feat(gallery): add Spark 1.7B BF16 variant

docs/content/features/model-gallery.md

@@ -45,8 +45,8 @@ catalog; the entry updates in place when the operation finishes.
 
 ## Spark-X2.5-1.7B
 
-Install Spark-X2.5-1.7B with automatic selection between its Q4_K_M and Q8_0
-GGUF builds:
+Install Spark-X2.5-1.7B with automatic selection among its Q4_K_M, Q8_0,
+and BF16 GGUF builds:
 
 ```bash
 local-ai models install spark-x2.5-1.7b-q4
@@ -58,6 +58,16 @@ To select the Q8_0 build explicitly:
 local-ai models install spark-x2.5-1.7b-q4 --variant spark-x2.5-1.7b-q8
 ```
 
+To select the BF16 build explicitly:
+
+```bash
+local-ai models install spark-x2.5-1.7b-q4 --variant spark-x2.5-1.7b-bf16
+```
+
+The BF16 weights require a 3.42 GB download. Automatic selection can choose
+BF16 when it fits available memory. Use `--variant spark-x2.5-1.7b-q4` or
+`--variant spark-x2.5-1.7b-q8` to keep a smaller build.
+
 These text-only builds use the llama.cpp backend and the embedded Jinja chat
 template. The gallery defaults to a 32,768-token context to limit memory use.
 The [source model](https://huggingface.co/XHToken/Spark-X2.5-1.7B) supports up to

gallery/index.yaml

@@ -4629,6 +4629,7 @@
   url: "github:mudler/LocalAI/gallery/virtual.yaml@master"
   variants:
     - model: spark-x2.5-1.7b-q8
+    - model: spark-x2.5-1.7b-bf16
   urls:
     - https://huggingface.co/XHToken/Spark-X2.5-1.7B
     - https://huggingface.co/XHToken/Spark-X2.5-1.7B-GGUF
@@ -4695,6 +4696,35 @@
       uri: huggingface://XHToken/Spark-X2.5-1.7B-GGUF/Spark-X2.5-1.7B-Q8_0.gguf
       sha256: cd77c03185a834bb1162a4b7713520be5838058bfc54873645beff470bb24442
 
+- !!merge <<: *spark-x2-5-1-7b
+  name: "spark-x2.5-1.7b-bf16"
+  variants: []
+  last_checked: "2026-09-15"
+  description: |
+    Spark-X2.5-1.7B is XHToken's 1.7B text model for conversation, reasoning,
+    coding, and multilingual tasks. This build uses BF16 GGUF weights,
+    the embedded Jinja chat template, and a 32K-token default context.
+  overrides:
+    backend: llama-cpp
+    context_size: 32768
+    known_usecases:
+      - chat
+    options:
+      - use_jinja:true
+    parameters:
+      model: llama-cpp/models/spark-x2.5-1.7b/Spark-X2.5-1.7B.gguf
+      temperature: 1
+      top_p: 0.95
+      top_k: -1
+      min_p: 0
+      repeat_penalty: 1
+    template:
+      use_tokenizer_template: true
+  files:
+    - filename: llama-cpp/models/spark-x2.5-1.7b/Spark-X2.5-1.7B.gguf
+      uri: https://huggingface.co/XHToken/Spark-X2.5-1.7B-GGUF/resolve/06c93e4e998a8e53e39b7b53d9bb842c9fbb5674/Spark-X2.5-1.7B.gguf
+      sha256: 67d5f2f06e6d898efcf0dc40cab8528bc82b871c8dafb0936784183d2c10cdd9
+
 - &spark-x2-5-4b
   name: "spark-x2.5-4b-q4"
   url: "github:mudler/LocalAI/gallery/virtual.yaml@master"