Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Nemotron fix #61

Closed
wants to merge 4 commits into from
Closed
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 25 additions & 0 deletions nemo_skills/inference/prompt/nemotron/eval_map.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,25 @@
# Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# need to contain an EVAL_MAP dictionary mapping model version name to prompts for different benchmarks
# can have a default key that will be used if benchmark name is not in dict

EVAL_MAP = {
'base': {
'default': 'nemotron/fewshot',
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That's not really for the base model as it still uses the extra id tokens. Why do you need to add this file? If you only eval on math benchmarks, you don't need to specify the eval map stuff and our previous commands should work. If you want to eval on both math and other benchmarks, then you need to add other prompts

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This file was added because Nemotron-340B math evaluations required this eval_map file. Or we can do without it as well?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It should work without it (there was a bug, but it should be fixed now). Let me know if you're running into problems

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If you do want to merge this, then please also add the prompts for code benchmarks + a proper base prompt (or just remove base). If not, let's just not merge this for now to not confuse people since we only want to have that file when multi-area eval is actually possible

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's not merge for now then since I haven't tested prompts with Nemotron models for different tasks.

},
'instruct': { # nemotron-instruct
'default': 'nemotron/zeroshot',
},
}
Loading