Upload README.md with huggingface_hub

README.md

@@ -1,77 +1,72 @@
 ---
 library_name: peft
 base_model: meta-llama/Llama-2-7b-hf
+license: cc-by-nc-4.0
+datasets:
+- meta-math/MetaMathQA
+- open-web-math/open-web-math
+- bigcode/starcoderdata
+- ise-uiuc/Magicoder-Evol-Instruct-110K
+language:
+- en
 ---
 
-# Model Card for Model ID
+# LoRA Learns Less and Forgets Less
 
-<!-- Provide a quick summary of what the model is/does. -->
 
+These are model checkpoints and LoRA adapters from the research paper ["LoRA Learns Less and Forgets Less"](https://arxiv.org/abs/2405.09673) (Biderman et al. TMLR, 2024). This work was done in collaboration with [Databricks Mosaic AI Research](https://www.databricks.com/research/mosaic).
 
 
 ## Model Details
 
-### Model Description
 
-<!-- Provide a longer summary of what this model is. -->
+- **Developed by:** Databricks Mosaic AI
+- **Model type:** Research Artifacts
+- **Language(s) (NLP):** English
+- **License:** cc-by-nc-4.0
+- **Finetuned from model:** Llama-2-7b
 
+We trained [Llama-2-7B](https://huggingface.co/meta-llama/Llama-2-7b-hf) using full finetuning and LoRA. Model checkpoints and LoRA adapters can be found on HuggingFace here: [LoRA-TMLR-2024](https://huggingface.co/LoRA-TMLR-2024). Intermediate checkpoints can be found in the branches of the respective models.
 
 
-- **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]
+| Setting | Dataset | HuggingFace Collection |
+| --------| ------| ------ |
+| Continued Pretraining - Code | [StarCoder-Python](https://huggingface.co/datasets/bigcode/starcoderdata) | [LoRA-TMLR-2024/continued-pretraining-code-starcoder-python](https://huggingface.co/collections/LoRA-TMLR-2024/continued-pretraining-code-starcoder-python-66f22ce3b26f416f21f58142) |
+| Continued Pretraing - Math | [OpenWebMath](https://huggingface.co/datasets/open-web-math/open-web-math) | [LoRA-TMLR-2024/continued-pretraining-math-openwebmath](https://huggingface.co/collections/LoRA-TMLR-2024/continued-pretraining-math-openwebmath-66f31d12f55fb27de05b2e3f) |
+| Instruction Finetuning - Code | [Magicoder-Evol-Instruct-110K](https://huggingface.co/datasets/ise-uiuc/Magicoder-Evol-Instruct-110K)| [LoRA-TMLR-2024/instruction-finetuning-code-magicoder-evol-instruct-110k](https://huggingface.co/collections/LoRA-TMLR-2024/instruction-finetuning-code-magicoder-evol-instruct-110k-66f224a800152f31e4942a3b) |
+| Instruction Finetuning - Math | [MetaMathQA](https://huggingface.co/datasets/meta-math/MetaMathQA) | [LoRA-TMLR-2024/instruction-finetuning-math-metamathqa](https://huggingface.co/collections/LoRA-TMLR-2024/instruction-finetuning-math-metamathqa-66f31cc40fda6b6b938d33e2) |
 
-### Model Sources [optional]
+All training was done using the Databricks MosaicML
+[composer](https://github.com/mosaicml/composer), [streaming](https://github.com/mosaicml/streaming), and [llm-foundry](https://github.com/mosaicml/llm-foundry) repositories, as well as the HuggingFace [peft](https://huggingface.co/docs/peft/en/index) library.
+
+### Model Sources
 
 <!-- Provide the basic links for the model. -->
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+- **Repository:** [https://github.com/danbider/lora-tradeoffs](https://github.com/danbider/lora-tradeoffs)
+- **Paper:** [LoRA Learns Less and Forgets Less](https://arxiv.org/abs/2405.09673)
+
+### Abstract
+
+Low-Rank Adaptation (LoRA) is a widely-used parameter-efficient finetuning method for
+large language models. LoRA saves memory by training only low rank perturbations to
+selected weight matrices. In this work, we compare the performance of LoRA and full
+finetuning on two target domains, programming and mathematics. We consider both the
+instruction finetuning (≈100K prompt-response pairs) and continued pretraining (≈20B
+unstructured tokens) data regimes. Our results show that, in the standard low-rank settings,
+LoRA substantially underperforms full finetuning. Nevertheless, LoRA better maintains the
+base model’s performance on tasks outside the target domain. We show that LoRA mitigates
+forgetting more than common regularization techniques such as weight decay and dropout;
+it also helps maintain more diverse generations. Finally, we show that full finetuning learns
+perturbations with a rank that is 10-100× greater than typical LoRA configurations, possibly
+explaining some of the reported gaps. We conclude by proposing best practices for finetuning
+with LoRA.
 
 ## 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. -->
+These are research artifacts that are intended for research purposes only.
 
-### 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
 
@@ -79,124 +74,147 @@ Use the code below to get started with the model.
 
 <!-- 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. -->
+The following datasets were used:
 
-#### Preprocessing [optional]
+| Setting | Dataset | 
+| --------| ------| 
+| Continued Pretraining - Code | [StarCoder-Python](https://huggingface.co/datasets/bigcode/starcoderdata) |
+| Continued Pretraing - Math | [OpenWebMath](https://huggingface.co/datasets/open-web-math/open-web-math) | 
+| Instruction Finetuning - Code | [Magicoder-Evol-Instruct-110K](https://huggingface.co/datasets/ise-uiuc/Magicoder-Evol-Instruct-110K)| 
+| Instruction Finetuning - Math | [MetaMathQA](https://huggingface.co/datasets/meta-math/MetaMathQA) | 
 
-[More Information Needed]
 
+### Training Procedure
 
-#### 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. -->
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
 
-[More Information Needed]
+In all four scenarios below, we use the Llama-2-7B base model [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf). For
+the CPT runs, we use the [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) tokenizer, while for the IFT runs we use the
+[meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) tokenizer.
+
+
+## Code CPT (StarCoder-Python)
+
+[StarCoder-Python](https://huggingface.co/datasets/bigcode/starcoderdata) (Li et al., 2023a) This dataset consists of permissively licensed repositories from GitHub, including Git commits, in 80+ programming languages. We chose the Python
+subset and sub-sampled it to 20B tokens.
+
+| Parameter                    | Value                                                                                   |
+|------------------------------|-----------------------------------------------------------------------------------------|
+| seq_len                      | 4096                                                                                    |
+| optimizer                    | decoupled_lionw (betas=[0.9, 0.95])                                                     |
+| learning_rate                | 1.0e-05 for LoRA and Full Finetuning                                                    |
+| scheduler                    | inv_sqrt_with_warmup (t_scale=1000ba, t_warmup=1000ba, t_cooldown=5086ba, alpha_f_decay=1, alpha_f_cooldown=0) |
+| weight_decay                 | 1.0e-06                                                                                 |
+| precision                    | amp_bf16                                                                                |
+| global_train_batch_size      | 192                                                                                     |
+| device_train_microbatch_size | 6                                                                                       |
+| gradient_clipping            | norm (threshold=1)                                                                      |
+| num_gpus                     | 32                                                                                      |
+
+We trained models for 0.25B, 0.5B, 1B, 2B, 4B, 8B, 16B and 20B tokens. These checkpoints can be found for each LoRA and full finetuning setting in the HuggingFace model branches.
+
+## Math CPT (OpenWebMath)
+
+[OpenWebMath](https://huggingface.co/datasets/open-web-math/open-web-math) (Paster et al., 2023) - This dataset contains 14.7B tokens derived from mathematical web pages from Common Crawl, correctly formatted to preserve mathematical content such as LaTeX equations. To match with the StarCoder-Python dataset, we trained on up to 20B tokens, repeating tokens beyond the first 14.7B. An analysis of this dataset shows that it contains a considerable amount of full English sentences.
+
+| Parameter                    | Value                                                                                   |
+|------------------------------|-----------------------------------------------------------------------------------------|
+| max_seq_len                  | 4096                                                                                    |
+| optimizer                    | decoupled_lionw (betas=[0.9, 0.95])                                                     |
+| learning_rate                | 1.0e-05 for full finetuning, 4.0e-05 for LoRA                                           |
+| scheduler                    | inv_sqrt_with_warmup (t_scale=1000ba, t_warmup=1000ba, t_cooldown=5086ba, alpha_f_decay=1, alpha_f_cooldown=0) |
+| weight_decay                 | 0                                                                                       |
+| precision                    | amp_bf16                                                                                |
+| global_train_batch_size      | 192                                                                                     |
+| device_train_microbatch_size | 6                                                                                       |
+| gradient_clipping            | norm (threshold=1)                                                                      |
+| num_gpus                     | 32                                                                                      |
+
+
+We trained models for 0.25B, 0.5B, 1B, 2B, 4B, 8B, 16B and 20B tokens. These checkpoints can be found for each LoRA and full finetuning setting in the HuggingFace model branches.
+
+## Code IFT (Magicoder-Evol-Instruct-110K)
+
+[Magicoder-Evol-Instruct-110K](https://huggingface.co/datasets/ise-uiuc/Magicoder-Evol-Instruct-110K) (Wei et al., 2023) This dataset contains 72.97M tokens
+of programming questions and answers. It reproduces the “Evol-Instruct” dataset of WizardCoder (Luo et al., 2023b) by iteratively prompting an LLM (GPT-4) to increase the difficulty of a set of question-answer pairs
+from Code Alpaca (Chaudhary, 2023).
+
+| Parameter                    | Value                                                                                   |
+|------------------------------|-----------------------------------------------------------------------------------------|
+| max_seq_len                  | 4096                                                                                    |
+| optimizer                    | decoupled_lionw (betas=[0.9, 0.95])                                                     |
+| learning_rate                | 5e-5 for full finetuning; 2e-4 for rank r = 16, 64 and 1e-4 for r = 256 α = 2r = 512 (due to instabilities/loss spikes at 2e-4) |
+| scheduler                    | cosine_with_warmup (alpha_f=0.01, t_warmup=0.1dur)                                      |
+| weight_decay                 | 0                                                                                       |
+| precision                    | amp_bf16                                                                                |
+| global_train_batch_size      | 192                                                                                     |
+| device_train_microbatch_size | 6                                                                                       |
+| gradient_clipping            | norm (threshold=1)                                                                      |
+| num_gpus                     | 32                                                                                      |
+
+Each model was finetuned separately for 1, 2, 4, 8 and 16 epochs.
+
+| Epoch    | Number of Batches | Estimated Tokens |   
+| -------- | ---------- | ----------------|
+| 1        | 193        | 72,970,000      |
+| 2        | 386        | 145,940,000     |
+| 4        | 772        | 291,880,000     |
+| 8        | 1544       | 583,760,000     |
+| 16       | 3088       | 1,167,520,000   |
+
+## Math IFT (MetaMathQA)
+
+[MetaMathQA](https://huggingface.co/datasets/meta-math/MetaMathQA) (Yu et al., 2023) This dataset was built by bootstrapping mathematical
+word problems from the training sets of GSM8K (Cobbe et al., 2021) and MATH (Hendrycks et al., 2021) by
+rewriting the questions with variations using GPT-3.5. This dataset contains 395K question-answer pairs and
+roughly 103M tokens.
+
+| Parameter                    | Value                                                                                   |
+|------------------------------|-----------------------------------------------------------------------------------------|
+| seq_len                      | 1024                                                                                    |
+| optimizer                    | decoupled_lionw (betas=[0.9, 0.95])                                                     |
+| learning_rate                | Full finetuning: 1e-5, LoRA: 1e-4 for r = 16, 64, 5e-5 for r = 256 due to instabilities |
+| scheduler                    | cosine_with_warmup (alpha_f=0.01, t_warmup=0.1dur)                                      |
+| weight_decay                 | 0                                                                                       |
+| precision                    | amp_bf16                                                                                |
+| global_train_batch_size      | 768                                                                                     |
+| device_train_microbatch_size | 24                                                                                      |
+| gradient_clipping            | norm (threshold=1)                                                                      |
+| num_gpus                     | 32                                                                                      |
+
+Each model was finetuned separately for 1, 2, 4, 8 and 16 epochs.
+
+| Epoch    | Estimated Tokens |   
+| -------- | ----------------|
+| 1        |  103,000,000  |
+| 2        | 206,000,000   |
+| 4        | 412,000,000   |
+| 8        | 824,000,000   |
+| 16       | 1,648,000,000 |
 
 ## 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]
+Model performance can be found in the paper [LoRA Learns Less and Forgets Less](https://arxiv.org/pdf/2405.09673). See Appendix for relevant tables.
 
-#### 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. -->
+## Citation
 
 **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]
+```
+@article{
+biderman2024lora,
+title={Lo{RA} Learns Less and Forgets Less},
+author={Dan Biderman and Jacob Portes and Jose Javier Gonzalez Ortiz and Mansheej Paul and Philip Greengard and Connor Jennings and Daniel King and Sam Havens and Vitaliy Chiley and Jonathan Frankle and Cody Blakeney and John Patrick Cunningham},
+journal={Transactions on Machine Learning Research},
+issn={2835-8856},
+year={2024},
+url={https://openreview.net/forum?id=aloEru2qCG},
+note={Featured Certification}
+}
+```
 
-## Model Card Contact
 
-[More Information Needed]
 ### Framework versions
 
 - PEFT 0.11.1