StarCoder2-3B-GGUF / README.md
apepkuss79's picture
Update README.md
61e18f5 verified
metadata
base_model: bigcode/starcoder2-3b
inference: false
license: bigcode-openrail-m
library_name: transformers
model_creator: bigcode
model_name: StarCoder2 3B
pipeline_tag: text-generation
quantized_by: Second State Inc.
tags:
  - code

StarCoder2-3B-GGUF

Original Model

bigcode/starcoder2-3b

Run with LlamaEdge

  • LlamaEdge version: coming soon

  • Context size: 3072

Quantized GGUF Models

Name Quant method Bits Size Use case
starcoder2-3b-Q2_K.gguf Q2_K 2 1.15 GB smallest, significant quality loss - not recommended for most purposes
starcoder2-3b-Q3_K_L.gguf Q3_K_L 3 1.68 GB small, substantial quality loss
starcoder2-3b-Q3_K_M.gguf Q3_K_M 3 1.51 GB very small, high quality loss
starcoder2-3b-Q3_K_S.gguf Q3_K_S 3 1.31 GB very small, high quality loss
starcoder2-3b-Q4_0.gguf Q4_0 4 1.71 GB legacy; small, very high quality loss - prefer using Q3_K_M
starcoder2-3b-Q4_K_M.gguf Q4_K_M 4 1.85 GB medium, balanced quality - recommended
starcoder2-3b-Q4_K_S.gguf Q4_K_S 4 1.74 GB small, greater quality loss
starcoder2-3b-Q5_0.gguf Q5_0 5 2.09 GB legacy; medium, balanced quality - prefer using Q4_K_M
starcoder2-3b-Q5_K_M.gguf Q5_K_M 5 2.16 GB large, very low quality loss - recommended
starcoder2-3b-Q5_K_S.gguf Q5_K_S 5 2.09 GB large, low quality loss - recommended
starcoder2-3b-Q6_K.gguf Q6_K 6 2.49 GB very large, extremely low quality loss
starcoder2-3b-Q8_0.gguf Q8_0 8 3.22 GB very large, extremely low quality loss - not recommended

Quantized with llama.cpp b2308