Vuanhngo11 commited on
Commit
ec24bd7
Β·
verified Β·
1 Parent(s): d3c8cde

Upload folder using huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -20,9 +20,9 @@ A collection of optimized GGUF quantized models derived from [bloomvn-0.5b-ppo](
20
 
21
  | Variant | Use Case | Download |
22
  |---------|-----------|------------|
23
- | base | This is the base model in FP16 format, suitable for general use cases where model size is not a concern, providing the full capabilities of the bloomvn-0.5b-ppo model. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/base.gguf)
24
- | q2_k | The 2-bit quantized variant is extremely compressed, making it ideal for highly constrained environments such as low-memory devices or applications where storage and bandwidth are limited, though it may come at the cost of some model performance. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/q2_k.gguf)
25
- | q3_k_m | The 3-bit quantized variant offers a very compressed model, suitable for memory-limited devices that still require a balance between model performance and size, making it a good choice for applications where some compression is necessary but performance cannot be heavily compromised. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/q3_k_m.gguf)
26
 
27
  ## 🀝 Contributors
28
 
 
20
 
21
  | Variant | Use Case | Download |
22
  |---------|-----------|------------|
23
+ | base | This FP16 format model is ideal for applications where high accuracy is crucial and storage space is not a concern, such as in data centers or high-performance computing environments. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/base.gguf)
24
+ | q2_k | The 2-bit quantized model is suitable for extremely constrained environments, such as low-power edge devices or those with very limited memory, where a balance between size and performance is necessary. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/q2_k.gguf)
25
+ | q3_k_m | The 3-bit quantized model offers a good balance between compression and performance, making it suitable for memory-limited devices that still require a reasonable level of accuracy, such as mid-range smartphones or embedded systems. | [πŸ“₯](https://huggingface.co/Vuanhngo11/bloomvn-0.5b-ppo-gguf/resolve/main/q3_k_m.gguf)
26
 
27
  ## 🀝 Contributors
28