Skip to content

Expose the context parameter during creation of TensorRT engines #1367

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jakstein wants to merge 2 commits into
base: master
Choose a base branch
from
+8 −5
Open

Expose the context parameter during creation of TensorRT engines #1367

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4695c99
add context to TensorRT building
jakstein May 3, 2026
089830f
Merge branch 'master' of https://github.com/mcmonkeyprojects/SwarmUI
jakstein May 3, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions src/BuiltinExtensions/ComfyUIBackend/ComfyUIWebAPI.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -266,7 +266,7 @@ public static async Task<JObject> ComfyInstallFeatures(Session session, string f
};

/// <summary>API route to create a TensorRT model.</summary>
public static async Task<JObject> DoTensorRTCreateWS(Session session, WebSocket ws, string model, string aspect, string aspectRange, int optBatch, int maxBatch)
public static async Task<JObject> DoTensorRTCreateWS(Session session, WebSocket ws, string model, string aspect, string aspectRange, int optBatch, int maxBatch, int contextLen = 75)
{
if (ModelsAPI.TryGetRefusalForModel(session, model, out JObject refusal))
{
Expand Down Expand Up @@ -336,7 +336,7 @@ public static async Task<JObject> DoTensorRTCreateWS(Session session, WebSocket
["batch_size_opt"] = optBatch,
["height_opt"] = prefY,
["width_opt"] = prefX,
["context_opt"] = 1,
["context_opt"] = contextLen / 75,
["num_video_frames"] = 25
}
};
Expand All @@ -360,8 +360,8 @@ public static async Task<JObject> DoTensorRTCreateWS(Session session, WebSocket
["width_opt"] = prefX,
["width_max"] = Math.Clamp(maxX, 256, 4096),
["context_min"] = 1,
["context_opt"] = 1,
["context_max"] = 128,
["context_opt"] = contextLen / 75,
["context_max"] = Math.Max(contextLen / 75, 128),
["num_video_frames"] = 25
}
};
Expand Down
1 change: 1 addition & 0 deletions src/Pages/_Generate/GenTabModals.cshtml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -247,6 +247,7 @@
</select>
<br>Preferred Batch Size: <input type="number" class="auto-number" id="tensorrt_batch_size" value="1" min="1" max="64" />
<br>Max Batch Size: <input type="number" class="auto-number" id="tensorrt_max_batch_size" value="4" min="1" max="64" />
<br>CLIP Context Length (tokens): <input type="number" class="auto-number" id="tensorrt_context" value="75" min="75" max="300" />
<div id="tensorrt_create_result_box">
</div>
</div>
Expand Down
4 changes: 3 additions & 1 deletion src/wwwroot/js/genpage/gentab/models.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1036,14 +1036,16 @@ function trt_modal_create() {
let rangeSelect = getRequiredElementById('tensorrt_aspect_range');
let batchSize = getRequiredElementById('tensorrt_batch_size');
let maxBatch = getRequiredElementById('tensorrt_max_batch_size');
let contextLen = getRequiredElementById('tensorrt_context');
let createButton = getRequiredElementById('trt_create_button');
let resultBox = getRequiredElementById('tensorrt_create_result_box');
let data = {
'model': modelSelect.value,
'aspect': aspectSelect.value,
'aspectRange': rangeSelect.value,
'optBatch': batchSize.value,
'maxBatch': maxBatch.value
'maxBatch': maxBatch.value,
'contextLen': contextLen.value
};
createButton.disabled = true;
resultBox.innerText = 'Creating TensorRT engine, please wait...';
Expand Down
Loading