forked from vllm-project/vllm
-
Notifications
You must be signed in to change notification settings - Fork 77
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
Extend accuracy tests for models that we support #824
Open
AnetaKaczynska
wants to merge
9
commits into
habana_main
Choose a base branch
from
dev/akaczynska/tests
base: habana_main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Changes from all commits
Commits
Show all changes
9 commits
Select commit
Hold shift + click to select a range
5df25da
Extend accuracy tests for models that we support
AnetaKaczynska 9a1c5fe
Remove Llama-2-7B
AnetaKaczynska 6d88711
Move models to existing test suites
AnetaKaczynska 3326256
Limit number of samples to 500 to speed up tests
AnetaKaczynska ecb863a
Run new small models only on G3 with TP=1
AnetaKaczynska d5e7947
Run new fp8 models only on G3 with TP=1
AnetaKaczynska cefd270
Limit number of samples to 250 for Mixtral
AnetaKaczynska a57179c
Run Mixtral only on G3 with TP=2
AnetaKaczynska 9216947
Split long tests into smaller parts
AnetaKaczynska File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
11 changes: 11 additions & 0 deletions
11
.jenkins/lm-eval-harness/configs/Mistral-7B-Instruct-v0.3.yaml
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| model_name: "/mnt/weka/data/pytorch/mistral/Mistral-7B-Instruct-v0.3" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.4905 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.5284 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" | ||
11 changes: 11 additions & 0 deletions
11
.jenkins/lm-eval-harness/configs/Mixtral-8x7B-Instruct-v0.1.yaml
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| model_name: "/mnt/weka/data/mlperf_models/Mixtral-8x7B-Instruct-v0.1" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.6967 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.6952 | ||
| limit: 250 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| model_name: "/mnt/weka/data/pytorch/Qwen/Qwen2-7b-Instruct" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.6565 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.7778 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,12 @@ | ||
| model_name: "/mnt/weka/data/pytorch/granite/granite-20b" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.5291 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.5564 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" | ||
| fp8: true |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| model_name: "/mnt/weka/data/pytorch/granite/granite-20b" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.5443 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.5670 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,12 @@ | ||
| model_name: "/mnt/weka/data/pytorch/granite/granite-8b" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.6376 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.6497 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" | ||
| fp8: true |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,11 @@ | ||
| model_name: "/mnt/weka/data/pytorch/granite/granite-8b" | ||
| tasks: | ||
| - name: "gsm8k_cot" | ||
| metrics: | ||
| - name: "exact_match,strict-match" | ||
| value: 0.6542 | ||
| - name: "exact_match,flexible-extract" | ||
| value: 0.6686 | ||
| limit: 500 | ||
| num_fewshot: 8 | ||
| dtype: "bfloat16" |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,3 @@ | ||
| Meta-Llama-3.1-8B-Instruct-fp8.yaml | ||
| granite-8b-fp8.yaml | ||
| granite-20b-fp8.yaml |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1 @@ | ||
| Mixtral-8x7B-Instruct-v0.1.yaml |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,2 @@ | ||
| granite-8b.yaml | ||
| granite-20b.yaml |
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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,2 @@ | ||
| Qwen2-7b-Instruct.yaml | ||
| Mistral-7B-Instruct-v0.3.yaml |
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
Oops, something went wrong.
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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@AnetaKaczynska I was rather more interested in the time taken by the test itself, those measurements are for whole suite including environment preparation (which can vary for many reasons, and will vary also for existing suites). The exact time for the suite can be measured by taking the "============================= test session starts ==============================" time and test finish
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, for clarity I gathered two values: