diff --git a/docs/source/en/_toctree.yml b/docs/source/en/_toctree.yml index ae632376f946..36588a2fcd78 100644 --- a/docs/source/en/_toctree.yml +++ b/docs/source/en/_toctree.yml @@ -410,6 +410,8 @@ title: Gemma - local: model_doc/gemma2 title: Gemma2 + - local: model_doc/glm + title: GLM - local: model_doc/openai-gpt title: GPT - local: model_doc/gpt_neo diff --git a/docs/source/en/index.md b/docs/source/en/index.md index 0a5518fd71c8..7884913785b4 100644 --- a/docs/source/en/index.md +++ b/docs/source/en/index.md @@ -136,6 +136,7 @@ Flax), PyTorch, and/or TensorFlow. | [ErnieM](model_doc/ernie_m) | ✅ | ❌ | ❌ | | [ESM](model_doc/esm) | ✅ | ✅ | ❌ | | [FairSeq Machine-Translation](model_doc/fsmt) | ✅ | ❌ | ❌ | +| [GLM](model_doc/glm) | ✅ | ❌ | ❌ | | [Falcon](model_doc/falcon) | ✅ | ❌ | ❌ | | [FalconMamba](model_doc/falcon_mamba) | ✅ | ❌ | ❌ | | [FastSpeech2Conformer](model_doc/fastspeech2_conformer) | ✅ | ❌ | ❌ | diff --git a/docs/source/en/model_doc/glm.md b/docs/source/en/model_doc/glm.md new file mode 100644 index 000000000000..d54c37f66513 --- /dev/null +++ b/docs/source/en/model_doc/glm.md @@ -0,0 +1,99 @@ + + +# GLM + +## Overview + +The GLM Model was proposed +in [ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools](https://arxiv.org/html/2406.12793v1) +by GLM Team, THUDM & ZhipuAI. + +The abstract from the paper is the following: + +*We introduce ChatGLM, an evolving family of large language models that we have been developing over time. This report +primarily focuses on the GLM-4 language series, which includes GLM-4, GLM-4-Air, and GLM-4-9B. They represent our most +capable models that are trained with all the insights and lessons gained from the preceding three generations of +ChatGLM. To date, the GLM-4 models are pre-trained on ten trillions of tokens mostly in Chinese and English, along with +a small set of corpus from 24 languages, and aligned primarily for Chinese and English usage. The high-quality alignment +is achieved via a multi-stage post-training process, which involves supervised fine-tuning and learning from human +feedback. Evaluations show that GLM-4 1) closely rivals or outperforms GPT-4 in terms of general metrics such as MMLU, +GSM8K, MATH, BBH, GPQA, and HumanEval, 2) gets close to GPT-4-Turbo in instruction following as measured by IFEval, 3) +matches GPT-4 Turbo (128K) and Claude 3 for long context tasks, and 4) outperforms GPT-4 in Chinese alignments as +measured by AlignBench. The GLM-4 All Tools model is further aligned to understand user intent and autonomously decide +when and which tool(s) to use—including web browser, Python interpreter, text-to-image model, and user-defined +functions—to effectively complete complex tasks. In practical applications, it matches and even surpasses GPT-4 All +Tools in tasks like accessing online information via web browsing and solving math problems using Python interpreter. +Over the course, we have open-sourced a series of models, including ChatGLM-6B (three generations), GLM-4-9B (128K, 1M), +GLM-4V-9B, WebGLM, and CodeGeeX, attracting over 10 million downloads on Hugging face in the year 2023 alone.* + +Tips: + +- This model was contributed by [THUDM](https://huggingface.co/THUDM). The most recent code can be + found [here](https://github.com/thudm/GLM-4). + + +## Usage tips + +`GLM-4` can be found on the [Huggingface Hub](https://huggingface.co/collections/THUDM/glm-4-665fcf188c414b03c2f7e3b7) + +In the following, we demonstrate how to use `glm-4-9b-chat` for the inference. Note that we have used the ChatML format for dialog, in this demo we show how to leverage `apply_chat_template` for this purpose. + +```python +>>> from transformers import AutoModelForCausalLM, AutoTokenizer +>>> device = "cuda" # the device to load the model onto + +>>> model = AutoModelForCausalLM.from_pretrained("THUDM/glm-4-9b-chat", device_map="auto") +>>> tokenizer = AutoTokenizer.from_pretrained("THUDM/glm-4-9b-chat") + +>>> prompt = "Give me a short introduction to large language model." + +>>> messages = [{"role": "user", "content": prompt}] + +>>> text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) + +>>> model_inputs = tokenizer([text], return_tensors="pt").to(device) + +>>> generated_ids = model.generate(model_inputs.input_ids, max_new_tokens=512, do_sample=True) + +>>> generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)] + +>>> response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0] +``` + +## GLMConfig + +[[autodoc]] GLMConfig + +## GLMModel + +[[autodoc]] GLMModel + - forward + +## GLMForCausalLM + +[[autodoc]] GLMForCausalLM + - forward + +## GLMForSequenceClassification + +[[autodoc]] GLMForSequenceClassification + - forward + +## GLMForTokenClassification + +[[autodoc]] GLMForTokenClassification + - forward diff --git a/docs/source/en/perf_infer_gpu_one.md b/docs/source/en/perf_infer_gpu_one.md index 346759aa2b25..577e39e88567 100644 --- a/docs/source/en/perf_infer_gpu_one.md +++ b/docs/source/en/perf_infer_gpu_one.md @@ -42,6 +42,7 @@ FlashAttention-2 is currently supported for the following architectures: * [Chameleon](https://huggingface.co/docs/transformers/model_doc/chameleon#transformers.Chameleon) * [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPModel) * [Cohere](https://huggingface.co/docs/transformers/model_doc/cohere#transformers.CohereModel) +* [GLM](https://huggingface.co/docs/transformers/model_doc/glm#transformers.GLMModel) * [Dbrx](https://huggingface.co/docs/transformers/model_doc/dbrx#transformers.DbrxModel) * [DistilBert](https://huggingface.co/docs/transformers/model_doc/distilbert#transformers.DistilBertModel) * [Gemma](https://huggingface.co/docs/transformers/model_doc/gemma#transformers.GemmaModel) @@ -214,6 +215,7 @@ For now, Transformers supports SDPA inference and training for the following arc * [CamemBERT](https://huggingface.co/docs/transformers/model_doc/camembert#transformers.CamembertModel) * [Chameleon](https://huggingface.co/docs/transformers/model_doc/chameleon#transformers.Chameleon) * [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPModel) +* [GLM](https://huggingface.co/docs/transformers/model_doc/glm#transformers.GLMModel) * [Cohere](https://huggingface.co/docs/transformers/model_doc/cohere#transformers.CohereModel) * [data2vec_audio](https://huggingface.co/docs/transformers/main/en/model_doc/data2vec#transformers.Data2VecAudioModel) * [Dbrx](https://huggingface.co/docs/transformers/model_doc/dbrx#transformers.DbrxModel) diff --git a/docs/source/en/quantization/hqq.md b/docs/source/en/quantization/hqq.md old mode 100644 new mode 100755 index 11489808aecb..34608cd64fd8 --- a/docs/source/en/quantization/hqq.md +++ b/docs/source/en/quantization/hqq.md @@ -30,13 +30,13 @@ To quantize a model, you need to create an [`HqqConfig`]. There are two ways of from transformers import AutoModelForCausalLM, AutoTokenizer, HqqConfig # Method 1: all linear layers will use the same quantization config -quant_config = HqqConfig(nbits=8, group_size=64, quant_zero=False, quant_scale=False, axis=0) #axis=0 is used by default +quant_config = HqqConfig(nbits=8, group_size=64) ``` ``` Python # Method 2: each linear layer with the same tag will use a dedicated quantization config -q4_config = {'nbits':4, 'group_size':64, 'quant_zero':False, 'quant_scale':False} -q3_config = {'nbits':3, 'group_size':32, 'quant_zero':False, 'quant_scale':False} +q4_config = {'nbits':4, 'group_size':64} +q3_config = {'nbits':3, 'group_size':32} quant_config = HqqConfig(dynamic_config={ 'self_attn.q_proj':q4_config, 'self_attn.k_proj':q4_config, diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index 078e4d0e4abd..0f4565508358 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -502,6 +502,7 @@ "Kosmos2Config", "Kosmos2Processor", ], + "models.glm": ["GlmConfig"], "models.layoutlm": [ "LayoutLMConfig", "LayoutLMTokenizer", @@ -2549,6 +2550,15 @@ "LlamaPreTrainedModel", ] ) + _import_structure["models.glm"].extend( + [ + "GlmForCausalLM", + "GlmForSequenceClassification", + "GlmForTokenClassification", + "GlmModel", + "GlmPreTrainedModel", + ] + ) _import_structure["models.llava"].extend( [ "LlavaForConditionalGeneration", @@ -5266,6 +5276,7 @@ GitProcessor, GitVisionConfig, ) + from .models.glm import GlmConfig from .models.glpn import GLPNConfig from .models.gpt2 import ( GPT2Config, @@ -6977,6 +6988,13 @@ GitPreTrainedModel, GitVisionModel, ) + from .models.glm import ( + GlmForCausalLM, + GlmForSequenceClassification, + GlmForTokenClassification, + GlmModel, + GlmPreTrainedModel, + ) from .models.glpn import ( GLPNForDepthEstimation, GLPNModel, diff --git a/src/transformers/integrations/hqq.py b/src/transformers/integrations/hqq.py index 10a6d06a3f9f..162b365668a0 100755 --- a/src/transformers/integrations/hqq.py +++ b/src/transformers/integrations/hqq.py @@ -66,6 +66,10 @@ def _prepare_for_hqq_linear(model, patch_params, has_been_replaced, current_key_ has_been_replaced = True + # Add these fake parameters to avoid loading fail + for att in ["W_q", "meta"]: + setattr(module, att, None) + if len(list(module.children())) > 0: _, has_been_replaced = _prepare_for_hqq_linear( module, @@ -97,7 +101,7 @@ def prepare_for_hqq_linear(model, quantization_config=None, modules_to_not_conve # Convert quantization_config to layer-wise config skip_modules = quantization_config.skip_modules - quant_config = quantization_config.to_dict() + quant_config = quantization_config.quant_config linear_tags = list(set(linear_tags) - set(skip_modules) - set(modules_to_not_convert)) if any(key in linear_tags for key in quant_config.keys()): @@ -113,7 +117,11 @@ def prepare_for_hqq_linear(model, quantization_config=None, modules_to_not_conve ) # We store quantization config as linear_tag -> hqq quant config - model.config.quantization_config = patch_params + model.config.quantization_config = { + "quant_config": quant_config, + "quant_method": quantization_config.quant_method, + "skip_modules": skip_modules, + } if not has_been_replaced: logger.warning("No linear modules were found in your model for quantization.") diff --git a/src/transformers/modeling_utils.py b/src/transformers/modeling_utils.py index fc0d6748cd1d..df0519566766 100755 --- a/src/transformers/modeling_utils.py +++ b/src/transformers/modeling_utils.py @@ -934,12 +934,17 @@ def _load_state_dict_into_meta_model( # For compatibility with PyTorch load_state_dict which converts state dict dtype to existing dtype in model, and which # uses `param.copy_(input_param)` that preserves the contiguity of the parameter in the model. # Reference: https://github.com/pytorch/pytorch/blob/db79ceb110f6646523019a59bbd7b838f43d4a86/torch/nn/modules/module.py#L2040C29-L2040C29 + old_param = model splits = param_name.split(".") for split in splits: old_param = getattr(old_param, split) + # Not all the attributes of a module are Parameters/Tensor + if not isinstance(old_param, (torch.nn.Parameter, torch.Tensor)): + old_param = None if old_param is None: break + if old_param is not None: if dtype is None: param = param.to(old_param.dtype) @@ -3819,6 +3824,7 @@ def from_pretrained( from_pt = not (from_tf | from_flax) # load pt weights early so that we know which dtype to init the model under + if from_pt: if not is_sharded and state_dict is None: # Time to load the checkpoint @@ -4176,6 +4182,9 @@ def _load_pretrained_model( expected_keys = list(model_state_dict.keys()) prefix = model.base_model_prefix + if hf_quantizer is not None: + expected_keys = hf_quantizer.update_expected_keys(model, expected_keys, loaded_keys) + def _fix_key(key): if "beta" in key: return key.replace("beta", "bias") @@ -4290,7 +4299,7 @@ def _fix_key(key): value = torch.empty(*param.size(), dtype=target_dtype) if ( not is_quantized - or getattr(hf_quantizer, "requires_parameters_quantization", False) + or (getattr(hf_quantizer, "requires_parameters_quantization", False)) or not hf_quantizer.check_quantized_param( model, param_value=value, param_name=key, state_dict={} ) diff --git a/src/transformers/models/__init__.py b/src/transformers/models/__init__.py index e47a4ed9c342..0b2c20f6eb80 100644 --- a/src/transformers/models/__init__.py +++ b/src/transformers/models/__init__.py @@ -97,6 +97,7 @@ gemma, gemma2, git, + glm, glpn, gpt2, gpt_bigcode, diff --git a/src/transformers/models/auto/configuration_auto.py b/src/transformers/models/auto/configuration_auto.py index 6d55f87d60ac..4e4e6603b404 100644 --- a/src/transformers/models/auto/configuration_auto.py +++ b/src/transformers/models/auto/configuration_auto.py @@ -114,6 +114,7 @@ ("gemma", "GemmaConfig"), ("gemma2", "Gemma2Config"), ("git", "GitConfig"), + ("glm", "GlmConfig"), ("glpn", "GLPNConfig"), ("gpt-sw3", "GPT2Config"), ("gpt2", "GPT2Config"), @@ -413,6 +414,7 @@ ("gemma", "Gemma"), ("gemma2", "Gemma2"), ("git", "GIT"), + ("glm", "GLM"), ("glpn", "GLPN"), ("gpt-sw3", "GPT-Sw3"), ("gpt2", "OpenAI GPT-2"), diff --git a/src/transformers/models/auto/modeling_auto.py b/src/transformers/models/auto/modeling_auto.py index 6e730e848db7..37598380dfab 100644 --- a/src/transformers/models/auto/modeling_auto.py +++ b/src/transformers/models/auto/modeling_auto.py @@ -111,6 +111,7 @@ ("gemma", "GemmaModel"), ("gemma2", "Gemma2Model"), ("git", "GitModel"), + ("glm", "GlmModel"), ("glpn", "GLPNModel"), ("gpt-sw3", "GPT2Model"), ("gpt2", "GPT2Model"), @@ -483,6 +484,7 @@ ("gemma", "GemmaForCausalLM"), ("gemma2", "Gemma2ForCausalLM"), ("git", "GitForCausalLM"), + ("glm", "GlmForCausalLM"), ("gpt-sw3", "GPT2LMHeadModel"), ("gpt2", "GPT2LMHeadModel"), ("gpt_bigcode", "GPTBigCodeForCausalLM"), @@ -909,6 +911,7 @@ ("funnel", "FunnelForSequenceClassification"), ("gemma", "GemmaForSequenceClassification"), ("gemma2", "Gemma2ForSequenceClassification"), + ("glm", "GlmForSequenceClassification"), ("gpt-sw3", "GPT2ForSequenceClassification"), ("gpt2", "GPT2ForSequenceClassification"), ("gpt_bigcode", "GPTBigCodeForSequenceClassification"), @@ -1093,6 +1096,7 @@ ("funnel", "FunnelForTokenClassification"), ("gemma", "GemmaForTokenClassification"), ("gemma2", "Gemma2ForTokenClassification"), + ("glm", "GlmForTokenClassification"), ("gpt-sw3", "GPT2ForTokenClassification"), ("gpt2", "GPT2ForTokenClassification"), ("gpt_bigcode", "GPTBigCodeForTokenClassification"), diff --git a/src/transformers/models/auto/tokenization_auto.py b/src/transformers/models/auto/tokenization_auto.py index 6a5cba11f094..3be273d012a4 100644 --- a/src/transformers/models/auto/tokenization_auto.py +++ b/src/transformers/models/auto/tokenization_auto.py @@ -204,6 +204,7 @@ ), ), ("git", ("BertTokenizer", "BertTokenizerFast" if is_tokenizers_available() else None)), + ("glm", (None, "PreTrainedTokenizerFast" if is_tokenizers_available() else None)), ("gpt-sw3", ("GPTSw3Tokenizer" if is_sentencepiece_available() else None, None)), ("gpt2", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)), ("gpt_bigcode", ("GPT2Tokenizer", "GPT2TokenizerFast" if is_tokenizers_available() else None)), diff --git a/src/transformers/models/glm/__init__.py b/src/transformers/models/glm/__init__.py new file mode 100644 index 000000000000..263f43a5ff9c --- /dev/null +++ b/src/transformers/models/glm/__init__.py @@ -0,0 +1,60 @@ +# Copyright 2020 The HuggingFace 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 +from typing import TYPE_CHECKING + +from ...utils import ( + OptionalDependencyNotAvailable, + _LazyModule, + is_torch_available, +) + + +_import_structure = { + "configuration_glm": ["GlmConfig"], +} + +try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_glm"] = [ + "GlmForCausalLM", + "GlmModel", + "GlmPreTrainedModel", + "GlmForSequenceClassification", + "GlmForTokenClassification", + ] + +if TYPE_CHECKING: + from .configuration_glm import GlmConfig + + try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_glm import ( + GlmForCausalLM, + GlmForSequenceClassification, + GlmForTokenClassification, + GlmModel, + GlmPreTrainedModel, + ) + +else: + import sys + + sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure, module_spec=__spec__) \ No newline at end of file diff --git a/src/transformers/models/glm/configuration_glm.py b/src/transformers/models/glm/configuration_glm.py new file mode 100644 index 000000000000..02041feaafd6 --- /dev/null +++ b/src/transformers/models/glm/configuration_glm.py @@ -0,0 +1,152 @@ +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# 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 modular. If any change should be done, please apply the change to the +# modular_xxx.py file directly. One of our CI enforces this +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# coding=utf-8 +# Copyright 2024 The GLM & ZhipuAI team and 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 ...configuration_utils import PretrainedConfig + + +class GlmConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`GlmModel`]. It is used to instantiate an Glm + 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 Glm-7B. + e.g. [google/glm-7b](https://huggingface.co/google/glm-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 Glm model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`GlmModel`] + 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_act (`str` or `function`, *optional*, defaults to `"gelu_pytorch_tanh"`): + The legacy activation function. It is overwritten by the `hidden_activation`. + hidden_activation (`str` or `function`, *optional*): + The non-linear activation function (function or string) in the decoder. Will default to `"gelu_pytorch_tanh"` + if not specified. `"gelu_pytorch_tanh"` uses an approximation of the `"gelu"` activation function. + 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. + ```python + >>> from transformers import GlmModel, GlmConfig + >>> # Initializing a Glm glm-7b style configuration + >>> configuration = GlmConfig() + >>> # Initializing a model from the glm-7b style configuration + >>> model = GlmModel(configuration) + >>> # Accessing the model configuration + >>> configuration = model.config + ``` + resid_pdrop (`float`, *optional*, defaults to `0.0`): + Dropout ratio in the decoder layers. + linear_bias (`bool`, *optional*, defaults to `False`): + Whether to use a bias in the MLP layers, as well as the query, key, value and output projection layers during self-attention. + """ + + model_type = "glm" + keys_to_ignore_at_inference = ["past_key_values"] + + def __init__( + self, + vocab_size=151552, + hidden_size=4096, + intermediate_size=13696, + num_hidden_layers=40, + num_attention_heads=32, + num_key_value_heads=2, + head_dim=128, + hidden_act="silu", + resid_pdrop=0.0, + attention_dropout=0.0, + max_position_embeddings=131072, + initializer_range=0.02, + rms_norm_eps=0.00000015625, + use_cache=True, + tie_word_embeddings=False, + rope_theta=10000.0, + pad_token_id=151329, + eos_token_id=[151329, 151336, 151338], + bos_token_id=None, + attention_bias=True, + linear_bias=False, + **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_act = hidden_act + self.initializer_range = initializer_range + self.rms_norm_eps = rms_norm_eps + self.use_cache = use_cache + self.rope_theta = rope_theta + self.attention_bias = attention_bias + self.attention_dropout = attention_dropout + self.resid_pdrop = resid_pdrop + self.linear_bias = linear_bias + + 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, + ) diff --git a/src/transformers/models/glm/convert_glm_weights_to_hf.py b/src/transformers/models/glm/convert_glm_weights_to_hf.py new file mode 100644 index 000000000000..ef5eda117593 --- /dev/null +++ b/src/transformers/models/glm/convert_glm_weights_to_hf.py @@ -0,0 +1,166 @@ +import argparse +import json +import os + +import torch +from safetensors.torch import load_file +from tokenizers import Regex, Tokenizer, decoders, pre_tokenizers, processors + +from transformers import AutoTokenizer, GlmConfig, GlmForCausalLM, PreTrainedTokenizerFast +from transformers.convert_slow_tokenizer import TikTokenConverter + + +STATE_DICT_MAPPING = { + "transformer.output_layer.": "lm_head.", + "transformer.": "model.", + ".embedding.word_embeddings.": ".embed_tokens.", + ".encoder.final_layernorm.": ".norm.", + ".encoder.layers.": ".layers.", + "rotary_pos_embed.": "rotary_emb.", + "self_attention.": "self_attn.", + "query_key_value.": "qkv_proj.", + "dense.": "o_proj.", + "dense_h_to_4h.": "gate_up_proj.", + "dense_4h_to_h.": "down_proj.", +} + + +class GlmConverter(TikTokenConverter): + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + + def converted(self) -> Tokenizer: + tokenizer = self.tokenizer() + tokenizer.pre_tokenizer = pre_tokenizers.Sequence( + [ + pre_tokenizers.Split(Regex(self.pattern), behavior="isolated", invert=False), + pre_tokenizers.ByteLevel(add_prefix_space=self.add_prefix_space, use_regex=False), + ] + ) + tokenizer.decoder = decoders.ByteLevel() + tokenizer.add_special_tokens(self.additional_special_tokens) + + tokenizer.post_processor = processors.Sequence( + [ + processors.ByteLevel(trim_offsets=False), + processors.TemplateProcessing( + single="[gMASK]:0 :0 $A:0", + pair="[gMASK]:0 :0 $A:0 $B:1", + special_tokens=[("[gMASK]", 151331), ("", 151333)], + ), + ], + ) + + return tokenizer + + +def merge_safetensors(input_dir: str): + all_files = [os.path.join(input_dir, x) for x in os.listdir(input_dir) if x.endswith(".safetensors")] + all_files = sorted(all_files, key=lambda x: int(x.rsplit("-", 3)[1])) + + all_weights = {} + for file in all_files: + tensors = load_file(file) + all_weights.update(tensors) + + return all_weights + + +def convert_state_dict(original_state_dict: dict): + new_dict = {} + + for key, value in original_state_dict.items(): + # Should not be part of the state dict + if "rotary_pos_emb.inv_freq" in key: + continue + + new_key = key + for old, new in STATE_DICT_MAPPING.items(): + new_key = new_key.replace(old, new) + + new_dict[new_key] = value + return new_dict + + +def convert_config(original_config: dict): + num_attention_heads = original_config.pop("num_attention_heads") + + new_config = GlmConfig( + vocab_size=original_config.pop("padded_vocab_size"), + hidden_size=original_config.pop("hidden_size"), + intermediate_size=original_config.pop("ffn_hidden_size"), + num_hidden_layers=original_config.pop("num_layers"), + num_attention_heads=num_attention_heads, + num_key_value_heads=( + num_attention_heads + if not original_config.pop("multi_query_attention") + else original_config.pop("multi_query_group_num") + ), + resid_pdrop=original_config.pop("hidden_dropout"), + attention_dropout=original_config.pop("attention_dropout"), + max_position_embeddings=original_config.pop("seq_length"), + rms_norm_eps=original_config.pop("layernorm_epsilon"), + rope_theta=10000.0 * original_config.pop("rope_ratio", 1), + use_cache=original_config.pop("use_cache"), + head_dim=original_config.pop("kv_channels"), + attention_bias=original_config.pop("add_qkv_bias"), + linear_bias=original_config.pop("add_bias_linear"), + eos_token_id=original_config.pop("eos_token_id"), + pad_token_id=original_config.pop("pad_token_id"), + tie_word_embeddings=original_config.pop("tie_word_embeddings"), + ) + print(f"Unused config keys: {original_config.keys(),}") + return new_config + + +def convert_glm_tokenizer(input_dir): + fast_tok = GlmConverter(os.path.join(input_dir, "tokenizer.model"), additional_special_tokens=[]).converted() + tokenizer = AutoTokenizer.from_pretrained("THUDM/glm-4-9b", trust_remote_code=True) + new_tok = PreTrainedTokenizerFast( + tokenizer_object=fast_tok, + bos_token=tokenizer.bos_token, + eos_token=tokenizer.eos_token, + pad_token=tokenizer.pad_token, + clean_up_tokenization_spaces=tokenizer.clean_up_tokenization_spaces, + additional_special_tokens=tokenizer.additional_special_tokens, + padding_side=tokenizer.padding_side, + ) + + return new_tok + + +def convert_glm_model(input_dir, output_dir): + # Load and convert config + with open(os.path.join(input_dir, "config.json")) as f: + original_config = json.load(f) + config = convert_config(original_config) + config.save_pretrained(output_dir) + + # Load and convert weights + original_state_dict = merge_safetensors(input_dir) + new_dict = convert_state_dict(original_state_dict) + with torch.device("meta"): + model = GlmForCausalLM(config) + model.load_state_dict(new_dict, strict=True, assign=True) + model.save_pretrained(output_dir) + + # Load and convert tokenizer + tokenizer = convert_glm_tokenizer(input_dir) + tokenizer.save_pretrained(output_dir) + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument( + "input_dir", + type=str, + help="Location of the local folder copied from the Hub.", + ) + parser.add_argument( + "output_dir", + type=str, + help="Location to write HF model and tokenizer", + ) + + args = parser.parse_args() + convert_glm_model(args.input_dir, args.output_dir) diff --git a/src/transformers/models/glm/modeling_glm.py b/src/transformers/models/glm/modeling_glm.py new file mode 100644 index 000000000000..91909da2b563 --- /dev/null +++ b/src/transformers/models/glm/modeling_glm.py @@ -0,0 +1,1389 @@ +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# 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 modular. If any change should be done, please apply the change to the +# modular_xxx.py file directly. One of our CI enforces this +# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨 +# coding=utf-8 +# Copyright 2024 The GLM & ZhipuAI team and 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. +import math +from typing import List, Optional, Tuple, Union + +import torch +import torch.nn as nn +import torch.utils.checkpoint +from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss + +from ...activations import ACT2FN +from ...cache_utils import Cache, DynamicCache, StaticCache +from ...generation import GenerationMixin +from ...modeling_attn_mask_utils import AttentionMaskConverter +from ...modeling_flash_attention_utils import _flash_attention_forward +from ...modeling_outputs import ( + BaseModelOutputWithPast, + CausalLMOutputWithPast, + SequenceClassifierOutputWithPast, + TokenClassifierOutput, +) +from ...modeling_utils import PreTrainedModel +from ...utils import ( + add_start_docstrings, + add_start_docstrings_to_model_forward, + is_flash_attn_2_available, + is_flash_attn_greater_or_equal_2_10, + is_torchdynamo_compiling, + logging, + replace_return_docstrings, +) +from .configuration_glm import GlmConfig + + +if is_flash_attn_2_available(): + from ...modeling_flash_attention_utils import _flash_attention_forward + + +logger = logging.get_logger(__name__) + + +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 + + +# Copied from transformers.models.llama.modeling_llama.LlamaRMSNorm with Llama->Glm +class GlmRMSNorm(nn.Module): + def __init__(self, hidden_size, eps=1e-6): + """ + GlmRMSNorm is equivalent to T5LayerNorm + """ + super().__init__() + self.weight = nn.Parameter(torch.ones(hidden_size)) + self.variance_epsilon = eps + + def forward(self, hidden_states): + input_dtype = hidden_states.dtype + hidden_states = hidden_states.to(torch.float32) + variance = hidden_states.pow(2).mean(-1, keepdim=True) + hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon) + return self.weight * hidden_states.to(input_dtype) + + def extra_repr(self): + return f"{tuple(self.weight.shape)}, eps={self.variance_epsilon}" + + +# Copied from transformers.models.gemma.modeling_gemma.GemmaRotaryEmbedding with gemma->glm, Gemma->Glm +class GlmRotaryEmbedding(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[..., 0::2] + x2 = x[..., 1::2] + return torch.stack((-x2, x1), dim=-1).flatten(-2) + + +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) + + # Interleave them instead of usual shape + cos = cos[..., : cos.shape[-1] // 2].repeat_interleave(2, dim=-1) + sin = sin[..., : sin.shape[-1] // 2].repeat_interleave(2, dim=-1) + + # Keep half for later concatenation + q, q_pass = q[..., : q.shape[-1] // 2], q[..., q.shape[-1] // 2 :] + k, k_pass = k[..., : k.shape[-1] // 2], k[..., k.shape[-1] // 2 :] + + # Apply rotary embeddings on the first half + q_embed = (q * cos) + (rotate_half(q) * sin) + k_embed = (k * cos) + (rotate_half(k) * sin) + + # Concatenate back to full shape + q_embed = torch.cat([q_embed, q_pass], dim=-1) + k_embed = torch.cat([k_embed, k_pass], dim=-1) + return q_embed, k_embed + + +class GlmMLP(nn.Module): + def __init__(self, config): + super().__init__() + + self.config = config + + self.gate_up_proj = nn.Linear(config.hidden_size, 2 * config.intermediate_size, bias=self.config.linear_bias) + self.down_proj = nn.Linear(config.intermediate_size, config.hidden_size, bias=self.config.linear_bias) + + self.activation_fn = ACT2FN[config.hidden_act] + + def forward(self, hidden_states: torch.FloatTensor) -> torch.FloatTensor: + up_states = self.gate_up_proj(hidden_states) + + gate, up_states = up_states.chunk(2, dim=-1) + up_states = up_states * self.activation_fn(gate) + + return self.down_proj(up_states) + + +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 GlmAttention(nn.Module): + def __init__(self, config: GlmConfig, layer_idx: Optional[int] = None): + 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_bias = config.attention_bias + self.attention_dropout = config.attention_dropout + self.hidden_size = config.hidden_size + self.num_heads = config.num_attention_heads + self.head_dim = self.hidden_size // self.num_heads + self.num_key_value_heads = config.num_key_value_heads + self.num_key_value_groups = self.num_heads // self.num_key_value_heads + self.is_causal = True + + if (self.head_dim * self.num_heads) != self.hidden_size: + raise ValueError( + f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}" + f" and `num_heads`: {self.num_heads})." + ) + + op_size = self.num_heads * self.head_dim + 2 * (self.num_key_value_heads * self.head_dim) + self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=self.config.linear_bias) + self.qkv_proj = nn.Linear(self.hidden_size, op_size, bias=self.attention_bias or self.config.linear_bias) + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + logger.warning_once("You are not running the flash-attention implementation, expect numerical differences.") + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + # repeat k/v heads if n_kv_heads < n_heads + 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)) / math.sqrt(self.head_dim) + + if attention_mask is not None: + causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] + attn_weights += causal_mask + + # upcast attention to fp32 + attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(value_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.reshape(bsz, q_len, self.hidden_size) + + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class GlmFlashAttention2(GlmAttention): + """ + GLM flash attention module. This module inherits from `GlmAttention` 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.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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + output_attentions = False + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + # 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + # repeat k/v heads if n_kv_heads < n_heads + key_states = repeat_kv(key_states, self.num_key_value_groups) + value_states = repeat_kv(value_states, self.num_key_value_groups) + + attn_dropout = 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. + + if query_states.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.qkv_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) + + # Reashape to the expected shape for Flash Attention + query_states = query_states.transpose(1, 2) + key_states = key_states.transpose(1, 2) + value_states = value_states.transpose(1, 2) + + attn_output = _flash_attention_forward( + query_states, + key_states, + value_states, + attention_mask, + q_len, + position_ids=position_ids, + dropout=attn_dropout, + sliding_window=getattr(self.config, "sliding_window", None), + use_top_left_mask=self._flash_attn_uses_top_left_mask, + is_causal=self.is_causal, + ) + + attn_output = attn_output.reshape(bsz, q_len, self.hidden_size).contiguous() + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class GlmSdpaAttention(GlmAttention): + """ + GLM attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from + `GlmAttention` as the weights of the module stays untouched. The only changes are on the forward pass to adapt to + SDPA API. + """ + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + if output_attentions: + logger.warning_once( + "GlmModel is using GlmSdpaAttention, 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, + ) + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + 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 = attention_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 attention_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. + # The q_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case q_len == 1. + 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, + ) + + attn_output = attn_output.transpose(1, 2).contiguous() + attn_output = attn_output.view(bsz, q_len, self.hidden_size) + + attn_output = self.o_proj(attn_output) + + return attn_output, None, past_key_value + + +GLM_ATTENTION_CLASSES = { + "eager": GlmAttention, + "flash_attention_2": GlmFlashAttention2, + "sdpa": GlmSdpaAttention, +} + + +class GlmDecoderLayer(nn.Module): + def __init__(self, config: GlmConfig, layer_idx: int): + super().__init__() + + self.config = config + self.self_attn = GLM_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx=layer_idx) + + self.mlp = GlmMLP(config) + self.input_layernorm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + + self.resid_attn_dropout = nn.Dropout(config.resid_pdrop) + self.resid_mlp_dropout = nn.Dropout(config.resid_pdrop) + self.post_attention_layernorm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, # will become mandatory in v4.46 + **kwargs, + ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]: + """ + Args: + hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` + attention_mask (`torch.FloatTensor`, *optional*): + attention mask of size `(batch_size, sequence_length)` if flash attention is used or `(batch_size, 1, + query_sequence_length, key_sequence_length)` if default attention is used. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + 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`). + past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states + cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*): + Indices depicting the position of the input sequence tokens in the sequence + position_embeddings (`Tuple[torch.FloatTensor, torch.FloatTensor]`, *optional*): + Tuple containing the cosine and sine positional embeddings of shape `(batch_size, seq_len, head_dim)`, + with `head_dim` being the embedding dimension of each attention head. + kwargs (`dict`, *optional*): + Arbitrary kwargs to be ignored, used for FSDP and other methods that injects code + into the model + """ + + residual = hidden_states + hidden_states = self.input_layernorm(hidden_states) + + # Self Attention + attn_outputs, 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, + position_embeddings=position_embeddings, + ) + + hidden_states = residual + self.resid_attn_dropout(attn_outputs) + residual = hidden_states + + hidden_states = self.post_attention_layernorm(hidden_states) + + hidden_states = self.mlp(hidden_states) + hidden_states = residual + self.resid_mlp_dropout(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (self_attn_weights,) + + if use_cache: + outputs += (present_key_value,) + + return outputs + + +GLM_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 ([`GlmConfig`]): + 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 Glm Model outputting raw hidden-states without any specific head on top.", + GLM_START_DOCSTRING, +) +class GlmPreTrainedModel(PreTrainedModel): + config_class = GlmConfig + base_model_prefix = "model" + supports_gradient_checkpointing = True + _no_split_modules = ["GlmDecoderLayer"] + _skip_keys_device_placement = ["past_key_values"] + _supports_flash_attn_2 = True + _supports_sdpa = True + _supports_cache_class = True + _supports_quantized_cache = True + _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 = "GlmConfig" + + +GLM_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, see our + [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache); + - 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 Glm Model outputting raw hidden-states without any specific head on top.", + GLM_START_DOCSTRING, +) +class GlmModel(GlmPreTrainedModel): + """ + Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`GlmDecoderLayer`] + + Args: + config: GlmConfig + """ + + def __init__(self, config: GlmConfig): + 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.layers = nn.ModuleList( + [GlmDecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)] + ) + self.norm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.rotary_emb = GlmRotaryEmbedding( + dim=config.head_dim // 2, max_position_embeddings=config.max_position_embeddings, base=config.rope_theta + ) + 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(GLM_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) + + # kept for BC (non `Cache` `past_key_values` inputs) + return_legacy_cache = False + if use_cache and not isinstance(past_key_values, Cache): + return_legacy_cache = True + if past_key_values is None: + past_key_values = DynamicCache() + else: + past_key_values = DynamicCache.from_legacy_cache(past_key_values) + logger.warning_once( + "We detected that you are passing `past_key_values` as a tuple of tuples. This is deprecated and " + "will be removed in v4.47. Please convert your cache or use an appropriate `Cache` class " + "(https://huggingface.co/docs/transformers/kv_cache#legacy-cache-format)" + ) + + if cache_position is None: + past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 + cache_position = torch.arange( + past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device + ) + if position_ids is None: + position_ids = cache_position.unsqueeze(0) + + causal_mask = self._update_causal_mask( + attention_mask, inputs_embeds, cache_position, past_key_values, output_attentions + ) + hidden_states = inputs_embeds + + # create position embeddings to be shared across the decoder layers + position_embeddings = self.rotary_emb(hidden_states, position_ids) + + # decoder layers + all_hidden_states = () if output_hidden_states else None + all_self_attns = () if output_attentions else None + next_decoder_cache = 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, + position_embeddings, + ) + 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, + position_embeddings=position_embeddings, + ) + + hidden_states = layer_outputs[0] + + if use_cache: + next_decoder_cache = layer_outputs[2 if output_attentions else 1] + + 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 = next_decoder_cache if use_cache else None + if return_legacy_cache: + next_cache = next_cache.to_legacy_cache() + + 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_causal_mask( + self, + attention_mask: torch.Tensor, + input_tensor: torch.Tensor, + cache_position: torch.Tensor, + past_key_values: Cache, + output_attentions: bool, + ): + if self.config._attn_implementation == "flash_attention_2": + if attention_mask is not None and 0.0 in attention_mask: + return attention_mask + return None + + # For SDPA, when possible, we will rely on its `is_causal` argument instead of its `attn_mask` argument, in + # order to dispatch on Flash Attention 2. This feature is not compatible with static cache, as SDPA will fail + # to infer the attention mask. + past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0 + using_static_cache = isinstance(past_key_values, StaticCache) + + # When output attentions is True, sdpa implementation's forward method calls the eager implementation's forward + if self.config._attn_implementation == "sdpa" and not using_static_cache and not output_attentions: + if AttentionMaskConverter._ignore_causal_mask_sdpa( + attention_mask, + inputs_embeds=input_tensor, + past_key_values_length=past_seen_tokens, + is_training=self.training, + ): + return None + + dtype, device = input_tensor.dtype, input_tensor.device + min_dtype = torch.finfo(dtype).min + sequence_length = input_tensor.shape[1] + if using_static_cache: + target_length = past_key_values.get_max_length() + else: + target_length = ( + attention_mask.shape[-1] + if isinstance(attention_mask, torch.Tensor) + else past_seen_tokens + sequence_length + 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], + ) + + if ( + self.config._attn_implementation == "sdpa" + and attention_mask is not None + and attention_mask.device.type == "cuda" + and not output_attentions + ): + # Attend to all tokens in fully masked rows in the causal_mask, for example the relevant first rows when + # using left padding. This is required by F.scaled_dot_product_attention memory-efficient attention path. + # Details: https://github.com/pytorch/pytorch/issues/110213 + causal_mask = AttentionMaskConverter._unmask_unattended(causal_mask, min_dtype) + + return causal_mask + + +class GlmForCausalLM(GlmPreTrainedModel, GenerationMixin): + _tied_weights_keys = ["lm_head.weight"] + + def __init__(self, config): + super().__init__(config) + self.model = GlmModel(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(GLM_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, + num_logits_to_keep: int = 0, + ) -> 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]`. + + num_logits_to_keep (`int`, *optional*): + Calculate logits for the last `num_logits_to_keep` tokens. If `0`, calculate logits for all + `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that + token can save memory, which becomes pretty significant for long sequences or large vocabulary size. + + Returns: + + Example: + + ```python + >>> from transformers import AutoTokenizer, GlmForCausalLM + + >>> model = GlmForCausalLM.from_pretrained("google/glm-7b") + >>> tokenizer = AutoTokenizer.from_pretrained("google/glm-7b") + + >>> 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?" + ```""" + 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] + if labels is None and not is_torchdynamo_compiling(): + logger.warning_once( + "Starting from v4.46, the `logits` model output will have the same type as the model (except at train time, where it will always be FP32)" + ) + # Only compute necessary logits, and do not upcast them to float if we are not computing the loss + # TODO: remove the float() operation in v4.46 + logits = self.lm_head(hidden_states[:, -num_logits_to_keep:, :]).float() + + loss = None + if labels is not None: + # Upcast to float if we need to compute the loss to avoid potential precision issues + logits = logits.float() + # 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, + num_logits_to_keep=None, + **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, "input_ids": None} + 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), "inputs_embeds": None} + + if isinstance(past_key_values, StaticCache) and attention_mask.ndim == 2: + if model_inputs["inputs_embeds"] is not None: + batch_size, sequence_length, _ = model_inputs["inputs_embeds"].shape + device = model_inputs["inputs_embeds"].device + else: + batch_size, sequence_length = model_inputs["input_ids"].shape + device = model_inputs["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, + ) + + if num_logits_to_keep is not None: + model_inputs["num_logits_to_keep"] = num_logits_to_keep + + 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 Glm Model transformer with a sequence classification head on top (linear layer). + + [`GlmForSequenceClassification`] 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). + """, + GLM_START_DOCSTRING, +) +class GlmForSequenceClassification(GlmPreTrainedModel): + def __init__(self, config): + super().__init__(config) + self.num_labels = config.num_labels + self.model = GlmModel(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(GLM_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[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 Glm 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. + """, + GLM_START_DOCSTRING, +) +class GlmForTokenClassification(GlmPreTrainedModel): + def __init__(self, config): + super().__init__(config) + self.num_labels = config.num_labels + self.model = GlmModel(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(GLM_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, + ) diff --git a/src/transformers/models/glm/modular_glm.py b/src/transformers/models/glm/modular_glm.py new file mode 100644 index 000000000000..ba3af15eec05 --- /dev/null +++ b/src/transformers/models/glm/modular_glm.py @@ -0,0 +1,575 @@ +# coding=utf-8 +# Copyright 2024 The GLM & ZhipuAI team and 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. +import math +from typing import Optional, Tuple + +import torch +import torch.nn as nn +import torch.utils.checkpoint + +from ...cache_utils import Cache +from ...modeling_flash_attention_utils import _flash_attention_forward +from ...utils import ( + is_flash_attn_2_available, + is_flash_attn_greater_or_equal_2_10, + logging, +) +from ..gemma.configuration_gemma import GemmaConfig +from ..gemma.modeling_gemma import ( + GemmaForCausalLM, + GemmaForSequenceClassification, + GemmaForTokenClassification, +) +from ..llama.modeling_llama import ( + LlamaModel, +) +from ..phi3.modeling_phi3 import ( + Phi3MLP, + Phi3RMSNorm, + Phi3RotaryEmbedding, +) + + +if is_flash_attn_2_available(): + from ...modeling_flash_attention_utils import _flash_attention_forward + + +logger = logging.get_logger(__name__) + + +class GlmConfig(GemmaConfig): + """ + resid_pdrop (`float`, *optional*, defaults to `0.0`): + Dropout ratio in the decoder layers. + linear_bias (`bool`, *optional*, defaults to `False`): + Whether to use a bias in the MLP layers, as well as the query, key, value and output projection layers during self-attention. + """ + + model_type = "glm" + + def __init__( + self, + vocab_size=151552, + hidden_size=4096, + intermediate_size=13696, + num_hidden_layers=40, + num_attention_heads=32, + num_key_value_heads=2, + head_dim=128, + hidden_act="silu", + resid_pdrop=0.0, + attention_dropout=0.0, + max_position_embeddings=131072, + initializer_range=0.02, + rms_norm_eps=0.00000015625, + use_cache=True, + tie_word_embeddings=False, + rope_theta=10000.0, + pad_token_id=151329, + eos_token_id=[151329, 151336, 151338], + bos_token_id=None, + attention_bias=True, + linear_bias=False, + **kwargs, + ): + super().__init__( + **kwargs, + ) + self.resid_pdrop = resid_pdrop + self.linear_bias = linear_bias + del self.hidden_activation + + +class GlmRMSNorm(Phi3RMSNorm): + pass + + +class GlmRotaryEmbedding(Phi3RotaryEmbedding): + pass + + +class GlmMLP(Phi3MLP): + def __init__(self, config): + super().__init__(config) + + self.gate_up_proj = nn.Linear(config.hidden_size, 2 * config.intermediate_size, bias=self.config.linear_bias) + self.down_proj = nn.Linear(config.intermediate_size, config.hidden_size, bias=self.config.linear_bias) + + +def rotate_half(x): + """Rotates half the hidden dims of the input.""" + x1 = x[..., 0::2] + x2 = x[..., 1::2] + return torch.stack((-x2, x1), dim=-1).flatten(-2) + + +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) + + # Interleave them instead of usual shape + cos = cos[..., : cos.shape[-1] // 2].repeat_interleave(2, dim=-1) + sin = sin[..., : sin.shape[-1] // 2].repeat_interleave(2, dim=-1) + + # Keep half for later concatenation + q, q_pass = q[..., : q.shape[-1] // 2], q[..., q.shape[-1] // 2 :] + k, k_pass = k[..., : k.shape[-1] // 2], k[..., k.shape[-1] // 2 :] + + # Apply rotary embeddings on the first half + q_embed = (q * cos) + (rotate_half(q) * sin) + k_embed = (k * cos) + (rotate_half(k) * sin) + + # Concatenate back to full shape + q_embed = torch.cat([q_embed, q_pass], dim=-1) + k_embed = torch.cat([k_embed, k_pass], dim=-1) + return q_embed, k_embed + + +class GlmAttention(nn.Module): + def __init__(self, config: GlmConfig, layer_idx: Optional[int] = None): + 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_bias = config.attention_bias + self.attention_dropout = config.attention_dropout + self.hidden_size = config.hidden_size + self.num_heads = config.num_attention_heads + self.head_dim = self.hidden_size // self.num_heads + self.num_key_value_heads = config.num_key_value_heads + self.num_key_value_groups = self.num_heads // self.num_key_value_heads + self.is_causal = True + + if (self.head_dim * self.num_heads) != self.hidden_size: + raise ValueError( + f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}" + f" and `num_heads`: {self.num_heads})." + ) + + op_size = self.num_heads * self.head_dim + 2 * (self.num_key_value_heads * self.head_dim) + self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=self.config.linear_bias) + self.qkv_proj = nn.Linear(self.hidden_size, op_size, bias=self.attention_bias or self.config.linear_bias) + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + logger.warning_once("You are not running the flash-attention implementation, expect numerical differences.") + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + # repeat k/v heads if n_kv_heads < n_heads + 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)) / math.sqrt(self.head_dim) + + if attention_mask is not None: + causal_mask = attention_mask[:, :, :, : key_states.shape[-2]] + attn_weights += causal_mask + + # upcast attention to fp32 + attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(value_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.reshape(bsz, q_len, self.hidden_size) + + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class GlmFlashAttention2(GlmAttention): + """ + GLM flash attention module. This module inherits from `GlmAttention` 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.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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + output_attentions = False + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + # 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs) + + # repeat k/v heads if n_kv_heads < n_heads + key_states = repeat_kv(key_states, self.num_key_value_groups) + value_states = repeat_kv(value_states, self.num_key_value_groups) + + attn_dropout = 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. + + if query_states.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.qkv_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) + + # Reashape to the expected shape for Flash Attention + query_states = query_states.transpose(1, 2) + key_states = key_states.transpose(1, 2) + value_states = value_states.transpose(1, 2) + + attn_output = _flash_attention_forward( + query_states, + key_states, + value_states, + attention_mask, + q_len, + position_ids=position_ids, + dropout=attn_dropout, + sliding_window=getattr(self.config, "sliding_window", None), + use_top_left_mask=self._flash_attn_uses_top_left_mask, + is_causal=self.is_causal, + ) + + attn_output = attn_output.reshape(bsz, q_len, self.hidden_size).contiguous() + attn_output = self.o_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + +class GlmSdpaAttention(GlmAttention): + """ + GLM attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from + `GlmAttention` as the weights of the module stays untouched. The only changes are on the forward pass to adapt to + SDPA API. + """ + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + if output_attentions: + logger.warning_once( + "GlmModel is using GlmSdpaAttention, 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, + ) + + bsz, q_len, _ = hidden_states.size() + + qkv = self.qkv_proj(hidden_states) + query_pos = self.num_heads * self.head_dim + query_states = qkv[..., :query_pos] + key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim] + value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :] + + 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 = position_embeddings + query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin) + + if past_key_value is not None: + cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position} # Specific to RoPE models + 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 = attention_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 attention_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. + # The q_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case q_len == 1. + 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, + ) + + attn_output = attn_output.transpose(1, 2).contiguous() + attn_output = attn_output.view(bsz, q_len, self.hidden_size) + + attn_output = self.o_proj(attn_output) + + return attn_output, None, past_key_value + + +GLM_ATTENTION_CLASSES = { + "eager": GlmAttention, + "flash_attention_2": GlmFlashAttention2, + "sdpa": GlmSdpaAttention, +} + + +class GlmDecoderLayer(nn.Module): + def __init__(self, config: GlmConfig, layer_idx: int): + super().__init__() + + self.config = config + self.self_attn = GLM_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx=layer_idx) + + self.mlp = GlmMLP(config) + self.input_layernorm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + + self.resid_attn_dropout = nn.Dropout(config.resid_pdrop) + self.resid_mlp_dropout = nn.Dropout(config.resid_pdrop) + self.post_attention_layernorm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + + 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, + position_embeddings: Optional[Tuple[torch.Tensor, torch.Tensor]] = None, # will become mandatory in v4.46 + **kwargs, + ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]: + """ + Args: + hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` + attention_mask (`torch.FloatTensor`, *optional*): + attention mask of size `(batch_size, sequence_length)` if flash attention is used or `(batch_size, 1, + query_sequence_length, key_sequence_length)` if default attention is used. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + 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`). + past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states + cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*): + Indices depicting the position of the input sequence tokens in the sequence + position_embeddings (`Tuple[torch.FloatTensor, torch.FloatTensor]`, *optional*): + Tuple containing the cosine and sine positional embeddings of shape `(batch_size, seq_len, head_dim)`, + with `head_dim` being the embedding dimension of each attention head. + kwargs (`dict`, *optional*): + Arbitrary kwargs to be ignored, used for FSDP and other methods that injects code + into the model + """ + + residual = hidden_states + hidden_states = self.input_layernorm(hidden_states) + + # Self Attention + attn_outputs, 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, + position_embeddings=position_embeddings, + ) + + hidden_states = residual + self.resid_attn_dropout(attn_outputs) + residual = hidden_states + + hidden_states = self.post_attention_layernorm(hidden_states) + + hidden_states = self.mlp(hidden_states) + hidden_states = residual + self.resid_mlp_dropout(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (self_attn_weights,) + + if use_cache: + outputs += (present_key_value,) + + return outputs + + +class GlmModel(LlamaModel): + def __init__(self, config: GlmConfig): + super().__init__(config) + self.layers = nn.ModuleList( + [GlmDecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)] + ) + self.norm = GlmRMSNorm(config.hidden_size, eps=config.rms_norm_eps) + self.rotary_emb = GlmRotaryEmbedding( + dim=config.head_dim // 2, max_position_embeddings=config.max_position_embeddings, base=config.rope_theta + ) + self.gradient_checkpointing = False + + # Initialize weights and apply final processing + self.post_init() + + +class GlmForCausalLM(GemmaForCausalLM): + def __init__(self, config): + super().__init__(config) + self.model = GlmModel(config) + self.post_init() + + +class GlmForSequenceClassification(GemmaForSequenceClassification): + def __init__(self, config): + super().__init__(config) + self.model = GlmModel(config) + self.post_init() + + +class GlmForTokenClassification(GemmaForTokenClassification): + def __init__(self, config): + super().__init__(config) + self.model = GlmModel(config) + self.post_init() diff --git a/src/transformers/quantizers/base.py b/src/transformers/quantizers/base.py old mode 100644 new mode 100755 index 73b3dbd8b259..015c0015cf7e --- a/src/transformers/quantizers/base.py +++ b/src/transformers/quantizers/base.py @@ -109,6 +109,18 @@ def update_missing_keys(self, model, missing_keys: List[str], prefix: str) -> Li """ return missing_keys + def update_expected_keys(self, model, expected_keys: List[str], loaded_keys: List[str]) -> List[str]: + """ + Override this method if you want to adjust the `update_expected_keys`. + + Args: + expected_keys (`List[str]`, *optional*): + The list of the expected keys in the initialized model. + loaded_keys (`List[str]`, *optional*): + The list of the loaded keys in the checkpoint. + """ + return expected_keys + def get_special_dtypes_update(self, model, torch_dtype: "torch.dtype") -> Dict[str, "torch.dtype"]: """ returns dtypes for modules that are not quantized - used for the computation of the device_map in case diff --git a/src/transformers/quantizers/quantizer_hqq.py b/src/transformers/quantizers/quantizer_hqq.py index cd32a99c00ac..775fea8f4901 100755 --- a/src/transformers/quantizers/quantizer_hqq.py +++ b/src/transformers/quantizers/quantizer_hqq.py @@ -62,7 +62,7 @@ def __init__(self, quantization_config, **kwargs): def validate_environment(self, *args, **kwargs): if not (is_hqq_available()): raise ImportError( - "HQQ is not available. Please follow the instructions to install it: `https://github.com/mobiusml/hqq/`" + "A valid HQQ version (>=0.2.1) is not available. Please follow the instructions to install it: `https://github.com/mobiusml/hqq/`." ) if kwargs.get("from_tf", False) or kwargs.get("from_flax", False): @@ -91,6 +91,65 @@ def validate_environment(self, *args, **kwargs): else: self.using_multi_gpu = len(set(device_map.values())) > 1 + def update_missing_keys( + self, model: "PreTrainedModel", missing_keys: List[str], prefix: str, **kwargs + ) -> List[str]: + if self.pre_quantized: + return [key for key in missing_keys if ("weight" not in key)] + else: + return missing_keys + + # Adds missing keys for HQQLinear modules that are loaded but the model with initialized with torch.nn.Linear + def update_expected_keys( + self, model: "PreTrainedModel", expected_keys: List[str], loaded_keys: List[str] + ) -> List[str]: + if not self.pre_quantized: + return expected_keys + + # Collects all quantizable (linear) layers + def _find_hqq_quantizable_layers(model, layers): + for name, module in model.named_children(): + if isinstance(module, (torch.nn.Linear)): + layers.add(module.name) + _find_hqq_quantizable_layers(module, layers) + + new_keys = set(expected_keys) + if is_hqq_available(): + from hqq.core.quantize import HQQLinear + + # Name modules + for name, module in model.named_modules(): + module.name = name + + # valid modules are Linear layers that have HQQLinear state_dict. We ignore skip_modules and any layers with Linear state_dict() params + _valid_modules = set() + _find_hqq_quantizable_layers(model, _valid_modules) + _valid_modules -= set(model.config.quantization_config["skip_modules"]) + + # Append new expected layers based on _ref_keys + _ref_keys = HQQLinear( + linear_layer=None, quant_config=None, compute_dtype=torch.float16, device="cpu" + ).state_dict_keys() - {"bias"} + + # Clean-up + _rm_keys = set() + for key in new_keys: + if any(_module in key for _module in _valid_modules): + _rm_keys.add(key) + new_keys -= _rm_keys + # At this point, new_keys contains all the keys of the layers that are NOT HQQLinear or torch.nn.Linear + + # Re-populate Linear/HQQLinear + for _module in _valid_modules: + if _module + ".weight" in loaded_keys: + new_keys.add(_module + ".weight") + else: + new_keys.update({_module + "." + _ref_key for _ref_key in _ref_keys}) + if _module + ".bias" in loaded_keys: + new_keys.add(_module + ".bias") + + return list(new_keys) + def check_quantized_param( self, model: "PreTrainedModel", @@ -99,9 +158,18 @@ def check_quantized_param( state_dict: Dict[str, Any], **kwargs, ) -> bool: + if is_hqq_available(): + from hqq.core.quantize import HQQLinear module, tensor_name = get_module_from_name(model, param_name) - return isinstance(module, torch.nn.Linear) and (tensor_name == "weight") + if self.pre_quantized: + return ( + (isinstance(module, torch.nn.Linear) or isinstance(module, HQQLinear)) + and tensor_name != "weight" + and tensor_name != "bias" + ) + else: + return isinstance(module, torch.nn.Linear) and tensor_name == "weight" def create_quantized_param( self, @@ -122,13 +190,43 @@ def create_quantized_param( from hqq.core.quantize import HQQLinear module, tensor_name = get_module_from_name(model, param_name) - - layer_name = param_name.replace(".weight", "").replace(".bias", "") + layer_name = ".".join(param_name.split(".")[:-1]) parent_module = find_parent(model, layer_name) node = layer_name.split(".")[-1] - # Step 0: set module state_dict - module_state_dict = {key.split(".")[-1]: state_dict[key] for key in state_dict if layer_name in key} + # set module state_dict + module_state_dict = {} + for k, v in state_dict.items(): + if layer_name + "." in k: + module_state_dict[k.split(".")[-1]] = v + if unexpected_keys is not None and k in unexpected_keys: + unexpected_keys.remove(k) + + if self.pre_quantized: + if isinstance(module, HQQLinear): + return + else: + hqq_layer = HQQLinear( + linear_layer=None, + quant_config=None, + compute_dtype=self.torch_dtype, + device=target_device, + ) + + hqq_layer.load_state_dict(module_state_dict) + + if hqq_layer.bias is not None and isinstance(hqq_layer.bias, torch.Tensor): + hqq_layer.bias = torch.nn.Parameter(hqq_layer.bias) + + if self.using_multi_gpu: + hqq_layer = self._patch_layer_for_multigpu(hqq_layer) + + setattr(parent_module, node, hqq_layer) + + # cleanup + del module.__dict__, module + torch.cuda.empty_cache() + return # Step 1: populate module with weight/bias from module state dict for key in module_state_dict: @@ -136,7 +234,6 @@ def create_quantized_param( # Step 2: Replace module with either HQQLinear or move it to device. We do this via setattr on the parent as doing on it on the module # directly doesn't work. - if hasattr(module, "quant_config"): hqq_layer = HQQLinear( module, @@ -192,7 +289,7 @@ def _process_model_after_weight_loading(self, model: "PreTrainedModel", **kwargs return model def is_serializable(self, safe_serialization=None): - return False + return True @property def is_trainable(self) -> bool: diff --git a/src/transformers/utils/dummy_pt_objects.py b/src/transformers/utils/dummy_pt_objects.py index f4e471ee7ab5..0de12317fe7a 100644 --- a/src/transformers/utils/dummy_pt_objects.py +++ b/src/transformers/utils/dummy_pt_objects.py @@ -1404,6 +1404,41 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["torch"]) +class GLMForCausalLM(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class GLMForSequenceClassification(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class GLMForTokenClassification(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class GLMModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class GLMPreTrainedModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + class BigBirdPegasusForQuestionAnswering(metaclass=DummyObject): _backends = ["torch"] diff --git a/src/transformers/utils/import_utils.py b/src/transformers/utils/import_utils.py index 169d3491053e..a98b17e4bd57 100755 --- a/src/transformers/utils/import_utils.py +++ b/src/transformers/utils/import_utils.py @@ -92,6 +92,7 @@ def _is_package_available(pkg_name: str, return_version: bool = False) -> Union[ FSDP_MIN_VERSION = "1.12.0" GGUF_MIN_VERSION = "0.10.0" XLA_FSDPV2_MIN_VERSION = "2.2.0" +HQQ_MIN_VERSION = "0.2.1" _accelerate_available, _accelerate_version = _is_package_available("accelerate", return_version=True) @@ -181,7 +182,7 @@ def _is_package_available(pkg_name: str, return_version: bool = False) -> Union[ _torchdistx_available = _is_package_available("torchdistx") _torchvision_available = _is_package_available("torchvision") _mlx_available = _is_package_available("mlx") -_hqq_available = _is_package_available("hqq") +_hqq_available, _hqq_version = _is_package_available("hqq", return_version=True) _tiktoken_available = _is_package_available("tiktoken") _blobfile_available = _is_package_available("blobfile") _liger_kernel_available = _is_package_available("liger_kernel") @@ -323,8 +324,8 @@ def is_torch_deterministic(): return True -def is_hqq_available(): - return _hqq_available +def is_hqq_available(min_version: str = HQQ_MIN_VERSION): + return _hqq_available and version.parse(_hqq_version) >= version.parse(min_version) def is_pygments_available(): diff --git a/src/transformers/utils/quantization_config.py b/src/transformers/utils/quantization_config.py index 19166f9ed92a..8be0bb672e51 100755 --- a/src/transformers/utils/quantization_config.py +++ b/src/transformers/utils/quantization_config.py @@ -193,15 +193,9 @@ class HqqConfig(QuantizationConfigMixin): Number of bits. Supported values are (8, 4, 3, 2, 1). group_size (`int`, *optional*, defaults to 64): Group-size value. Supported values are any value that is divisble by weight.shape[axis]). - quant_zero (`bool`, *optional*, defaults to `True`): - Quantize the zero-point if set to `True`. - quant_scale (`bool`, *optional*, defaults to `False`): - Quantize the scaling if set to `True`. - offload_meta (`bool`, *optional*, defaults to `False`): - Offload the meta-data to the CPU if set to `True`. view_as_float (`bool`, *optional*, defaults to `False`): View the quantized weight as float (used in distributed training) if set to `True`. - axis (`int`, *optional*, defaults to 0): + axis (`Optional[int]`, *optional*): Axis along which grouping is performed. Supported values are 0 or 1. dynamic_config (dict, *optional*): Parameters for dynamic configuration. The key is the name tag of the layer and the value is a quantization config. @@ -216,11 +210,8 @@ def __init__( self, nbits: int = 4, group_size: int = 64, - quant_zero: bool = True, - quant_scale: bool = False, - offload_meta: bool = False, view_as_float: bool = False, - axis: int = 0, + axis: Optional[int] = None, dynamic_config: Optional[dict] = None, skip_modules: List[str] = ["lm_head"], **kwargs, @@ -228,6 +219,16 @@ def __init__( if is_hqq_available(): from hqq.core.quantize import BaseQuantizeConfig as HQQBaseQuantizeConfig + for deprecated_key in ["quant_zero", "quant_scale", "offload_meta"]: + if deprecated_key in kwargs: + logger.info( + deprecated_key + " is deprecated. This parameter will be ignored in quantization settings." + ) + + if axis is None: + axis = 1 + logger.info("Setting axis=1 as faster backends such as TorchAO or BitBlas are only compatible with it.") + if axis not in [0, 1]: raise ValueError("Invalid axis value. Only 0 and 1 are allowed.") @@ -240,9 +241,6 @@ def __init__( **{ "nbits": nbits, "group_size": group_size, - "quant_zero": quant_zero, - "quant_scale": quant_scale, - "offload_meta": offload_meta, "view_as_float": view_as_float, "axis": axis, } @@ -259,12 +257,26 @@ def post_init(self): """ pass + @classmethod + def from_dict(cls, config: Dict[str, Any]): + """ + Override from_dict, used in AutoQuantizationConfig.from_dict in quantizers/auto.py + """ + instance = cls() + instance.quant_config = config["quant_config"] + instance.skip_modules = config["skip_modules"] + return instance + def to_dict(self) -> Dict[str, Any]: """ Serializes this instance to a Python dictionary. Returns: `Dict[str, Any]`: Dictionary of all the attributes that make up this configuration instance. """ - return self.quant_config + return { + "quant_config": self.quant_config, + "quant_method": self.quant_method, + "skip_modules": self.skip_modules, + } def __repr__(self): config_dict = self.to_dict() diff --git a/tests/models/glm/__init__.py b/tests/models/glm/__init__.py new file mode 100644 index 000000000000..e69de29bb2d1 diff --git a/tests/models/glm/test_modeling_glm.py b/tests/models/glm/test_modeling_glm.py new file mode 100644 index 000000000000..57be6d28bfe2 --- /dev/null +++ b/tests/models/glm/test_modeling_glm.py @@ -0,0 +1,508 @@ +# coding=utf-8 +# Copyright 2024 The 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. +"""Testing suite for the PyTorch Glm model.""" + +import unittest + +import pytest + +from transformers import AutoModelForCausalLM, AutoTokenizer, GlmConfig, is_torch_available +from transformers.testing_utils import ( + require_flash_attn, + require_torch, + require_torch_accelerator, + require_torch_sdpa, + slow, + torch_device, +) + +from ...generation.test_utils import GenerationTesterMixin +from ...test_configuration_common import ConfigTester +from ...test_modeling_common import ModelTesterMixin, ids_tensor +from ...test_pipeline_mixin import PipelineTesterMixin + + +if is_torch_available(): + import torch + + from transformers import ( + GlmForCausalLM, + GlmForSequenceClassification, + GlmForTokenClassification, + GlmModel, + ) + + +@require_torch +class GlmModelTester: + config_class = GlmConfig + if is_torch_available(): + model_class = GlmModel + for_causal_lm_class = GlmForCausalLM + for_sequence_class = GlmForSequenceClassification + for_token_class = GlmForTokenClassification + + def __init__( + self, + parent, + batch_size=13, + seq_length=7, + is_training=True, + use_input_mask=True, + use_token_type_ids=False, + use_labels=True, + vocab_size=99, + hidden_size=32, + num_hidden_layers=2, + num_attention_heads=4, + num_key_value_heads=2, + intermediate_size=37, + hidden_act="silu", + attention_dropout=0.1, + resid_pdrop=0.1, + max_position_embeddings=512, + initializer_range=0.02, + num_labels=3, + num_choices=4, + pad_token_id=0, + scope=None, + ): + self.parent = parent + self.batch_size = batch_size + self.seq_length = seq_length + self.is_training = is_training + self.use_input_mask = use_input_mask + self.use_token_type_ids = use_token_type_ids + self.use_labels = use_labels + self.vocab_size = vocab_size + self.hidden_size = hidden_size + self.num_hidden_layers = num_hidden_layers + self.num_attention_heads = num_attention_heads + self.num_key_value_heads = num_key_value_heads + self.intermediate_size = intermediate_size + self.hidden_act = hidden_act + self.resid_pdrop = resid_pdrop + self.attention_dropout = attention_dropout + self.max_position_embeddings = max_position_embeddings + self.initializer_range = initializer_range + self.num_labels = num_labels + self.num_choices = num_choices + self.pad_token_id = pad_token_id + self.scope = scope + self.head_dim = self.hidden_size // self.num_attention_heads + + # Copied from tests.models.mistral.test_modeling_mistral.MistralModelTester.prepare_config_and_inputs + def prepare_config_and_inputs(self): + input_ids = ids_tensor([self.batch_size, self.seq_length], self.vocab_size) + + input_mask = None + if self.use_input_mask: + input_mask = torch.tril(torch.ones(self.batch_size, self.seq_length)).to(torch_device) + + token_type_ids = None + if self.use_token_type_ids: + token_type_ids = ids_tensor([self.batch_size, self.seq_length], self.type_vocab_size) + + sequence_labels = None + token_labels = None + choice_labels = None + if self.use_labels: + sequence_labels = ids_tensor([self.batch_size], self.type_sequence_label_size) + token_labels = ids_tensor([self.batch_size, self.seq_length], self.num_labels) + choice_labels = ids_tensor([self.batch_size], self.num_choices) + + config = self.get_config() + + return config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels + + def get_config(self): + return self.config_class( + vocab_size=self.vocab_size, + hidden_size=self.hidden_size, + num_hidden_layers=self.num_hidden_layers, + num_attention_heads=self.num_attention_heads, + num_key_value_heads=self.num_key_value_heads, + intermediate_size=self.intermediate_size, + hidden_act=self.hidden_act, + attention_dropout=self.attention_dropout, + resid_pdrop=self.resid_pdrop, + max_position_embeddings=self.max_position_embeddings, + initializer_range=self.initializer_range, + pad_token_id=self.pad_token_id, + head_dim=self.head_dim, + ) + + def create_and_check_model( + self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels + ): + model = self.model_class(config=config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=input_mask) + result = model(input_ids) + self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size)) + + def create_and_check_model_as_decoder( + self, + config, + input_ids, + token_type_ids, + input_mask, + sequence_labels, + token_labels, + choice_labels, + encoder_hidden_states, + encoder_attention_mask, + ): + model = self.model_class(config) + model.to(torch_device) + model.eval() + result = model( + input_ids, + attention_mask=input_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + ) + result = model( + input_ids, + attention_mask=input_mask, + encoder_hidden_states=encoder_hidden_states, + ) + result = model(input_ids, attention_mask=input_mask) + self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size)) + + def create_and_check_for_causal_lm( + self, + config, + input_ids, + token_type_ids, + input_mask, + sequence_labels, + token_labels, + choice_labels, + encoder_hidden_states, + encoder_attention_mask, + ): + model = self.for_causal_lm_class(config=config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=input_mask, labels=token_labels) + self.parent.assertEqual(result.logits.shape, (self.batch_size, self.seq_length, self.vocab_size)) + + def create_and_check_decoder_model_past_large_inputs( + self, + config, + input_ids, + token_type_ids, + input_mask, + sequence_labels, + token_labels, + choice_labels, + encoder_hidden_states, + encoder_attention_mask, + ): + model = self.for_causal_lm_class(config=config) + model.to(torch_device) + model.eval() + + # first forward pass + outputs = model( + input_ids, + attention_mask=input_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + use_cache=True, + ) + past_key_values = outputs.past_key_values + + # create hypothetical multiple next token and extent to next_input_ids + next_tokens = ids_tensor((self.batch_size, 3), config.vocab_size) + next_mask = ids_tensor((self.batch_size, 3), vocab_size=2) + + # append to next input_ids and + next_input_ids = torch.cat([input_ids, next_tokens], dim=-1) + next_attention_mask = torch.cat([input_mask, next_mask], dim=-1) + + output_from_no_past = model( + next_input_ids, + attention_mask=next_attention_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + output_hidden_states=True, + )["hidden_states"][0] + output_from_past = model( + next_tokens, + attention_mask=next_attention_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + past_key_values=past_key_values, + output_hidden_states=True, + )["hidden_states"][0] + + # select random slice + random_slice_idx = ids_tensor((1,), output_from_past.shape[-1]).item() + output_from_no_past_slice = output_from_no_past[:, -3:, random_slice_idx].detach() + output_from_past_slice = output_from_past[:, :, random_slice_idx].detach() + + self.parent.assertTrue(output_from_past_slice.shape[1] == next_tokens.shape[1]) + + # test that outputs are equal for slice + self.parent.assertTrue(torch.allclose(output_from_past_slice, output_from_no_past_slice, atol=1e-3)) + + # Copied from tests.models.llama.test_modeling_llama.LlamaModelTester.prepare_config_and_inputs_for_common with Llama->Glm + def prepare_config_and_inputs_for_common(self): + config_and_inputs = self.prepare_config_and_inputs() + ( + config, + input_ids, + token_type_ids, + input_mask, + sequence_labels, + token_labels, + choice_labels, + ) = config_and_inputs + inputs_dict = {"input_ids": input_ids, "attention_mask": input_mask} + return config, inputs_dict + + +@require_torch +class GlmModelTest(ModelTesterMixin, GenerationTesterMixin, PipelineTesterMixin, unittest.TestCase): + all_model_classes = ( + (GlmModel, GlmForCausalLM, GlmForSequenceClassification, GlmForTokenClassification) + if is_torch_available() + else () + ) + all_generative_model_classes = (GlmForCausalLM,) if is_torch_available() else () + pipeline_model_mapping = ( + { + "feature-extraction": GlmModel, + "text-classification": GlmForSequenceClassification, + "token-classification": GlmForTokenClassification, + "text-generation": GlmForCausalLM, + } + if is_torch_available() + else {} + ) + test_headmasking = False + test_pruning = False + + # Need to remove 0.9 in `test_cpu_offload` + # This is because we are hitting edge cases with the causal_mask buffer + model_split_percents = [0.5, 0.6] + + # used in `test_torch_compile` + _torch_compile_test_ckpt = "THUDM/glm-4-9b" + + # TODO (ydshieh): Check this. See https://app.circleci.com/pipelines/github/huggingface/transformers/79245/workflows/9490ef58-79c2-410d-8f51-e3495156cf9c/jobs/1012146 + def is_pipeline_test_to_skip( + self, pipeline_test_casse_name, config_class, model_architecture, tokenizer_name, processor_name + ): + return True + + def setUp(self): + self.model_tester = GlmModelTester(self) + self.config_tester = ConfigTester(self, config_class=GlmConfig, hidden_size=37) + + def test_config(self): + self.config_tester.run_common_tests() + + def test_model(self): + config_and_inputs = self.model_tester.prepare_config_and_inputs() + self.model_tester.create_and_check_model(*config_and_inputs) + + def test_model_various_embeddings(self): + config_and_inputs = self.model_tester.prepare_config_and_inputs() + for type in ["absolute", "relative_key", "relative_key_query"]: + config_and_inputs[0].position_embedding_type = type + self.model_tester.create_and_check_model(*config_and_inputs) + + def test_Glm_sequence_classification_model(self): + config, input_dict = self.model_tester.prepare_config_and_inputs_for_common() + print(config) + config.num_labels = 3 + input_ids = input_dict["input_ids"] + attention_mask = input_ids.ne(1).to(torch_device) + sequence_labels = ids_tensor([self.model_tester.batch_size], self.model_tester.type_sequence_label_size) + model = self.model_tester.for_sequence_class(config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=attention_mask, labels=sequence_labels) + self.assertEqual(result.logits.shape, (self.model_tester.batch_size, self.model_tester.num_labels)) + + def test_Glm_sequence_classification_model_for_single_label(self): + config, input_dict = self.model_tester.prepare_config_and_inputs_for_common() + config.num_labels = 3 + config.problem_type = "single_label_classification" + input_ids = input_dict["input_ids"] + attention_mask = input_ids.ne(1).to(torch_device) + sequence_labels = ids_tensor([self.model_tester.batch_size], self.model_tester.type_sequence_label_size) + model = self.model_tester.for_sequence_class(config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=attention_mask, labels=sequence_labels) + self.assertEqual(result.logits.shape, (self.model_tester.batch_size, self.model_tester.num_labels)) + + def test_Glm_sequence_classification_model_for_multi_label(self): + config, input_dict = self.model_tester.prepare_config_and_inputs_for_common() + config.num_labels = 3 + config.problem_type = "multi_label_classification" + input_ids = input_dict["input_ids"] + attention_mask = input_ids.ne(1).to(torch_device) + sequence_labels = ids_tensor( + [self.model_tester.batch_size, config.num_labels], self.model_tester.type_sequence_label_size + ).to(torch.float) + model = self.model_tester.for_sequence_class(config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=attention_mask, labels=sequence_labels) + self.assertEqual(result.logits.shape, (self.model_tester.batch_size, self.model_tester.num_labels)) + + def test_Glm_token_classification_model(self): + config, input_dict = self.model_tester.prepare_config_and_inputs_for_common() + config.num_labels = 3 + input_ids = input_dict["input_ids"] + attention_mask = input_ids.ne(1).to(torch_device) + token_labels = ids_tensor([self.model_tester.batch_size, self.model_tester.seq_length], config.num_labels) + model = self.model_tester.for_token_class(config=config) + model.to(torch_device) + model.eval() + result = model(input_ids, attention_mask=attention_mask, labels=token_labels) + self.assertEqual( + result.logits.shape, + (self.model_tester.batch_size, self.model_tester.seq_length, self.model_tester.num_labels), + ) + + @unittest.skip(reason="Glm uses GQA on all models so the KV cache is a non standard format") + def test_past_key_values_format(self): + pass + + +@slow +@require_torch_accelerator +class GlmIntegrationTest(unittest.TestCase): + input_text = ["Hello I am doing", "Hi today"] + # This variable is used to determine which CUDA device are we using for our runners (A10 or T4) + # Depending on the hardware we get different logits / generations + cuda_compute_capability_major_version = None + + @classmethod + def setUpClass(cls): + if is_torch_available() and torch.cuda.is_available(): + # 8 is for A100 / A10 and 7 for T4 + cls.cuda_compute_capability_major_version = torch.cuda.get_device_capability()[0] + + def test_model_9b_fp16(self): + model_id = "THUDM/glm-4-9b" + EXPECTED_TEXTS = [ + "Hello I am doing a project on the history of the internetSolution:\n\nStep 1: Introduction\nThe history of the", + "Hi today I am going to show you how to make a simple and easy to make a DIY paper flower.", + ] + + model = AutoModelForCausalLM.from_pretrained(model_id, low_cpu_mem_usage=True, torch_dtype=torch.float16).to( + torch_device + ) + + tokenizer = AutoTokenizer.from_pretrained(model_id) + inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device) + + output = model.generate(**inputs, max_new_tokens=20, do_sample=False) + output_text = tokenizer.batch_decode(output, skip_special_tokens=True) + + self.assertEqual(output_text, EXPECTED_TEXTS) + + def test_model_9b_bf16(self): + model_id = "THUDM/glm-4-9b" + + EXPECTED_TEXTS = [ + "Hello I am doing a project on the history of the internetSolution:\n\nStep 1: Introduction\nThe history of the", + "Hi today I am going to show you how to make a simple and easy to make a DIY paper flower.", + ] + + model = AutoModelForCausalLM.from_pretrained(model_id, low_cpu_mem_usage=True, torch_dtype=torch.bfloat16).to( + torch_device + ) + + tokenizer = AutoTokenizer.from_pretrained(model_id) + inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device) + + output = model.generate(**inputs, max_new_tokens=20, do_sample=False) + output_text = tokenizer.batch_decode(output, skip_special_tokens=True) + + self.assertEqual(output_text, EXPECTED_TEXTS) + + def test_model_9b_eager(self): + model_id = "THUDM/glm-4-9b" + + EXPECTED_TEXTS = [ + "Hello I am doing a project on the history of the internetSolution:\n\nStep 1: Introduction\nThe history of the", + "Hi today I am going to show you how to make a simple and easy to make a DIY paper flower.", + ] + + model = AutoModelForCausalLM.from_pretrained( + model_id, low_cpu_mem_usage=True, torch_dtype=torch.bfloat16, attn_implementation="eager" + ) + model.to(torch_device) + + tokenizer = AutoTokenizer.from_pretrained(model_id) + inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device) + + output = model.generate(**inputs, max_new_tokens=20, do_sample=False) + output_text = tokenizer.batch_decode(output, skip_special_tokens=True) + + self.assertEqual(output_text, EXPECTED_TEXTS) + + @require_torch_sdpa + def test_model_9b_sdpa(self): + model_id = "THUDM/glm-4-9b" + + EXPECTED_TEXTS = [ + "Hello I am doing a project on the history of the internetSolution:\n\nStep 1: Introduction\nThe history of the", + "Hi today I am going to show you how to make a simple and easy to make a DIY paper flower.", + ] + + model = AutoModelForCausalLM.from_pretrained( + model_id, low_cpu_mem_usage=True, torch_dtype=torch.bfloat16, attn_implementation="sdpa" + ) + model.to(torch_device) + + tokenizer = AutoTokenizer.from_pretrained(model_id) + inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device) + + output = model.generate(**inputs, max_new_tokens=20, do_sample=False) + output_text = tokenizer.batch_decode(output, skip_special_tokens=True) + + self.assertEqual(output_text, EXPECTED_TEXTS) + + @require_flash_attn + @pytest.mark.flash_attn_test + def test_model_9b_flash_attn(self): + model_id = "THUDM/glm-4-9b" + EXPECTED_TEXTS = [ + "Hello I am doing a project on the history of the internetSolution:\n\nStep 1: Introduction\nThe history of the", + "Hi today I am going to show you how to make a simple and easy to make a DIY paper flower.", + ] + + model = AutoModelForCausalLM.from_pretrained( + model_id, low_cpu_mem_usage=True, torch_dtype=torch.bfloat16, attn_implementation="flash_attention_2" + ) + model.to(torch_device) + + tokenizer = AutoTokenizer.from_pretrained(model_id) + inputs = tokenizer(self.input_text, return_tensors="pt", padding=True).to(torch_device) + + output = model.generate(**inputs, max_new_tokens=20, do_sample=False) + output_text = tokenizer.batch_decode(output, skip_special_tokens=True) + + self.assertEqual(output_text, EXPECTED_TEXTS) diff --git a/tests/quantization/hqq/test_hqq.py b/tests/quantization/hqq/test_hqq.py index 45c64676a7e4..6d08a0f0e669 100755 --- a/tests/quantization/hqq/test_hqq.py +++ b/tests/quantization/hqq/test_hqq.py @@ -94,8 +94,7 @@ def test_to_dict(self): quantization_config = HqqConfig() hqq_orig_config = quantization_config.to_dict() - for key in hqq_orig_config: - self.assertEqual(quantization_config.quant_config[key], hqq_orig_config[key]) + self.assertEqual(quantization_config.quant_config, hqq_orig_config["quant_config"]) @slow @@ -109,7 +108,7 @@ def test_fp16_quantized_model(self): """ Simple LLM model testing fp16 """ - quant_config = HqqConfig(nbits=8, group_size=64, quant_zero=False, quant_scale=False, axis=0) + quant_config = HqqConfig(nbits=8, group_size=64) hqq_runner = HQQLLMRunner( model_id=MODEL_ID, quant_config=quant_config, compute_dtype=torch.float16, device=torch_device @@ -118,26 +117,24 @@ def test_fp16_quantized_model(self): check_hqqlayer(self, hqq_runner.model.model.layers[0].self_attn.v_proj) check_forward(self, hqq_runner.model) - def test_f16_quantized_model_with_offloading(self): + +@slow +@require_torch_gpu +@require_torch_multi_gpu +@require_accelerate +class HQQTestMultiGPU(unittest.TestCase): + def tearDown(self): + cleanup() + + def test_fp16_quantized_model_multipgpu(self): """ - Simple LLM model testing bfp16 with meta-data offloading + Simple LLM model testing fp16 with multi-gpu """ - q4_config = {"nbits": 4, "group_size": 64, "quant_zero": False, "quant_scale": False} - q3_config = {"nbits": 3, "group_size": 32, "quant_zero": False, "quant_scale": False, "offload_meta": True} - quant_config = HqqConfig( - dynamic_config={ - "self_attn.q_proj": q4_config, - "self_attn.k_proj": q4_config, - "self_attn.v_proj": q4_config, - "self_attn.o_proj": q4_config, - "mlp.gate_proj": q3_config, - "mlp.up_proj": q3_config, - "mlp.down_proj": q3_config, - } - ) + + quant_config = HqqConfig(nbits=8, group_size=64) hqq_runner = HQQLLMRunner( - model_id=MODEL_ID, quant_config=quant_config, compute_dtype=torch.float16, device=torch_device + model_id=MODEL_ID, quant_config=quant_config, compute_dtype=torch.float16, device="auto" ) check_hqqlayer(self, hqq_runner.model.model.layers[0].self_attn.v_proj) @@ -146,22 +143,40 @@ def test_f16_quantized_model_with_offloading(self): @slow @require_torch_gpu -@require_torch_multi_gpu @require_accelerate -class HQQTestMultiGPU(unittest.TestCase): +class HQQSerializationTest(unittest.TestCase): def tearDown(self): cleanup() - def test_fp16_quantized_model_multipgpu(self): + def test_model_serialization(self): """ - Simple LLM model testing fp16 with multi-gpu + Simple HQQ LLM save/load test """ - - quant_config = HqqConfig(nbits=8, group_size=64, quant_zero=False, quant_scale=False, axis=0) + quant_config = HqqConfig(nbits=4, group_size=64) hqq_runner = HQQLLMRunner( - model_id=MODEL_ID, quant_config=quant_config, compute_dtype=torch.float16, device="auto" + model_id=MODEL_ID, quant_config=quant_config, compute_dtype=torch.float16, device=torch_device ) - check_hqqlayer(self, hqq_runner.model.model.layers[0].self_attn.v_proj) - check_forward(self, hqq_runner.model) + input_tensor = torch.zeros((1, 8), dtype=torch.int32, device=torch_device) + + with torch.no_grad(): + logits_ref = hqq_runner.model.forward(input_tensor).logits + + # Save + saved_model_id = "quant_model" + hqq_runner.model.save_pretrained(saved_model_id) + + # Remove old model + del hqq_runner.model + torch.cuda.empty_cache() + + # Load and check if the logits match + model_loaded = AutoModelForCausalLM.from_pretrained( + "quant_model", torch_dtype=torch.float16, device_map=torch_device, low_cpu_mem_usage=True + ) + + with torch.no_grad(): + logits_loaded = model_loaded.forward(input_tensor).logits + + self.assertEqual((logits_loaded - logits_ref).abs().mean().item(), 0) diff --git a/utils/modular_model_converter.py b/utils/modular_model_converter.py index c5bf769f9288..63967309c40d 100644 --- a/utils/modular_model_converter.py +++ b/utils/modular_model_converter.py @@ -537,7 +537,7 @@ def __init__(self, python_module, new_name, given_old_name=None, given_new_name= "feature_extractor": {}, } self.match_patterns = "|".join(self.files.keys()) - self.all_functions = {} + self.all_definitions = {} def visit_ImportFrom(self, node: cst.ImportFrom) -> None: """When visiting imports from `transformers.models.xxx` we need to: @@ -590,6 +590,11 @@ def leave_SimpleStatementLine(self, original_node, updated_node): self.global_scope_index += 100 return updated_node + @m.visit(m.ClassDef() | m.FunctionDef() | m.AddAssign()) + def create_global_node(self, node): + name = re.search(r"(?:def|class)\s+([a-zA-Z_]\w*)|([A-Z_]\w*)", self.python_module.code_for_node(node)) + self.all_nodes[name] = node + def leave_ClassDef(self, original_node, updated_node): """ 1. Filter the `base` classes of this class @@ -647,9 +652,12 @@ def leave_ClassDef(self, original_node, updated_node): node = class_finder.global_nodes.get(dependency, None) if node is not None: if dependency not in file_to_update: + node = self.all_definitions.get(dependency, node) start_insert_idx -= 1 file_to_update[dependency] = {"insert_idx": start_insert_idx, "node": node} elif dependency not in self.inserted_deps: + print("processing :", dependency) + # make sure the node is written after its dependencies start_insert_idx = file_to_update[dependency]["insert_idx"] - 1 if ( @@ -683,6 +691,12 @@ def leave_ClassDef(self, original_node, updated_node): self.files["modeling"][class_name] = {"insert_idx": self.global_scope_index, "node": updated_node} return updated_node + def leave_FunctionDef(self, original_node, node): + parent_node = self.get_metadata(cst.metadata.ParentNodeProvider, original_node) + if m.matches(parent_node, m.Module()): + self.all_definitions[node.name.value] = node + return node + def leave_If(self, original_node, node): parent_node = self.get_metadata(cst.metadata.ParentNodeProvider, original_node) if m.matches(parent_node, m.Module()): @@ -757,7 +771,7 @@ def save_modeling_file(modular_file, converted_file): parser = argparse.ArgumentParser() parser.add_argument( "--files_to_parse", - default=["all"], + default=["src/transformers/models/glm/modular_glm.py"], nargs="+", help="A list of `modular_xxxx` files that should be converted to single model file", )