Feat: Support for `falcon-mamba` architecture #9074
ggerganov
merged
18 commits into
ggml-org:master
from
younesbelkada:add-fm-support
feat: initial support for llama.cpp
59a08be7
fix: lint
bfa02868
refactor: better refactor
b97704c9
Update src/llama.cpp
343b5836
Update src/llama.cpp
a8109e35
fix: address comments
184a4c67
Update convert_hf_to_gguf.py
34942654
fix: add more cleanup and harmonization
f7d2e910
fix: lint
9e22bb7e
compilade
commented
on 2024-08-18
Update gguf-py/gguf/gguf_writer.py
4553502d
fix: change name
d637bb97
Apply suggestions from code review
57c3eb41
add in operator
bf5e3440
fix: add `dt_b_c_rms` in `llm_load_print_meta`
ca4db9e5
fix: correct printf format for bool
78ad84f0
fix: correct print format
7aeccbb7
compilade
approved these changes
on 2024-08-18
Update src/llama.cpp
5c0f108e
ggerganov
approved these changes
on 2024-08-19
llama : quantize more Mamba tensors
3491291a
Login to write a write a comment.
Login via GitHub