Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Index your code
hub
/
github.com/ml-explore/mlx-lm
/ types & classes
Types & classes
816 in github.com/ml-explore/mlx-lm
⨍
Functions
2,622
◇
Types & classes
816
↳
Endpoints
9
↓ 58 callers
Class
KVCache
mlx_lm/models/cache.py:325
↓ 41 callers
Class
ArraysCache
mlx_lm/models/cache.py:594
↓ 38 callers
Class
RotatingKVCache
mlx_lm/models/cache.py:410
↓ 32 callers
Class
SwitchGLU
mlx_lm/models/switch_layers.py:160
↓ 28 callers
Class
MockCache
tests/test_server.py:75
↓ 18 callers
Class
BatchGenerator
A batch generator implements continuous batching. This class provides automatic management of prompt processing and generation batches,
mlx_lm/generate.py:1486
↓ 13 callers
Class
ScaleConfig
mlx_lm/quant/awq.py:25
↓ 12 callers
Class
CacheList
mlx_lm/models/cache.py:814
↓ 10 callers
Class
MultiLinear
mlx_lm/models/mla.py:9
↓ 9 callers
Class
BatchRotatingKVCache
mlx_lm/models/cache.py:1133
↓ 9 callers
Class
LRUPromptCache
mlx_lm/models/cache.py:1623
↓ 8 callers
Class
BitLinear
BitLinear module with memory-efficient weight handling.
mlx_lm/models/bitlinear_layers.py:92
↓ 7 callers
Class
BatchKVCache
mlx_lm/models/cache.py:912
↓ 7 callers
Class
RMSNorm
mlx_lm/models/gemma3_text.py:104
↓ 7 callers
Class
SwitchLinear
mlx_lm/models/switch_layers.py:93
↓ 6 callers
Class
SequenceStateMachine
A state machine that uses one Aho-Corasick trie per state to efficiently track state across a generated sequence. The transitions are provide
mlx_lm/generate.py:943
↓ 5 callers
Class
RMSNorm
mlx_lm/models/gemma2.py:30
↓ 5 callers
Class
RMSNorm
mlx_lm/models/plamo2.py:40
↓ 5 callers
Class
ZeroCenteredRMSNorm
mlx_lm/models/step3p5.py:66
↓ 4 callers
Class
LoRA
mlx_lm/models/rwkv7.py:162
↓ 4 callers
Class
PromptTrieResult
mlx_lm/models/cache.py:1524
↓ 4 callers
Class
SuScaledRoPE
mlx_lm/models/rope_utils.py:10
↓ 3 callers
Class
AWQConfig
mlx_lm/quant/awq.py:34
↓ 3 callers
Class
CacheDataset
mlx_lm/tuner/datasets.py:158
↓ 3 callers
Class
ChunkedKVCache
mlx_lm/models/cache.py:731
↓ 3 callers
Class
DeepseekV3Model
mlx_lm/models/deepseek_v3.py:319
↓ 3 callers
Class
FusedLinear
mlx_lm/models/afm7.py:123
↓ 3 callers
Class
NemotronLayerNorm1P
mlx_lm/models/nemotron.py:54
↓ 3 callers
Class
QuantizedKVCache
mlx_lm/models/cache.py:232
↓ 3 callers
Class
RMSNorm
mlx_lm/models/recurrent_gemma.py:39
↓ 3 callers
Class
RMSNorm
mlx_lm/models/gemma.py:27
↓ 3 callers
Class
Response
mlx_lm/server.py:226
↓ 3 callers
Class
ResponseGenerator
mlx_lm/server.py:440
↓ 3 callers
Class
ShortConv1d
mlx_lm/models/kimi_linear.py:235
↓ 2 callers
Class
APIHandler
mlx_lm/server.py:1059
↓ 2 callers
Class
Attention
mlx_lm/models/gpt2.py:29
↓ 2 callers
Class
BailingMoeMLP
mlx_lm/models/bailing_moe.py:60
↓ 2 callers
Class
CompletionRequest
mlx_lm/server.py:198
↓ 2 callers
Class
DeepseekMLP
mlx_lm/models/deepseek.py:108
↓ 2 callers
Class
DeepseekV2MLP
mlx_lm/models/deepseek_v2.py:248
↓ 2 callers
Class
DeepseekV32MLP
mlx_lm/models/deepseek_v32.py:262
↓ 2 callers
Class
DeepseekV3MLP
mlx_lm/models/deepseek_v3.py:172
↓ 2 callers
Class
DoRALinear
mlx_lm/tuner/dora.py:9
↓ 2 callers
Class
Dots1MLP
mlx_lm/models/dots1.py:162
↓ 2 callers
Class
DummyModelProvider
tests/test_server.py:24
↓ 2 callers
Class
Ernie4_5_MLP
mlx_lm/models/ernie4_5_moe.py:94
↓ 2 callers
Class
FusedLoRALinear
mlx_lm/models/afm7.py:32
↓ 2 callers
Class
GenerationContext
mlx_lm/server.py:209
↓ 2 callers
Class
GenerationResponse
The output of :func:`stream_generate`. Args: text (str): The next segment of decoded text. This can be an empty string. toke
mlx_lm/generate.py:270
↓ 2 callers
Class
Glm4MoeLiteMLP
mlx_lm/models/glm4_moe_lite.py:177
↓ 2 callers
Class
KimiMLP
mlx_lm/models/kimi_linear.py:57
↓ 2 callers
Class
KlearMLP
mlx_lm/models/Klear.py:110
↓ 2 callers
Class
LayerNorm2D
mlx_lm/models/cohere.py:30
↓ 2 callers
Class
LayerNormPerHead
mlx_lm/models/stablelm.py:30
↓ 2 callers
Class
LinearSubblockReplacement
A simple linear layer used to replace Attention or MLP blocks.
mlx_lm/models/nemotron-nas.py:238
↓ 2 callers
Class
LongcatFlashDecoderLayer
mlx_lm/models/longcat_flash.py:280
↓ 2 callers
Class
MLP
mlx_lm/models/bailing_moe_linear.py:114
↓ 2 callers
Class
MLP
mlx_lm/models/gpt2.py:71
↓ 2 callers
Class
MLP
mlx_lm/models/glm4_moe.py:111
↓ 2 callers
Class
MLP
mlx_lm/models/mimo_v2_flash.py:127
↓ 2 callers
Class
MLP
mlx_lm/models/exaone_moe.py:113
↓ 2 callers
Class
MLP
mlx_lm/models/afm7.py:266
↓ 2 callers
Class
MLP
mlx_lm/models/llama4.py:137
↓ 2 callers
Class
MLP
mlx_lm/models/hunyuan.py:144
↓ 2 callers
Class
MLP
mlx_lm/models/afmoe.py:137
↓ 2 callers
Class
MLXLM
mlx_lm/evaluate.py:72
↓ 2 callers
Class
NemotronHMLP
mlx_lm/models/nemotron_h.py:297
↓ 2 callers
Class
Qwen3NextMLP
mlx_lm/models/qwen3_next.py:161
↓ 2 callers
Class
SparseMoeBlock
mlx_lm/models/dbrx.py:125
↓ 2 callers
Class
Step3p5MLP
mlx_lm/models/step3p5.py:76
↓ 2 callers
Class
SwiGLU
mlx_lm/models/gpt_oss.py:62
↓ 2 callers
Class
SwitchMLP
mlx_lm/models/switch_layers.py:202
↓ 2 callers
Class
TokenShift
mlx_lm/models/rwkv7.py:198
↓ 2 callers
Class
TokenizerWrapper
A wrapper that combines an HF tokenizer and a detokenizer. Accessing any attribute other than the ``detokenizer`` is forwarded to the hugging
mlx_lm/tokenizer_utils.py:287
↓ 2 callers
Class
TrainingArgs
mlx_lm/tuner/trainer.py:42
↓ 1 callers
Class
AFMModel
mlx_lm/models/afm7.py:330
↓ 1 callers
Class
AfmoeMoE
mlx_lm/models/afmoe.py:167
↓ 1 callers
Class
AfmoeModel
mlx_lm/models/afmoe.py:278
↓ 1 callers
Class
ApertusAttention
mlx_lm/models/apertus.py:51
↓ 1 callers
Class
ApertusDecoderLayer
mlx_lm/models/apertus.py:117
↓ 1 callers
Class
ApertusMLP
mlx_lm/models/apertus.py:36
↓ 1 callers
Class
ApertusModel
mlx_lm/models/apertus.py:137
↓ 1 callers
Class
Attention
mlx_lm/models/qwen2.py:32
↓ 1 callers
Class
Attention
mlx_lm/models/ernie4_5.py:31
↓ 1 callers
Class
Attention
mlx_lm/models/cohere2.py:33
↓ 1 callers
Class
Attention
mlx_lm/models/gemma4_text.py:176
↓ 1 callers
Class
Attention
mlx_lm/models/bailing_moe_linear.py:137
↓ 1 callers
Class
Attention
mlx_lm/models/lfm2_moe.py:54
↓ 1 callers
Class
Attention
mlx_lm/models/qwen3_moe.py:37
↓ 1 callers
Class
Attention
mlx_lm/models/openelm.py:57
↓ 1 callers
Class
Attention
mlx_lm/models/plamo.py:28
↓ 1 callers
Class
Attention
mlx_lm/models/granite.py:36
↓ 1 callers
Class
Attention
mlx_lm/models/glm4_moe.py:49
↓ 1 callers
Class
Attention
mlx_lm/models/mimo_v2_flash.py:54
↓ 1 callers
Class
Attention
mlx_lm/models/gemma.py:37
↓ 1 callers
Class
Attention
mlx_lm/models/exaone_moe.py:164
↓ 1 callers
Class
Attention
mlx_lm/models/minicpm.py:45
↓ 1 callers
Class
Attention
mlx_lm/models/afm7.py:165
↓ 1 callers
Class
Attention
mlx_lm/models/llama.py:45
↓ 1 callers
Class
Attention
mlx_lm/models/minicpm3.py:39
next →
1–100 of 816, ranked by callers