add autoround._generate_recipe() #758
Closed
Conversation
This file contains hidden or 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
Signed-off-by: xinhe3 <[email protected]>
xin3he
force-pushed
the
xinhe/mix-precision
branch
from
027b2c2 to
d8b831e
Compare
xin3he
commented
Aug 24, 2025
Signed-off-by: xinhe3 <[email protected]>
Signed-off-by: xinhe3 <[email protected]>
wenhuach21
reviewed
Aug 25, 2025
wenhuach21
reviewed
Aug 25, 2025
wenhuach21
reviewed
Aug 25, 2025
Signed-off-by: xinhe3 <[email protected]>
xin3he
force-pushed
the
xinhe/mix-precision
branch
from
419b63c to
79323d6
Compare
|
Have you tested it on an MoE model? It might require some special handling. |
Signed-off-by: xinhe3 <[email protected]>
Signed-off-by: xinhe3 <[email protected]>
wenhuach21
reviewed
Aug 25, 2025
| combination_list = [] | ||
| numel_list = [] | ||
| loss_list = [] | ||
| for hp_layers in combinations(quantizable_layers, quantizable_num): |
There was a problem hiding this comment.
How about creating a new file in the experimental folder and moving these there?
wenhuach21
reviewed
Aug 25, 2025
| clear_memory() | ||
| input_ids = to_device(input_ids, self.cache_device) | ||
| input_others = to_device(input_others, self.cache_device) | ||
| if self.recipe_mode: |
There was a problem hiding this comment.
It would be better to wrap this new code into a function and call it as early as possible.
wenhuach21
reviewed
Aug 25, 2025
auto_round/autoround.py
Outdated
|
|
||
| return current_input_ids, current_input_others | ||
|
|
||
| def _dump_average_bits(self, layer_config=None): |
There was a problem hiding this comment.
This function cannot be used by AutoRound, since layers are converted to QuantizedLinear after quantization. If the function can correctly dump average bits in typical scenarios such as INT4, I’d prefer to keep it in the class. Otherwise, it would be better to move it elsewhere for now.
Signed-off-by: xinhe3 <[email protected]>
Signed-off-by: xinhe3 <[email protected]>
Signed-off-by: xinhe3 <[email protected]>
Signed-off-by: xinhe3 <[email protected]>
xin3he
force-pushed
the
xinhe/mix-precision
branch
3 times, most recently
from
580ec8b to
f02331d
Compare
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.
This code is used for INC accuracy tuning, currently only
mx_fp8is supported for mixing withmx_fp4andnv_fp4.