Upload folder using huggingface_hub

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "kajdun/iubaris-13b-v3_GPTQ",
+  "_name_or_path": "kajdun/iubaris-13b-v3-GPTQ",
   "architectures": [
     "LlamaForCausalLM"
   ],
@@ -12,7 +12,6 @@
   "max_length": 4096,
   "max_position_embeddings": 4096,
   "model_type": "llama",
-  "model_id": "iubaris-13b-v3_GPTQ", 
   "num_attention_heads": 40,
   "num_hidden_layers": 40,
   "num_key_value_heads": 40,

gptq_model-4bit-128g.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22fc368b0f95f28d645019d5ac1273aa5d0b0ddca032fb36c1afac5d05015b93
+size 7259449576

handler.py

@@ -14,7 +14,7 @@ model_directory = "/repository/"
 
 tokenizer_path = f"{model_directory}tokenizer.model"
 model_config_path = f"{model_directory}config.json"
-model_path = f"{model_directory}model.safetensors"
+model_path = f"{model_directory}gptq_model-4bit-128g.safetensors"
 
 class EndpointHandler():
     def __init__(self, path=""):

quantize_config.json

@@ -7,5 +7,5 @@
   "sym": true,
   "true_sequential": true,
   "model_name_or_path": null,
-  "model_file_base_name": "model"
+  "model_file_base_name": null
 }