From 9b52f0e2dc858c3461c13770628d5dac71a67e69 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Dani=C3=ABl=20de=20Kok?= Date: Tue, 4 Jun 2024 14:26:07 +0200 Subject: [PATCH] Fix Phi-2 with `tp>1` (#2003) # What does this PR do? We were using the wrong parallelism in the up-projection. ## Before submitting - [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case). - [x] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests), Pull Request section? - [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link to it if that's the case. - [ ] Did you make sure to update the documentation with your changes? Here are the [documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and [here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation). - [ ] Did you write any new necessary tests? ## Who can review? Anyone in the community is free to review the PR once the tests have passed. Feel free to tag members/contributors who may be interested in your PR. --- .../models/custom_modeling/flash_phi_modeling.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/server/text_generation_server/models/custom_modeling/flash_phi_modeling.py b/server/text_generation_server/models/custom_modeling/flash_phi_modeling.py index af3206dd..53d3ea42 100644 --- a/server/text_generation_server/models/custom_modeling/flash_phi_modeling.py +++ b/server/text_generation_server/models/custom_modeling/flash_phi_modeling.py @@ -238,7 +238,7 @@ class PhiMLP(nn.Module): ) # llama weights are up_proj and down_proj and bias=False - self.up_proj = TensorParallelRowLinear.load( + self.up_proj = TensorParallelColumnLinear.load( config, prefix=f"{prefix}.fc1", weights=weights,