enqueue is there. testing inference side.

crates/cli/dev-database-seed/src/seeding/model_weights.rs

@@ -971,8 +971,8 @@ pub async fn seed_weights(mysql_pool: &Pool<MySql>) -> AnyhowResult<()> {
     //original_seed_weights(mysql_pool,user_token).await?;
     //seed_weights_for_user_token(mysql_pool, user_token).await?;
     //seed_weights_for_paging(mysql_pool,user_token).await?;
-    //seed_weights_for_testing_inference(mysql_pool,user_token).await?;
-    println!("TESTING DOWLOAD");
-    test_seed_weights_files().await?;
+    seed_weights_for_testing_inference(mysql_pool,user_token).await?;
+    //println!("TESTING DOWLOAD");
+    //test_seed_weights_files().await?;
     Ok(())
 }

crates/lib/mysql_queries/src/payloads/generic_inference_args/image_generation_payload.rs

@@ -73,9 +73,6 @@ pub struct StableDiffusionArgs {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub maybe_number_of_samples: Option<i32>,
 
-    #[serde(rename = "bz")]
-    pub maybe_batch_size: Option<i32>,
-
     #[serde(rename = "bc")]
     pub maybe_batch_count: Option<i32>,
 

crates/service/job/inference_job/config/inference-job.development.env

@@ -104,6 +104,14 @@ VITS_INFERENCE_MAYBE_DOCKER_IMAGE_SHA="03017896367b"
 RERENDER_INFERENCE_ROOT_DIRECTORY="/home/salt/clone/storyteller-ml/animation/rerender-lora/"
 RERENDER_INFERENCE_EXECUTABLE_OR_COMMAND="venv/bin/python rerender.py"
 
+# ================================= #
+#                                   #
+#          STABLE DIFFUSION         #
+#                                   #
+# ================================= #
+
+STABLE_DIFFUSION_INFERENCE_COMMAND="/home/tensor/code/storyteller/storyteller-ml/image-gen"
+
 # ================================= #
 #                                   #
 #          CACHE SETTINGS           #

crates/service/web/storyteller_web/src/http_server/endpoints/image_gen/enqueue_image_generation.rs

@@ -119,7 +119,6 @@ pub struct EnqueueImageGenRequest {
     maybe_lora_upload_path: Option<String>,
     maybe_cfg_scale: Option<i32>,
     maybe_number_of_samples: Option<i32>,
-    maybe_batch_size: Option<i32>,
     maybe_batch_count: Option<i32>,
 }
 
@@ -189,6 +188,7 @@ pub async fn enqueue_image_generation_request(
     server_state: web::Data<Arc<ServerState>>
 ) -> Result<HttpResponse, EnqueueImageGenRequestError> {
 
+    // TODO: Brandon need to figure out premium vs not premium
     let mut maybe_user_token: Option<UserToken> = None;
     let  visbility =  enums::common::visibility::Visibility::Public;
     let mut mysql_connection = server_state.mysql_pool.acquire().await.map_err(|err| {
@@ -235,6 +235,8 @@ pub async fn enqueue_image_generation_request(
         }
     }
 
+    // DETECT premium user and queue
+
     // ==================== RATE LIMIT ==================== //
 
     let rate_limiter = match maybe_user_session {
@@ -299,13 +301,9 @@ pub async fn enqueue_image_generation_request(
         None => 20,
     };
 
-    let batch_size = match request.maybe_batch_size {
-        Some(val) => if val > 4 { 4 } else { val },
-        None => 1,
-    };
 
     let batch_count = match request.maybe_batch_count {
-        Some(val) => if val > 2 { 2 } else { val },
+        Some(val) => if val > 4 { 4 } else { val },
         None => 1,
     };
 
@@ -336,7 +334,6 @@ pub async fn enqueue_image_generation_request(
     let type_of_inference = request.type_of_inference.to_string().clone();
 
 
-
     let inference_args = StableDiffusionArgs {
         maybe_sd_model_token: Some(sd_weight_token),
         maybe_lora_model_token: Some(lora_token),
@@ -350,7 +347,6 @@ pub async fn enqueue_image_generation_request(
         maybe_cfg_scale: Some(cfg_scale),
         maybe_number_of_samples: Some(number_of_samples),
         maybe_batch_count: Some(batch_count),
-        maybe_batch_size: Some(batch_size),
         maybe_width: Some(width),
         maybe_height: Some(height),
         maybe_sampler: Some(sampler)

crates/service/web/storyteller_web/src/http_server/endpoints/image_gen/json.txt

@@ -0,0 +1,18 @@
+{
+  "uuid_idempotency_token": "123",
+  "type_of_inference": "checkpoint",
+  "maybe_sd_model_token": "weight_dmmthavhawqc2hj7yqyemcbf8",
+  "maybe_lora_model_token": "weight_t7gz78fjg27m0wtw6r33gafxs",
+  "maybe_prompt": "a very well drawn, anime girl, with pink and purple hair sitting down on a chair relaxed.",
+  "maybe_a_prompt": "a anime girl, with pink and purple hair sitting down on a chair relaxed.",
+  "maybe_n_prompt": "nsfw, black and white, low quality, pixelated",
+  "maybe_seed": "-1",
+  "maybe_width": "512",
+  "maybe_height": "512",
+  "maybe_sampler": "DPM++ 2M SDE Heun",
+  "maybe_upload_path": "https://drive.google.com/file/d/1WRgR2pn0Ky8ls5_9Zq6tQlHTBvyWeach/view?usp=sharing",
+  "maybe_lora_upload_path": "https://drive.google.com/file/d/1WRgR2pn0Ky8ls5_9Zq6tQlHTBvyWeach/view?usp=sharing",
+  "maybe_cfg_scale": "7",
+  "maybe_number_of_samples": "32",
+  "maybe_batch_count": "4"
+}

crates/service/web/storyteller_web/src/routes.rs

@@ -195,10 +195,11 @@ pub fn add_routes<T, B> (app: App<T>, server_environment: ServerEnvironment) ->
   app = add_user_rating_routes(app); /* /v1/user_rating/... */
   app = add_subscription_routes(app); /* /v1/subscriptions/... */
   app = add_voice_designer_routes(app); /* /v1/voice_designer */
-
-  if server_environment == ServerEnvironment::Development {
-     app = add_weights_routes(app); /* /v1/stubs/... */
-  }
+    app = add_image_gen_routes(app);
+    app = add_weights_routes(app);
+  // if server_environment == ServerEnvironment::Development {
+  //
+  // }
   // ==================== Comments ====================
 
   let mut app = RouteBuilder::from_app(app)
@@ -217,6 +218,7 @@ pub fn add_routes<T, B> (app: App<T>, server_environment: ServerEnvironment) ->
       .add_get("/v1/user_bookmarks/list/entity/{entity_type}/{entity_token}", list_user_bookmarks_for_entity_handler)
       .into_app();
 
+
   // ==================== Animations ====================
 
   let mut app = RouteBuilder::from_app(app)