xenova/transformers.js

Pull Requests Commits

xenova committed 168 days ago

c8aea33c

add support for aggregation strategy

xenova committed 168 days ago

59fb238d

Add support for openai_privacy_filter

xenova committed 168 days ago

bc4fce18

use tsconfig.json for type tests

xenova committed 168 days ago

077ceac5

Update types.test.js

xenova committed 168 days ago

6d07670a

Update token-classification pipeline output type tests

xenova committed 168 days ago

acd80bec

Update test_pipelines_token_classification.js

xenova committed 168 days ago

5d7c8080

[version] Update to 4.1.0

xenova committed 175 days ago

Verified ca10e4d1

Update default generation parameters (#1649)

xenova committed 175 days ago

Verified fbab88f6

Re-enable SmolVLM (#1648)

pdufour committed 175 days ago

Verified 39bec702

chore: bump doc-builder SHA for PR upload workflow (#1644)

rtrompier committed 175 days ago

Verified 78d4ec3d

Add support for q1, q1f16, q2, q2f16 data types (#1647)

xenova committed 175 days ago

Verified fdd496e9

Improve tokenizer types based on input function parameters (#1641)

xenova committed 176 days ago

Verified ed84c15f

Cached generation improvements (+ past_key_values via pipeline function) (#1638)

xenova committed 177 days ago

Verified 8d615c5a

🔒 Pin GitHub Actions to commit SHAs (#1626)

paulinebm committed 183 days ago

Verified 60379e22

[version] Update to 4.0.1

xenova committed 188 days ago

Verified 64da6359

casually dropping the most capable open weights on the planet (#1627)

xenova committed 188 days ago

Verified 3cee1ac2

[version] Update to 4.0.0 (#1620)

xenova committed 191 days ago

Verified 364ebd41

[version] Update to 4.0.0-next.11

xenova committed 191 days ago

Verified 3458fa5e

Upgrade onnxruntime-web version (#1618)

xenova committed 191 days ago

Verified 28f92674

Fix custom cache in Node.js when cache returns non-paths (e.g., Response) (#1617)

xenova committed 191 days ago

Verified ab22b27e

Emit `progress_total` events from `PreTrainedModel.from_pretrained()` (#1615)

xenova committed 191 days ago

Verified d8340931

Fix ModelRegistry calls when loading multimodal models with text-generation pipeline (#1614)

xenova committed 191 days ago

Verified a0d86d50

Overdue Whisper fixes (#1594)

xenova committed 191 days ago

Verified 43b26629

[version] Update to 4.0.0-next.10

xenova committed 193 days ago

Verified d6280f73

Add support for Cohere ASR (#1610)

xenova committed 194 days ago

Verified b368d75e

Fix typo in WebGPU documentation (#1607)

mattsoulanille committed 195 days ago

Verified e94d5592

[version] Update to 4.0.0-next.9

xenova committed 196 days ago

Verified d5610ad6

feat: add ModelRegistry.get_available_dtypes() to detect quantization (#1580)

sroussey committed 197 days ago

Verified b8228dad

Add support for qwen3_5_text <-> Qwen3_5ForCausalLM (#1602)

xenova committed 197 days ago

Verified 99a8eebc

Older