Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/andrewkchan/deepseek.cpp
/ functions
Functions
195 in github.com/andrewkchan/deepseek.cpp
⨍
Functions
195
◇
Types & classes
36
↓ 1 callers
Function
moe_gate
src/infer.cpp:493
↓ 1 callers
Method
ndim
src/codec.cpp:236
↓ 1 callers
Function
per_expert_blockwise_quantize
(expert_weights: torch.Tensor, block_size: torch.Tensor, dtype: torch.dtype)
convert.py:277
↓ 1 callers
Function
per_expert_k_quantize
(expert_weights: torch.Tensor, method: Literal["q2_k", "q3_k"])
convert.py:288
↓ 1 callers
Function
per_tensor_dequantize
(qweight: torch.Tensor, scale: torch.Tensor)
convert.py:246
↓ 1 callers
Function
quant_to_codec_dtype
src/codec.cpp:55
↓ 1 callers
Function
quantize_row_q2_K_ref
src/quant.cpp:147
↓ 1 callers
Function
quantize_row_q3_K_ref
src/quant.cpp:308
↓ 1 callers
Method
rms_kv_a_weight
src/model.h:314
↓ 1 callers
Method
rms_kv_a_weight
src/model.h:404
↓ 1 callers
Method
rms_q_a_weight
src/model.h:313
↓ 1 callers
Method
rms_q_a_weight
src/model.h:403
↓ 1 callers
Function
run_interactive
src/main.cpp:514
↓ 1 callers
Method
sample_prob
src/sampler.cpp:12
↓ 1 callers
Function
sigmoid
src/infer.cpp:489
↓ 1 callers
Function
string_to_codec_dtype
src/codec.cpp:85
↓ 1 callers
Function
test_attn
src/test.cpp:70
↓ 1 callers
Function
test_matmul
src/test.cpp:128
↓ 1 callers
Method
to_dict
(self)
convert.py:123
Method
Block
src/model.cpp:149
Method
BlockMHA
src/model.cpp:354
Method
BlockMLA
src/model.cpp:518
Method
DebugTensor
src/model.h:488
Method
DebugTensor
src/model.cpp:904
Method
InferenceState
src/model.cpp:677
Method
Model
src/model.cpp:756
Function
PYBIND11_MODULE
quantizer.cpp:68
Method
ProfileDisabledScope
src/profile.cpp:48
Method
ProfileScope
src/profile.cpp:22
Method
QTensor
src/codec.h:113
Method
Sampler
src/sampler.cpp:7
Method
Session
src/main.cpp:72
Method
Tokenizer
src/tokenizer.cpp:3
Method
YALMData
src/codec.cpp:255
Method
__init__
(self, config, tokenizer_config, quant, n_layers, use_mla, bsize)
convert.py:47
Method
_attention_impl
src/infer.cpp:935
Method
_block_cpu
src/infer.cpp:811
Method
_copy_embedding
src/infer.cpp:1217
Method
_forward_cpu
src/infer.cpp:1265
Method
attention_impl
src/model.cpp:493
Function
clip
src/infer.cpp:644
Function
copy_debug_tensor
src/infer.cpp:19
Function
dequantize_row_q8_K
src/quant.cpp:655
Function
dump_debug_map
src/infer.cpp:30
Function
dump_debug_map_as_safetensors
src/infer.cpp:73
Function
ffn_cpu
src/infer.cpp:1166
Method
from_codec_tensor
src/codec.cpp:166
Method
from_directory
src/codec.cpp:333
Method
has_ext_modules
(self)
setup.py:7
Function
layernorm
src/infer.cpp:613
Function
main
src/test.cpp:312
Function
main
src/main.cpp:594
Method
max_err
src/model.cpp:913
Function
mem_bench2_thread
src/test.cpp:257
Function
mha_cpu
src/infer.cpp:1143
Method
mode
src/model.h:143
Method
moegate
src/model.h:210
Method
moegate_bias
src/model.h:211
Method
n_elements
src/codec.cpp:245
Method
parse_args
Returns true if args are valid, false otherwise
src/main.cpp:91
Method
parse_args
Returns true if args are valid, false otherwise
src/main.cpp:203
Function
quantize_q2_k
quantizer.cpp:4
Function
quantize_q3_k
quantizer.cpp:36
Method
rms_att_weight
src/model.h:205
Method
rms_ffn_weight
src/model.h:206
Method
sample_argmax
src/sampler.cpp:28
Method
save
src/debug.cpp:5
Function
save_debug_tensor
src/infer.cpp:27
Method
set_mode
src/model.h:144
Method
shared_w1
src/model.h:212
Method
shared_w2
src/model.h:213
Method
shared_w3
src/model.h:214
Function
string_to_quant
src/codec.cpp:24
Method
update_from_file
src/codec.cpp:262
Method
w1
src/model.h:207
Method
w2
src/model.h:208
Method
w3
src/model.h:209
Method
wc
src/model.h:408
Method
wkv_a
src/model.h:318
Method
wkv_a
src/model.h:406
Method
wkv_b
src/model.h:319
Method
wo
src/model.h:320
Method
wo
src/model.h:407
Method
wq
src/model.h:315
Method
wq_a
src/model.h:316
Method
wq_a
src/model.h:405
Method
wq_b
src/model.h:317
Method
wq_rope_b
src/model.h:409
Method
wv_b
src/model.h:410
Method
~Block
src/model.cpp:288
Method
~BlockMHA
src/model.cpp:463
Method
~BlockMLA
src/model.cpp:622
Method
~InferenceState
src/model.cpp:728
Method
~ProfileDisabledScope
src/profile.cpp:53
Method
~ProfileScope
src/profile.cpp:30
← previous
101–195 of 195, ranked by callers