3

Source document

[2025-04-04 15:38] lisamegawatts: conversation_id=None messages=[USER: how to export, ASSISTANT: To export your resume, you can use the `render` command provided by the `resumed` tool. Here's how you can do it: 1. **Install the necessary packages**: Make sure you have `resumed` and a JSON Resume theme installed. For example: ```shell npm install resumed jsonresume-theme-even ``` 2. **Use the `render` command**: Run the following command in your terminal: ```shell resumed render [filename] --theme [theme-name] ``` Replace `[filename]` with the path to your resume JSON file (default is `resume.json`), and `[theme-name]` with the name of the theme you installed (e.g., `jsonresume-theme-even`). 3. **Options**: - `-o`, `--output`: Specify the output filename (default is `resume.html`). - `-t`, `--theme`: Specify the theme to use. Example command: ```shell resumed render resume.json --theme jsonresume-theme-even ``` This will generate an **resume.html** file by default. If you want to specify a different output file, you can do so by using the `-o` or `--output` option: ```shell resumed render resume.json --theme jsonresume-theme-even --output custom-output.html ``` This will create a file named `custom-output.html` by default.] metadata={} [2025-04-05 06:03] lisamegawatts: https://colab.research.google.com/drive/1KoVg0OrMioZihTwkqCxrpmNxMeZeCZfc?usp=sharing [2025-04-06 00:59] lisamegawatts: <@164501800613969920> “These models are our best yet thanks to distillation from Llama 4 Behemoth, a 288 billion active parameter model with 16 experts that is our most powerful yet and among the world’s smartest LLMs. Llama 4 Behemoth outperforms GPT-4.5, Claude Sonnet 3.7, and Gemini 2.0 Pro on several STEM benchmarks. Llama 4 Behemoth is still training, and we’re excited to share more details about it even while it’s still in flight” [2025-04-06 01:00] lisamegawatts: Llama4 is distillation in action [2025-04-06 01:05] traves_theberge: <@1211062099137265723> https://huggingface.co/zed-industries/zeta Have you seen how they are doing there training with SFT Fine-Tuning and DPO Fine-Tuning [2025-04-06 01:06] lisamegawatts: Yea the one we tried was sft [2025-04-06 01:07] lisamegawatts: Wonder if the note book works out of box, we can try 😄 [2025-04-06 01:07] traves_theberge: The example here is using standard SFT but the DPO is really interacting to me. [2025-04-06 01:08] traves_theberge: (files: 11yGILqsbZ0PKuGKQySvmHQ.png) [2025-04-06 01:10] traves_theberge: https://medium.com/@mauryaanoop3/detailed-guide-on-dpo-fine-tuning-027815d15837 [2025-04-06 01:12] lisamegawatts: This one compares sft, dpo, and gdpo https://github.com/BigBinnie/GDPO [2025-04-06 01:14] lisamegawatts: The reasoning gemma i made used grpo, https://pub.towardsai.net/group-relative-policy-optimization-grpo-illustrated-breakdown-explanation-684e71b8a3f2