Adding Compute-Context-Length (CCL) #576
Open
+2,213
−211
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
vjanfaza
requested review from
ochougul,
quic-amitraj,
quic-hemagnih and
quic-rishinr
as code owners
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
quic#557) Updated the run_vlm_kv_model_on_pytorch and run_vlm_kv_model_on_ort methods to run for the latest dual QPC setup. Along with the required changes to be made in the Input Handler of VLMs. Also updated the way head_dim is calculated for past_key_value creation as certain models now provide specific head_dim. We fallback to previous method if the parameter isn't found in the config. Signed-off-by: Dhiraj Kumar Sah <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Abukhoyer Shaik <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
Signed-off-by: Vahid Janfaza <[email protected]>
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Outdated suggestions cannot be applied.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.
Compute-Context-Length (CCL) technique optimizes the throughput of large language models (LLMs) on Qualcomm devices when handling very large context lengths. The current Ahead Of Time (AOT) compilation on Qualcomm devices doesn't predict the number of tokens needed, leading to significant throughput drops during the prefilling and the decoding phases. This happens because the system performs attention calculations based on large context length. To address this issue, we introduce Compute Context Length (CCL), an additional ONNX variable that allows for dynamic context-length specialization. By generating tokens using smaller, more manageable context lengths (CCL), we optimize memory reads and attention calculations, thereby improving throughput.