Add new files to repository

config.json

@@ -25,7 +25,8 @@
   "rope_scaling": null,
   "rope_theta": 10000.0,
   "tie_word_embeddings": false,
-  "transformers_version": "4.48.3",
-  "use_cache": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.49.0",
+  "use_cache": true,
   "vocab_size": 32000
 }

generation_config.json

@@ -3,5 +3,5 @@
   "bos_token_id": 1,
   "eos_token_id": 2,
   "pad_token_id": 0,
-  "transformers_version": "4.48.3"
+  "transformers_version": "4.49.0"
 }

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9aa4bfd0970e36a19a1b2e6127a188d0a803398f0d29ab3e3e9b030acdae2139
+size 649920885

onnx/model_bnb4.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adc7059722bae4fb925d0a73f409c97666ece55f17fe2914aad51d2acc8e36ba
+size 176170279

onnx/model_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a4a785ae43577058a2ac5414c8dd3001f74d6945894b9cd1d4324e54fb69e85b
+size 325091713

onnx/model_int8.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb442954e76e56b333c0e2f56624f541bbb527fdc1a8277a20893874404386db
+size 162818537

onnx/model_q4.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2daffba27d3c9e8660b28a815802e406402b08b4b603ca2ecc358d70d4b043af
+size 184783537

onnx/model_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:385ae25e0416b73086b90bf7cf4687ae4b552f026a89da2e3a1af573fe0132fc
+size 126984987

onnx/model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb442954e76e56b333c0e2f56624f541bbb527fdc1a8277a20893874404386db
+size 162818537

onnx/model_uint8.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48d4f752716773d086c8507f315ef5390bcc18efff53d38e7981e225a25f66d9
+size 162818574

quantize_config.json

@@ -0,0 +1,18 @@
+{
+    "modes": [
+        "fp16",
+        "q8",
+        "int8",
+        "uint8",
+        "q4",
+        "q4f16",
+        "bnb4"
+    ],
+    "per_channel": false,
+    "reduce_range": false,
+    "block_size": null,
+    "is_symmetric": true,
+    "accuracy_level": null,
+    "quant_type": 1,
+    "op_block_list": null
+}