Fixes configuration default values#43592
Conversation
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
zucchini-nlp
added
the
for patch
|
Adding fix to |
zucchini-nlp
changed the title
|
run-slow: cohere2, deformable_detr, emu3, exaone4, falcon_mamba, fast_vlm, flava, florence2, glm46v, got_ocr2, gpt_bigcode, gpt_neox, gptj, internvl, jetmoe, mamba |
|
This comment contains models: ["models/cohere2", "models/deformable_detr", "models/emu3", "models/exaone4", "models/falcon_mamba", "models/fast_vlm", "models/flava", "models/florence2", "models/glm46v", "models/got_ocr2", "models/gpt_bigcode", "models/gpt_neox", "models/gptj", "models/internvl", "models/jetmoe", "models/mamba"] |
CI Results✅ No failing test specific to this PR 🎉 ! |
|
@bot /style |
|
Deformable detr is flaky now, apparently related to the random order of tests 😢 Not reproducible locally if I run a single testcase |
|
@bot /repo |
|
Repo. Consistency bot fixed some files and pushed the changes. |
|
[For maintainers] Suggested jobs to run (before merge) run-slow: cohere2, cohere2_vision, deepseek_vl, deepseek_vl_hybrid, deformable_detr, emu3, exaone4, falcon_mamba, fast_vlm, flava, florence2, glm46v, got_ocr2, gpt_bigcode, gpt_neox, gptj |
|
run-slow: llava_onevision, llava_next_video |
|
This comment contains models: ["models/llava_next_video", "models/llava_onevision"] |
CI Results✅ No failing test specific to this PR 🎉 ! |
3 participants
What does this PR do?
Fixes #43525
Fixes #43572
Adds missing
pad_token_idandtie_word_embeddingsto config classes with their defaults