Don't use GGUF but convert T5 directly from Hugging Face #967
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sogartar
force-pushed
the
users/sogartar/t5-irpa-directly-from-hf
branch
2 times, most recently
from
7559cc5 to
8af8997
Compare
sogartar
requested review from
KyleHerndon,
rsuderman,
archana-ramalingam and
IanNod
sogartar
force-pushed
the
users/sogartar/t5-irpa-directly-from-hf
branch
from
8af8997 to
e39df9b
Compare
sogartar
force-pushed
the
users/sogartar/t5-irpa-directly-from-hf
branch
from
e39df9b to
b0d3637
Compare
KyleHerndon
approved these changes
Feb 17, 2025
|
|
||
| gguf_to_optional_config_names_map = { | ||
| "t5.decoder_start_token_id": ["decoder_start_token_id"], | ||
| def from_hugging_face_config( |
There was a problem hiding this comment.
Not blocking on this, but can these keys be discovered automatically?
| @@ -498,19 +493,19 @@ def testCompareAgainstTransformers( | |||
|
|
|||
| theta = Theta( | |||
| { | |||
| "attn_q.weight": DefaultPrimitiveTensor( | |||
| "q.weight": DefaultPrimitiveTensor( | |||
There was a problem hiding this comment.
What is the motivation with all the renames? Is it just the corresponding names from huggingface's layout instead of gguf's?
We don't want the stack to depend on the conversion tool from the lamma.cpp repo. Also the conversion to GGUF would not convert all tensors to bf16, but leave some in f32. We would like to control that ourselves if needed. This change makes any previously generated IRPA files obsolete.
sogartar
force-pushed
the
users/sogartar/t5-irpa-directly-from-hf
branch
from
e0286ac to
26b0ad7
Compare
IanNod
pushed a commit
that referenced
this pull request
Feb 22, 2025
We don't want the stack to depend on the conversion tool from the lamma.cpp repo. Also the conversion to GGUF would not convert all tensors to bf16, but leave some in f32. We would like to control that ourselves if needed. This change makes any previously generated IRPA files obsolete.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
We don't want the stack to depend on the conversion tool from the lamma.cpp repo. Also the conversion to GGUF would not convert all tensors to bf16, but leave some in f32. We would like to control that ourselves if needed.
This change makes any previously generated IRPA files obsolete.