huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-06-19 15:52:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

[Gaudi] Fix the integration-test issues (#3265)

Browse Source 
Signed-off-by: yuanwu <yuan.wu@intel.com>
This commit is contained in:
Yuan Wu 2025-06-13 20:47:06 +08:00 committed by GitHub
parent ded4cb52ac
commit 3752143b39
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 5 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
backends/gaudi/Makefile
View File

@ -50,6 +50,7 @@ local-dev-install: install-dependencies
# In order to run the integration tests, you need to first build the image (make -C backends/gaudi image) # In order to run the integration tests, you need to first build the image (make -C backends/gaudi image)
run-integration-tests: run-integration-tests:
pip install -U pip uv
uv pip install -r ${root_dir}/backends/gaudi/server/integration-tests/requirements.txt uv pip install -r ${root_dir}/backends/gaudi/server/integration-tests/requirements.txt
DOCKER_VOLUME=${root_dir}/data \ DOCKER_VOLUME=${root_dir}/data \
HF_TOKEN=`cat ${HOME}/.cache/huggingface/token` \ HF_TOKEN=`cat ${HOME}/.cache/huggingface/token` \
@ -57,6 +58,7 @@ run-integration-tests:
# This is used to capture the expected outputs for the integration tests offering an easy way to add more models to the integration tests # This is used to capture the expected outputs for the integration tests offering an easy way to add more models to the integration tests
capture-expected-outputs-for-integration-tests: capture-expected-outputs-for-integration-tests:
pip install -U pip uv
DOCKER_VOLUME=${root_dir}/data \ DOCKER_VOLUME=${root_dir}/data \
HF_TOKEN=`cat ${HOME}/.cache/huggingface/token` \ HF_TOKEN=`cat ${HOME}/.cache/huggingface/token` \
uv run pytest --durations=0 -sv ${root_dir}/backends/gaudi/server/integration-tests/capture_expected_outputs.py uv run pytest --durations=0 -sv ${root_dir}/backends/gaudi/server/integration-tests/capture_expected_outputs.py

18
backends/gaudi/server/integration-tests/test_model.py
View File

@ -9,8 +9,8 @@ TEST_CONFIGS = {
"meta-llama/Llama-3.1-8B-Instruct-shared": { "meta-llama/Llama-3.1-8B-Instruct-shared": {
"model_id": "meta-llama/Llama-3.1-8B-Instruct", "model_id": "meta-llama/Llama-3.1-8B-Instruct",
"input": "What is Deep Learning?", "input": "What is Deep Learning?",
"expected_greedy_output": " A Beginner’s Guide\nDeep learning is a subset of machine learning that involves the use", "expected_greedy_output": " A Beginner’s Guide\nDeep learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. It is a type of",
"expected_batch_output": " A Beginner’s Guide\nDeep learning is a subset of machine learning that involves the use", "expected_batch_output": " A Beginner’s Guide\nDeep learning is a subset of machine learning that involves the use of artificial neural networks to analyze and interpret data. It is a type of",
"args": [ "args": [
"--sharded", "--sharded",
"true", "true",
@ -165,20 +165,6 @@ TEST_CONFIGS = {
"4", "4",
], ],
}, },
"facebook/opt-125m": {
"model_id": "facebook/opt-125m",
"input": "What is Deep Learning?",
"expected_greedy_output": "\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout",
"expected_batch_output": "\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout the Author\n\nAbout",
"args": [
"--max-input-tokens",
"512",
"--max-total-tokens",
"1024",
"--max-batch-size",
"4",
],
},
"EleutherAI/gpt-j-6b": { "EleutherAI/gpt-j-6b": {
"model_id": "EleutherAI/gpt-j-6b", "model_id": "EleutherAI/gpt-j-6b",
"input": "What is Deep Learning?", "input": "What is Deep Learning?",

1
backends/gaudi/server/text_generation_server/models/custom_modeling/flash_mistral_modeling.py
View File

@ -111,6 +111,7 @@ class MistralAttention(torch.nn.Module):
) )
self.num_heads = config.num_attention_heads self.num_heads = config.num_attention_heads
self.hidden_size = config.hidden_size self.hidden_size = config.hidden_size
if getattr(config, "head_dim", None) is not None: if getattr(config, "head_dim", None) is not None:
self.head_size = config.head_dim self.head_size = config.head_dim
else: else: