Skip to content

Actions: ml-explore/mlx-lm

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
867 workflow runs
867 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Add MTP support for Step 3.5 Flash Build and Test #1942: Pull request #901 synchronize by janhilgard
Action required janhilgard:feat/step3p5-model
janhilgard:feat/step3p5-model
Action required
Fix ChunkedKVCache dropping valid tokens during decode (llama4) Build and Test #1941: Pull request #1673 opened by binaydhakal
Action required binaydhakal:fix-chunked-kv-cache-decode
binaydhakal:fix-chunked-kv-cache-decode
Action required
Raise clear error when LoRA config uses 'alpha' instead of 'scale' Build and Test #1940: Pull request #1671 opened by arjunneema09
Action required arjunneema09:arjunneema09-patch-1
arjunneema09:arjunneema09-patch-1
Action required
Validate per-prompt max_tokens in batch_generate Build and Test #1939: Pull request #1668 synchronize by kupilikula
Action required kupilikula:validate-batch-generate-max-tokens
kupilikula:validate-batch-generate-max-tokens
Action required
Validate per-prompt max_tokens in batch_generate Build and Test #1938: Pull request #1668 opened by kupilikula
Action required kupilikula:validate-batch-generate-max-tokens
kupilikula:validate-batch-generate-max-tokens
Action required
Add packed gated delta kernel, bitwise-pinned by an explicit-tree comparator Build and Test #1937: Pull request #1559 synchronize by wyanzhao
Action required wyanzhao:mlx-lm-pr-packed-gdn
wyanzhao:mlx-lm-pr-packed-gdn
Action required
Add prompt checkpoints for hybrid caches Build and Test #1936: Pull request #1667 opened by stale2000
Action required stale2000:hybrid-prompt-checkpoints
stale2000:hybrid-prompt-checkpoints
Action required
Save the best-validation-loss adapter, and add optional early stopping Build and Test #1935: Pull request #1665 opened by axiom-of-choice
Action required axiom-of-choice:lora-best-checkpoint-early-stopping
axiom-of-choice:lora-best-checkpoint-early-stopping
Action required
Fix Falcon-Mamba long prompt resource exhaustion (#1656) Build and Test #1934: Commit 254d153 pushed by michalk8
4m 46s main
main
4m 46s
Fix Falcon-Mamba long prompt resource exhaustion Build and Test #1933: Pull request #1656 synchronize by michalk8
4m 57s mrjoema:joema-1st-pr
mrjoema:joema-1st-pr
4m 57s
Honor seed=0 in iterate_batches (#1661) Build and Test #1932: Commit 00aeaea pushed by michalk8
4m 46s main
main
4m 46s
Fix evaluate crash on tasks with no stop sequences Build and Test #1931: Pull request #1664 opened by choipilkyu
4m 44s choipilkyu:fix-rstrip-until-empty
choipilkyu:fix-rstrip-until-empty
4m 44s
Add Apertus tool parser Build and Test #1930: Pull request #1663 opened by gabriben
Action required gabriben:apertus-tool-parser
gabriben:apertus-tool-parser
Action required
Honor seed=0 in iterate_batches Build and Test #1929: Pull request #1661 opened by axiom-of-choice
5m 10s axiom-of-choice:fix-iterate-batches-seed-zero
axiom-of-choice:fix-iterate-batches-seed-zero
5m 10s
Add optional gradient clipping and non-finite guard to the LoRA trainer Build and Test #1928: Pull request #1659 opened by axiom-of-choice
Action required axiom-of-choice:lora-grad-clip
axiom-of-choice:lora-grad-clip
Action required
Add support for Kanana-2 Tiny (kanana2_tiny) Build and Test #1927: Pull request #1655 synchronize by choipilkyu
Action required choipilkyu:add-kanana2-tiny
choipilkyu:add-kanana2-tiny
Action required
fix(generate): guard per-sequence logits_processors against None in batched generation Build and Test #1926: Pull request #1657 opened by chronoshift
Action required chronoshift:fix-batched-logits-processors-none
chronoshift:fix-batched-logits-processors-none
Action required
Fix Falcon-Mamba long prompt resource exhaustion Build and Test #1925: Pull request #1656 opened by mrjoema
4m 56s mrjoema:joema-1st-pr
mrjoema:joema-1st-pr
4m 56s
Add Granite tool parser Build and Test #1924: Pull request #1264 synchronize by jonpspri
Action required jonpspri:granite
jonpspri:granite
Action required
Add support for Kanana-2 Tiny (kanana2_tiny) Build and Test #1923: Pull request #1655 opened by choipilkyu
Action required choipilkyu:add-kanana2-tiny
choipilkyu:add-kanana2-tiny
Action required
Enable readline for the chat prompt and persist history Build and Test #1922: Pull request #1654 opened by Vlor999
Action required Vlor999:feat/chat-cli-readline
Vlor999:feat/chat-cli-readline
Action required
Align quantized_kv_start defaults with the CLI Build and Test #1921: Pull request #1652 opened by Vlor999
Action required Vlor999:fix/quantized-kv-start-default-v2
Vlor999:fix/quantized-kv-start-default-v2
Action required
Serve fully cached repos without a Hub round-trip Build and Test #1920: Pull request #1650 opened by Vlor999
Action required Vlor999:fix/load-local-files-only
Vlor999:fix/load-local-files-only
Action required
Add Gear model support Build and Test #1919: Pull request #1648 opened by crinex
Action required crinex:add-gear-model
crinex:add-gear-model
Action required
Fix MiniMax M2 nullable tool arguments Build and Test #1918: Pull request #1647 opened by Mr-Neutr0n
Action required Mr-Neutr0n:fix/minimax-m2-nullable-arguments
Mr-Neutr0n:fix/minimax-m2-nullable-arguments
Action required