Neat, I like how that one includes token counting – though frustratingly it can't count Claude tokens (just OpenAI ones) because @AnthropicAI haven't released a public version of the Claude tokenizer yet
LLMS
-
Sweeps Training System for Axolotl Open Source
By
–
Built a simple sweeps system for training with Axolotl… would anyone want this/find this useful? cc @winglian
-
ChatGPT Web Interface and Code Interpreter Usage Preferences
By
–
I still do most of my casual LLM usage via the ChatGPT web interface partly out of convenience but also because that gives me Code Interpreter which I lean on for all sorts of things
-
Logging LLM Prompts to SQLite Database for Analysis
By
–
The thing I like most about using the terminal for this is that my LLM tool logs everything to a SQLite database – I have more than 2,000 logged prompt/responses now from over 100 different models Would be neat to get a notebook to log prompts globally like that as well
-
Using LLMs and ChatGPT tools: practical guide series
By
–
This is the 11th post in my ongoing series about how I use LLMs and tools like ChatGPT https://
simonwillison.net/series/using-l
lms/
… -
Files-to-Prompt Tool: Building with Claude 3 Opus
By
–
I released a new tool called files-to-prompt, which concatenates together files and directories to help pipe them into an LLM I built it almost entirely through prompting Claude 3 Opus. Here's a detailed write-up of how I did that and what I learned:
-
Stable LM 2 1.6B Update: Enhanced Conversations and Tool Usage
By
–
This release includes an update to Stable LM 2 1.6B, which improves its conversational skills in all of the seven aforementioned languages and incorporates tool usage and function calling. (2/3)
-
Stable LM 2 12B Multilingual Language Model Released
By
–
Stable LM 2 12B is a pair of powerful 12 billion parameter language models trained on multilingual data in English, Spanish, German, Italian, French, Portuguese, and Dutch, featuring a base and instruction-tuned model. You can now try the model here:
-
Memory Allocation Strategy in LLM Training Implementation
By
–
You can look at the raw training implementation here: https://
github.com/karpathy/llm.c
/blob/master/train_gpt2.c
… You'll see that we allocate all the required memory a single time in the beginning in one large block of 1D memory. From there on during training, no memory gets created or destroyed, so we stay at -
llm.c: Train GPT-2 in Pure C Without Heavy Dependencies
By
–
Have you ever wanted to train LLMs in pure C without 245MB of PyTorch and 107MB of cPython? No? Well now you can! With llm.c: https://
github.com/karpathy/llm.c To start, implements GPT-2 training on CPU/fp32 in only ~1,000 lines of clean code. It compiles and runs instantly, and exactly