llama.cpp
Add support for BitnetForCausalLM (new model / new datatype)
#7931
Merged

Add support for BitnetForCausalLM (new model / new datatype) #7931

ggerganov merged 38 commits into ggml-org:master from Eddie-Wang1120:bitnet
Eddie-Wang1120
Eddie-Wang1120 hf bitnet v1
076b4a19
Eddie-Wang1120 hf bitnet e2e v2
57dfc3bc
Eddie-Wang1120 finish bitnet e2e
1f2e0ee0
Eddie-Wang1120 finish f16 hf bitnet e2e
5e596601
Eddie-Wang1120 remove unsed
2a01a7ce
Eddie-Wang1120 finish bitnet i2 e2e
4e1ab506
Eddie-Wang1120 move i2s to quantize v1
ca090855
move i2 to quantize
dbee0a86
clean code
1c5a8b7f
clean code 2
3a0f8b06
Eddie-Wang1120 fix codestyle
97d22be5
Eddie-Wang1120 fix code
344467f2
Eddie-Wang1120 fix
65ac3a36
Eddie-Wang1120 fix code
abd798d7
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
841c903f
Eddie-Wang1120 fix merge
c0fd4df8
Eddie-Wang1120 remove unused
de1d5073
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
2322e9db
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
c0cd08d4
Eddie-Wang1120 change table name
f395dd9c
Eddie-Wang1120 fix whitespace
5e5eee7b
github-actions github-actions added examples
github-actions github-actions added python
github-actions github-actions added ggml
compilade
compilade commented on 2024-06-14
Eddie-Wang1120 delete redundant
7a8961ff
bartowski1182
Dampfinchen
JackCloudman
compilade
compilade commented on 2024-06-14
Eddie-Wang1120
Eddie-Wang1120
Eddie-Wang1120 i2_s to absmax
95dced07
mofosyne mofosyne added Tensor Encoding Scheme
Green-Sky
Eddie-Wang1120 finish i2_s/i8_s vec_dot x86 simd
569a03ed
Eddie-Wang1120
Green-Sky
ggerganov
Eddie-Wang1120
ggerganov
Eddie-Wang1120
ggerganov
Eddie-Wang1120 i2s->q22
a03eff31
compilade
compilade commented on 2024-06-17
Eddie-Wang1120 fix code
4edc958f
compilade
compilade commented on 2024-06-18
Eddie-Wang1120 remove block scale
89c7e4c1
compilade
compilade commented on 2024-06-18
compilade
compilade commented on 2024-06-19
Eddie-Wang1120 add dequantize
fcf2da46
Eddie-Wang1120 fix seq
fa9a742b
slaren
compilade
compilade commented on 2024-06-19
Eddie-Wang1120 update avx2
230396bc
Green-Sky
Green-Sky commented on 2024-06-20
Eddie-Wang1120 remove q2_2
2b097682
Eddie-Wang1120 remove q22_grid
a58cf0d6
Eddie-Wang1120
Eddie-Wang1120
Green-Sky
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
abcdc503
Eddie-Wang1120 fix whitespace
c6ddfa7e
Eddie-Wang1120
slaren
slaren commented on 2024-06-21
slaren
slaren commented on 2024-06-21
Eddie-Wang1120 reuse llm_build_kv
55a57a50
Eddie-Wang1120
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
0520d88e
gonzalo-santamaria-iic
slaren
slaren approved these changes on 2024-06-21
Eddie-Wang1120
flatsiedatsie
ggerganov
ggerganov approved these changes on 2024-06-23
ggerganov
ggerganov commented on 2024-06-23
Eddie-Wang1120 Eddie-Wang1120 requested a review from Green-Sky Green-Sky 1 year ago
Eddie-Wang1120 Eddie-Wang1120 requested a review from compilade compilade 1 year ago
Green-Sky
Green-Sky approved these changes on 2024-06-23
ggerganov
ggerganov commented on 2024-06-23
Eddie-Wang1120 Merge branch 'ggerganov:master' into bitnet
16f0c30d
Eddie-Wang1120 fix bo
226c5eed
flatsiedatsie
ggerganov ggerganov merged e112b610 into master 1 year ago
Dampfinchen

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone