feat: support Qwen2-VL & GME-Qwen2-VL model on npu device. #399
Conversation
This file contains hidden or 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
xanecdotex
requested review from
dongxianzhe,
liutongxuan,
xiao-yu-chen and
yq33victor
xanecdotex
force-pushed
the
qwen2vl_embedding
branch
from
6d904ed to
03edd98
Compare
xanecdotex
changed the title
xanecdotex
force-pushed
the
qwen2vl_embedding
branch
from
03edd98 to
59dac9c
Compare
xanecdotex
force-pushed
the
qwen2vl_embedding
branch
from
59dac9c to
ae16a2c
Compare
yq33victor
reviewed
Dec 12, 2025
| namespace layer { | ||
|
|
||
| #if defined(USE_NPU) | ||
| class Qwen2VisionEncoderLayer |
Collaborator
There was a problem hiding this comment.
please refer to the code: xllm/core/layers/llama_decoder_layer.h, don't use #ifdefine here.
| WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| See the License for the specific language governing permissions and | ||
| limitations under the License. | ||
| ==============================================================================*/ |
Collaborator
There was a problem hiding this comment.
what's the difference to npu/qwen2_vl.h npu/qwen2_vl_embedding.h ?
Collaborator
Author
There was a problem hiding this comment.
qwen2_vl_embedding.h是为了支持vlm embedding模型gme-qwen2-vl
xanecdotex
force-pushed
the
qwen2vl_embedding
branch
5 times, most recently
from
131db27 to
7bc6598
Compare
XuZhang99
reviewed
Dec 15, 2025
xllm/models/vlm/qwen2_vl.h
Outdated
|
|
||
| namespace xllm { | ||
|
|
||
| #define PrintTensor(tensor) print_tensor(tensor, #tensor, 10, true, false); |
Collaborator
There was a problem hiding this comment.
what does PrintTensor do?
Collaborator
Author
There was a problem hiding this comment.
沿用qwen2_5_vl.h风格，已删
xllm/models/vlm/qwen2_vl.h
Outdated
| } | ||
|
|
||
| private: | ||
| int dim_ = 0; |
Collaborator
There was a problem hiding this comment.
specify int is int32_t or int64_t.
xllm/models/vlm/qwen2_vl.h
Outdated
|
|
||
| int64_t d_model = model_args.mm_projection_dim(); // out_hidden_size | ||
| int context_dim = model_args.mm_hidden_size(); | ||
| int spatial_merge_size = model_args.mm_spatial_merge_size(); |
xanecdotex
force-pushed
the
qwen2vl_embedding
branch
from
7bc6598 to
71832a6
Compare
XuZhang99
approved these changes
Dec 16, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
No description provided.