RickyChen / 陳昭儒 c455b771fd [Bugfix][CPU] Fix RotaryEmbedding fallback causing gibberish with --enforce-eager (#31643) ...

Signed-off-by: rickychen-infinirc <ricky.chen@infinirc.com>

2026-01-06 01:25:38 +08:00

..

__init__.py

Standardise get_rope to use rope_parameters["partial_rotary_factor"], not rotary_dim (#30389)

2025-12-11 20:45:23 +00:00

base.py

[Bugfix][CPU] Fix RotaryEmbedding fallback causing gibberish with --enforce-eager (#31643)

2026-01-06 01:25:38 +08:00

common.py

[CustomOp][Refactor] Extract common methods for ApplyRotaryEmb CustomOp (#31021)

2025-12-19 22:16:09 +08:00

deepseek_scaling_rope.py

[Perf] enable flashinfer rotary_embedding custom ops in DeepSeek rotary (#30729)

2025-12-18 14:31:18 -05:00

dual_chunk_rope.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

dynamic_ntk_alpha_rope.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

dynamic_ntk_scaling_rope.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

ernie45_vl_rope.py

[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)

2025-12-15 19:08:16 -08:00

linear_scaling_rope.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

llama3_rope.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

llama4_vision_rope.py

[XPU][bugfix] fix rope for llama4 and deepseek (#25145)

2025-10-30 09:43:13 +08:00

mrope.py

[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)

2025-12-15 19:08:16 -08:00

ntk_scaling_rope.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

phi3_long_rope_scaled_rope.py

[Bugfix] In LongRoPE, decide short vs long based on max_model_len (#27431)

2025-10-28 12:00:56 +00:00

xdrope.py

[CustomOp] Extract ApplyRotaryEmb as CustomOp and unify the dispatch logic (#29873)

2025-12-15 19:08:16 -08:00

yarn_scaling_rope.py

Add truncate arg to yarn to match openai implementation of gpt-oss (#28244)

2025-11-20 18:53:50 +08:00