Skip to content

Pull requests: ml-explore/mlx-lm

New pull request New
139 Open 641 Closed
139 Open 641 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Support for Zyphra/ZAYA1-base
#1261 opened May 9, 2026 by kyr0 Loading…
Fix LFM2.5 tool parser inference
#1260 opened May 8, 2026 by blairhudson Loading…
Fix server XTC crash from heterogeneous xtc_special_tokens
#1258 opened May 7, 2026 by odysa • Draft
3 of 5 tasks
Fix ArraysCache missing is_trimmable/trim for hybrid model prompt cache
#1254 opened May 6, 2026 by EagerofLight Loading…
1
Fix BatchRotatingKVCache rotated flag deserializing to True
#1251 opened May 6, 2026 by odysa Loading…
1
Fix mlx_lm.server --adapter-path silently ignored at startup
#1249 opened May 6, 2026 by odysa Loading…
1
Drop redundant lm_head AWQ quant triple in load_model
#1247 opened May 6, 2026 by scyyh11 Loading…
add: lfm2/2.5 tool parser
#1246 opened May 5, 2026 by jbuchananr Loading…
3 tasks done
2
chore: remove unused imports and variables
#1244 opened May 5, 2026 by odysa Loading…
fix: wrap ast.literal_eval in try/except for Qwen3 tool parser
#1239 opened May 3, 2026 by lawcontinue Loading…
1
fix(gemma4): add stop_gradient on MoE router top_k_indices
#1238 opened May 2, 2026 by TrentCarter Loading…
Add PLaMo 3 model support
#1234 opened Apr 30, 2026 by mitmul Loading…
1
[transformers-to-mlx skill] Add Talkie (TalkieForCausalLM) model
#1231 opened Apr 30, 2026 by warshanks Loading…
2
fix(generate): avoid None entries in merged logits_processors
#1230 opened Apr 29, 2026 by BLuchterhand Loading…
1
Support loading a system prompt from a file
#1229 opened Apr 29, 2026 by Mottl Loading…
Skip lm_head on non-rank-0 pipeline-parallel ranks
#1228 opened Apr 29, 2026 by lawcontinue Loading…
1
[transformers-to-mlx skill] Add bailing_hybrid (Ling-2.6-flash) model
#1227 opened Apr 29, 2026 by ivanfioravanti Contributor Loading…
12
generate.py: GenerationBatch.filter — add else branches so logits_processors / samplers length stays in lockstep with uids
#1225 opened Apr 28, 2026 by mloiterman Loading…
1
Support per-expert MoE checkpoints in qwen3_5_moe.sanitize, plus FP8 dequant
#1224 opened Apr 28, 2026 by sdayal Loading…
Add Laguna XS.2
#1223 opened Apr 28, 2026 by Blaizzy Contributor Loading…
2
Add Talkie model support
#1220 opened Apr 28, 2026 by ZimengXiong Loading…
Add MiMo V2.5
#1219 opened Apr 28, 2026 by kernelpool Contributor Loading…
2
feat(server): opt-in disk-backed L2 prompt cache (--prompt-cache-disk-dir)
#1218 opened Apr 27, 2026 by freddyhaddad Loading…
Add Metal VJP kernel for gated_delta_update (trainable Qwen3.5 / Qwen3-Next LoRA on Apple Silicon)
#1217 opened Apr 27, 2026 by SudarkinV Loading…
fix(utils): skip already-quantized layers in load_model._quantize predicate
#1216 opened Apr 27, 2026 by adurham Contributor Loading…
ProTip! Exclude everything labeled bug with -label:bug.
⚡