Upload ChatGLMForConditionalGeneration

README.md

@@ -0,0 +1,200 @@
+---
+library_name: transformers
+tags:
+- llama-factory
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]

config.json

@@ -0,0 +1,50 @@
+{
+ "_name_or_path": "THUDM/glm-4-9b-chat",
+ "add_bias_linear": false,
+ "add_qkv_bias": true,
+ "apply_query_key_layer_scaling": true,
+ "apply_residual_connection_post_layernorm": false,
+ "architectures": [
+ "ChatGLMForConditionalGeneration"
+ ],
+ "attention_dropout": 0.0,
+ "attention_softmax_in_fp32": true,
+ "auto_map": {
+ "AutoConfig": "configuration_chatglm.ChatGLMConfig",
+ "AutoModel": "modeling_chatglm.ChatGLMForConditionalGeneration",
+ "AutoModelForCausalLM": "THUDM/glm-4-9b-chat--modeling_chatglm.ChatGLMForConditionalGeneration",
+ "AutoModelForSeq2SeqLM": "THUDM/glm-4-9b-chat--modeling_chatglm.ChatGLMForConditionalGeneration",
+ "AutoModelForSequenceClassification": "THUDM/glm-4-9b-chat--modeling_chatglm.ChatGLMForSequenceClassification"
+ },
+ "bias_dropout_fusion": true,
+ "classifier_dropout": null,
+ "eos_token_id": [
+ 151329,
+ 151336,
+ 151338
+ ],
+ "ffn_hidden_size": 13696,
+ "fp32_residual_connection": false,
+ "hidden_dropout": 0.0,
+ "hidden_size": 4096,
+ "kv_channels": 128,
+ "layernorm_epsilon": 1.5625e-07,
+ "model_type": "chatglm",
+ "multi_query_attention": true,
+ "multi_query_group_num": 2,
+ "num_attention_heads": 32,
+ "num_hidden_layers": 40,
+ "num_layers": 40,
+ "original_rope": true,
+ "pad_token_id": 151329,
+ "padded_vocab_size": 151552,
+ "post_layer_norm": true,
+ "rmsnorm": true,
+ "rope_ratio": 500,
+ "seq_length": 131072,
+ "tie_word_embeddings": false,
+ "torch_dtype": "bfloat16",
+ "transformers_version": "4.44.2",
+ "use_cache": true,
+ "vocab_size": 151552
+}

configuration_chatglm.py

@@ -0,0 +1,58 @@
+from transformers import PretrainedConfig
+
+
+class ChatGLMConfig(PretrainedConfig):
+ model_type = "chatglm"
+
+ def __init__(
+ self,
+ num_layers=28,
+ padded_vocab_size=65024,
+ hidden_size=4096,
+ ffn_hidden_size=13696,
+ kv_channels=128,
+ num_attention_heads=32,
+ seq_length=2048,
+ hidden_dropout=0.0,
+ classifier_dropout=None,
+ attention_dropout=0.0,
+ layernorm_epsilon=1e-5,
+ rmsnorm=True,
+ apply_residual_connection_post_layernorm=False,
+ post_layer_norm=True,
+ add_bias_linear=False,
+ add_qkv_bias=False,
+ bias_dropout_fusion=True,
+ multi_query_attention=False,
+ multi_query_group_num=1,
+ rope_ratio=1,
+ apply_query_key_layer_scaling=True,
+ attention_softmax_in_fp32=True,
+ fp32_residual_connection=False,
+ **kwargs
+ ):
+ self.num_layers = num_layers
+ self.vocab_size = padded_vocab_size
+ self.padded_vocab_size = padded_vocab_size
+ self.hidden_size = hidden_size
+ self.ffn_hidden_size = ffn_hidden_size
+ self.kv_channels = kv_channels
+ self.num_attention_heads = num_attention_heads
+ self.seq_length = seq_length
+ self.hidden_dropout = hidden_dropout
+ self.classifier_dropout = classifier_dropout
+ self.attention_dropout = attention_dropout
+ self.layernorm_epsilon = layernorm_epsilon
+ self.rmsnorm = rmsnorm
+ self.apply_residual_connection_post_layernorm = apply_residual_connection_post_layernorm
+ self.post_layer_norm = post_layer_norm
+ self.add_bias_linear = add_bias_linear
+ self.add_qkv_bias = add_qkv_bias
+ self.bias_dropout_fusion = bias_dropout_fusion
+ self.multi_query_attention = multi_query_attention
+ self.multi_query_group_num = multi_query_group_num
+ self.rope_ratio = rope_ratio
+ self.apply_query_key_layer_scaling = apply_query_key_layer_scaling
+ self.attention_softmax_in_fp32 = attention_softmax_in_fp32
+ self.fp32_residual_connection = fp32_residual_connection
+ super().__init__(**kwargs)

generation_config.json

@@ -0,0 +1,13 @@
+{
+ "do_sample": true,
+ "eos_token_id": [
+ 151329,
+ 151336,
+ 151338
+ ],
+ "max_length": 128000,
+ "pad_token_id": 151329,
+ "temperature": 0.8,
+ "top_p": 0.8,
+ "transformers_version": "4.44.2"
+}

model-00001-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2466853a5560e02b5d4a9c3f3d55869b2bdd41fe51132121f714a944cafc2fd
+size 8992052232

model-00002-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2ad8a25c7382f9416ace14734862a9da2437e5a4ce76e5dcbfe19e7cab542a0
+size 8566374880

model-00003-of-00003.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07ac1878a23796a69dbc48efd0cd45a1fc8f2b80c23b761ada772e6046578cf1
+size 1241514128

model.safetensors.index.json

@@ -0,0 +1,291 @@
+{
+ "metadata": {
+ "total_size": 18799902784
+ },
+ "weight_map": {
+ "transformer.embedding.word_embeddings.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.final_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.0.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.1.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.10.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.11.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.12.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.13.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.14.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.15.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.16.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.17.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.18.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.19.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.19.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.19.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.19.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.19.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.2.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.20.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.21.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.22.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.23.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.24.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.25.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.26.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.27.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.28.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.29.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.3.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.30.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.31.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.32.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.33.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.34.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.35.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.36.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.37.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.38.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.input_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.mlp.dense_4h_to_h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.mlp.dense_h_to_4h.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.self_attention.dense.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.self_attention.query_key_value.bias": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.39.self_attention.query_key_value.weight": "model-00002-of-00003.safetensors",
+ "transformer.encoder.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.4.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.5.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.6.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.7.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.8.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.mlp.dense_4h_to_h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.mlp.dense_h_to_4h.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.self_attention.dense.weight": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.self_attention.query_key_value.bias": "model-00001-of-00003.safetensors",
+ "transformer.encoder.layers.9.self_attention.query_key_value.weight": "model-00001-of-00003.safetensors",
+ "transformer.output_layer.weight": "model-00003-of-00003.safetensors",
+ "transformer.rotary_pos_emb.inv_freq": "model-00001-of-00003.safetensors"
+ }
+}

modeling_chatglm.py

@@ -0,0 +1,1138 @@
+""" PyTorch ChatGLM model. """
+
+import math
+import sys
+import torch
+import torch.utils.checkpoint
+import torch.nn.functional as F
+from torch import nn
+from torch.nn import CrossEntropyLoss, LayerNorm, MSELoss, BCEWithLogitsLoss
+from torch.nn.utils import skip_init
+from typing import Optional, Tuple, Union, List, Dict, Any
+
+from transformers.modeling_outputs import (
+ BaseModelOutputWithPast,
+ CausalLMOutputWithPast,
+ SequenceClassifierOutputWithPast,
+)
+from transformers.modeling_utils import PreTrainedModel
+from transformers.utils import logging, is_torch_npu_available
+from transformers.generation.logits_process import LogitsProcessor
+from transformers.generation.utils import ModelOutput
+
+from .configuration_chatglm import ChatGLMConfig
+
+try:
+ from transformers.utils import is_flash_attn_greater_or_equal_2_10, is_flash_attn_2_available
+
+ if is_flash_attn_2_available():
+ from flash_attn import flash_attn_func, flash_attn_varlen_func
+ from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa
+except:
+ pass
+
+# flags required to enable jit fusion kernels
+
+if sys.platform != 'darwin' and not is_torch_npu_available():
+ torch._C._jit_set_profiling_mode(False)
+ torch._C._jit_set_profiling_executor(False)
+ torch._C._jit_override_can_fuse_on_cpu(True)
+ torch._C._jit_override_can_fuse_on_gpu(True)
+
+logger = logging.get_logger(__name__)
+
+_CHECKPOINT_FOR_DOC = "THUDM/ChatGLM"
+_CONFIG_FOR_DOC = "ChatGLMConfig"
+
+
+def default_init(cls, *args, **kwargs):
+ return cls(*args, **kwargs)
+
+
+class InvalidScoreLogitsProcessor(LogitsProcessor):
+ def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> torch.FloatTensor:
+ if torch.isnan(scores).any() or torch.isinf(scores).any():
+ scores.zero_()
+ scores[..., 198] = 5e4
+ return scores
+
+
+def split_tensor_along_last_dim(
+ tensor: torch.Tensor,
+ num_partitions: int,
+ contiguous_split_chunks: bool = False,
+) -> List[torch.Tensor]:
+ """Split a tensor along its last dimension.
+
+ Arguments:
+ tensor: input tensor.
+ num_partitions: number of partitions to split the tensor
+ contiguous_split_chunks: If True, make each chunk contiguous
+ in memory.
+
+ Returns:
+ A list of Tensors
+ """
+ # Get the size and dimension.
+ last_dim = tensor.dim() - 1
+ last_dim_size = tensor.size()[last_dim] // num_partitions
+ # Split.
+ tensor_list = torch.split(tensor, last_dim_size, dim=last_dim)
+ # Note: torch.split does not create contiguous tensors by default.
+ if contiguous_split_chunks:
+ return tuple(chunk.contiguous() for chunk in tensor_list)
+
+ return tensor_list
+
+
+class RotaryEmbedding(nn.Module):
+ def __init__(self, dim, rope_ratio=1, original_impl=False, device=None, dtype=None):
+ super().__init__()
+ inv_freq = 1.0 / (10000 ** (torch.arange(0, dim, 2, device=device).to(dtype=dtype) / dim))
+ self.register_buffer("inv_freq", inv_freq)
+ self.dim = dim
+ self.original_impl = original_impl
+ self.rope_ratio = rope_ratio
+
+ def forward_impl(
+ self, seq_len: int, n_elem: int, dtype: torch.dtype, device: torch.device, base: int = 10000
+ ):
+ """Enhanced Transformer with Rotary Position Embedding.
+
+ Derived from: https://github.com/labmlai/annotated_deep_learning_paper_implementations/blob/master/labml_nn/
+ transformers/rope/__init__.py. MIT License:
+ https://github.com/labmlai/annotated_deep_learning_paper_implementations/blob/master/license.
+ """
+ # $\Theta = {\theta_i = 10000^{\frac{2(i-1)}{d}}, i \in [1, 2, ..., \frac{d}{2}]}$
+ base = base * self.rope_ratio
+ theta = 1.0 / (base ** (torch.arange(0, n_elem, 2, dtype=torch.float, device=device) / n_elem))
+
+ # Create position indexes `[0, 1, ..., seq_len - 1]`
+ seq_idx = torch.arange(seq_len, dtype=torch.float, device=device)
+
+ # Calculate the product of position index and $\theta_i$
+ idx_theta = torch.outer(seq_idx, theta).float()
+
+ cache = torch.stack([torch.cos(idx_theta), torch.sin(idx_theta)], dim=-1)
+
+ # this is to mimic the behaviour of complex32, else we will get different results
+ if dtype in (torch.float16, torch.bfloat16, torch.int8):
+ cache = cache.bfloat16() if dtype == torch.bfloat16 else cache.half()
+ return cache
+
+ def forward(self, max_seq_len, offset=0):
+ return self.forward_impl(
+ max_seq_len, self.dim, dtype=self.inv_freq.dtype, device=self.inv_freq.device
+ )
+
+
+@torch.jit.script
+def apply_rotary_pos_emb(x: torch.Tensor, rope_cache: torch.Tensor) -> torch.Tensor:
+ # x: [b, np, sq, hn]
+ b, np, sq, hn = x.size(0), x.size(1), x.size(2), x.size(3)
+ rot_dim = rope_cache.shape[-2] * 2
+ x, x_pass = x[..., :rot_dim], x[..., rot_dim:]
+ # truncate to support variable sizes
+ rope_cache = rope_cache[:, :sq]
+ xshaped = x.reshape(b, np, sq, rot_dim // 2, 2)
+ rope_cache = rope_cache.view(-1, 1, sq, xshaped.size(3), 2)
+ x_out2 = torch.stack(
+ [
+ xshaped[..., 0] * rope_cache[..., 0] - xshaped[..., 1] * rope_cache[..., 1],
+ xshaped[..., 1] * rope_cache[..., 0] + xshaped[..., 0] * rope_cache[..., 1],
+ ],
+ -1,
+ )
+ x_out2 = x_out2.flatten(3)
+ return torch.cat((x_out2, x_pass), dim=-1)
+
+
+class RMSNorm(torch.nn.Module):
+ def __init__(self, normalized_shape, eps=1e-5, device=None, dtype=None, **kwargs):
+ super().__init__()
+ self.weight = torch.nn.Parameter(torch.empty(normalized_shape, device=device, dtype=dtype))
+ self.eps = eps
+
+ def forward(self, hidden_states: torch.Tensor):
+ input_dtype = hidden_states.dtype
+ variance = hidden_states.to(torch.float32).pow(2).mean(-1, keepdim=True)
+ hidden_states = hidden_states * torch.rsqrt(variance + self.eps)
+
+ return (self.weight * hidden_states).to(input_dtype)
+
+
+class CoreAttention(torch.nn.Module):
+ def __init__(self, config: ChatGLMConfig, layer_number):
+ super(CoreAttention, self).__init__()
+ self.config = config
+ self.apply_query_key_layer_scaling = config.apply_query_key_layer_scaling
+ self.attention_softmax_in_fp32 = config.attention_softmax_in_fp32
+ if self.apply_query_key_layer_scaling:
+ self.attention_softmax_in_fp32 = True
+ self.layer_number = max(1, layer_number)
+ self.is_causal = True
+
+ projection_size = config.kv_channels * config.num_attention_heads
+
+ # Per attention head and per partition values.
+ self.hidden_size_per_partition = projection_size
+ self.hidden_size_per_attention_head = projection_size // config.num_attention_heads
+ self.num_attention_heads_per_partition = config.num_attention_heads
+
+ coeff = None
+ self.norm_factor = math.sqrt(self.hidden_size_per_attention_head)
+ if self.apply_query_key_layer_scaling:
+ coeff = self.layer_number
+ self.norm_factor *= coeff
+ self.coeff = coeff
+
+ self.attention_dropout = torch.nn.Dropout(config.attention_dropout)
+
+ def forward(self, query_layer, key_layer, value_layer, attention_mask):
+ # [b, np, sq, sk]
+ output_size = (query_layer.size(0), query_layer.size(1), query_layer.size(2), key_layer.size(2))
+
+ # [b, np, sq, hn] -> [b * np, sq, hn]
+ query_layer = query_layer.view(output_size[0] * output_size[1], output_size[2], -1)
+ # [b, np, sk, hn] -> [b * np, sk, hn]
+ key_layer = key_layer.view(output_size[0] * output_size[1], output_size[3], -1)
+
+ # preallocting input tensor: [b * np, sq, sk]
+ matmul_input_buffer = torch.empty(
+ output_size[0] * output_size[1], output_size[2], output_size[3], dtype=query_layer.dtype,
+ device=query_layer.device
+ )
+
+ # Raw attention scores. [b * np, sq, sk]
+ matmul_result = torch.baddbmm(
+ matmul_input_buffer,
+ query_layer, # [b * np, sq, hn]
+ key_layer.transpose(1, 2), # [b * np, hn, sk]
+ beta=0.0,
+ alpha=(1.0 / self.norm_factor),
+ )
+
+ # change view to [b, np, sq, sk]
+ attention_scores = matmul_result.view(*output_size)
+
+ # ===========================
+ # Attention probs and dropout
+ # ===========================
+
+ # attention scores and attention mask [b, np, sq, sk]
+ if self.attention_softmax_in_fp32:
+ attention_scores = attention_scores.float()
+ if self.coeff is not None:
+ attention_scores = attention_scores * self.coeff
+ if attention_mask is None and attention_scores.shape[2] == attention_scores.shape[3]:
+ attention_mask = torch.ones(output_size[0], 1, output_size[2], output_size[3],
+ device=attention_scores.device, dtype=torch.bool)
+ attention_mask.tril_()
+ attention_mask = ~attention_mask
+ if attention_mask is not None:
+ attention_scores = attention_scores.masked_fill(attention_mask, float("-inf"))
+ attention_probs = F.softmax(attention_scores, dim=-1)
+ attention_probs = attention_probs.type_as(value_layer)
+
+ # This is actually dropping out entire tokens to attend to, which might
+ # seem a bit unusual, but is taken from the original Transformer paper.
+ attention_probs = self.attention_dropout(attention_probs)
+
+ # query layer shape: [b * np, sq, hn]
+ # value layer shape: [b, np, sk, hn]
+ # attention shape: [b, np, sq, sk]
+ # context layer shape: [b, np, sq, hn]
+ output_size = (value_layer.size(0), value_layer.size(1), query_layer.size(1), value_layer.size(3))
+ # change view [b * np, sk, hn]
+ value_layer = value_layer.view(output_size[0] * output_size[1], value_layer.size(2), -1)
+ # change view [b * np, sq, sk]
+ attention_probs = attention_probs.view(output_size[0] * output_size[1], output_size[2], -1)
+ # matmul: [b * np, sq, hn]
+ context_layer = torch.bmm(attention_probs, value_layer)
+ # change view [b, np, sq, hn]
+ context_layer = context_layer.view(*output_size)
+ # [b, np, sq, hn] --> [b, sq, np, hn]
+ context_layer = context_layer.transpose(1, 2).contiguous()
+ # [b, sq, np, hn] --> [b, sq, hp]
+ new_context_layer_shape = context_layer.size()[:-2] + (self.hidden_size_per_partition,)
+ context_layer = context_layer.reshape(*new_context_layer_shape)
+
+ return context_layer
+
+
+class SdpaAttention(CoreAttention):
+ def forward(self, query_layer, key_layer, value_layer, attention_mask):
+ if attention_mask is None and query_layer.shape[2] == key_layer.shape[2]:
+ context_layer = torch.nn.functional.scaled_dot_product_attention(query_layer, key_layer, value_layer,
+ is_causal=True,
+ dropout_p=self.config.attention_dropout if self.training else 0.0)
+ else:
+ if attention_mask is not None:
+ attention_mask = ~attention_mask
+ context_layer = torch.nn.functional.scaled_dot_product_attention(query_layer, key_layer, value_layer,
+ attention_mask,
+ dropout_p=self.config.attention_dropout if self.training else 0.0)
+ context_layer = context_layer.transpose(1, 2).contiguous()
+ new_context_layer_shape = context_layer.size()[:-2] + (self.hidden_size_per_partition,)
+ context_layer = context_layer.reshape(*new_context_layer_shape)
+ return context_layer
+
+
+def _get_unpad_data(attention_mask):
+ seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32)
+ indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten()
+ max_seqlen_in_batch = seqlens_in_batch.max().item()
+ cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.int32), (1, 0))
+ return (
+ indices,
+ cu_seqlens,
+ max_seqlen_in_batch,
+ )
+
+
+# Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2
+class FlashAttention2(CoreAttention):
+ def __init__(self, *args, **kwargs):
+ super().__init__(*args, **kwargs)
+ self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10()
+
+ def forward(self, query_states, key_states, value_states, attention_mask):
+ query_states = query_states.transpose(1, 2)
+ key_states = key_states.transpose(1, 2)
+ value_states = value_states.transpose(1, 2)
+ batch_size, query_length = query_states.shape[:2]
+ if not self._flash_attn_uses_top_left_mask:
+ causal = self.is_causal
+ else:
+ # TODO: Remove the `query_length != 1` check once Flash Attention for RoCm is bumped to 2.1. For details, please see the comment in LlamaFlashAttention2 __init__.
+ causal = self.is_causal and query_length != 1
+ dropout = self.config.attention_dropout if self.training else 0.0
+ # Contains at least one padding token in the sequence
+ if attention_mask is not None:
+ query_states, key_states, value_states, indices_q, cu_seq_lens, max_seq_lens = self._upad_input(
+ query_states, key_states, value_states, attention_mask, query_length
+ )
+
+ cu_seqlens_q, cu_seqlens_k = cu_seq_lens
+ max_seqlen_in_batch_q, max_seqlen_in_batch_k = max_seq_lens
+
+ attn_output_unpad = flash_attn_varlen_func(
+ query_states,
+ key_states,
+ value_states,
+ cu_seqlens_q=cu_seqlens_q,
+ cu_seqlens_k=cu_seqlens_k,
+ max_seqlen_q=max_seqlen_in_batch_q,
+ max_seqlen_k=max_seqlen_in_batch_k,
+ dropout_p=dropout,
+ softmax_scale=None,
+ causal=causal,
+ )
+
+ attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length)
+ else:
+ attn_output = flash_attn_func(
+ query_states, key_states, value_states, dropout, softmax_scale=None, causal=causal
+ )
+ attn_output = attn_output.reshape(batch_size, query_length, self.hidden_size_per_partition).contiguous()
+ return attn_output
+
+ def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length):
+ indices_k, cu_seqlens_k, max_seqlen_in_batch_k = _get_unpad_data(attention_mask)
+ batch_size, kv_seq_len, num_key_value_heads, head_dim = key_layer.shape
+
+ key_layer = index_first_axis(
+ key_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+ )
+ value_layer = index_first_axis(
+ value_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k
+ )
+ if query_length == kv_seq_len:
+ query_layer = index_first_axis(
+ query_layer.reshape(batch_size * kv_seq_len, self.num_attention_heads_per_partition, head_dim),
+ indices_k
+ )
+ cu_seqlens_q = cu_seqlens_k
+ max_seqlen_in_batch_q = max_seqlen_in_batch_k
+ indices_q = indices_k
+ elif query_length == 1:
+ max_seqlen_in_batch_q = 1
+ cu_seqlens_q = torch.arange(
+ batch_size + 1, dtype=torch.int32, device=query_layer.device
+ ) # There is a memcpy here, that is very bad.
+ indices_q = cu_seqlens_q[:-1]
+ query_layer = query_layer.squeeze(1)
+ else:
+ # The -q_len: slice assumes left padding.
+ attention_mask = attention_mask[:, -query_length:]
+ query_layer, indices_q, cu_seqlens_q, max_seqlen_in_batch_q = unpad_input(query_layer, attention_mask)
+
+ return (
+ query_layer,
+ key_layer,
+ value_layer,
+ indices_q,
+ (cu_seqlens_q, cu_seqlens_k),
+ (max_seqlen_in_batch_q, max_seqlen_in_batch_k),
+ )
+
+
+CORE_ATTENTION_CLASSES = {
+ "eager": CoreAttention,
+ "sdpa": SdpaAttention,
+ "flash_attention_2": FlashAttention2
+}
+
+
+class SelfAttention(torch.nn.Module):
+ """Parallel self-attention layer abstract class.
+
+ Self-attention layer takes input with size [s, b, h]
+ and returns output of the same size.
+ """
+
+ def __init__(self, config: ChatGLMConfig, layer_number, device=None):
+ super(SelfAttention, self).__init__()
+ self.layer_number = max(1, layer_number)
+
+ self.projection_size = config.kv_channels * config.num_attention_heads
+
+ # Per attention head and per partition values.
+ self.hidden_size_per_attention_head = self.projection_size // config.num_attention_heads
+ self.num_attention_heads_per_partition = config.num_attention_heads
+
+ self.multi_query_attention = config.multi_query_attention
+ self.qkv_hidden_size = 3 * self.projection_size
+ if self.multi_query_attention:
+ self.num_multi_query_groups_per_partition = config.multi_query_group_num
+ self.qkv_hidden_size = (
+ self.projection_size + 2 * self.hidden_size_per_attention_head * config.multi_query_group_num
+ )
+ self.query_key_value = nn.Linear(config.hidden_size, self.qkv_hidden_size,
+ bias=config.add_bias_linear or config.add_qkv_bias,
+ device=device, **_config_to_kwargs(config)
+ )
+
+ self.core_attention = CORE_ATTENTION_CLASSES[config._attn_implementation](config, self.layer_number)
+
+ # Output.
+ self.dense = nn.Linear(self.projection_size, config.hidden_size, bias=config.add_bias_linear,
+ device=device, **_config_to_kwargs(config)
+ )
+
+ def _allocate_memory(self, inference_max_sequence_len, batch_size, device=None, dtype=None):
+ if self.multi_query_attention:
+ num_attention_heads = self.num_multi_query_groups_per_partition
+ else:
+ num_attention_heads = self.num_attention_heads_per_partition
+ return torch.empty(
+ inference_max_sequence_len,
+ batch_size,
+ num_attention_heads,
+ self.hidden_size_per_attention_head,
+ dtype=dtype,
+ device=device,
+ )
+
+ def forward(
+ self, hidden_states, attention_mask, rotary_pos_emb, kv_cache=None, use_cache=True
+ ):
+ # hidden_states: [b, sq, h]
+
+ # =================================================
+ # Pre-allocate memory for key-values for inference.
+ # =================================================
+ # =====================
+ # Query, Key, and Value
+ # =====================
+
+ # Attention heads [b, sq, h] --> [b, sq, (np * 3 * hn)]
+ mixed_x_layer = self.query_key_value(hidden_states)
+
+ if self.multi_query_attention:
+ (query_layer, key_layer, value_layer) = mixed_x_layer.split(
+ [
+ self.num_attention_heads_per_partition * self.hidden_size_per_attention_head,
+ self.num_multi_query_groups_per_partition * self.hidden_size_per_attention_head,
+ self.num_multi_query_groups_per_partition * self.hidden_size_per_attention_head,
+ ],
+ dim=-1,
+ )
+ query_layer = query_layer.view(
+ query_layer.size()[:-1] + (self.num_attention_heads_per_partition, self.hidden_size_per_attention_head)
+ )
+ key_layer = key_layer.view(
+ key_layer.size()[:-1] + (self.num_multi_query_groups_per_partition, self.hidden_size_per_attention_head)
+ )
+ value_layer = value_layer.view(
+ value_layer.size()[:-1]
+ + (self.num_multi_query_groups_per_partition, self.hidden_size_per_attention_head)
+ )
+ else:
+ new_tensor_shape = mixed_x_layer.size()[:-1] + \
+ (self.num_attention_heads_per_partition,
+ 3 * self.hidden_size_per_attention_head)
+ mixed_x_layer = mixed_x_layer.view(*new_tensor_shape)
+
+ # [b, sq, np, 3 * hn] --> 3 [b, sq, np, hn]
+ (query_layer, key_layer, value_layer) = split_tensor_along_last_dim(mixed_x_layer, 3)
+
+ # [b, sq, np, hn] -> [b, np, sq, hn]
+ query_layer, key_layer, value_layer = [k.transpose(1, 2) for k in [query_layer, key_layer, value_layer]]
+
+ # apply relative positional encoding (rotary embedding)
+ if rotary_pos_emb is not None:
+ query_layer = apply_rotary_pos_emb(query_layer, rotary_pos_emb)
+ key_layer = apply_rotary_pos_emb(key_layer, rotary_pos_emb)
+
+ # adjust key and value for inference
+ if kv_cache is not None:
+ cache_k, cache_v = kv_cache
+ key_layer = torch.cat((cache_k, key_layer), dim=2)
+ value_layer = torch.cat((cache_v, value_layer), dim=2)
+ if use_cache:
+ if kv_cache is None:
+ kv_cache = torch.cat((key_layer.unsqueeze(0).unsqueeze(0), value_layer.unsqueeze(0).unsqueeze(0)),
+ dim=1)
+ else:
+ kv_cache = (key_layer, value_layer)
+ else:
+ kv_cache = None
+
+ if self.multi_query_attention:
+ key_layer = key_layer.unsqueeze(2)
+ key_layer = key_layer.expand(
+ -1, -1, self.num_attention_heads_per_partition // self.num_multi_query_groups_per_partition, -1, -1
+ )
+ key_layer = key_layer.contiguous().view(
+ key_layer.size()[:1] + (self.num_attention_heads_per_partition,) + key_layer.size()[3:]
+ )
+ value_layer = value_layer.unsqueeze(2)
+ value_layer = value_layer.expand(
+ -1, -1, self.num_attention_heads_per_partition // self.num_multi_query_groups_per_partition, -1, -1
+ )
+ value_layer = value_layer.contiguous().view(
+ value_layer.size()[:1] + (self.num_attention_heads_per_partition,) + value_layer.size()[3:]
+ )
+
+ # ==================================
+ # core attention computation
+ # ==================================
+
+ context_layer = self.core_attention(query_layer, key_layer, value_layer, attention_mask)
+
+ # =================
+ # Output. [sq, b, h]
+ # =================
+
+ output = self.dense(context_layer)
+
+ return output, kv_cache
+
+
+def _config_to_kwargs(args):
+ common_kwargs = {
+ "dtype": args.torch_dtype,
+ }
+ return common_kwargs
+
+
+class MLP(torch.nn.Module):
+ """MLP.
+
+ MLP will take the input with h hidden state, project it to 4*h
+ hidden dimension, perform nonlinear transformation, and project the
+ state back into h hidden dimension.
+ """
+
+ def __init__(self, config: ChatGLMConfig, device=None):
+ super(MLP, self).__init__()
+
+ self.add_bias = config.add_bias_linear
+
+ # Project to 4h. If using swiglu double the output width, see https://arxiv.org/pdf/2002.05202.pdf
+ self.dense_h_to_4h = nn.Linear(
+ config.hidden_size,
+ config.ffn_hidden_size * 2,
+ bias=self.add_bias,
+ device=device,
+ **_config_to_kwargs(config)
+ )
+
+ def swiglu(x):
+ x = torch.chunk(x, 2, dim=-1)
+ return F.silu(x[0]) * x[1]
+
+ self.activation_func = swiglu
+
+ # Project back to h.
+ self.dense_4h_to_h = nn.Linear(
+ config.ffn_hidden_size,
+ config.hidden_size,
+ bias=self.add_bias,
+ device=device,
+ **_config_to_kwargs(config)
+ )
+
+ def forward(self, hidden_states):
+ # [s, b, 4hp]
+ intermediate_parallel = self.dense_h_to_4h(hidden_states)
+ intermediate_parallel = self.activation_func(intermediate_parallel)
+ # [s, b, h]
+ output = self.dense_4h_to_h(intermediate_parallel)
+ return output
+
+
+class GLMBlock(torch.nn.Module):
+ """A single transformer layer.
+
+ Transformer layer takes input with size [s, b, h] and returns an
+ output of the same size.
+ """
+
+ def __init__(self, config: ChatGLMConfig, layer_number, device=None):
+ super(GLMBlock, self).__init__()
+ self.layer_number = layer_number
+
+ self.apply_residual_connection_post_layernorm = config.apply_residual_connection_post_layernorm
+
+ self.fp32_residual_connection = config.fp32_residual_connection
+
+ LayerNormFunc = RMSNorm if config.rmsnorm else LayerNorm
+ # Layernorm on the input data.
+ self.input_layernorm = LayerNormFunc(config.hidden_size, eps=config.layernorm_epsilon, device=device,
+ dtype=config.torch_dtype)
+
+ # Self attention.
+ self.self_attention = SelfAttention(config, layer_number, device=device)
+ self.hidden_dropout = config.hidden_dropout
+
+ # Layernorm on the attention output
+ self.post_attention_layernorm = LayerNormFunc(config.hidden_size, eps=config.layernorm_epsilon, device=device,
+ dtype=config.torch_dtype)
+
+ # MLP
+ self.mlp = MLP(config, device=device)
+
+ def forward(
+ self, hidden_states, attention_mask, rotary_pos_emb, kv_cache=None, use_cache=True,
+ ):
+ # hidden_states: [s, b, h]
+
+ # Layer norm at the beginning of the transformer layer.
+ layernorm_output = self.input_layernorm(hidden_states)
+ # Self attention.
+ attention_output, kv_cache = self.self_attention(
+ layernorm_output,
+ attention_mask,
+ rotary_pos_emb,
+ kv_cache=kv_cache,
+ use_cache=use_cache
+ )
+
+ # Residual connection.
+ if self.apply_residual_connection_post_layernorm:
+ residual = layernorm_output
+ else:
+ residual = hidden_states
+
+ layernorm_input = torch.nn.functional.dropout(attention_output, p=self.hidden_dropout, training=self.training)
+ layernorm_input = residual + layernorm_input
+
+ # Layer norm post the self attention.
+ layernorm_output = self.post_attention_layernorm(layernorm_input)
+
+ # MLP.
+ mlp_output = self.mlp(layernorm_output)
+
+ # Second residual connection.
+ if self.apply_residual_connection_post_layernorm:
+ residual = layernorm_output
+ else:
+ residual = layernorm_input
+
+ output = torch.nn.functional.dropout(mlp_output, p=self.hidden_dropout, training=self.training)
+ output = residual + output
+
+ return output, kv_cache
+
+
+class GLMTransformer(torch.nn.Module):
+ """Transformer class."""
+
+ def __init__(self, config: ChatGLMConfig, device=None):
+ super(GLMTransformer, self).__init__()
+
+ self.fp32_residual_connection = config.fp32_residual_connection
+ self.post_layer_norm = config.post_layer_norm
+
+ # Number of layers.
+ self.num_layers = config.num_layers
+
+ # Transformer layers.
+ def build_layer(layer_number):
+ return GLMBlock(config, layer_number, device=device)
+
+ self.layers = torch.nn.ModuleList([build_layer(i + 1) for i in range(self.num_layers)])
+
+ if self.post_layer_norm:
+ LayerNormFunc = RMSNorm if config.rmsnorm else LayerNorm
+ # Final layer norm before output.
+ self.final_layernorm = LayerNormFunc(config.hidden_size, eps=config.layernorm_epsilon, device=device,
+ dtype=config.torch_dtype)
+
+ self.gradient_checkpointing = False
+
+ def _get_layer(self, layer_number):
+ return self.layers[layer_number]
+
+ def forward(
+ self, hidden_states, attention_mask, rotary_pos_emb, kv_caches=None,
+ use_cache: Optional[bool] = True,
+ output_hidden_states: Optional[bool] = False,
+ ):
+ if not kv_caches:
+ kv_caches = [None for _ in range(self.num_layers)]
+ presents = () if use_cache else None
+ if self.gradient_checkpointing and self.training:
+ if use_cache:
+ logger.warning_once(
+ "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+ )
+ use_cache = False
+
+ all_self_attentions = None
+ all_hidden_states = () if output_hidden_states else None
+ for index in range(self.num_layers):
+ if output_hidden_states:
+ all_hidden_states = all_hidden_states + (hidden_states,)
+
+ layer = self._get_layer(index)
+ if self.gradient_checkpointing and self.training:
+ layer_ret = torch.utils.checkpoint.checkpoint(
+ layer,
+ hidden_states,
+ attention_mask,
+ rotary_pos_emb,
+ kv_caches[index],
+ use_cache,
+ use_reentrant=False
+ )
+ else:
+ layer_ret = layer(
+ hidden_states,
+ attention_mask,
+ rotary_pos_emb,
+ kv_cache=kv_caches[index],
+ use_cache=use_cache
+ )
+ hidden_states, kv_cache = layer_ret
+ if use_cache:
+ # token by token decoding, use tuple format
+ if kv_caches[0] is not None:
+ presents = presents + (kv_cache,)
+ # prefilling in decoding, use tensor format to save cuda memory
+ else:
+ if len(presents) == 0:
+ presents = kv_cache
+ else:
+ presents = torch.cat((presents, kv_cache.to(presents.device)), dim=0)
+
+ if output_hidden_states:
+ all_hidden_states = all_hidden_states + (hidden_states,)
+
+ # Final layer norm.
+ if self.post_layer_norm:
+ hidden_states = self.final_layernorm(hidden_states)
+
+ return hidden_states, presents, all_hidden_states, all_self_attentions
+
+
+class ChatGLMPreTrainedModel(PreTrainedModel):
+ """
+ An abstract class to handle weights initialization and
+ a simple interface for downloading and loading pretrained models.
+ """
+
+ is_parallelizable = False
+ supports_gradient_checkpointing = True
+ config_class = ChatGLMConfig
+ base_model_prefix = "transformer"
+ _no_split_modules = ["GLMBlock"]
+ _supports_flash_attn_2 = True
+ _supports_sdpa = True
+
+ def _init_weights(self, module: nn.Module):
+ """Initialize the weights."""
+ return
+
+ def get_masks(self, input_ids, past_key_values, padding_mask=None):
+ if self.config._attn_implementation == "flash_attention_2":
+ if padding_mask is not None and not padding_mask.all():
+ return padding_mask
+ return None
+ batch_size, seq_length = input_ids.shape
+ full_attention_mask = torch.ones(batch_size, seq_length, seq_length, device=input_ids.device)
+ full_attention_mask.tril_()
+ past_length = 0
+ if past_key_values:
+ past_length = past_key_values[0][0].shape[2]
+ if past_length:
+ full_attention_mask = torch.cat((torch.ones(batch_size, seq_length, past_length,
+ device=input_ids.device), full_attention_mask), dim=-1)
+ if padding_mask is not None:
+ full_attention_mask = full_attention_mask * padding_mask.unsqueeze(1)
+ if not past_length and padding_mask is not None:
+ full_attention_mask -= padding_mask.unsqueeze(-1) - 1
+ full_attention_mask = (full_attention_mask < 0.5).bool()
+ full_attention_mask.unsqueeze_(1)
+ return full_attention_mask
+
+ def get_position_ids(self, input_ids, device):
+ batch_size, seq_length = input_ids.shape
+ position_ids = torch.arange(seq_length, dtype=torch.long, device=device).unsqueeze(0).repeat(batch_size, 1)
+ return position_ids
+
+class Embedding(torch.nn.Module):
+ """Language model embeddings."""
+
+ def __init__(self, config: ChatGLMConfig, device=None):
+ super(Embedding, self).__init__()
+
+ self.hidden_size = config.hidden_size
+ # Word embeddings (parallel).
+ self.word_embeddings = nn.Embedding(
+ config.padded_vocab_size,
+ self.hidden_size,
+ dtype=config.torch_dtype,
+ device=device
+ )
+ self.fp32_residual_connection = config.fp32_residual_connection
+
+ def forward(self, input_ids):
+ # Embeddings.
+ words_embeddings = self.word_embeddings(input_ids)
+ embeddings = words_embeddings
+ # If the input flag for fp32 residual connection is set, convert for float.
+ if self.fp32_residual_connection:
+ embeddings = embeddings.float()
+ return embeddings
+
+
+class ChatGLMModel(ChatGLMPreTrainedModel):
+ def __init__(self, config: ChatGLMConfig, device=None, empty_init=True):
+ super().__init__(config)
+ if empty_init:
+ init_method = skip_init
+ else:
+ init_method = default_init
+ init_kwargs = {}
+ if device is not None:
+ init_kwargs["device"] = device
+ self.embedding = init_method(Embedding, config, **init_kwargs)
+ self.num_layers = config.num_layers
+ self.multi_query_group_num = config.multi_query_group_num
+ self.kv_channels = config.kv_channels
+
+ # Rotary positional embeddings
+ self.seq_length = config.seq_length
+ rotary_dim = (
+ config.hidden_size // config.num_attention_heads if config.kv_channels is None else config.kv_channels
+ )
+
+ self.rotary_pos_emb = RotaryEmbedding(rotary_dim // 2, rope_ratio=config.rope_ratio,
+ original_impl=config.original_rope,
+ device=device, dtype=config.torch_dtype)
+ self.encoder = init_method(GLMTransformer, config, **init_kwargs)
+ self.output_layer = init_method(nn.Linear, config.hidden_size, config.padded_vocab_size, bias=False,
+ dtype=config.torch_dtype, **init_kwargs)
+
+ def get_input_embeddings(self):
+ return self.embedding.word_embeddings
+
+ def set_input_embeddings(self, value):
+ self.embedding.word_embeddings = value
+
+ def forward(
+ self,
+ input_ids,
+ position_ids: Optional[torch.Tensor] = None,
+ attention_mask: Optional[torch.BoolTensor] = None,
+ full_attention_mask: Optional[torch.BoolTensor] = None,
+ past_key_values: Optional[Tuple[Tuple[torch.Tensor, torch.Tensor], ...]] = None,
+ inputs_embeds: Optional[torch.Tensor] = None,
+ use_cache: Optional[bool] = None,
+ output_attentions: Optional[bool] = None,
+ output_hidden_states: Optional[bool] = None,
+ return_dict: Optional[bool] = None,
+ ):
+ 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
+
+ batch_size, seq_length = input_ids.shape
+
+ if inputs_embeds is None:
+ inputs_embeds = self.embedding(input_ids)
+
+ if full_attention_mask is None:
+ if (attention_mask is not None and not attention_mask.all()) or (past_key_values and seq_length != 1):
+ full_attention_mask = self.get_masks(input_ids, past_key_values, padding_mask=attention_mask)
+
+ # Rotary positional embeddings
+ rotary_pos_emb = self.rotary_pos_emb(self.seq_length)
+ if position_ids is not None:
+ rotary_pos_emb = rotary_pos_emb[position_ids]
+ else:
+ rotary_pos_emb = rotary_pos_emb[None, :seq_length]
+
+ # Run encoder.
+ hidden_states, presents, all_hidden_states, all_self_attentions = self.encoder(
+ inputs_embeds, full_attention_mask, rotary_pos_emb=rotary_pos_emb,
+ kv_caches=past_key_values, use_cache=use_cache, output_hidden_states=output_hidden_states
+ )
+ if presents is not None and type(presents) is torch.Tensor:
+ presents = presents.split(1, dim=0)
+ presents = list(presents)
+ presents = [list(x.squeeze(0).split(1, dim=0)) for x in presents]
+ presents = [tuple([x.squeeze(0) for x in y]) for y in presents]
+ presents = tuple(presents)
+
+ if not return_dict:
+ return tuple(v for v in [hidden_states, presents, all_hidden_states, all_self_attentions] if v is not None)
+
+ return BaseModelOutputWithPast(
+ last_hidden_state=hidden_states,
+ past_key_values=presents,
+ hidden_states=all_hidden_states,
+ attentions=all_self_attentions,
+ )
+
+
+class ChatGLMForConditionalGeneration(ChatGLMPreTrainedModel):
+ def __init__(self, config: ChatGLMConfig, empty_init=True, device=None):
+ super().__init__(config)
+
+ self.max_sequence_length = config.max_length
+ self.transformer = ChatGLMModel(config, empty_init=empty_init, device=device)
+ self.config = config
+
+ def _update_model_kwargs_for_generation(
+ self,
+ outputs: ModelOutput,
+ model_kwargs: Dict[str, Any],
+ is_encoder_decoder: bool = False,
+ ) -> Dict[str, Any]:
+ # update past_key_values
+ cache_name, cache = self._extract_past_from_model_output(outputs)
+ model_kwargs[cache_name] = cache
+
+ # update attention mask
+ if "attention_mask" in model_kwargs:
+ attention_mask = model_kwargs["attention_mask"]
+ model_kwargs["attention_mask"] = torch.cat(
+ [attention_mask, attention_mask.new_ones((attention_mask.shape[0], 1))], dim=-1
+ )
+
+ # update position ids
+ if "position_ids" in model_kwargs:
+ position_ids = model_kwargs["position_ids"]
+ new_position_id = position_ids[..., -1:].clone()
+ new_position_id += 1
+ model_kwargs["position_ids"] = torch.cat(
+ [position_ids, new_position_id], dim=-1
+ )
+
+ model_kwargs["is_first_forward"] = False
+ return model_kwargs
+
+ def prepare_inputs_for_generation(
+ self,
+ input_ids: torch.LongTensor,
+ past_key_values: Optional[torch.Tensor] = None,
+ attention_mask: Optional[torch.Tensor] = None,
+ position_ids: Optional[torch.Tensor] = None,
+ use_cache: Optional[bool] = None,
+ is_first_forward: bool = True,
+ **kwargs
+ ) -> dict:
+ # only last token for input_ids if past is not None
+ if position_ids is None:
+ position_ids = self.get_position_ids(input_ids, device=input_ids.device)
+ if not is_first_forward:
+ if past_key_values is not None:
+ position_ids = position_ids[..., -1:]
+ input_ids = input_ids[:, -1:]
+ return {
+ "input_ids": input_ids,
+ "past_key_values": past_key_values,
+ "position_ids": position_ids,
+ "attention_mask": attention_mask,
+ "return_last_logit": True,
+ "use_cache": use_cache
+ }
+
+ def forward(
+ self,
+ input_ids: Optional[torch.Tensor] = None,
+ position_ids: Optional[torch.Tensor] = None,
+ attention_mask: Optional[torch.Tensor] = None,
+ past_key_values: Optional[Tuple[torch.FloatTensor]] = None,
+ inputs_embeds: Optional[torch.Tensor] = None,
+ labels: Optional[torch.Tensor] = None,
+ use_cache: Optional[bool] = None,
+ output_attentions: Optional[bool] = None,
+ output_hidden_states: Optional[bool] = None,
+ return_dict: Optional[bool] = None,
+ return_last_logit: Optional[bool] = False,
+ ):
+ 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
+
+ transformer_outputs = self.transformer(
+ input_ids=input_ids,
+ position_ids=position_ids,
+ attention_mask=attention_mask,
+ past_key_values=past_key_values,
+ inputs_embeds=inputs_embeds,
+ use_cache=use_cache,
+ output_hidden_states=output_hidden_states,
+ return_dict=return_dict,
+ )
+
+ hidden_states = transformer_outputs[0]
+ if return_last_logit:
+ hidden_states = hidden_states[:, -1:]
+ lm_logits = self.transformer.output_layer(hidden_states)
+
+ loss = None
+ if labels is not None:
+ lm_logits = lm_logits.to(torch.float32)
+
+ # Shift so that tokens < n predict n
+ shift_logits = lm_logits[..., :-1, :].contiguous()
+ shift_labels = labels[..., 1:].contiguous()
+ # Flatten the tokens
+ loss_fct = CrossEntropyLoss(ignore_index=-100)
+ loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))
+
+ lm_logits = lm_logits.to(hidden_states.dtype)
+ loss = loss.to(hidden_states.dtype)
+
+ if not return_dict:
+ output = (lm_logits,) + transformer_outputs[1:]
+ return ((loss,) + output) if loss is not None else output
+
+ return CausalLMOutputWithPast(
+ loss=loss,
+ logits=lm_logits,
+ past_key_values=transformer_outputs.past_key_values,
+ hidden_states=transformer_outputs.hidden_states,
+ attentions=transformer_outputs.attentions,
+ )
+
+ @staticmethod
+ def _reorder_cache(
+ past: Tuple[Tuple[torch.Tensor, torch.Tensor], ...], beam_idx: torch.LongTensor
+ ) -> Tuple[Tuple[torch.Tensor, torch.Tensor], ...]:
+ """
+ This function is used to re-order the `past_key_values` cache if [`~PreTrainedModel.beam_search`] or
+ [`~PreTrainedModel.beam_sample`] is called. This is required to match `past_key_values` with the correct
+ beam_idx at every generation step.
+
+ Output shares the same memory storage as `past`.
+ """
+ return tuple(
+ (
+ layer_past[0].index_select(0, beam_idx.to(layer_past[0].device)),
+ layer_past[1].index_select(0, beam_idx.to(layer_past[1].device)),
+ )
+ for layer_past in past
+ )
+
+
+class ChatGLMForSequenceClassification(ChatGLMPreTrainedModel):
+ def __init__(self, config: ChatGLMConfig, empty_init=True, device=None):
+ super().__init__(config)
+
+ self.num_labels = config.num_labels
+ self.transformer = ChatGLMModel(config, empty_init=empty_init, device=device)
+
+ self.classifier_head = nn.Linear(config.hidden_size, config.num_labels, bias=True, dtype=config.torch_dtype)
+ if config.classifier_dropout is not None:
+ self.dropout = nn.Dropout(config.classifier_dropout)
+ else:
+ self.dropout = None
+ self.config = config
+
+ def forward(
+ self,
+ input_ids: Optional[torch.LongTensor] = None,
+ position_ids: Optional[torch.LongTensor] = None,
+ attention_mask: Optional[torch.Tensor] = None,
+ full_attention_mask: Optional[torch.Tensor] = None,
+ past_key_values: Optional[Tuple[Tuple[torch.Tensor, torch.Tensor], ...]] = None,
+ inputs_embeds: Optional[torch.LongTensor] = 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[torch.Tensor, ...], SequenceClassifierOutputWithPast]:
+ return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+ transformer_outputs = self.transformer(
+ input_ids=input_ids,
+ position_ids=position_ids,
+ attention_mask=attention_mask,
+ full_attention_mask=full_attention_mask,
+ 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]
+ pooled_hidden_states = hidden_states[:, -1]
+ if self.dropout is not None:
+ pooled_hidden_states = self.dropout(pooled_hidden_states)
+ logits = self.classifier_head(pooled_hidden_states)
+
+ loss = None
+ if labels is not None:
+ 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(logits.squeeze().float(), labels.squeeze())
+ else:
+ loss = loss_fct(logits.float(), labels)
+ elif self.config.problem_type == "single_label_classification":
+ loss_fct = CrossEntropyLoss()
+ loss = loss_fct(logits.view(-1, self.num_labels).float(), labels.view(-1))
+ elif self.config.problem_type == "multi_label_classification":
+ loss_fct = BCEWithLogitsLoss()
+ loss = loss_fct(logits.float(), labels.view(-1, self.num_labels))
+
+ if not return_dict:
+ output = (logits,) + transformer_outputs[1:]
+ return ((loss,) + output) if loss is not None else output
+
+ return SequenceClassifierOutputWithPast(
+ loss=loss,
+ logits=logits,
+ past_key_values=transformer_outputs.past_key_values,
+ hidden_states=transformer_outputs.hidden_states,
+ attentions=transformer_outputs.attentions,
+ )