ b/modeling_gemma2.py @@ -0,0 +1,1398 @@ +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# This file was automatically generated from . +# Do NOT edit this file manually as any edits will be overwritten by the generation of +# the file from the diff. If any change should be done, please apply the change to the +# diff.py file directly. +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# coding=utf-8 +# Copyright 2024 Google Inc. HuggingFace Inc. team. All rights reserved. +# +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +from typing import List, Optional, Tuple, Union + +import torch +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, HybridCache +from transformers.modeling_outputs import ( + BaseModelOutputWithPast, + CausalLMOutputWithPast, + SequenceClassifierOutputWithPast, + TokenClassifierOutput, +) +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, + is_flash_attn_greater_or_equal_2_10, + logging, + replace_return_docstrings, +) +from transformers.modeling_attn_mask_utils import _prepare_4d_attention_mask_for_sdpa +from .configuration_gemma2 import CodeXEmbedConfig +from transformers import AutoTokenizer, AutoModel +import torch +import logging +import numpy as np +from typing import List, Dict + + +if is_flash_attn_2_available(): + from transformers.modeling_flash_attention_utils import _flash_attention_forward + + +logger = logging.getLogger(__name__) + + +# Copied from transformers.models.llama.modeling_llama._prepare_4d_causal_attention_mask_with_cache_position +def _prepare_4d_causal_attention_mask_with_cache_position( + attention_mask: torch.Tensor, + sequence_length: int, + target_length: int, + dtype: torch.dtype, + device: torch.device, + min_dtype: float, + cache_position: torch.Tensor, + batch_size: int, +): + """ + Creates a causal 4D mask of shape `(batch_size, 1, query_length, key_value_length)` from a 2D mask of shape + `(batch_size, key_value_length)`, or if the input `attention_mask` is already 4D, do nothing. + + Args: + attention_mask (`torch.Tensor`): + A 2D attention mask of shape `(batch_size, key_value_length)` or a 4D attention mask of shape `(batch_size, 1, query_length, key_value_length)`. + sequence_length (`int`): + The sequence length being processed. + target_length (`int`): + The target length: when generating with static cache, the mask should be as long as the static cache, to account for the 0 padding, the part of the cache that is not filled yet. + dtype (`torch.dtype`): + The dtype to use for the 4D attention mask. + device (`torch.device`): + The device to plcae the 4D attention mask on. + min_dtype (`float`): + The minimum value representable with the dtype `dtype`. + cache_position (`torch.Tensor`): + Indices depicting the position of the input sequence tokens in the sequence. + batch_size (`torch.Tensor`): + Batch size. + """ + if attention_mask is not None and attention_mask.dim() == 4: + # In this case we assume that the mask comes already in inverted form and requires no inversion or slicing. + causal_mask = attention_mask + else: + causal_mask = torch.full((sequence_length, target_length), fill_value=min_dtype, dtype=dtype, device=device) + if sequence_length != 1: + causal_mask = torch.triu(causal_mask, diagonal=1) + causal_mask *= torch.arange(target_length, device=device) > cache_position.reshape(-1, 1) + causal_mask = causal_mask[None, None, :, :].expand(batch_size, 1, -1, -1) + if attention_mask is not None: + causal_mask = causal_mask.clone() # copy to contiguous memory for in-place edit + mask_length = attention_mask.shape[-1] + padding_mask = causal_mask[:, :, :, :mask_length] + attention_mask[:, None, None, :] + padding_mask = padding_mask == 0 + causal_mask[:, :, :, :mask_length] = causal_mask[:, :, :, :mask_length].masked_fill( + padding_mask, min_dtype + ) + + return causal_mask + + +class Gemma2RMSNorm(nn.Module): + def __init__(self, dim: int, eps: float = 1e-6): + super().__init__() + self.eps = eps + self.weight = nn.Parameter(torch.zeros(dim)) + + def _norm(self, x): + return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) + + def forward(self, x): + output = self._norm(x.float()) + # Llama does x.to(float16) * w whilst Gemma2 is (x * w).to(float16) + # See https://github.com/huggingface/transformers/pull/29402 + output = output * (1.0 + self.weight.float()) + return output.type_as(x) + + def extra_repr(self): + return f"{tuple(self.weight.shape)}, eps={self.eps}" + + +class Gemma2RotaryEmbedding(nn.Module): + def __init__(self, dim, max_position_embeddings=2048, base=10000, device=None): + super().__init__() + + self.dim = dim + self.max_position_embeddings = max_position_embeddings + self.base = base + inv_freq = 1.0 / (self.base ** (torch.arange(0, self.dim, 2, dtype=torch.int64).float() / self.dim)) + self.register_buffer("inv_freq", tensor=inv_freq, persistent=False) + + @torch.no_grad() + def forward(self, x, position_ids, seq_len=None): + # x: [bs, num_attention_heads, seq_len, head_size] + self.inv_freq.to(x.device) + inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1) + position_ids_expanded = position_ids[:, None, :].float() + # Force float32 since bfloat16 loses precision on long contexts + # See https://github.com/huggingface/transformers/pull/29285 + device_type = x.device.type + device_type = device_type if isinstance(device_type, str) and device_type != "mps" else "cpu" + with torch.autocast(device_type=device_type, enabled=False): + freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2) + emb = torch.cat((freqs, freqs), dim=-1) + cos = emb.cos() + sin = emb.sin() + return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype) + + +def rotate_half(x): + """Rotates half the hidden dims of the input.""" + x1 = x[..., : x.shape[-1] // 2] + x2 = x[..., x.shape[-1] // 2 :] + return torch.cat((-x2, x1), dim=-1) + + +def apply_rotary_pos_emb(q, k, cos, sin, position_ids=None, unsqueeze_dim=1): + """Applies Rotary Position Embedding to the query and key tensors. + + Args: + q (`torch.Tensor`): The query tensor. + k (`torch.Tensor`): The key tensor. + cos (`torch.Tensor`): The cosine part of the rotary embedding. + sin (`torch.Tensor`): The sine part of the rotary embedding. + position_ids (`torch.Tensor`, *optional*): + Deprecated and unused. + unsqueeze_dim (`int`, *optional*, defaults to 1): + The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and + sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note + that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and + k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes + cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have + the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2. + Returns: + `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding. + """ + cos = cos.unsqueeze(unsqueeze_dim) + sin = sin.unsqueeze(unsqueeze_dim) + q_embed = (q * cos) + (rotate_half(q) * sin) + k_embed = (k * cos) + (rotate_half(k) * sin) + return q_embed, k_embed + + +class Gemma2MLP(nn.Module): + def __init__(self, config): + super().__init__() + self.config = config + self.hidden_size = config.hidden_size + self.intermediate_size = config.intermediate_size + self.gate_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False) + self.up_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False) + self.down_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=False) + self.act_fn = ACT2FN[config.hidden_activation] + + def forward(self, x): + return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x)) + + +def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor: + """ + This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch, + num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim) + """ + batch, num_key_value_heads, slen, head_dim = hidden_states.shape + if n_rep == 1: + return hidden_states + hidden_states = hidden_states[:, :, None, :, :].expand(batch, num_key_value_heads, n_rep, slen, head_dim) + return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim) + + +class Gemma2Attention(nn.Module): + """Multi-headed attention from 'Attention Is All You Need' paper""" + + def __init__(self, config: CodeXEmbedConfig, layer_idx: Optional[int] = None, is_causal: bool=False): + super().__init__() + self.config = config + self.layer_idx = layer_idx + if layer_idx is None: + logger.warning_once( + f"Instantiating {self.__class__.__name__} without passing a `layer_idx` is not recommended and will " + "lead to errors during the forward call if caching is used. Please make sure to provide a `layer_idx` " + "when creating this class." + ) + + self.attention_dropout = config.attention_dropout + self.hidden_size = config.hidden_size + self.num_heads = config.num_attention_heads + self.head_dim = config.head_dim + self.num_key_value_heads = config.num_key_value_heads + self.num_key_value_groups = self.num_heads // self.num_key_value_heads + self.max_position_embeddings = config.max_position_embeddings + self.rope_theta = config.rope_theta + self.is_causal = is_causal + self.scaling = config.query_pre_attn_scalar**-0.5 + + if self.hidden_size % self.num_heads != 0: + raise ValueError( + f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}" + f" and `num_heads`: {self.num_heads})." + ) + + self.q_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.attention_bias) + self.k_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=config.attention_bias) + self.v_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=config.attention_bias) + self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=config.attention_bias) + self.rotary_emb = Gemma2RotaryEmbedding( + self.head_dim, + max_position_embeddings=self.max_position_embeddings, + base=self.rope_theta, + ) + self.sliding_window = config.sliding_window if not bool(layer_idx % 2) else None + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_value: Optional[Cache] = None, + output_attentions: bool = False, + use_cache: bool = False, + cache_position: Optional[torch.LongTensor] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + bsz, q_len, _ = hidden_states.size() + + query_states = self.q_proj(hidden_states) + key_states = self.k_proj(hidden_states) + value_states = self.v_proj(hidden_states) + + query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2) + key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + + cos, sin = self.rotary_emb(value_states, position_ids) + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + # sin and cos are specific to RoPE models; cache_position needed for the static cache + cache_kwargs = { + "sin": sin, + "cos": cos, + "sliding_window": self.sliding_window, + "cache_position": cache_position, + } + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + key_states = repeat_kv(key_states, self.num_key_value_groups) + value_states = repeat_kv(value_states, self.num_key_value_groups) + + attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) * self.scaling + + if self.config.attn_logit_softcapping is not None: + attn_weights = attn_weights / self.config.attn_logit_softcapping + attn_weights = torch.tanh(attn_weights) + attn_weights = attn_weights * self.config.attn_logit_softcapping + + if attention_mask is not None: # no matter the length, we just slice it + causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] + attn_weights = attn_weights + causal_mask + + # upcast attention to fp32 + attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query_states.dtype) + attn_weights = nn.functional.dropout(attn_weights, p=self.attention_dropout, training=self.training) + attn_output = torch.matmul(attn_weights, value_states) + + if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.transpose(1, 2).contiguous() + + attn_output = attn_output.view(bsz, q_len, -1) + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class Gemma2FlashAttention2(Gemma2Attention): + """ + Gemma2 flash attention module. This module inherits from `Gemma2Attention` as the weights of the module stays + untouched. The only required change would be on the forward pass where it needs to correctly call the public API of + flash attention and deal with padding tokens in case the input contains any of them. + """ + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + + # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1. + # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0. + # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left). + self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10() + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.LongTensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_value: Optional[Cache] = None, + output_attentions: bool = False, + use_cache: bool = False, + cache_position: Optional[torch.LongTensor] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + output_attentions = False + + bsz, q_len, _ = hidden_states.size() + + query_states = self.q_proj(hidden_states) + key_states = self.k_proj(hidden_states) + value_states = self.v_proj(hidden_states) + + # Flash attention requires the input to have the shape + # batch_size x seq_length x head_dim x hidden_dim + # therefore we just need to keep the original shape + query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2) + key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + + cos, sin = self.rotary_emb(value_states, position_ids) + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + # sin and cos are specific to RoPE models; cache_position needed for the static cache + cache_kwargs = { + "sin": sin, + "cos": cos, + "sliding_window": self.sliding_window, + "cache_position": cache_position, + } + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + if attention_mask is not None: + seq_len = attention_mask.shape[1] + key_states = key_states[:, :, :seq_len] + value_states = value_states[:, :, :seq_len] + + # TODO: These transpose are quite inefficient but Flash Attention requires the layout [batch_size, sequence_length, num_heads, head_dim]. We would need to refactor the KV cache + # to be able to avoid many of these transpose/reshape/view. + query_states = query_states.transpose(1, 2) + key_states = key_states.transpose(1, 2) + value_states = value_states.transpose(1, 2) + + dropout_rate = self.attention_dropout if self.training else 0.0 + + # In PEFT, usually we cast the layer norms in float32 for training stability reasons + # therefore the input hidden states gets silently casted in float32. Hence, we need + # cast them back in the correct dtype just to be sure everything works as expected. + # This might slowdown training & inference so it is recommended to not cast the LayerNorms + # in fp32. (Gemma2RMSNorm handles it correctly) + + input_dtype = query_states.dtype + if input_dtype == torch.float32: + if torch.is_autocast_enabled(): + target_dtype = torch.get_autocast_gpu_dtype() + # Handle the case where the model is quantized + elif hasattr(self.config, "_pre_quantization_dtype"): + target_dtype = self.config._pre_quantization_dtype + else: + target_dtype = self.q_proj.weight.dtype + + logger.warning_once( + f"The input hidden states seems to be silently casted in float32, this might be related to" + f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in" + f" {target_dtype}." + ) + + query_states = query_states.to(target_dtype) + key_states = key_states.to(target_dtype) + value_states = value_states.to(target_dtype) + + attn_output = _flash_attention_forward( + query_states, + key_states, + value_states, + attention_mask, + q_len, + dropout=dropout_rate, + softmax_scale=self.scaling, + is_causal=self.is_causal, + use_top_left_mask=self._flash_attn_uses_top_left_mask, + softcap=self.config.attn_logit_softcapping if is_flash_attn_greater_or_equal("2.6.0") else None, + ) + + attn_output = attn_output.reshape(bsz, q_len, -1).contiguous() + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class Gemma2SdpaAttention(Gemma2Attention): + """ + Gemma2 attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from + `Gemma2Attention` as the weights of the module stays untouched. The only changes are on the forward pass to adapt to + SDPA API. + """ + + # Adapted from Gemma2Attention.forward + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_value: Optional[Cache] = None, + output_attentions: bool = False, + use_cache: bool = False, + cache_position: Optional[torch.LongTensor] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + if output_attentions: + # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented. + logger.warning_once( + "Gemma2Model is using Gemma2SdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, " + 'but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.' + ) + return super().forward( + hidden_states=hidden_states, + attention_mask=attention_mask, + position_ids=position_ids, + past_key_value=past_key_value, + output_attentions=output_attentions, + use_cache=use_cache, + cache_position=cache_position, + ) + + bsz, q_len, _ = hidden_states.size() + + query_states = self.q_proj(hidden_states) + key_states = self.k_proj(hidden_states) + value_states = self.v_proj(hidden_states) + + query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2) + key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2) + + cos, sin = self.rotary_emb(value_states, position_ids) + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + # sin and cos are specific to RoPE models; cache_position needed for the static cache + cache_kwargs = { + "sin": sin, + "cos": cos, + "sliding_window": self.sliding_window, + "cache_position": cache_position, + } + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + key_states = repeat_kv(key_states, self.num_key_value_groups) + value_states = repeat_kv(value_states, self.num_key_value_groups) + + causal_mask = attention_mask + if attention_mask is not None: + causal_mask = causal_mask[:, :, :, : key_states.shape[-2]] + + # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask, + # Reference: https://github.com/pytorch/pytorch/issues/112577. + if query_states.device.type == "cuda" and causal_mask is not None: + query_states = query_states.contiguous() + key_states = key_states.contiguous() + value_states = value_states.contiguous() + + # We dispatch to SDPA's Flash Attention or Efficient kernels via this `is_causal` if statement instead of an inline conditional assignment + # in SDPA to support both torch.compile's dynamic shapes and full graph options. An inline conditional prevents dynamic shapes from compiling. + # We pass our own mask, so is_causal must be False + is_causal = True if causal_mask is None and q_len > 1 else False + + attn_output = torch.nn.functional.scaled_dot_product_attention( + query_states, + key_states, + value_states, + attn_mask=causal_mask, + dropout_p=self.attention_dropout if self.training else 0.0, + is_causal=is_causal, + scale=self.scaling, + ) + + attn_output = attn_output.transpose(1, 2).contiguous() + attn_output = attn_output.view(bsz, q_len, -1) + + attn_output = self.o_proj(attn_output) + + return attn_output, None, past_key_value + + +GEMMA2_ATTENTION_CLASSES = { + "eager": Gemma2Attention, + "flash_attention_2": Gemma2FlashAttention2, + "sdpa": Gemma2SdpaAttention, +} + + +class Gemma2DecoderLayer(nn.Module): + def __init__(self, config: CodeXEmbedConfig, layer_idx: int, is_causal: bool): + super().__init__() + self.config = config + self.hidden_size = config.hidden_size + + self.self_attn = GEMMA2_ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx, is_causal=is_causal) + + self.mlp = Gemma2MLP(config) + self.input_layernorm = Gemma2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.post_attention_layernorm = Gemma2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) + + self.is_sliding = not bool(layer_idx % 2) + self.pre_feedforward_layernorm = Gemma2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.post_feedforward_layernorm = Gemma2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.sliding_window = config.sliding_window + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_value: Optional[Cache] = None, + output_attentions: Optional[bool] = False, + use_cache: Optional[bool] = False, + cache_position: Optional[torch.LongTensor] = None, + ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]: + if self.is_sliding and attention_mask is not None: # efficient SDPA and no padding + # Flash-attn is a 2D tensor + if self.config._attn_implementation == "flash_attention_2": + attention_mask = attention_mask[:, -self.sliding_window :] + else: + min_dtype = torch.finfo(attention_mask.dtype).min + sliding_window_mask = torch.tril( + torch.ones_like(attention_mask, dtype=torch.bool), diagonal=-self.sliding_window + ) + attention_mask = torch.where(sliding_window_mask, min_dtype, attention_mask) + if attention_mask.shape[-1] <= 1: # when decoding + attention_mask = attention_mask[:, :, :, -self.sliding_window :] + + residual = hidden_states + + hidden_states = self.input_layernorm(hidden_states) + + # Self Attention + hidden_states, self_attn_weights, present_key_value = self.self_attn( + hidden_states=hidden_states, + attention_mask=attention_mask, + position_ids=position_ids, + past_key_value=past_key_value, + output_attentions=output_attentions, + use_cache=use_cache, + cache_position=cache_position, + ) + hidden_states = self.post_attention_layernorm(hidden_states) + hidden_states = residual + hidden_states + + residual = hidden_states + hidden_states = self.pre_feedforward_layernorm(hidden_states) + hidden_states = self.mlp(hidden_states) + hidden_states = self.post_feedforward_layernorm(hidden_states) + hidden_states = residual + hidden_states + + outputs = (hidden_states,) + + if output_attentions: + outputs += (self_attn_weights,) + + if use_cache: + outputs += (present_key_value,) + + return outputs + + +GEMMA2_START_DOCSTRING = r""" + This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the + library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads + etc.) + + This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass. + Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage + and behavior. + + Parameters: + config ([`CodeXEmbedConfig`]): + Model configuration class with all the parameters of the model. Initializing with a config file does not + load the weights associated with the model, only the configuration. Check out the + [`~PreTrainedModel.from_pretrained`] method to load the model weights. +""" + + +@add_start_docstrings( + "The bare Gemma2 Model outputting raw hidden-states without any specific head on top.", + GEMMA2_START_DOCSTRING, +) +class Gemma2PreTrainedModel(PreTrainedModel): + config_class = CodeXEmbedConfig + base_model_prefix = "model" + supports_gradient_checkpointing = True + _no_split_modules = ["Gemma2DecoderLayer"] + _skip_keys_device_placement = ["past_key_values"] + _supports_flash_attn_2 = True + _supports_sdpa = True + _supports_cache_class = True + _supports_quantized_cache = False + _supports_static_cache = True + + def _init_weights(self, module): + std = self.config.initializer_range + if isinstance(module, nn.Linear): + module.weight.data.normal_(mean=0.0, std=std) + if module.bias is not None: + module.bias.data.zero_() + elif isinstance(module, nn.Embedding): + module.weight.data.normal_(mean=0.0, std=std) + if module.padding_idx is not None: + module.weight.data[module.padding_idx].zero_() + + +_CONFIG_FOR_DOC = "CodeXEmbedConfig" + + +GEMMA2_INPUTS_DOCSTRING = r""" + Args: + input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): + Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide + it. + + Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and + [`PreTrainedTokenizer.__call__`] for details. + + [What are input IDs?](../glossary#input-ids) + attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*): + Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`: + + - 1 for tokens that are **not masked**, + - 0 for tokens that are **masked**. + + [What are attention masks?](../glossary#attention-mask) + + Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and + [`PreTrainedTokenizer.__call__`] for details. + + If `past_key_values` is used, optionally only the last `input_ids` have to be input (see + `past_key_values`). + + If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`] + and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more + information on the default strategy. + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): + Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, + config.n_positions - 1]`. + + [What are position IDs?](../glossary#position-ids) + past_key_values (`Cache` or `tuple(tuple(torch.FloatTensor))`, *optional*): + Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention + blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` + returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`. + + Two formats are allowed: + - a [`~cache_utils.Cache`] instance; + - Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of + shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`). This is also known as the legacy + cache format. + + The model will output the same cache format that is fed as input. If no `past_key_values` are passed, the + legacy cache format will be returned. + + If `past_key_values` are used, the user can optionally input only the last `input_ids` (those that don't + have their past key value states given to this model) of shape `(batch_size, 1)` instead of all `input_ids` + of shape `(batch_size, sequence_length)`. + inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*): + Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This + is useful if you want more control over how to convert `input_ids` indices into associated vectors than the + model's internal embedding lookup matrix. + use_cache (`bool`, *optional*): + If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see + `past_key_values`). + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned + tensors for more detail. + output_hidden_states (`bool`, *optional*): + Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for + more detail. + return_dict (`bool`, *optional*): + Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. + cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*): + Indices depicting the position of the input sequence tokens in the sequence. Contrarily to `position_ids`, + this tensor is not affected by padding. It is used to update the cache in the correct position and to infer + the complete sequence length. +""" + + +@add_start_docstrings( + "The bare Gemma2 Model outputting raw hidden-states without any specific head on top.", + GEMMA2_START_DOCSTRING, +) +class Gemma2Model(Gemma2PreTrainedModel): + """ + Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`Gemma2DecoderLayer`] + + Args: + config: CodeXEmbedConfig + """ + + def __init__(self, config: CodeXEmbedConfig, **kwargs): + super().__init__(config) + self.padding_idx = config.pad_token_id + self.vocab_size = config.vocab_size + + self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx) + self.is_causal = getattr(kwargs, 'is_causal', False) + self.layers = nn.ModuleList( + [Gemma2DecoderLayer(config, layer_idx, self.is_causal) for layer_idx in range(config.num_hidden_layers)] + ) + self.norm = Gemma2RMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.gradient_checkpointing = False + + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.embed_tokens + + def set_input_embeddings(self, value): + self.embed_tokens = value + + @add_start_docstrings_to_model_forward(GEMMA2_INPUTS_DOCSTRING) + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + cache_position: Optional[torch.LongTensor] = None, + ) -> Union[Tuple, BaseModelOutputWithPast]: + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + use_cache = use_cache if use_cache is not None else self.config.use_cache + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + if (input_ids is None) ^ (inputs_embeds is not None): + raise ValueError( + "You cannot specify both input_ids and inputs_embeds at the same time, and must specify either one" + ) + + if self.gradient_checkpointing and self.training and use_cache: + logger.warning_once( + "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`." + ) + use_cache = False + + if inputs_embeds is None: + inputs_embeds = self.embed_tokens(input_ids) + + if cache_position is None: + cache_position = torch.arange(0, inputs_embeds.shape[1], device=inputs_embeds.device) + + if position_ids is None: + position_ids = cache_position.unsqueeze(0) + + if self.is_causal: + causal_mask = self._update_attention_mask( + attention_mask, inputs_embeds, cache_position, past_key_values, output_attentions + ) + else: + causal_mask = _prepare_4d_attention_mask_for_sdpa( + attention_mask, inputs_embeds.dtype + ) + + # embed positions + hidden_states = inputs_embeds + + # normalized + # Gemma2 downcasts the below to float16, causing sqrt(3072)=55.4256 to become 55.5 + # See https://github.com/huggingface/transformers/pull/29402 + normalizer = torch.tensor(self.config.hidden_size**0.5, dtype=hidden_states.dtype) + hidden_states = hidden_states * normalizer + + all_hidden_states = () if output_hidden_states else None + all_self_attns = () if output_attentions else None + + for decoder_layer in self.layers: + if output_hidden_states: + all_hidden_states += (hidden_states,) + + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + decoder_layer.__call__, + hidden_states, + causal_mask, + position_ids, + past_key_values, + output_attentions, + use_cache, + cache_position, + ) + else: + layer_outputs = decoder_layer( + hidden_states, + attention_mask=causal_mask, + position_ids=position_ids, + past_key_value=past_key_values, + output_attentions=output_attentions, + use_cache=use_cache, + cache_position=cache_position, + ) + + hidden_states = layer_outputs[0] + + if output_attentions: + all_self_attns += (layer_outputs[1],) + + hidden_states = self.norm(hidden_states) + + # add hidden states from the last decoder layer + if output_hidden_states: + all_hidden_states += (hidden_states,) + + next_cache = past_key_values if use_cache else None + + if not return_dict: + return tuple(v for v in [hidden_states, next_cache, all_hidden_states, all_self_attns] if v is not None) + return BaseModelOutputWithPast( + last_hidden_state=hidden_states, + past_key_values=next_cache, + hidden_states=all_hidden_states, + attentions=all_self_attns, + ) + + def _update_attention_mask( + self, + attention_mask: torch.Tensor, + input_tensor: torch.Tensor, + cache_position: torch.Tensor, + past_key_values: Cache, + output_attentions: bool, + ): + # Flash Attention currently doesn't support static cache but Gemma2 work only with static cache. + # So we will pass in attention mask as is in any case, not only when ther's padding. Then we'll use its shape + # to cut out keys/values trailing 0 used in static cache. This workaround should be compile compatible + # as it doesn't cause dynamic control issues. + if self.config._attn_implementation == "flash_attention_2": + return attention_mask + + dtype, device = input_tensor.dtype, input_tensor.device + min_dtype = torch.finfo(dtype).min + sequence_length = input_tensor.shape[1] + if isinstance(past_key_values, HybridCache): + target_length = past_key_values.get_max_length() + else: + target_length = attention_mask.shape[-1] if attention_mask is not None else input_tensor.shape[1] + + # In case the provided `attention` mask is 2D, we generate a causal mask here (4D). + causal_mask = _prepare_4d_causal_attention_mask_with_cache_position( + attention_mask, + sequence_length=sequence_length, + target_length=target_length, + dtype=dtype, + device=device, + min_dtype=min_dtype, + cache_position=cache_position, + batch_size=input_tensor.shape[0], + ) + return causal_mask + + +class Gemma2ForCausalLM(Gemma2PreTrainedModel): + _tied_weights_keys = ["lm_head.weight"] + + def __init__(self, config): + super().__init__(config) + self.model = Gemma2Model(config) + self.vocab_size = config.vocab_size + self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False) + + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.model.embed_tokens + + def set_input_embeddings(self, value): + self.model.embed_tokens = value + + def get_output_embeddings(self): + return self.lm_head + + def set_output_embeddings(self, new_embeddings): + self.lm_head = new_embeddings + + def set_decoder(self, decoder): + self.model = decoder + + def get_decoder(self): + return self.model + + @add_start_docstrings_to_model_forward(GEMMA2_INPUTS_DOCSTRING) + @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + labels: Optional[torch.LongTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + cache_position: Optional[torch.LongTensor] = None, + ) -> Union[Tuple, CausalLMOutputWithPast]: + r""" + Args: + labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): + Labels for computing the masked language modeling loss. Indices should either be in `[0, ..., + config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored + (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`. + + Returns: + + Example: + + ```python + >>> from transformers import AutoTokenizer, GemmaForCausalLM + + >>> model = GemmaForCausalLM.from_pretrained("google/gemma-2-9b") + >>> tokenizer = AutoTokenizer.from_pretrained("google/gemma-2-9b") + + >>> prompt = "What is your favorite condiment?" + >>> inputs = tokenizer(prompt, return_tensors="pt") + + >>> # Generate + >>> generate_ids = model.generate(inputs.input_ids, max_length=30) + >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0] + "What is your favorite condiment?" + ```""" + if self.training and self.config._attn_implementation != "eager": + logger.warning_once( + "It is strongly recommended to train Gemma2 models with the `eager` attention implementation " + f"instead of `{self.config._attn_implementation}`. Use `eager` with `AutoModelForCausalLM.from_pretrained('', attn_implementation='eager')`." + ) + output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions + output_hidden_states = ( + output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states + ) + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn) + outputs = self.model( + input_ids=input_ids, + attention_mask=attention_mask, + position_ids=position_ids, + past_key_values=past_key_values, + inputs_embeds=inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + cache_position=cache_position, + ) + + hidden_states = outputs[0] + logits = self.lm_head(hidden_states) + if self.config.final_logit_softcapping is not None: + logits = logits / self.config.final_logit_softcapping + logits = torch.tanh(logits) + logits = logits * self.config.final_logit_softcapping + + logits = logits.float() + loss = None + if labels is not None: + # Shift so that tokens < n predict n + shift_logits = logits[..., :-1, :].contiguous() + shift_labels = labels[..., 1:].contiguous() + # Flatten the tokens + loss_fct = CrossEntropyLoss() + shift_logits = shift_logits.view(-1, self.config.vocab_size) + shift_labels = shift_labels.view(-1) + # Enable model parallelism + shift_labels = shift_labels.to(shift_logits.device) + loss = loss_fct(shift_logits, shift_labels) + + if not return_dict: + output = (logits,) + outputs[1:] + return (loss,) + output if loss is not None else output + + return CausalLMOutputWithPast( + loss=loss, + logits=logits, + past_key_values=outputs.past_key_values, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + + def prepare_inputs_for_generation( + self, + input_ids, + past_key_values=None, + attention_mask=None, + inputs_embeds=None, + cache_position=None, + position_ids=None, + use_cache=True, + **kwargs, + ): + # If we have cache: let's slice `input_ids` through `cache_position`, to keep only the unprocessed tokens + # Exception 1: when passing input_embeds, input_ids may be missing entries + # Exception 2: some generation methods do special slicing of input_ids, so we don't need to do it here + if past_key_values is not None: + if inputs_embeds is not None: # Exception 1 + input_ids = input_ids[:, -cache_position.shape[0] :] + elif input_ids.shape[1] != cache_position.shape[0]: # Default case (the "else", a no op, is Exception 2) + input_ids = input_ids[:, cache_position] + + if attention_mask is not None and position_ids is None: + # create position_ids on the fly for batch generation + position_ids = attention_mask.long().cumsum(-1) - 1 + position_ids.masked_fill_(attention_mask == 0, 1) + if past_key_values: + position_ids = position_ids[:, -input_ids.shape[1] :] + # This `clone` call is needed to avoid recapturing cuda graphs with `torch.compile`'s + # `mode="reduce-overhead`, as otherwise the input `position_ids` would have various stride + # during the decoding. Here, simply using `.contiguous()` is not sufficient as in the + # batch size = 1 case, `position_ids` is already contiguous but with varying stride + # which retriggers a capture. + position_ids = position_ids.clone(memory_format=torch.contiguous_format) + + # if `inputs_embeds` are passed, we only want to use them in the 1st generation step + if inputs_embeds is not None and cache_position[0] == 0: + model_inputs = {"inputs_embeds": inputs_embeds} + else: + # The clone here is for the same reason as for `position_ids`. + model_inputs = {"input_ids": input_ids.clone(memory_format=torch.contiguous_format)} + + if isinstance(past_key_values, HybridCache) and attention_mask.ndim == 2: + if inputs_embeds is not None: + batch_size, sequence_length = inputs_embeds.shape + device = inputs_embeds.device + else: + batch_size, sequence_length = input_ids.shape + device = input_ids.device + + dtype = self.lm_head.weight.dtype + min_dtype = torch.finfo(dtype).min + + attention_mask = _prepare_4d_causal_attention_mask_with_cache_position( + attention_mask, + sequence_length=sequence_length, + target_length=past_key_values.get_max_length(), + dtype=dtype, + device=device, + min_dtype=min_dtype, + cache_position=cache_position, + batch_size=batch_size, + ) + + model_inputs.update( + { + "position_ids": position_ids, + "cache_position": cache_position, + "past_key_values": past_key_values, + "use_cache": use_cache, + "attention_mask": attention_mask, + } + ) + return model_inputs + + +@add_start_docstrings( + """ + The Gemma2 Model transformer with a sequence classification head on top (linear layer). + + [`Gemma2ForSequenceClassification`] uses the last token in order to do the classification, as other causal models + (e.g. GPT-2) do. + + Since it does classification on the last token, it requires to know the position of the last token. If a + `pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If + no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the + padding tokens when `inputs_embeds` are passed instead of `input_ids`, it does the same (take the last value in + each row of the batch). + """, + GEMMA2_START_DOCSTRING, +) +class Gemma2ForSequenceClassification(Gemma2PreTrainedModel): + def __init__(self, config): + super().__init__(config) + self.num_labels = config.num_labels + self.model = Gemma2Model(config) + self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False) + + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.model.embed_tokens + + def set_input_embeddings(self, value): + self.model.embed_tokens = value + + @add_start_docstrings_to_model_forward(GEMMA2_INPUTS_DOCSTRING) + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + labels: Optional[torch.LongTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, SequenceClassifierOutputWithPast]: + r""" + labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): + Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., + config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If + `config.num_labels > 1` a classification loss is computed (Cross-Entropy). + """ + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + transformer_outputs = self.model( + input_ids, + attention_mask=attention_mask, + position_ids=position_ids, + past_key_values=past_key_values, + inputs_embeds=inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + hidden_states = transformer_outputs[0] + logits = self.score(hidden_states) + + if input_ids is not None: + batch_size = input_ids.shape[0] + else: + batch_size = inputs_embeds.shape[0] + + if self.config.pad_token_id is None and batch_size != 1: + raise ValueError("Cannot handle batch sizes > 1 if no padding token is defined.") + if self.config.pad_token_id is None: + sequence_lengths = -1 + else: + if input_ids is not None: + # if no pad token found, use modulo instead of reverse indexing for ONNX compatibility + sequence_lengths = torch.eq(input_ids, self.config.pad_token_id).int().argmax(-1) - 1 + sequence_lengths = sequence_lengths % input_ids.shape[-1] + sequence_lengths = sequence_lengths.to(logits.device) + else: + sequence_lengths = -1 + + pooled_logits = logits[torch.arange(batch_size, device=logits.device), sequence_lengths] + + loss = None + if labels is not None: + labels = labels.to(logits.device) + if self.config.problem_type is None: + if self.num_labels == 1: + self.config.problem_type = "regression" + elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int): + self.config.problem_type = "single_label_classification" + else: + self.config.problem_type = "multi_label_classification" + + if self.config.problem_type == "regression": + loss_fct = MSELoss() + if self.num_labels == 1: + loss = loss_fct(pooled_logits.squeeze(), labels.squeeze()) + else: + loss = loss_fct(pooled_logits, labels) + elif self.config.problem_type == "single_label_classification": + loss_fct = CrossEntropyLoss() + loss = loss_fct(pooled_logits.view(-1, self.num_labels), labels.view(-1)) + elif self.config.problem_type == "multi_label_classification": + loss_fct = BCEWithLogitsLoss() + loss = loss_fct(pooled_logits, labels) + if not return_dict: + output = (pooled_logits,) + transformer_outputs[1:] + return ((loss,) + output) if loss is not None else output + + return SequenceClassifierOutputWithPast( + loss=loss, + logits=pooled_logits, + past_key_values=transformer_outputs.past_key_values, + hidden_states=transformer_outputs.hidden_states, + attentions=transformer_outputs.attentions, + ) + + +@add_start_docstrings( + """ + The Gemma2 Model transformer with a token classification head on top (a linear layer on top of the hidden-states + output) e.g. for Named-Entity-Recognition (NER) tasks. + """, + GEMMA2_START_DOCSTRING, +) +class Gemma2ForTokenClassification(Gemma2PreTrainedModel): + def __init__(self, config): + super().__init__(config) + self.num_labels = config.num_labels + self.model = Gemma2Model(config) + if getattr(config, "classifier_dropout", None) is not None: + classifier_dropout = config.classifier_dropout + elif getattr(config, "hidden_dropout", None) is not None: + classifier_dropout = config.hidden_dropout + else: + classifier_dropout = 0.1 + self.dropout = nn.Dropout(classifier_dropout) + self.score = nn.Linear(config.hidden_size, config.num_labels) + + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.model.embed_tokens + + def set_input_embeddings(self, value): + self.model.embed_tokens = value + + @add_start_docstrings_to_model_forward(GEMMA2_INPUTS_DOCSTRING) + def forward( + self, + input_ids: Optional[torch.LongTensor] = None, + attention_mask: Optional[torch.Tensor] = None, + position_ids: Optional[torch.LongTensor] = None, + past_key_values: Optional[List[torch.FloatTensor]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + labels: Optional[torch.LongTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, TokenClassifierOutput]: + r""" + labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*): + Labels for computing the sequence classification/regression loss. Indices should be in `[0, ..., + config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If + `config.num_labels > 1` a classification loss is computed (Cross-Entropy). + """ + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + outputs = self.model( + input_ids, + attention_mask=attention_mask, + position_ids=position_ids, + past_key_values=past_key_values, + inputs_embeds=inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + sequence_output = outputs[0] + sequence_output = self.dropout(sequence_output) + logits = self.score(sequence_output) + + loss = None + if labels is not None: + loss_fct = CrossEntropyLoss() + loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1)) + + if not return_dict: + output = (logits,) + outputs[2:] + return ((loss,) + output) if loss is not None else output + + return TokenClassifierOutput( + loss=loss, + logits=logits, + hidden_states=outputs.hidden_states, + attentions=outputs.attentions, + ) + +def get_detailed_instruct(task_description: str, query: str) -> str: + return f'Instruct: {task_description}\nQuery: {query}' + +class CodeXEmbedModel2B(PreTrainedModel): + config_class = CodeXEmbedConfig + base_model_prefix = 'model' + def __init__(self, config, **kwargs): + super().__init__(config) + self.model = Gemma2Model.from_pretrained(config._name_or_path, trust_remote_code=True, is_causal=False, device_map="auto") + self.tokenizer = AutoTokenizer.from_pretrained(config._name_or_path, trust_remote_code=True, device_map="auto") + + if not self.tokenizer.pad_token: + self.tokenizer.pad_token = self.tokenizer.eos_token + self.tokenizer.padding_side = 'right' + + def forward(self, **kwargs): + return self.model(**kwargs) + + def last_token_pool(self, model_output, attention_mask): + last_hidden_states = model_output.last_hidden_state + sequence_lengths = attention_mask.sum(dim=1) - 1 + batch_size = last_hidden_states.shape[0] + return last_hidden_states[torch.arange(batch_size, device=last_hidden_states.device), sequence_lengths] + + def encode_text(self, texts: List[str], max_length: int) -> np.ndarray: + logging.info(f"Encoding {len(texts)} texts...") + + # Tokenize all texts + encoded_input = self.tokenizer( + texts, + padding=True, + truncation=True, + max_length=max_length, + return_tensors="pt" + ).to('cuda') + + # Generate embeddings + with torch.no_grad(): + model_output = self.model(**encoded_input) + embeddings = self.last_token_pool(model_output, encoded_input['attention_mask']) + + if embeddings is None: + logging.error("Embeddings are None.") + else: + logging.info(f"Encoded {len(embeddings)} embeddings.") + + return embeddings.cpu() + + def encode_queries(self, queries: List[str], max_length: int, instruction: str, **kwargs) -> np.ndarray: + all_queries = [get_detailed_instruct(instruction, query) for query in queries] + return self.encode_text(all_queries, max_length) + + def encode_corpus(self, corpus: List[str], max_length: int, + **kwargs) -> np.ndarray: + return self.encode_text(corpus, max_length) + +## AutoModel Register +AutoModel.register(CodeXEmbedConfig, CodeXEmbedModel2B) + +## Register for auto class +CodeXEmbedModel2B.register_for_auto_class("AutoModel") diff --git a/modules.json b/modules.json new file mode 100644 index 0000000..f7640f9 --- /dev/null +++ b/modules.json @@ -0,0 +1,14 @@ +[ + { + "idx": 0, + "name": "0", + "path": "", + "type": "sentence_transformers.models.Transformer" + }, + { + "idx": 1, + "name": "1", + "path": "1_Pooling", + "type": "sentence_transformers.models.Pooling" + } +] \ No newline at end of file diff --git a/sentence_bert_config.json b/sentence_bert_config.json new file mode 100644 index 0000000..42dcdfc --- /dev/null +++ b/sentence_bert_config.json @@ -0,0 +1,4 @@ +{ + "max_seq_length": 4096, + "do_lower_case": false +} \ No newline at end of file diff --git a/special_tokens_map.json b/special_tokens_map.json new file mode 100644 index 0000000..8d6368f --- /dev/null +++ b/special_tokens_map.json @@ -0,0 +1,34 @@ +{ + "additional_special_tokens": [ + "", + "" + ], + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..af0eac5 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3f289bc05132635a8bc7aca7aa21255efd5e18f3710f43e3cdb96bcd41be4922 +size 17525357 diff --git a/tokenizer.model b/tokenizer.model new file mode 100644 index 0000000..796efe9 --- /dev/null +++ b/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61a7b147390c64585d6c3543dd6fc636906c9af3865a5548f27f31aee1d4c8e2 +size 4241003 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..14ef17b --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,2013 @@ +{ + "add_bos_token": true, + "add_eos_token": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "3": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "4": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "5": { + "content": "<2mass>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "6": { + "content": "[@BOS@]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "7": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "8": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "9": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "10": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "11": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "12": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "13": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "14": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "15": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "16": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "17": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "18": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "19": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "20": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "21": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "22": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "23": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "24": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "25": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "26": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "27": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "28": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "29": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "30": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "31": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "32": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "33": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "34": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "35": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "36": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "37": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "38": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "39": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "40": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "41": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "42": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "43": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "44": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "45": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "46": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "47": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "48": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "49": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "50": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "51": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "52": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "53": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "54": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "55": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "56": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "57": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "58": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "59": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "60": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "61": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "62": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "63": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "64": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "65": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "66": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "67": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "68": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "69": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "70": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "71": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "72": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "73": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "74": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "75": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "76": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "77": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "78": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "79": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "80": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "81": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "82": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "83": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "84": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "85": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "86": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "87": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "88": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "89": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "90": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "91": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "92": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "93": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "94": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "95": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "96": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "97": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "98": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "99": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "100": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "101": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "102": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "103": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "104": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "105": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "106": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "107": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "108": { + "content": "\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "109": { + "content": "\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "110": { + "content": "\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "111": { + "content": "\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "112": { + "content": "\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "113": { + "content": "\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "114": { + "content": "\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "115": { + "content": "\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "116": { + "content": "\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "117": { + "content": "\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "118": { + "content": "\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "119": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "120": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "121": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "122": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "123": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "124": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "125": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "126": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "127": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "128": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "129": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "130": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "131": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "132": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "133": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "134": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "135": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "136": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "137": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "138": { + "content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "139": { + "content": "▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "140": { + "content": "▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "141": { + "content": "▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "142": { + "content": "▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "143": { + "content": "▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "144": { + "content": "▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "145": { + "content": "▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "146": { + "content": "▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "147": { + "content": "▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "148": { + "content": "▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "149": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "150": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "151": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "152": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "153": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "154": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "155": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "156": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "157": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "158": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "159": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "160": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "161": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "162": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "163": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "164": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "165": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "166": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "167": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "168": { + "content": "▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "169": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "170": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "172": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "173": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "174": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "175": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "171": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "176": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "177": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "178": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "179": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "180": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "181": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "182": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "183": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "184": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "185": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "186": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "187": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "188": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "189": { + "content": "

", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "190": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "191": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "192": { + "content": "
", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "193": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "194": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "195": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "196": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "197": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "198": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "199": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "200": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "201": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "202": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "203": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "204": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "205": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "206": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "207": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "208": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "209": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "210": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "211": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "212": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "213": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "214": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "215": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "216": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255968": { + "content": "[toxicity=0]", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255969": { + "content": "\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255970": { + "content": "\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255971": { + "content": "\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255972": { + "content": "\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255973": { + "content": "\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255974": { + "content": "\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255975": { + "content": "\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255976": { + "content": "\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255977": { + "content": "\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255978": { + "content": "\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255979": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255980": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255981": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255982": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255983": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255984": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255985": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255986": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255987": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255988": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255989": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255990": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255991": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255992": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255993": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255994": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255995": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255996": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255997": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255998": { + "content": "\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + }, + "255999": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": false + } + }, + "additional_special_tokens": [ + "", + "" + ], + "bos_token": "", + "chat_template": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '' + role + '\n' + message['content'] | trim + '\n' }}{% endfor %}{% if add_generation_prompt %}{{'model\n'}}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "", + "model_max_length": 1000000000000000019884624838656, + "pad_token": "", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "GemmaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +}