The model is
very sensitive
to system message. Make sure you're using this system message (system role) at the beginning of the conversation:
You are MiniThinky, a helpful AI assistant. You always think before giving the answer. Use <|thinking|> before thinking and <|answer|> before giving the answer.
Q&A
Hardware used to trained it?
I used a HF space with 4xL40S, trained for 5 hours. Eval loss is about 0.8
Benchmark?
I don't have time to do it alone. If you can help, please open a discussion!
Can it count number of "r" in "raspberry"?
Unfortunately no
Other things that I can tune?
Maybe lower temperature, or set top_k=1
TODO: include more info here + maybe do some benchmarks? (Plz add a discussion if you're interested)
Runs of ngxson MiniThinky-1B-Llama-3.2 on huggingface.co
20
Total runs
0
24-hour runs
0
3-day runs
0
7-day runs
0
30-day runs
More Information About MiniThinky-1B-Llama-3.2 huggingface.co Model
MiniThinky-1B-Llama-3.2 huggingface.co
MiniThinky-1B-Llama-3.2 huggingface.co is an AI model on huggingface.co that provides MiniThinky-1B-Llama-3.2's model effect (), which can be used instantly with this ngxson MiniThinky-1B-Llama-3.2 model. huggingface.co supports a free trial of the MiniThinky-1B-Llama-3.2 model, and also provides paid use of the MiniThinky-1B-Llama-3.2. Support call MiniThinky-1B-Llama-3.2 model through api, including Node.js, Python, http.
MiniThinky-1B-Llama-3.2 huggingface.co is an online trial and call api platform, which integrates MiniThinky-1B-Llama-3.2's modeling effects, including api services, and provides a free online trial of MiniThinky-1B-Llama-3.2, you can try MiniThinky-1B-Llama-3.2 online for free by clicking the link below.
ngxson MiniThinky-1B-Llama-3.2 online free url in huggingface.co:
MiniThinky-1B-Llama-3.2 is an open source model from GitHub that offers a free installation service, and any user can find MiniThinky-1B-Llama-3.2 on GitHub to install. At the same time, huggingface.co provides the effect of MiniThinky-1B-Llama-3.2 install, users can directly use MiniThinky-1B-Llama-3.2 installed effect in huggingface.co for debugging and trial. It also supports api for free installation.
MiniThinky-1B-Llama-3.2 install url in huggingface.co: