minors

src/blog_multimodal_fundation.md

@@ -15,9 +15,14 @@
 - knowledge distill
 
 ---
-## Some tools
+## Some Popular Papers
+
+#### Efficiently Modeling Long Sequences with Structured State Spaces [[pdf]](https://arxiv.org/pdf/2111.00396)
+_Albert Gu, Karan Goel, and Christopher R´e_
+
+#### Mamba: Linear-Time Sequence Modeling with Selective State Spaces [[pdf]](https://arxiv.org/pdf/2312.00752)
+_Albert Gu and Tri Dao_
 
-- [run llama2 locally](https://www.linkedin.com/pulse/three-steps-run-llama-2-7b-chat-model-any-cpu-machine-nirmal-patel-3pw9f/)
 
 ---
 ## Fundation Models

src/running_with_llm_win11.md

@@ -0,0 +1,28 @@
+## Install Docker + WSL + Ubuntu + Nvidia on Win11
+[link](https://blog.csdn.net/godblesstao/article/details/135893429)
+1. Install Docker Desktop
+2. Install WSL2 (Ubuntu22.04)
+3. Update Win11 Nvidia driver and cuda>=12.0 via Nvidia GeForce Experience
+4. MobaXterm with session Ubuntu22.04
+5. Install Docker in Ubuntu22.04
+6. Install Nvidia Container Toolkit
+
+## Both run in docker!
+### [Ollama](https://ollama.com/)
+```
+sudo docker run -d --gpus=all -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
+sudo docker exec -it ollama ollama run llama3
+```
+### [OpenWebUI](https://github.com/open-webui/open-webui)
+```
+sudo docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main
+http://localhost:3000
+```
+
+## [Llama3](https://github.com/meta-llama/llama3)
+```
+torchrun --nproc_per_node 1 example_chat_completion.py \
+    --ckpt_dir Meta-Llama-3-8B-Instruct/ \
+    --tokenizer_path Meta-Llama-3-8B-Instruct/tokenizer.model \
+    --max_seq_len 512 --max_batch_size 6
+```