TEST: Add llama logits tests #30835
Merged
TEST: Add llama logits tests #30835
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
younesbelkada
commented
May 16, 2024
| EXPECTED_SLICE = torch.tensor([-8.1406, -8.0547, 2.7461, -1.2344, -0.1448, -1.8262, -1.0020, -1.8154, -1.6895, -1.8516, -2.3574, -0.9277, 3.7598, 6.5742, -1.2998, -0.1177, -8.1406, -2.9688, -2.9199, -3.1699, -3.5254, -2.3555, -2.7988, -3.4141, -2.8262, -4.5195, -3.3379, -3.3164, -2.7832, -3.0273]) # fmt: skip | ||
| torch.testing.assert_close(out[0, 0, :30], EXPECTED_SLICE, atol=1e-5, rtol=1e-5) | ||
| model = LlamaForCausalLM.from_pretrained( | ||
| "meta-llama/Llama-2-7b-hf", device_map="auto", torch_dtype=torch.bfloat16, attn_implementation="eager" |
There was a problem hiding this comment.
we need to use eager otherwise bf16 + SDPA fails
Comment on lines
667
to
668
| 8: torch.tensor([-12.8281, -7.4609, -0.4668, -8.0703, -7.2539, -8.0078, -6.4961, -7.7734, -7.8516, -7.0352, -6.2188, -7.1367, -1.8564, 1.9922, -8.6328]) # fmt: skip | ||
| } |
There was a problem hiding this comment.
fmt skip not needed in between fmt on and off
| @require_torch_accelerator | ||
| @slow | ||
| @unittest.skip("Model is too large") | ||
| def test_model_7b_logits(self): |
There was a problem hiding this comment.
As long as the tokenizer is tested with this, LGMT to remove
ArthurZucker
approved these changes
May 17, 2024
This reverts commit b66e01e.
younesbelkada
commented
May 17, 2024
itazap
pushed a commit
that referenced
this pull request
May 24, 2024
* add llama logits test * fix * fix tests " " * fix for a10 * format * format * fix * [run-slow] remove fmt: skip * Your commit message * test commit * Revert "test commit" This reverts commit b66e01e. * [run-slow]llama * Update tests/models/llama/test_modeling_llama.py * [run-slow]llama * empty commit
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.
What does this PR do?
Adds the llama logits tests that have been always skipped 馃槩
Tested on our A10 & T4 Runners !