|
|
--- |
|
|
license: other |
|
|
language: |
|
|
- en |
|
|
pipeline_tag: text-generation |
|
|
tags: |
|
|
- facebook |
|
|
- meta |
|
|
- llama |
|
|
- llama3.1 |
|
|
- llama 3.1 |
|
|
- llama 8b |
|
|
- gguf |
|
|
--- |
|
|
|
|
|
# With fixes applied for: |
|
|
- 3.1 rope scaling factors (#8676) |
|
|
- `llama-bpe` as tokenizer Proper Llama 3.1 Support in llama.cpp (#8650) |
|
|
- `<|python_tag|>` works for tool calls. |
|
|
|
|
|
# Following files are fixed and others are being replaced. |
|
|
- Meta-Llama-3.1-8B-Instruct.Q4_K_M.gguf |
|
|
- Meta-Llama-3.1-8B-Instruct.IQ4_XS.gguf |
|
|
|
|
|
|
|
|
# REF |
|
|
https://github.com/ggerganov/llama.cpp/issues/8650 |
|
|
https://github.com/ggerganov/llama.cpp/pull/8676 |
|
|
|
