fix Llava test-bwd failure #639

jp1924 · 2025-03-31T06:00:03Z

Summary

Testing Done

Hardware Type:
run make test to ensure correctness
run make checkstyle to ensure code style
run make test-convergence to ensure convergence

convergence-test log

HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/fp32/test_mini_models.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 13 items

test/convergence/fp32/test_mini_models.py::test_mini_model[mini_llama3-32-0.0001-dtype0-1e-08-2e-05-0.0001-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [  7%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_llava-32-0.0001-dtype1-1e-08-1e-05-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:267 Support for transformers versions < 4.49.0 will soon be discontinued due to issues with incorrect legacy processing. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/35526
PASSED                                                                   [ 15%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_mllama-32-0.0001-dtype2-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 23%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_qwen2-32-0.0001-dtype3-1e-08-1e-05-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:855 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 30%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_qwen2_vl-32-0.0001-dtype4-1e-05-0.1-0.005-1e-05-0.005-1e-05] SKIPPED [ 38%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_qwen2_5_vl-32-0.0001-dtype5-1e-05-0.1-0.005-1e-05-0.005-1e-05] SKIPPED [ 46%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_olmo2-32-0.0001-dtype6-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 53%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_phi3-32-0.0001-dtype7-1e-08-1e-05-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:1067 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 61%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_mistral-32-0.0001-dtype8-1e-08-1e-05-0.005-1e-05-0.005-1e-05] PASSED [ 69%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_gemma1-32-0.0001-dtype9-1e-08-0.0001-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:598 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 76%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_gemma1.1-32-0.0001-dtype10-1e-08-0.0001-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:598 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 84%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_gemma2-32-0.0001-dtype11-1e-08-0.0001-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:672 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 92%]
test/convergence/fp32/test_mini_models.py::test_mini_model[mini_granite3-32-0.0001-dtype12-1e-08-0.0001-0.005-1e-05-0.005-1e-05] SKIPPED [100%]

============== 8 passed, 5 skipped, 1 warning in 69.42s (0:01:09) ==============
HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/fp32/test_mini_models_multimodal.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 6 items

test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_qwen2_vl-32-0.0001-dtype0-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 16%]
test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_llava-32-0.0001-dtype1-1e-08-1e-05-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 33%]
test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_qwen2_5_vl-32-0.0001-dtype2-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 50%]
test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_mllama-32-0.0001-dtype3-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 66%]
test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_paligemma-32-0.0001-dtype4-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 83%]
test/convergence/fp32/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_paligemma2-32-0.0001-dtype5-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [100%]

================== 1 passed, 5 skipped, 2 warnings in 30.71s ===================
HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/fp32/test_mini_models_with_logits.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 13 items

test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_llama3-32-0.0001-dtype0-1e-08-2e-05-0.0001-1e-05-0.005-1e-05] PASSED [  7%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_llava-32-0.0001-dtype1-1e-08-1e-05-0.005-1e-05-0.005-1e-05] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 15%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_mllama-32-0.0001-dtype2-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 23%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_qwen2-32-0.0001-dtype3-1e-08-1e-05-0.005-1e-05-0.005-1e-05] PASSED [ 30%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_qwen2_vl-32-0.0001-dtype4-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 38%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_qwen2_5_vl-32-0.0001-dtype5-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 46%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_olmo2-32-0.0001-dtype6-1e-08-1e-05-0.005-1e-05-0.005-1e-05] SKIPPED [ 53%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_phi3-32-0.0001-dtype7-1e-08-1e-05-0.005-1e-05-0.005-1e-05] PASSED [ 61%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_mistral-32-0.0001-dtype8-1e-08-1e-05-0.005-1e-05-0.005-1e-05] PASSED [ 69%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_gemma1-32-0.0001-dtype9-1e-08-0.0001-0.005-1e-05-0.005-1e-05] PASSED [ 76%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_gemma1.1-32-0.0001-dtype10-1e-08-0.0001-0.005-1e-05-0.005-1e-05] PASSED [ 84%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_gemma2-32-0.0001-dtype11-1e-08-0.0001-0.005-1e-05-0.005-1e-05] PASSED [ 92%]
test/convergence/fp32/test_mini_models_with_logits.py::test_mini_model[mini_granite3-32-0.0001-dtype12-1e-08-0.0001-0.005-1e-05-0.005-1e-05] SKIPPED [100%]

============== 8 passed, 5 skipped, 1 warning in 68.41s (0:01:08) ==============
HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/bf16/test_mini_models.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 12 items

test/convergence/bf16/test_mini_models.py::test_mini_model[mini_llama3-32-0.0001-dtype0-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [  8%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_llava-32-0.0001-dtype1-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:267 Support for transformers versions < 4.49.0 will soon be discontinued due to issues with incorrect legacy processing. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/35526
PASSED                                                                   [ 16%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_granite3-32-0.0001-dtype2-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 25%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_mllama-32-0.0001-dtype3-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 33%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_qwen2-32-0.0001-dtype4-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:855 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 41%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_qwen2_vl-32-0.0001-dtype5-0.001-0.05-0.1-0.01-0.01-0.01] SKIPPED [ 50%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_qwen2_5_vl-32-0.0001-dtype6-0.001-0.05-0.1-0.01-0.01-0.01] SKIPPED [ 58%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_phi3-32-0.0001-dtype7-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:1067 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 66%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_mistral-32-0.0001-dtype8-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 75%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_olmo2-32-0.0001-dtype9-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 83%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_gemma1-32-0.0001-dtype10-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:598 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 91%]
test/convergence/bf16/test_mini_models.py::test_mini_model[mini_gemma1.1-32-0.0001-dtype11-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:598 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [100%]

=================== 7 passed, 5 skipped, 1 warning in 46.95s ===================
HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/bf16/test_mini_models_multimodal.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 6 items

test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_qwen2_vl-32-0.0001-dtype0-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 16%]
test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_llava-32-0.0001-dtype1-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 33%]
test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_qwen2_5_vl-32-0.0001-dtype2-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 50%]
test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_mllama-32-0.0001-dtype3-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 66%]
test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_paligemma-32-0.0001-dtype4-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 83%]
test/convergence/bf16/test_mini_models_multimodal.py::test_mini_model_multimodal[mini_paligemma2-32-0.0001-dtype5-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [100%]

================== 1 passed, 5 skipped, 2 warnings in 19.27s ===================
HF_DATASETS_OFFLINE=1 python -m pytest --disable-warnings test/convergence/bf16/test_mini_models_with_logits.py
============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.3.5, pluggy-1.5.0
rootdir: /root/workspace/jp-liger
configfile: pyproject.toml

----------------------------- live log collection ------------------------------
INFO     datasets:config.py:54 PyTorch version 2.5.1+cu121 available.
collected 12 items

test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_llama3-32-0.0001-dtype0-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [  8%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_llava-32-0.0001-dtype1-0.001-0.01-0.1-0.01-0.01-0.01] 
-------------------------------- live log call ---------------------------------
WARNING  liger_kernel.transformers.monkey_patch:monkey_patch.py:209 Support for transformers versions < 4.46.1 will soon be discontinued due to issues with incorrect gradient accumulation. 
 Please consider upgrading to avoid potential issues. See details: https://github.com/huggingface/transformers/pull/34191
PASSED                                                                   [ 16%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_granite3-32-0.0001-dtype2-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 25%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_mllama-32-0.0001-dtype3-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 33%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_qwen2-32-0.0001-dtype4-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 41%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_qwen2_vl-32-0.0001-dtype5-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 50%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_qwen2_5_vl-32-0.0001-dtype6-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [ 58%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_phi3-32-0.0001-dtype7-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 66%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_mistral-32-0.0001-dtype8-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 75%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_gemma1-32-0.0001-dtype9-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 83%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_gemma1.1-32-0.0001-dtype10-0.001-0.01-0.1-0.01-0.01-0.01] PASSED [ 91%]
test/convergence/bf16/test_mini_models_with_logits.py::test_mini_model[mini_olmo2-32-0.0001-dtype11-0.001-0.01-0.1-0.01-0.01-0.01] SKIPPED [100%]

=================== 7 passed, 5 skipped, 1 warning in 50.33s ===================

env

transformers             4.44.2
torch                    2.5.1+cu121
torchaudio               2.5.1+cu121
torchvision              0.20.1+cu121

…move deprecated parameters

jp1924 · 2025-03-31T08:06:37Z

╭─ Modal Deprecation Warning (2025-02-06) ─────────────────────────────────────╮
│ Using Python module paths will require using the -m flag in a future version │
│ of Modal.                                                                    │
│ Use `modal run -m dev.modal.tests_bwd` instead.                              │
╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Modal Deprecation Warning (2025-01-08) ─────────────────────────────────────╮
│ modal.Mount usage will soon be deprecated.                                   │
│                                                                              │
│ Use image.add_local_dir instead, which is functionally and performance-wise  │
│ equivalent.                                                                  │
│                                                                              │
│ See https://modal.com/docs/guide/modal-1-0-migration for more details.       │
│                                                                              │
│ Source:                                                                      │
│ /home/runner/work/Liger-Kernel/Liger-Kernel/dev/modal/tests_bwd.py:14        │
│   repo = modal.Mount.from_local_dir(ROOT_PATH, remote_path=REMOTE_ROOT_PATH) │
╰──────────────────────────────────────────────────────────────────────────────╯
╭─ Error ──────────────────────────────────────────────────────────────────────╮
│ Token missing. Could not authenticate client. If you have token credentials, │
│ see modal.com/docs/reference/modal.config for setup help. If you are a new   │
│ user, register an account at modal.com, then run modal token new.          │
╰──────────────────────────────────────────────────────────────────────────────╯

These errors occur, but all tests pass successfully.

Tcc0403

Thx

jp1924 added 3 commits March 31, 2025 05:57

Add model input names to fast_tokenizer in multimodal processor tests

56bced1

Refactor lce_forward_deprecated to streamline input processing and re…

b60bb61

…move deprecated parameters

Remove unused logger import from llava.py

bfacf89

Tcc0403 approved these changes Mar 31, 2025

View reviewed changes

Tcc0403 merged commit f248529 into linkedin:main Mar 31, 2025
4 of 8 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix Llava test-bwd failure #639

fix Llava test-bwd failure #639

Uh oh!

jp1924 commented Mar 31, 2025 •

edited

Loading

Uh oh!

jp1924 commented Mar 31, 2025

Uh oh!

Tcc0403 left a comment

Uh oh!

Uh oh!

Uh oh!

fix Llava test-bwd failure #639

fix Llava test-bwd failure #639

Uh oh!

Conversation

jp1924 commented Mar 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Testing Done

env

Uh oh!

jp1924 commented Mar 31, 2025

Uh oh!

Tcc0403 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

jp1924 commented Mar 31, 2025 •

edited

Loading