qubit999/llama3.2-3b-instruct 🔢📝✓ → 📝

▶️ 33 runs 📅 Dec 2024 ⚙️ Cog 0.13.6 🔗 GitHub ⚖️ License
code-generation multilingual text-generation

About

The Llama 3.2 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction-tuned generative models in 1B and 3B sizes (text in/text out).

Example Output

Prompt:

"Name 3 animals with wings"

Output

{"role":"assistant","content":"Here are 3 animals with wings:

  1. Butterfly
  2. Eagle
  3. Bat"}

Performance Metrics

0.70s Prediction Time
169.26s Total Time
All Input Parameters 
{
  "top_k": 0,
  "top_p": 0.9,
  "prompt": "Name 3 animals with wings",
  "do_sample": true,
  "temperature": 0.7,
  "system_prompt": "You are an AI chatbot.",
  "length_penalty": 1,
  "max_new_tokens": 4096,
  "min_new_tokens": 1,
  "repetition_penalty": 1
}
Input Parameters
top_k Type: integerDefault: 0
Top-k sampling
top_p Type: numberDefault: 0.9
Top-p (nucleus) sampling
prompt Type: stringDefault: Name 3 animals with wings
Question
do_sample Type: booleanDefault: true
Use sampling
temperature Type: numberDefault: 0.7
Sampling temperature
system_prompt Type: stringDefault: You are an AI chatbot.
System prompt
length_penalty Type: numberDefault: 1
Length penalty
max_new_tokens Type: integerDefault: 4096
Maximum number of new tokens to generate
min_new_tokens Type: integerDefault: 1
Minimum number of new tokens to generate
repetition_penalty Type: numberDefault: 1
Repetition penalty
Output Schema

Output

Type: string

Version Details
Version ID
80d7b7baea950c6069ff5093952857521cc9a10e7ac9b038d54fa2456b2fcd02
Version Created
December 5, 2024
Run on Replicate →