huggingface/transformers

Pull Requests Commits

Revert cleanup approach, restore forward-pass fix for embed_tokens

ydshieh committed 1 day ago

1e2900df

Fix TestToolCallGemma failures via setUp/tearDown GPU cleanup

ydshieh committed 1 day ago

60ab0f95

Fix: use embed_tokens forward instead of .weight to avoid meta tensor error

ydshieh committed 1 day ago

8f56a5d1

[Gemma4] Fix pad_embedding device mismatch when accelerate offloads embed_tokens to CPU

ydshieh committed 1 day ago

dc20417d

[Docs]: Update GLM 5.3 (#48401)

Dovis01 committed 2 days ago

Verified 42ca9701

Avoid print to stdout that fails the job `check_failed_tests` job (#48391)

ydshieh committed 2 days ago

Verified 805a9e93

Fix incorrect tuple return annotations on forward methods returning a Tensor (#48359)

Gronoxx committed 2 days ago

Verified 5f8ab9bb

Fix interval merge invariant in _find_disjoint (#47860)

sharmax-vikas committed 2 days ago

Verified 0b80c4ab

[VibeVoice] Skip generate export tests (flaky) (#48396)

ydshieh committed 2 days ago

Verified efed2433

skip mtp slow tests for now (#48328) (#48329)

tarekziade committed 2 days ago

Verified 9aafba63

Update Tailscale action version in workflow (#48394)

glegendre01 committed 2 days ago

Verified 721e4927

fix some failure in xpu (#48252)

sywangyi committed 2 days ago

Verified d4dc3f28

[Improvement] Make gated delta rule more explicit (#47625)

remi-or committed 3 days ago

Verified 83d46aa2

[CB] Fix wrong device scoping (#48370)

remi-or committed 3 days ago

Verified a8e8d144

Bump transformers-mlinter to 0.1.5 and clear the new findings (#48259)

tarekziade committed 3 days ago

Verified ba356d2b

Raise when a paged attention forward is called with no cache (#48297)

qgallouedec committed 3 days ago

Verified a19f0d7b

[docs] Pass ContinuousBatchingConfig and sliding window models (#48381)

stevhliu committed 3 days ago

Verified e972043b

fix: flash-attn fallback failing on torch2.13 (#48388)

NanoCode012 committed 3 days ago

Verified 281dd533

[LongcatFlash] Fix test_longcat_generation_cpu: use device_map="cpu" to avoid MoE disk offload issue (#48377)

ydshieh committed 3 days ago

Verified 83d024e1

[Qwen3VLMoe] Update `test_small_model_integration_test_batch` expected output (value drift) (#48376)

ydshieh committed 3 days ago

Verified 155b8993

[ONNX] Skip affected models on torch 2.13 (two dynamo regressions) (#48191)

ydshieh committed 3 days ago

Verified 19876312

Fix `safe_open` mmap memory exhaustion on Windows by using `pread` backend (#48341)

eryk-roch committed 3 days ago

Verified 8631167e

Retry get_daily_ci_runs on stale GitHub API cache (#48374)

ydshieh committed 3 days ago

Verified aa372615

Fix Zamba2 construction for num_mem_blocks > 1 checkpoints (#48325)

john-rocky committed 4 days ago

Verified 323f2434

Implement VibeVoice (#40546)

pengzhiliang committed 4 days ago

Verified 640a08a5

[Docs] Change 5.3 Flash pos in toc (#48366)

Dovis01 committed 4 days ago

Verified afa80816

[conftest] Use get_cpu_ram_total_gib for psutil patch (cgroup-aware) (#48290)

ydshieh committed 4 days ago

Verified 35a66fa6

Fix flaky test_training_gradient_checkpointing for BigBirdPegasus (fp noise filter) (#48332)

ydshieh committed 4 days ago

Verified 9b4f0a0c

Quiet continuous batching at default verbosity (#48314)

qgallouedec committed 4 days ago

Verified 2e7033ed

Wait for the first request in the async continuous batching bootstrap (#48304)

qgallouedec committed 4 days ago

Verified 01846b52

Older