Search results

[BUG] ChatSession should handle multiple tool calls

Status: Open.

#134 In ml-explore/mlx-swift-lm;

· davidkoski opened on Mar 4, 2026

MoE models 7.3x slower than Python mlx-lm (per-token sync + evalLock)

Status: Open.

#124 In ml-explore/mlx-swift-lm;

· dirvine opened on Feb 28, 2026

[BUG] KVCache can be reloaded but not used with ChatSession

Status: Open.

#123 In ml-explore/mlx-swift-lm;

· davidkoski opened on Feb 27, 2026

[BUG] MLXLMTests.ChatSessionTests testChatSessionSync are unreliable

Status: Open.

#119 In ml-explore/mlx-swift-lm;

· davidkoski opened on Feb 25, 2026

Add support for PaddleOCR-VL

model wantedDesire for a model to be portedDesire for a model to be ported

Status: Open.

#101 In ml-explore/mlx-swift-lm;

· lhr0909 opened on Feb 11, 2026

[Proposal] make this a pure lib, without HuggingFace dependency

Status: Open.

#98 In ml-explore/mlx-swift-lm;

· zallsold-lgtm opened on Feb 10, 2026

[BUG] MLXEmbedders need to respect max token counts

Status: Open.

#62 In ml-explore/mlx-swift-lm;

· davidkoski opened on Jan 15, 2026

GPT-oss requires Harmony format (render + parse)

Status: Open.

#61 In ml-explore/mlx-swift-lm;

· ronaldmannak opened on Jan 14, 2026

[BUG] Jinja template parsing fails with modulo error for FunctionGemma model

Status: Open.

#31 In ml-explore/mlx-swift-lm;

· agabeyalioglu opened on Dec 22, 2025

[BUG] Qwen3VL-4B-4bit is slow in generation

Status: Open.

#19 In ml-explore/mlx-swift-lm;

· jimmycmlo opened on Dec 10, 2025

Support for Hunyuan

model wantedDesire for a model to be portedDesire for a model to be ported

Status: Open.

#21 In ml-explore/mlx-swift-lm;

· wzqww23 opened on Dec 7, 2025

Support for Deepseek OCR

model wantedDesire for a model to be portedDesire for a model to be ported

Status: Open.

#15 In ml-explore/mlx-swift-lm;

· noorbhatia opened on Dec 5, 2025