Update README.md

README.md

@@ -21,8 +21,8 @@ The core models released in this batch include the following:
 
 | Size | Training Tokens | Layers | Hidden Size | Attention Heads | Context Length |
 |------|--------|---------|-------------|-----------------|----------------|
-| [OLMo2-7B July 2024](https://huggingface.co/allenai/OLMo-7B-0724-hf) | 4 Trillion   | 32     | 4096        | 32              |  4096  |
-| [OLMo2- 13B July 2024](https://huggingface.co/allenai/OLMo-1B-0724-hf) | 5 Trillion   | 40     | 5120        | 42              |  4096  |
+| [OLMo2-7B July 2024](https://huggingface.co/allenai/OLMo-2-7B-1124) | 4 Trillion   | 32     | 4096        | 32              |  4096  |
+| [OLMo2- 13B July 2024](https://huggingface.co/allenai/OLMo-2-13B-1124) | 5 Trillion   | 40     | 5120        | 42              |  4096  |
 
 ## Inference
 
@@ -99,15 +99,15 @@ For more documentation, see the [GitHub readme](https://github.com/allenai/OLMo?
     - Core repo (training, inference, fine-tuning etc.): https://github.com/allenai/OLMo
     - Evaluation code: https://github.com/allenai/OLMo-Eval
     - Further fine-tuning code: https://github.com/allenai/open-instruct
-- **Paper:** [Link](https://arxiv.org/abs/2402.00838)
-- **Technical blog post:** https://blog.allenai.org/olmo-1-7-7b-a-24-point-improvement-on-mmlu-92b43f7d269d 
-- **W&B Logs:** [pretraining](https://wandb.ai/ai2-llm/OLMo-7B/groups/OLMo-1.7-7B), [annealing](https://wandb.ai/ai2-llm/OLMo-7B/groups/OLMo-1.7-7B-anneal)
+<!-- - **Paper:** [Link](https://arxiv.org/abs/2402.00838) -->
+<!-- - **Technical blog post:** https://blog.allenai.org/olmo-1-7-7b-a-24-point-improvement-on-mmlu-92b43f7d269d  -->
+<!-- - **W&B Logs:** [pretraining](https://wandb.ai/ai2-llm/OLMo-7B/groups/OLMo-1.7-7B), [annealing](https://wandb.ai/ai2-llm/OLMo-7B/groups/OLMo-1.7-7B-anneal) -->
 
 
 <!-- TODO -->
 ## Evaluation
-
-Core model results for OLMo2 7B models are found below:
+`TODO`
+<!-- Core model results for OLMo2 7B models are found below:
 
 | Task              | Llama-7b | Llama2-7b | Falcon-7b | Mpt-7b | OLMo-7B | Llama2-13b | OLMo 7B April 2024 | **OLMo2 7B** |
 |-------------------|----------|-----------|-----------|--------|---------|------------|--------------------|-----------------------|
@@ -139,71 +139,17 @@ And for OLMo 13B model:
 | sciq          | 25     | 94.7              | 86        | 90.5                                   | 88.1    | 93.4 |
 | winogrande    | 50     | 64.9              | 53.3      | 58.9                                   | 58.9    | 61.4 |
 | Average       | 36.11  | 68.41             | 56.44     | 61.48                                  | 62.42   | 65.0 |
-
-\*Unlike OLMo, Pythia, and TinyLlama, StabilityAI has not disclosed yet the data StableLM was trained on, making comparisons with other efforts challenging.
+ -->
 
 ## Model Details
 
 ### Data
-For training data details, please see the [Dolma](https://huggingface.co/datasets/allenai/dolma) documentation.
-**This model uses the new 1.7 version with more data sources, better deduplication, and quality filtering**.
-During the annealing phase we use a higher quality subset of Dolma with a linearly decaying learning rate to 0.
+`TODO`
 
 ### Staged training / annealing
+`TODO`
+
 
-In contrast to OLMo 1.0, we trained OLMo 7B July with a two-stage curriculum: 
-* In the first stage, we trained the model from scratch on the Dolma 1.7 dataset. We set a cosine learning rate schedule with a warmup of 2500 steps, a peak learning rate of 3e-4, and a cosine decay to 3e-5 after 3T tokens. We cut off this stage after 2.7T tokens, when the learning rate is still somewhat high. 
-* At this point we switch to the second stage, in which we train on a higher-quality subset of Dolma 1.7 (see below) for another 50B tokens, while linearly decaying the learning rate to 0. Our high-quality subset includes (1) using all available Wikipedia, OpenWebMath and Flan data, (2) removing Dolma CC, CC News, and Megawika, and (3) rebalancing remaining sources to achieve approximately equal proportions of each. See exact token counts and relative proportions of this second stage mix below.
-Both stages contribute equally to the final performance of the OLMo model. After the first stage, OLMo 1.7 already outperforms OLMo 1.0. The second stage consistently adds 2 to 3 points of performance on top.
-
-
-<!-- ### Architecture
-
-OLMo2 7B architecture with peer models for comparison.
-
-|                        | **OLMo2 7B** | [OLMo2 13B](https://huggingface.co/allenai/OLMo2-13B-1124) | [Llama 2 7B](https://huggingface.co/meta-llama/Llama-2-7b) | [OpenLM 7B](https://laion.ai/blog/open-lm/) | [Falcon 7B](https://huggingface.co/tiiuae/falcon-7b) | PaLM 8B |
-|------------------------|-------------------|-------------------|---------------------|--------------------|--------------------|------------------|
-| d_model                | 4096              | 4096              | 4096                | 4096               | 4544               | 4096             |
-| num heads              | 32                | 42                | 32                  | 32                 | 71                 | 16               |
-| num layers             | 32                | 40                | 32                  | 32                 | 32                 | 32               |
-| MLP ratio              | ~8/3              | ~8/3              | ~8/3                | ~8/3               | 4                  | 4                |
-| LayerNorm type         | RMS Norm          | RMS Norm          | RMSNorm             | parametric LN      | parametric LN      | parametric LN    |
-| pos embeddings         | RoPE              | RoPE              | RoPE                | RoPE               | RoPE               | RoPE             |
-| attention variant      | full              | full              | GQA                 | full               | MQA                | MQA              |
-| biases                 | none              | none              | none                | in LN only         | in LN only         | none             |
-| block type             | sequential        | sequential        | sequential          | sequential         | parallel           | parallel         |
-| activation             | SwiGLU            | SwiGLU            | SwiGLU              | SwiGLU             | GeLU               | SwiGLU           |
-| sequence length        | 4096              | 4096              | 4096                | 2048               | 2048               | 2048             |
-| batch size (instances) | 1024              | 2048              | 1024                | 2048               | 2304               | 512              |
-| batch size (tokens)    | ~4M               | ~4M               | ~4M                 | ~4M                | ~4M                | ~1M              |
-| weight tying           | no                | no                | no                  | no                 | no                 | yes              |
-
-
-### Hyperparameters 
-
-AdamW optimizer parameters are shown below.
-
-| Size | Peak LR    | Betas           | Epsilon     | Weight Decay |
-|------|------------|-----------------|-------------|--------------|
-| 7B   | 3.0E-4   | (0.9, 0.95)   | 1.0E-8    | 0.1          |
-| 13B   | 9.0E-4   | (0.9, 0.95)   | 1.0E-8    | 0.1          |
-
-Optimizer settings comparison with peer models.
-
-|                       | **OLMo2 7B**  | [OLMo2 13B](https://huggingface.co/allenai/OLMo2-13B-1124) | [Llama 2 7B](https://huggingface.co/meta-llama/Llama-2-7b) | [OpenLM 7B](https://laion.ai/blog/open-lm/) | [Falcon 7B](https://huggingface.co/tiiuae/falcon-7b) |
-|-----------------------|------------------|------------------|---------------------|--------------------|--------------------|
-| warmup steps          | 2000             | 2000             | 2000                | 2000               | 1000               |
-| peak LR               | 3.0E-04          | 9.0E-04          | 3.0E-04             | 3.0E-04            | 6.0E-04            |
-| minimum LR            | 3.0E-05          | 3.0E-05          | 3.0E-05             | 3.0E-05            | 1.2E-05            |
-| weight decay          | 0.1              | 0.1              | 0.1                 | 0.1                | 0.1                |
-| beta1                 | 0.9              | 0.9              | 0.9                 | 0.9                | 0.99               |
-| beta2                 | 0.95             | 0.95             | 0.95                | 0.95               | 0.999              |
-| epsilon               | 1.0E-08          | 1.0E-08          | 1.0E-05             | 1.0E-05            | 1.0E-05            |
-| LR schedule           | cosine           | cosine           | cosine              | cosine             | cosine             |
-| gradient clipping     | global 1.0       | global 1.0       | global 1.0          | global 1.0         | global 1.0         |
-| gradient reduce dtype | FP32             | FP32             | FP32                | FP32               | BF16               |
-| optimizer state dtype | FP32             | FP32             | most likely FP32    | FP32               | FP32               |
- -->
 
 
 ## Bias, Risks, and Limitations
@@ -211,23 +157,9 @@ Optimizer settings comparison with peer models.
 Like any base language model or fine-tuned model without safety filtering, these models can easily be prompted by users to generate harmful and sensitive content. Such content may also be produced unintentionally, especially in cases involving bias, so we recommend that users consider the risks when applying this technology. Additionally, many statements from OLMo or any LLM are often inaccurate, so facts should be verified.
 
 
-<!-- ## Citation
-
-**BibTeX:**
+## Citation
+`TODO`
 
-```
-@article{Groeneveld2023OLMo,
-  title={OLMo: Accelerating the Science of Language Models},
-  author={Groeneveld, Dirk and Beltagy, Iz and Walsh, Pete and Bhagia, Akshita and Kinney, Rodney and Tafjord, Oyvind and Jha, Ananya Harsh and Ivison, Hamish and Magnusson, Ian and Wang, Yizhong and Arora, Shane and Atkinson, David and Authur, Russell and Chandu, Khyathi and Cohan, Arman and Dumas, Jennifer and Elazar, Yanai and Gu, Yuling and Hessel, Jack and Khot, Tushar and Merrill, William and Morrison, Jacob and Muennighoff, Niklas and Naik, Aakanksha and Nam, Crystal and Peters, Matthew E. and Pyatkin, Valentina and Ravichander, Abhilasha and Schwenk, Dustin and Shah, Saurabh and Smith, Will and Subramani, Nishant and Wortsman, Mitchell and Dasigi, Pradeep and Lambert, Nathan and Richardson, Kyle and Dodge, Jesse and Lo, Kyle and Soldaini, Luca and Smith, Noah A. and Hajishirzi, Hannaneh},
-  journal={Preprint},
-  year={2024}
-}
-```
-
-**APA:**
-
-Groeneveld, D., Beltagy, I., Walsh, P., Bhagia, A., Kinney, R., Tafjord, O., Jha, A., Ivison, H., Magnusson, I., Wang, Y., Arora, S., Atkinson, D., Authur, R., Chandu, K., Cohan, A., Dumas, J., Elazar, Y., Gu, Y., Hessel, J., Khot, T., Merrill, W., Morrison, J., Muennighoff, N., Naik, A., Nam, C., Peters, M., Pyatkin, V., Ravichander, A., Schwenk, D., Shah, S., Smith, W., Subramani, N., Wortsman, M., Dasigi, P., Lambert, N., Richardson, K., Dodge, J., Lo, K., Soldaini, L., Smith, N., & Hajishirzi, H. (2024). OLMo: Accelerating the Science of Language Models. Preprint.
- -->
 ## Model Card Contact