Skip to content

Pull requests: ml-explore/mlx-lm

New pull request New
191 Open 651 Closed
191 Open 651 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

[Cohere] Add cohere2_moe model support
#1340 opened Jun 2, 2026 by Terrencezzj Loading…
Add Mellum (Mellum 2) model support
#1339 opened Jun 2, 2026 by jedisct1 Loading…
feat(gpt-oss): support unsloth BF16 layout + shared MoE LoRA adapters
#1338 opened Jun 1, 2026 by electron-rare Loading…
5
Add deepseek_dsml tool parser for DeepSeek-V4's native DSML format
#1337 opened May 31, 2026 by snagnever Loading…
Fix json_tools: match "<tool_call" prefix to survive ">" token merge
#1336 opened May 31, 2026 by snagnever Loading…
1
Support for Poolside LagunaXS open source coding model in nvfp4
#1334 opened May 31, 2026 by tzachicohen Loading…
Fix LFM2.5 pythonic tool parser auto-detection
#1333 opened May 31, 2026 by ChristianWeyer Loading…
Fix seeded stochastic completions in the server
#1331 opened May 30, 2026 by lyonsno Contributor Loading…
1
Fix detokenizer for byte-level tokenizers with an SPM-style decoder (fixes #1041)
#1329 opened May 30, 2026 by robertlangdonn Loading…
1
Fix Qwen3-Coder tool parser and harden server against mid-stream client disconnects
#1328 opened May 30, 2026 by i1rr Loading…
4 tasks done
Fix mlx_lm.server 404 on short prompts (clamp negative start in think-token search)
#1327 opened May 29, 2026 by devYRPauli Loading…
1
Add Step 3.7 Flash
#1325 opened May 29, 2026 by kernelpool Contributor Loading…
Consolidate MLA kv_b_proj sanitize/shard into shared helpers
#1324 opened May 29, 2026 by scyyh11 Loading…
1
Guard MLA caches from KV quantization
#1323 opened May 28, 2026 by xxxkkw Loading…
Add BatchQuantizedKVCache
#1322 opened May 28, 2026 by xxxkkw Loading…
Keep model parameters for fuse GGUF export
#1321 opened May 28, 2026 by abnormal749 Loading…
Fix Qwen3.5 MTP sanitize norm shift
#1320 opened May 28, 2026 by xxxkkw Loading…
Add optional fused SwitchGLU gate-up projection
#1319 opened May 28, 2026 by xxxkkw Loading…
Add opt-in recurrent profiling instrumentation
#1318 opened May 28, 2026 by xxxkkw Loading…
Add MiniCPM5 XML tool-call parser
#1317 opened May 28, 2026 by scrappylabsai Loading…
2 tasks
Add olmo hybrid
#1315 opened May 27, 2026 by cmurray1105 Loading…
Add hybrid prefix cache restore coverage
#1314 opened May 27, 2026 by xxxkkw Loading…
Warn when speculative decoding may hurt throughput for MoE models
#1313 opened May 26, 2026 by Shylin26 Loading…
1
Expose tunable Metal memory limits in the trainer
#1312 opened May 26, 2026 by kru2710shna Loading…
1
Fix Qwen3_5 mixed-bit per-tensor quantization keys not remapped through sanitize
#1311 opened May 25, 2026 by ivaniguarans Loading…
1
ProTip! Updated in the last three days: updated:>2026-05-30.