Upload ONNX weights (#1)

- Upload ONNX weights (6edb617608adcf360f6f4e5f65b632a358c30134)
- Update config.json (edadf3efbc15a1d230249cef604f66e5fa8684ff)


Co-authored-by: Joshua <[email protected]>

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+onnx/model.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/model_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text

config.json

@@ -25,6 +25,12 @@
   "tie_word_embeddings": true,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.42.3",
+  "transformers.js_config": {
+    "kv_cache_dtype": {
+      "q4f16": "float16",
+      "fp16": "float16"
+    }
+  },
   "use_cache": true,
   "vocab_size": 49152
 }

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c08976e4c0b0dc39cc120d99e3787913bb664d44a3d2ed42ec44360b7f14f50
+size 179811

onnx/model.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54b9e26f56ca92b748dab4970b460346cfad9dd9a0703390a900c6a31f0c1c25
+size 7250256084

onnx/model_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dcaade33f176b7c70f063b815ad431337405d96f72c6969fc61595870ad9013
+size 1528144599

onnx/model_fp16.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f48c05c14ed97738f8dc5854c20c229ddc8661f43fa914085843901a4ba8740
+size 2097152000

onnx/model_q4.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:579c1a6e5a0036b98c0d8404e35635777ff2a03d2f179491742ed6af4b546596
+size 1474840244

onnx/model_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f59ee54c919da8d680a5ddcc4fefefdd55f4f216f90a1a8397e8eb106915c129
+size 1165363569