ml-explore mlx-examples issues

ml-explore / mlx-examples

Examples in the MLX framework

MIT License

6.27k stars 895 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Memory leak in mlx_lm.server?

#1124 ivanfioravanti opened 8 hours ago
3
Add `last_segment_tokens` to `stream_generate` GenerationResponse

#1123 mattjcly opened 11 hours ago
0
Put prompt processing in same stream

#1122 awni closed 12 hours ago
0
docs: update stream_generate return type annotation

#1121 madroidmaq closed 13 hours ago
0
mlx_lm.generate --system-prompt doesn't convert `\n` as new line.

#1120 chigkim closed 1 day ago
1
Fix object property value in mlx_lm.server chat completions response not matching OpenAI spec

#1119 kconner closed 1 day ago
2
Fix: Allow converting whisper models from local paths

#1118 remixer-dec closed 1 day ago
0
Allow loading from diffusers ckpt

#1117 angeloskath opened 3 days ago
0
Use other flux-based model files?

#1116 puppyapple opened 4 days ago
1
Fix format

#1115 angeloskath closed 5 days ago
0
Add seed argument to stable diffusion image to image

#1114 louen closed 5 days ago
0
Fix data_iter in prepare_dataset from speechcommands example

#1113 sakares opened 5 days ago
2
ValueError: [dequantize] The matrix should be given as a uint32

#1111 chaihahaha closed 5 days ago
4
Unable to load Qwen2-VL-1.5B-Instruct model using mlx_lm

#1110 dolphingarlic closed 1 day ago
1
[Feature Request] Unable to Unload or Swap adapters at runtime

#1108 chimezie opened 1 week ago
2
[MLX LM] Fix f-string formatting in memory warning message

#1105 skeetmtp closed 1 week ago
0
Significantly reduced inference speed after Lora finetunig

#1104 hschaeufler opened 1 week ago
1
Completion only fine-tuning of instruction models with collections of HF datasets

#1103 chimezie opened 2 weeks ago
4
libc++abi: Unable to build metal library from source error: invalid value 'metal3.1' in '-std=metal3.1'

#1102 maxlund closed 2 weeks ago
11
Whisper transcribing music sounds

#1101 aaa3334 opened 2 weeks ago
2
Tencent HunYuan MOE model

#1100 awni closed 2 days ago
3
Generation refactor: part 2

#1099 awni closed 2 days ago
0
Support for AI21 Jamba-1.5

#1097 vlbosch opened 2 weeks ago
1
FLUX: Add support for setup configuration to publish module

#1096 madroidmaq opened 2 weeks ago
0
LoRA trains on tokens with duplicated BOS for some models

#1095 chimezie closed 2 weeks ago
3
[MLX LM] Sampler refactor + a few improvements

#1094 awni closed 2 weeks ago
1
Fix rotating kv cache size

#1093 angeloskath closed 2 weeks ago
0
Fix spm decoder multi-byte

#1092 awni closed 2 weeks ago
0
"Your computer was restarted because of a problem" after QLoRa on Mistral Nemo fused and (later) quantized checkpoint

#1091 chimezie closed 2 weeks ago
9
Generalize HF datasets to a collection of HF datasets via `hf_datasets`

#1090 chimezie closed 1 week ago
1
chore(mlx-lm): add max token arg for mlx_lm.chat

#1089 mzbac closed 3 weeks ago
0
[Feature Request] Custom "chat" HF datasets

#1088 chimezie opened 3 weeks ago
0
load_custom_hf_dataset not handling the text_feature argument properly

#1087 chimezie opened 3 weeks ago
0
Custom local dataset features

#1085 chimezie opened 3 weeks ago
0
[mlx-whisper] Error while transcribing - Cannot infer the shape of an empty array

#1084 thelazyoxymoron closed 3 weeks ago
3
mlx-whisper Not installable on python 3.13

#1083 alper closed 3 weeks ago
2
Fix whisper edge case when decoding past limit

#1082 awni closed 3 weeks ago
0
Clear cache every now and then

#1081 awni closed 3 weeks ago
0
Whisper improvements

#1080 awni closed 3 weeks ago
0
Fix special tokens for deep seek

#1079 awni closed 3 weeks ago
0
Add Max Token Limit for Generation

#1078 N8python opened 3 weeks ago
29
Detokenizer crash with the mlx-lm 19.2 and DeepSeek-Coder-V2 EOS Token

#1077 mattjcly closed 3 weeks ago
0
[BUG] memory during fine tuning grows until MacBook crashes

#1076 hschaeufler closed 3 weeks ago
4
Quantized KV Cache

#1075 barronalex closed 3 weeks ago
1
Add support for falcon-mamba

#1074 ilyasch2 closed 3 weeks ago
1
Fix detokenizer space match for quote

#1072 awni closed 4 weeks ago
0
[BUG] missing spaces in response with mlx-lm 0.19.1 and 0.19.2

#1073 hschaeufler closed 4 weeks ago
2
Cannot train llama vision 3.2

#1071 pdufour closed 1 month ago
1
add timeout for generate

#1070 jincdream closed 4 weeks ago
1
Wire models in MLX LM

#1069 awni closed 3 weeks ago
0