diff --git a/_sql/migrations_squashed/model_weights/model_weights.sql b/_sql/migrations_squashed/model_weights/model_weights.sql
index 4ec5ccca3d..72205f0de7 100644
--- a/_sql/migrations_squashed/model_weights/model_weights.sql
+++ b/_sql/migrations_squashed/model_weights/model_weights.sql
@@ -95,6 +95,7 @@ CREATE TABLE model_weights (
   -- The filename that was used at upload time (if available)
   original_filename VARCHAR(255) DEFAULT NULL,
 
+  -- TODO: Make this a BIGINT so we can store more than 2GB.
   -- File characteristics
   file_size_bytes INT(10) NOT NULL DEFAULT 0,
 
diff --git a/crates/service/job/inference_job/src/job/job_types/image_generation/sd/process_job.rs b/crates/service/job/inference_job/src/job/job_types/image_generation/sd/process_job.rs
index 51abe752ea..c6d6507f14 100644
--- a/crates/service/job/inference_job/src/job/job_types/image_generation/sd/process_job.rs
+++ b/crates/service/job/inference_job/src/job/job_types/image_generation/sd/process_job.rs
@@ -334,7 +334,8 @@ pub async fn process_job_sd(
         maybe_last_update_user_token: None,
         original_download_url: Some(download_url),
         original_filename: None,
-        file_size_bytes: metadata.file_size_bytes,
+        // file_size_bytes: metadata.file_size_bytes, // TODO(bt,2024-02-03): We need to migrate the column to be BIGINT
+        file_size_bytes: 0,
         file_checksum_sha2: metadata.sha256_checksum,
         public_bucket_hash: bucket_details.object_hash,
         maybe_public_bucket_prefix: Some(bucket_details.prefix),